2ac heartjnl bcs abstracts 2011

June 2011 Volume 97 Supplement I

British Cardiovascular Society Annual Conference13–15 June 2011

Manchester Central

heart.bmj.com www.bcs.com

heart97

Volum

e 97 Supplement I Pages A

1–A104

HEA

RT

June 2011

A ENDOTHELIAL CELL NITRIC OXIDE BIOAVAILABILITY ANDINSULIN SENSITIVITY ARE REGULATED BY IGF-1 ANDINSULIN RECEPTOR LEVELS

doi:10.1136/heartjnl-2011-300110.1

A Abbas, H Viswambharan, H Imrie, A Rajwani, M Kahn, M Gage, R Cubbon, J Surr,S Wheatcroft, M Kearney. Leeds Institute of Genetics Health and Therapeutics, Leeds, UK

Background In a similar manner to insulin, the growth promotinghormone Insulin-like Growth Factor-1 (IGF-1), may be an importantregulator of endothelial nitric oxide (NO) bioavailability. We havepreviously reported evidence of increased basal NO production inthe vasculature in two murine models of reduced IGF-1 receptor(global hemizygous knockout (IGFRKO) and endothelial cell specificIGF-1R knockout (ECIGFRKO)). Augmentation of this increase inNO is relative to progressive decrease in IGF-1R number (WT vsECIGFRKO hemizygotes p¼0.01, WT vs ECIGFRKO homozygotesp¼0.001). Furthermore, by decreasing IGF-1R numbers in theinsulin resistant hemizygous insulin receptor knockout (IRKO)model (IRKO 3 IGFRKO) we have shown insulin sensitivity in thevasculature can be restored. In this study, we have investigatedfurther these receptor interactions with the generation of a mouseoverexpressing the human IGF-1R specifically in the endotheliumunder control of the Tie-2 promoter-enhancer (hIGFREO), and bytargeted knockdown of the IGF-1R in human umbilical vein endo-thelial cells (HUVECs).Methods Metabolic function was assessed in mice by tolerance testsusing whole-blood micro-sampling after insulin or glucose intra-peritoneal injection. Cardiovascular function was assessed bythoracic aortic vasomotion ex vivo in the organbath. Complimentaryin vitro studies were conducted by siRNA mediated downregulationof the IGF-1 receptor in HUVECs with and wihout insulin stim-ulation. Nitric oxide synthase activity was measured using an assaymeasuring conversion of [14C]-L-arginine to [14 C]-L-citrulline.Results Glucose and insulin tolerance testing showed no differencebetween hIGFREO mice and wild-type (WT) littermates. Murinethoracic aorta from hIGFREO mice were hypercontractile tophenylepherine (PE) compared to WT (Emax hIGFREO¼0.9160.045 g; WT¼0.6260.045 g, p¼0.0036) with decreasedresponse to LNMMA (Emax hIGFREO¼47.7069.87 g;WT¼106.1630.10 g, p¼0.048). These data indicate reduced endo-thelial NO bioavailability in hIGFREO mice compared to WT.HUVECs transfected with IGF1R-siRNA showed increased basaland insulin mediated eNOS phosphorylation in the presence ofinsulin (Ins: 16464.9% vs siRNA+Ins: 19260.7%, p<0.05). eNOSactivity (L-arginine, L-citrulline assay) was enhanced upon trans-

fection with IGF1R-siRNA (Scrambled siRNA: 95.7613.7% vsIGF1R-siRNA: 188.7648.3%, p<0.05).Implications These data demonstrate that increasing numbers ofIGF-1R specifically in murine endothelium leads to reduced NObioavailability. Complementary siRNA studies confirm results ofprevious murine studies that reducing IGF-1R numbers enhance NObioavailability. Therefore this raises the intriguing possibility thatmanipulation of IGF-1R numbers may represent a novel therapeuticAbstract A Figure 1

Abstract A Figure 2

Abstract A Figure 3

Abstract A Figure 4

Heart June 2011 Vol 97 Suppl 1 A1

Young Research Workers’ Prize

strategy by which to modify vascular NO bioavailability andendothelial cell insulin sensitivity.

B VH-IVUS FINDINGS PREDICT MAJOR ADVERSECARDIOVASCULAR EVENTS. THE VIVA STUDY (VIRTUALHISTOLOGY INTRAVASCULAR ULTRASOUND INVULNERABLE ATHEROSCLEROSIS)

doi:10.1136/heartjnl-2011-300110.2

1P A Calvert, 1D R Obaid, 2N E J West, 2L M Shapiro, 2D McNab, 2C G Densem,2P M Schofield, 2D Braganza, 2S C Clarke, 2M O’Sullivan, 3K K Ray, 1M R Bennett.1University of Cambridge, Cambridge, UK; 2Papworth Hospital NHS Foundation Trust,Cambridge, UK; 3St George’s University of London, London, UK

Background Identification of high-risk atherosclerotic plaques offersopportunities for risk stratification and targeted intensive treatmentof patients with coronary artery disease. Virtual Histology intra-vascular ultrasound (VH-IVUS) has been validated in human athe-rectomy and post-mortem studies and can classify plaques intopresumed high- and low-risk groups. However, VH-IVUS-definedplaques have not been shown to be associated with major adversecardiovascular events (MACE), or biomarkers that confer increasedcardiovascular risk, such as serum cytokines or shortened leukocytetelomere length (LTL).Methods 170 patients with stable angina or troponin-positive acutecoronary syndrome (ACS), referred for percutaneous coronary

intervention (PCI) were prospectively enrolled and underwent full3-vessel VH-IVUS pre-PCI. Troponin-I (cTnI), IL-6, IL-18, hsCRP,neopterin, MCP-1 and sICAM-1 were measured pre-PCI and 24-hpost-PCI. LTL was determined by qPCR. The combined primaryendpoint (MACE) included unplanned revascularisation, myocardialinfarction (MI) and death, with a secondary endpoint of post-PCIMI (MI 4a).Results 18 MACE occurred in 16 patients (median follow-up: 625(463e990) days). 30 372 mm of VH-IVUS were analysed and 1106plaques classified (Abstract B Figure 1) locally and via a core-lab.After multivariable regression:1. Total number of non-calcified VH-IVUS-identified thin

capped fibroatheromata (VHTCFA) was the only factorindependently associated with MACE (HR¼3.16, (95%CI¼1.16 to 8.64), p¼0.025).

2. Total VHTCFA number (OR¼1.26 (1.03 to 1.53) p¼0.021)and total stent length (OR¼1.04 (1.01 to 1.08), p¼0.01) werethe only factors independently associated with MI 4a.

3. A novel 3-vessel vulnerability index (necrotic core: fibroustissue ratio) and side branch losswere independently associatedwith stenting-related cTnI rise (standardised beta coefficient(sb)¼0.29, p¼0.004 and sb¼0.23, p¼0.019 respectively).

4. Necrotic core area at the minimum luminal area frame wasthe only factor independently associated with ACS presenta-tion (OR¼1.59, p¼0.030).

5. Stented vessel VHTCFA number (OR¼1.75 (1.22 to 2.51),p¼0.002) was independently associated with the lower LTLtertile (DNA-based cardiovascular risk predictor).

6. Stenting-related IL-6 rise was the only biomarker independ-ently associated with MACE (HR¼1.03 (1.01e1.05),p¼0.007).

Conclusion We present the first report of an association betweenVHTCFA and MACE. This provides novel evidence that VHTCFAdefinitions are important in their own right (rather than asanalogues of histological TFCA definitions). We also present the firstreport of associations between VHTCFA and MI 4a as well as anovel vulnerability index that is association with stenting-relatedtroponin rise. Finally, we report a novel association betweenVHTCFA and DNA-based cardiovascular risk prediction (LTL).

Abstract A Figure 5

Abstract A Figure 6

Abstract B Figure 1

A2 Heart June 2011 Vol 97 Suppl 1


C INSULIN RESISTANCE IMPAIRS ANGIOGENIC PROGENITORCELL FUNCTION AND DELAYS ENDOTHELIAL REPAIRFOLLOWING VASCULAR INJURY

doi:10.1136/heartjnl-2011-300110.3

M B Kahn, N Yuldasheva, R Cubbon, J Surr, S Rashid, H Viswambharan, H Imrie,A Abbas, A Rajwani, M Gage, M T Kearney, S Wheatcroft. Leeds University, Leeds, UK

Introduction Insulin-resistance, the primary metabolic abnormalityunderpinning type-2-diabetes mellitus (T2DM) and obesity, is animportant risk factor for the development of atherosclerotic cardi-ovascular disease. Circulating-angiogenic-progenitor-cells (APCs)participate in endothelial-repair following arterial injury. Type-2diabetes is associated with fewer circulating APCs, APC dysfunctionand impaired endothelial-repair. We set out to determine whetherinsulin-resistance per se adversely affects APCs and endothelial-regeneration.Research Design and Methods We quantified APCs and assessedAPC-mobilisation and function in mice hemizygous for knockout ofthe insulin receptor (IRKO) and wild-type (WT) littermate controls.Endothelial-regeneration following femoral artery wire-injury wasalso quantified at time intervals after denudation and following APCtransfusion.Results The metabolic phenotype of IRKO mice was consistentwith compensated insulin resistance, with hyperinsulinaemia after aglucose challenge but a normal blood glucose response to a glucosetolerance test. IRKO mice had fewer circulating Sca-1+/Flk-1+APCs than WT mice at baseline. Culture of mononuclear-cellsdemonstrated that IRKO mice had fewer APCs in peripheral-blood,but not in bone-marrow or spleen, suggestive of a mobilisationdefect. Defective VEGF-stimulated APC mobilisation was confirmedin IRKO mice, consistent with reduced eNOS expression in bonemarrow and impaired vascular eNOS activity. Paracrine-angiogenic-activity of APCs from IRKO mice was impaired compared to thosefrom WT animals. Endothelial-regeneration of the femoral arteryfollowing denuding wire-injury was delayed in IRKO micecompared to WT (re-endothelialised area 35.864.8% vs 66.665.2%at day 5 following injury and 35.664.8% vs 59.866.6% at day 7;P<0.05) (Abstract C Figure 1A). Transfusion of mononuclear-cellsfrom WT mice normalised the impaired endothelial-regeneration inIRKO mice (5764% vs 2565%; p<0.002). Transfusion of c-kit+bone-marrow cells from WT mice also restored endothelial-regen-eration in IRKO mice (6262% vs 2565%; p<0.002). However,transfusion of c-kit+ cells from IRKO mice was less effective atimproving endothelial-repair (6262% vs 4564%; p<0.02) (AbstractC Figure 1B).Conclusions Insulin-resistance impairs APC function and delaysendothelial-regeneration following arterial injury. These findingssupport the hypothesis that insulin-resistance per se is sufficient tojeopardise endogenous vascular repair. Defective endothelial-repair

may be normalised by transfusion of APCs from insulin-sensitiveanimals but not from insulin-resistant animals. These data mayhave important implications for the development of therapeuticstrategies for insulin-resistance associated cardiovascular disease.

D UPTAKE OF ULTRASMALL SUPERPARAMAGNETICPARTICLES OF IRON OXIDE PREDICTS GROWTH INABDOMINAL AORTIC ANEURYSMS: A PILOT STUDY

doi:10.1136/heartjnl-2011-300110.4

J M J Richards, S I Semple, T J Mac Gillivray, C Gray, J P Langrish, M Williams,M Dweck, W Wallace, G McKillop, R T A Chalmers, O J Garden, D E Newby. Universityof Edinburgh, Edinburgh, UK

Background Prediction of abdominal aortic aneurysm (AAA)expansion and rupture is challenging and currently relies on serialmeasurements of maximum aneurysm diameter. Using ultrasmallsuperparamagnetic particles of iron oxide (USPIO) and MRI, weaimed to assess whether areas of cellular inflammation correlatedwith the rate of abdominal aortic aneurysm expansion.Methods and Results An image acquisition and data analysis algo-rithm for the detection of focal USPIO accumulation in tissues wasdeveloped. Patients (n¼29; 27 male; aged 7065 years) with asymp-tomatic AAA (4.0e6.6 cm) were recruited from an outpatientsurveillance programme and underwent 3T MRI before and 24e36 hafter administration of USPIO. The change in T2* value on T2*-weighted imaging was used to detect accumulation of USPIO withintheabdominal aortic aneurysm.Histologyof aorticwall tissue samplesconfirmed co-localisation and uptake of USPIO in areas with macro-phage infiltration. Patients were classified into one of three groups onthe basis of imaging findings (Abstract D Figure 1). Group 1: peril-uminal USPIO uptake only. Group 2: USPIO uptake throughout thethrombus. Group 3: USPIO uptake in the aorticwall. Patients in group3 with distinct mural uptake of USPIO had a threefold higher growthrate (n¼13; 0.66 cm/yr; p¼0.020) than those with no (Group 1; n¼7;0.22 cm/yr) or non-specific USPIO uptake (Group 2; n¼9; 0.24 cm/yr)despite having similar aneurysm diameters (5.460.6, 5.160.5 and5.060.5 cm respectively; p>0.05) and patient characteristics(p>0.05). In one patient with an inflammatory aneurysm, USPIOuptake and inflammation extended beyond the aortic wall intosurrounding tissues.Conclusion USPIO uptake in the aortic wall detects cellular inflam-mation in patients with AAA and appears to predict more rapidlyprogressive AAA expansion. This technique therefore holds majorpromise as a new method of risk-stratifying patients with AAA thatextends beyond the simple anatomical measure of aneurysm diameter.

Abstract D Figure 1

Abstract C Figure 1 (A) Time-dependent endothelial regenerationfollowing vascular injury (n=5mice per group; *denotes p<0.05). (B) Effectson endothelial regeneration 5 days after wire-injury of transfusion of spleen-derived MNCs or BM-derived c-kit (CD117)+ve cells fromWT or IRKO mice(n=4 mice per group).



Abstract D Figure 2

E INTEGRATIVE GENOMICS APPROACHES IDENTIFY NEWGENES CONTROLLING HEART RATE

doi:10.1136/heartjnl-2011-300110.5

1,2J S Ware, 3H Dobrzynski, 4M Pravanec, 1P J Muckett, 1S Wilkinson,5Y Jamshidi, 1T J Aitma, 6N S Peters, 1,2S A Cook. 1MRC Clinical Sciences Centre,Imperial College London, London, UK; 2National Heart & Lung Institute, ImperialCollege London, London, UK; 3School of Medicine, University of Manchester,Manchester, UK; 4Institute of Physiology, Czech Academy of Science, Prague, UK;5Division of Clinical Developmental Sciences, St. George’s University ofLondon, London, UK; 6National Heart & Lung Institute, Imperial College London,London, UK

Introduction Heart rate (HR) is a fundamental measure of cardiacfunction, and is of prognostic and therapeutic significance. Weapplied genetic and genomic approaches to identify new loci andgenes controlling HR in a rat model that has previously been used tofind human cardiovascular disease genes.Methods Telemetric aortic pressure transducers were implanted into226 animals from 33 rat strains: the Brown Norway, the Sponta-neously Hypertensive Rat, and 31 strains from a recombinant inbredpanel derived from these parental strains and HR was measured overseveral weeks. Statistical analyses were carried out using the Rpackage, and quantitative trait loci (QTL) identified by linkagemapping using QTL Reaper. Potential covariates of HR wereanalysed in SPSS. The sinus node (SN) and right atria (RA) of 20 ratswere microdissected (Abstract E Figure 1). Gene expression datawere generated with the Affymetrix Rat Gene 1.0 ST microarrayand analysed using Bioconductor. Differentially expressed geneswere identified using SAM & Limma. Genes in the QTL that wereexpressed in the SN were resequenced to identify potential causativesequence variants.Results Narrow sense heritability of HR in this population was51%, suggesting a large genetic contribution to HR. Linkagemapping identified a region on rat chromosome 13 controlling HR,with peak LOD score 6.7 (Abstract E Figure 2A). This QTL has notpreviously been identified in human, rat or mouse. Mean nocturnalHR in strains carrying the SHR allele was 388, compared with 357in BN-like strains; an allelic effect of 31bpm (8.7%, p<0.00005)that is equivalent to 5e9 bpm in humans, corresponding to adecreased risk of cardiovascular death of 10%e29%. Two inde-pendent approaches (linear regression modelling & correlationanalysis) confirmed that this effect was independent of potentialphysiological covariates, suggesting that the effect may be intrinsicto the heart, rather than due to neurohumeral influences.We havegenerated the first genomewide transcript expression profile of theSN. Three genes at the new HR locus were enriched in the SN

(Abstract E Figure 2B) and potentially causative sequence variantsin these 3 candidate genes have been identified. We have translatedthese findings to humans using data from a genome-wide associ-ation study population.Conclusions We have identified a new genomic locus for HR, whichdoes not contain genes in pathways already known to determineHR. We prioritised three candidate genes at the locus, which may betargets for therapeutic modulation of HR in patients with heartdisease.

Abstract E Figure 1 Small (1 mm2) pieces of tissue were isolated fromthe rat SN and distant trabeculated RA and RNA extracted for geneexpression profiling.

Abstract E Figure 2 (A) Interval mapping revealed a linkage peak onchr13. Linkage was strongest for nocturnal HR, with a LOD score of 6.7(p>0.00005). The horizontal line approximates to genomewidesignificance. (B) A volcano plot showing 3 genes significantly enriched inSN compared with RA.



Abstracts1 ROUTE OF ADMISSION IN STEMI: DO PATIENTS WHO

PRESENT DIRECTLY TO A PCI-CAPABLE HOSPITAL DIFFERFROM INTER-HOSPITAL TRANSFERS?

doi:10.1136/heartjnl-2011-300198.1

D Austin, Z Adam, J Shome, M Awan, A G C Sutton, J A Hall, R A Wright, D F Muir,N M Swanson, J Carter, MA de Belder. James Cook University Hospital, Middles-brough, UK

Background Rapid delivery of reperfusion therapy with PPCI is thegold standard treatment in STEMI. Systems have been developed,such as direct admission to a PCI-capable hospital, to minimise thetime from diagnosis to PPCI. Despite this, a significant minority ofpatients are initially admitted to non-PCI capable hospitals. The aimof this study was to determine whether patients differed in theircharacteristics, time to PPCI, and outcome by route of admission.Methods The study was performed in a single tertiary centre inNorth England. Data are collected routinely on all patients under-going PPCI and include demographic, clinical and procedural varia-bles. In-hospital MACCE (death, re-infarction or CVA) and mortalityare collected providing relevant outcome measures. Baseline clinicalvariables by route of admission were compared and unadjusted in-hospital MACCE rates determined. One-year mortality by route ofadmission was calculated using the K-M product limit estimate. In-hospital and 1-year outcomes were analysed after adjustment forfactors known to be predictors of early mortality following STEMI(models 1 and 3). To determine the relative importance of delays intreatment, call-to-balloon time was added (models 2 and 4). Logisticregression was used for the adjusted in-hospital outcomes, and Cox-proportional regression for adjusted 1-year mortality.Results 2268 patients were included in the analysis. 510 patients(22.5%) were treated with PPCI following transfer from a non-PCIcapable centre. Analysis of baseline variables (Abstract 1 table 1)showed the transfer group were more likely to have an LADocclusion treated, and previous MI. Despite shorter DTB times, thetransfer group had a greater median CTB time (52 minutes longer)compared with direct admissions. Other baseline variables werestatistically no different between groups. There were 110 in-hospitalMACCE events, and 168 deaths within 1-year follow-up. The transfergroup had significantly higher unadjusted in-hospital MACCE rates(2.4% absolute, 58% relative increase (Abstract 1 table 2)). At 1 year,the transfer group had significantly higher unadjusted mortality(2.7% absolute, 48% relative increase (Abstract 1 table 2)). Afteradjustment for relevant co-variates (models 1 and 3) route ofadmission remained a significant predictor of in-hospital and 1-yearmortality. With the addition of call-to-balloon time, no significant

difference in outcome was noted by route of admission for either in-hospital or 1-year events.

Abstract 1 Table 2

Direct Transfer OR (±95% CI) p

In-hospital MACCE 4.3% 6.7% 1.58 (1.04 to 2.39) 0.03

Adjusted in-hospitalMACCE (model 1)

1.64 (1.00 to 2.28) 0.05

Adjusted in-hospitalMACCE (model 2)

1.34 (0.79 to 2.29) 0.27

Direct Transfer HR (±95% CI) p

1-year mortality 7% 9.7% 1.48 (1.06 to 2.07) 0.02

Adjusted 1-yearmortality (model 3)

1.41 (0.99 to 2.01) 0.05

Adjusted 1-yearmortality (model 4)

1.29 (0.87 to 1.89) 0.20

Conclusion In this study, patients who presented directly hadsuperior in-hospital and 1-year outcomes compared with those whorequired transfer from other hospitals. Adjustment for longer call-to-balloon times attenuated the finding of poorer outcomes in thesepatients, suggesting that delays in treatment are critical. Systems ofcare should be designed to avoid admission of STEMI patients tonon-PCI hospitals, and facilitate more rapid transfer of patientswhere this has not been possible.

2 A “DIRECT” TRANSFER PROTOCOL FOR PATIENTS WITHNON ST-ELEVATION MYOCARDIAL INFARCTION REDUCESTIME TO CORONARY ANGIOGRAPHY

doi:10.1136/heartjnl-2011-300198.2

S M Gallagher, M J Lovell, D Jones, E Ferguson, S Antoniou, S Mohiddin, M Westwood,A Mathur, R A Archbold, C Knight, A K Jain. London Chest Hospital, Barts and theLondon NHS Trust, London, UK

Introduction Patients with non-ST elevation myocardial infarctions(NSTEMI) are at high risk of further cardiac events. Nationalguidelines recommend “early” coronary angiography within 96 h ofpresentation. Most patients with NSTEMI present to their districtgeneral hospital (DGH), and await transfer to the regional cardiaccentre for angiography. This care model has inherent time delays,and delivery of early angiography is problematic.Methods A novel clinical care pathway for the management ofNSTEMI, known locally as the Heart Attack Centre-Extension orHAC-X, has been investigated. This pathway identifies patients withNSTEMI by clinical assessment and rapid point-of-care troponintesting while in the emergency department (ED). Patients meetingcriteria for urgent transfer receive evidence based medical therapy forNSTEMI (see Abstract 2 table 1) in the ED, and are transferred to thetertiary centre within 1 h without referral. All unstable patients aretaken straight to the cardiac catheterisation laboratory. For stablepatients, coronary angiography is undertaken on the same day, or ifpatients arrive after 17:00 on the next available routine list. The studygroup consists of 775 patients divided into two groups; 464 patientstreated before the instigation of the HAC-X pathway (Pre-HACX),and 311 patients treated via the novel pathway (Post-HACX). Wehave undertaken a prospective observational study of post-HAC-Xpatients, assessing need for angiography and or revascularisationalong with discharge diagnosis. We have also compared the waitingtime for angiography of pre-HAC-X and post-HAC-X groups.Results 250/311 (80.4%) of HACX patients underwent angiography.Following angiography, 144/250 (57.6%) were treated with coronaryrevascularisation (108 (75%) PCI and 36 (25%) CABG). 106/250

Abstract 1 Table 1

Direct Transfer p

Age (years6SD) 64.3 (12.7) 63.9 (12.4) 0.17

Male 1252 (71.2) 367 (72.0) 0.74

Diabetes 177 (10.1) 55 (10.8) 0.68

Previous MI 225 (12.6) 89 (17.3) 0.001

Treated vessel 0.001

LMS 24 (1.4) 13 (2.5)

LAD 630 (36.1) 218 (42.9)

LCx 249 (14.3) 83 (16.3)

RCA 812 (46.6) 188 (37.0)

Graft 28 (1.7) 5 (1.1)

Cardiogenic shock 28 (1.7) 35 (6.9) 0.61

Smoking (ex/current) 1331 (75.7) 377 (73.9) 0.42

Call-to-balloon time 102 (82e135) 154 (107e235) <0.001

Door-to-balloon time 44 (29e76) 34 (24e50) <0.001


BCS Abstracts 2011

(42.4%) of patients were treated with medical therapy alone. NSTE-ACS (encompassing NSTEMI and unstable angina) was the dischargediagnosis for 75.4% of HACX patients. 10% of patients had anothercause for chest pain symptoms (including pericarditis and, myocar-ditis); 14.6%had a non-cardiac diagnosis.Mean time frompresentationto angiography was pre-HAC-X 7349 mins (66836) and post HAC-X754 mins (6458) (p<0.0001) (see Abstract figure 1). Pre-HAC-X mean

wait for transfer to tertiary centre was 4.1 (64.7) days. Median lengthof stay for HACX patients was 3 days. HAC-X has reduced wait forcoronary angiography by 3.4 days per patient.Conclusions This novel care pathway allows delivery of earlyangiography to NSTEMI patients in accordance with nationalguidance. Importantly, the pathway allows accurate diagnosis ofNSTEMI, and inappropriate transfers are infrequent. Its intro-duction has resulted in a significant reduction in time to angiog-raphy for NSTEMI patients, and significant reductions in totalhospital bed occupancy for patients with NSTEMI.

3 SURVIVAL FOLLOWING ACUTE MYOCARDIAL INFARCTIONIN PATIENTS OF SOUTH ASIAN AND WHITE EUROPEANETHNICITY IN THE UK

doi:10.1136/heartjnl-2011-300198.3

1N N Gholap, 2R L Mehta, 2K Khunti, 2M J Davies, 2I B Squire. 1University Hospitals ofLeicester, Leicester, UK; 2University of Leicester, Leicester, UK

Introduction Some UK studies have suggested higher case-fatalityrates following acute myocardial infarction (AMI) in British SouthAsian (SA), compared to white European (WE) people, driven byhigher prevalence of diabetes in the SA ethnic group. However otherstudies have suggested similar or even better adjusted overall post-AMIsurvival for these ethnic groups. In patients with AMI, both priordiagnosis of diabetes as well as acutely elevated blood glucose regardlessof diabetes status are associated with adverse outcomes. The aim ofthis study was to compare survival rates following AMI in SA andWE patients drawn from a contemporary, multi-ethnic UK population.Methods: We conducted a retrospective cohort study of total 4111(SA 18%) consecutive patients with AMI admitted between October2002 and September 2008. Baseline differences between the ethnicsubgroups were examined using independent two-sample t tests forcontinuous and c2 tests for categorical variables. Cox regressionmodels were constructed to identify determinants of 30-days and 1-year mortality, entering ethnicity, random admission blood glucoseand antecedent diabetes individually and together along with otherrelevant variables.Results: SA patients were younger (62 vs 67 years, p<0.005) and lesslikely to have smoked (16% vs 40%, p<0.005) but more likely tohave hypertension (55% vs 49%, p¼0.004) or diabetes (40% vs 16%,p<0.005) at presentation compared to WE patients. All cause 30-dayand 1-year mortality proportions were 10.0 % and 15.2% in SAcompared to respectively 9.9 % and 16.7 % in WE patients. For SAethnicity, the univariate HR of 30-day mortality was 1.01 (95% CI0.79 to 1.30) compared to WE ethnicity. On multivariate analysis(excluding antecedent diabetes and admission blood glucose) thisassociation of SA ethnicity and mortality became significant (HR1.56, CI 1.10 to 2.23) and remained so when antecedent diabeteswas added to the analysis (HR 1.48, CI 1.03 to 2.13). However whenadmission blood glucose was added to the model, association ofethnicity with mortality became non-significant (HR 1.31, CI 0.86to 1.99). Conversely each unit (mmol/l) increase in admission bloodglucose was associated with 7% increase in mortality (HR 1.07, CI1.04 to 1.10) in this model, after adjusting for all the covariates.Furthermore exclusion of ethnicity and antecedent diabetes fromthe model did not alter the predictive value of admission bloodglucose (HR 1.08, CI 1.05 to 1.10). Similar associations wereobserved for 1-year mortality.Conclusions Despite higher prevalence of diabetes in SA patients,their mortality post AMI was similar to WE patients. Furthermore,admission hyperglycaemia more so than antecedent diabetes was animportant predictor of increased mortality post AMI. To improvesurvival, active management of admission hyperglycaemia should beconsidered in patients admitted with AMI, regardless of theirdiabetes status or ethnicity.

Abstract 2 Figure 1 Time from ED presentation to coronaryangiography.

Abstract 2 Table 1 Inclusion and exclusion criteria for HACX

Inclusion criteria Symptoms suggestive of myocardial ischaemia

With ECG changes including: ST depression; T waveinversion in V1-4; Dynamic T wave changes

OR positive troponin I assay

Exclusion criteria Unexplained anaemia (Hb<10)

Hypoxia

Acute renal failure

Loss of consciousness

Recent trauma

Overt sepsis

Immediate medical therapy includes Aspirin 300 mg

Clopidogrel 600 mg

Fondaparinux 2.5 mg

Eptifibatide bolus (180 mg/kg) as long as nobleeding contraindications


BCS Abstracts 2011

4 A RATIONAL APPROACH TO RAISED TROPONINS ON AHYPERACUTE STROKE UNIT: COPING WITH THE IMPACT ONCARDIOLOGY SERVICES

doi:10.1136/heartjnl-2011-300198.4

S S Nijjer, G Banerjee, J Barker, S Banerjee, S Connolly, K F Fox. Imperial CollegeHealthcare NHS Trust, London, UK

Introduction Troponin is frequently measured on admission tohyperacute stroke units (HASUs). Modest elevations in stroke arecommon but whether patient management changes in response tothe blood test is unclear. Raised troponin without chest pain ordynamic ECG changes create diagnostic dilemmas. Managementstrategies were assessed with the introduction of the Imperial HASUcovering North West London.Methods Consecutive HASU admissions over 6 months wereassessed for measurement of troponin, the result, and the cardiacinvestigations performed. Clinical parameters guided investigationslead by two Consultant Cardiologists (KF, SC) rather than strictresearch protocol.Results 412 patients were admitted: 245 patients had a total of 435troponin-I levels measured, without chest pain or dynamicischaemic ECG changes. 70 (29%) patients had positive levels(>0.032 ng/l): 53 (22%) were “low” (0.032e0.3 ng/l), 17 (7%) were“high” (>0.3 ng/l). 237 had diagnoses readily available: 170 hadstroke (ischaemic or haemorrhagic), 67 had non-stroke (eg, seizure).Troponin was more likely to be raised if stroke, OR 4.3 (2.0e9.7,p¼0.0001). Five patients with “high” troponins had non-invasivestress testing (1 perfusion scan and 4 stress echos): all were negative.All positive troponins had echocardiography and cardiology reviewwith no change in management in 91% of cases. 6 patients hadinvasive coronary angiography: 3 “high” and 3 “low” troponin. Only2 patients (3% of those with positive troponin) required percuta-neous coronary intervention (PCI); both had troponin >0.3 andmultiple cardiac risk factors. Patients with troponin <0.3 did notrequire PCIeall three had normal coronary arteries.Conclusions Every positive troponin necessitated a review andadditional tests, increasing demand on cardiology services withoutincrease in resources. However only 2 patients required PCI withthe majority medically managed. We propose a pragmatic pathwayfor when troponin is performed as a routine test. Raised troponins>0.3 ng/l should be assessed for chest pain and ECG changessuggesting true myocardial infarction. Without these, non-invasiveassessment and optimal medical therapy is sufficient in themajority. Minor troponin rise (0.032e0.3 ng/l) represents myocy-tolysis: cerebral insular damage causes sympathoadrenal activationand patchy myocyte damage. Without chest pain or ECG changes,optimal medical management without further investigation isappropriate. Since this does not represent true acute coronarysyndrome, an early invasive strategy confers no additional benefitover medical therapy. In contrast, aspirin and statins benefitboth stroke and any coronary disease present. The financial andmedical implications of performing non-indicated tests in a routinemanner when the result will be disregarded is significant. There-fore, we caution against routine measurement of troponin instroke.

5 IMPLEMENTATION OF A SENSITIVE TROPONIN I ASSAYREDUCES DEATH AND RECURRENT MYOCARDIALINFARCTION IN PATIENTS WITH SUSPECTED ACUTECORONARY SYNDROME

doi:10.1136/heartjnl-2011-300198.5

1K K Lee, 1N L Mills, 1A M D Churchhouse, 1A Anand, 1D Gamble, 1A Shah, 1E Paterson,2M MacLeod, 3C Graham, 4S Walker, 1M A Denvir, 1K A A Fox, 1D E Newby. 1Centre for

Cardiovascular Science, Edinburgh University, Edinburgh, UK; 2Edinburgh Heart Centre,Royal Infirmary of Edinburgh, Edinburgh, UK; 3Epidemiology and Statistics Core, Well-come Trust Clinical Research Facility, Edinburgh, UK; 4Department of Clinical Biochem-istry, Royal Infirmary of Edinburgh, Edinburgh, UK

Introduction Although troponin assays have become increasinglymore sensitive, it is unclear whether further reductions in thethreshold of detection for plasma troponin concentrations impacton clinical outcomes in patients with suspected acute coronarysyndrome. The aim of this study was to determine whetherlowering the diagnostic threshold for myocardial infarction with asensitive troponin assay will improve clinical outcomes.Methods Consecutive patients admitted with suspected acutecoronary syndrome before (n¼1038; validation phase) and after(n¼1054; implementation phase) lowering the threshold of detec-tion for myocardial necrosis from 0.20 to 0.05 ng/ml with a sensitivetroponin I assay were stratified into three groups: <0.05, 0.05e0.19and $0.20 ng/ml. During the validation phase, only concentrationsabove the original diagnostic threshold of $0.20 ng/ml werereported to clinicians. Event-free survival (reinfarction and death) at1 year were compared in patients grouped by plasma troponinconcentrations.Results Plasma troponin concentrations were <0.05 ng/ml in 1340(64%), 0.05e0.19 ng/ml in 170 (8%) and $0.20 ng/ml in 582 (28%)patients. During the validation phase, 39% of patients withundisclosed plasma troponin concentrations of 0.05e0.19 ng/mlwere dead or had recurrent myocardial infarction at 1 year,compared to 7% and 24% of those with troponin concentrations<0.05 ng/ml (p<0.001) or $0.20 ng/ml (p¼0.007) respectively.During the implementation phase, lowering the diagnosticthreshold to 0.05 ng/ml reduced 1-year death and recurrentmyocardial infarction from 39% to 21% in patients with troponinconcentrations of 0.05e0.19 ng/ml (OR 0.42, 95%CI 0.24 to 0.84,p¼0.013), whereas clinical outcomes were unchanged in patientswith troponin concentrations <0.05 ng/ml or $0.20 ng/ml(Abstract 5 figure 1).

Abstract 5 Figure 1

Conclusions In patients with suspected acute coronary syndrome,implementation of a sensitive troponin assay increases the diagnosisof myocardial infarction by a third, and identifies those at high-riskof reinfarction and death. Lowering the diagnostic threshold ofplasma troponin is associated with major reductions in morbidityand mortality.


BCS Abstracts 2011

6 CARDIAC MORBIDITY AND MORTALITY CAN BEACCURATELY PREDICTED IN PATIENTS PRESENTING WITHACS USING MULTIPLE BIOMARKERS MEASURED ON ANADMISSION BLOOD SAMPLE

doi:10.1136/heartjnl-2011-300198.6

1I R Pearson, 1K Viswanathan, 1N Kilcullen, 2A S Hall, 2C P Gale, 1U M Sivananthan,1J H Barth, 2C Morrell. 1Leeds Teaching Hospitals, Leeds, UK; 2University of Leeds,Leeds, UK

Background Rapid assessment of patients with suspected acutecoronary syndrome (ACS) allows the right patients to receive theright treatment at the right time. Discrimination of risk permitsclinical triage into pathways of immediate inpatient or deferredoutpatient care. It is known that a significant proportion of theACS patients sent home following an “MI screen”, based on anegative 12-h troponin level, are misdiagnosed as having non-cardiac chest pain when in fact they are at high risk of cardiacevents. It has been shown that the novel biomarker H-FABP candetect myocardial ischaemia even in the absence of myocytenecrosis. We hypothesise that a multi biomarker blood test incor-porating troponin I, CK-MB and H-FABP, taken on admission,can accurately discriminate those patients with a non-cardiaccause of chest pain who are at low risk of cardiac morbidity ormortality.Methods We studied 519 patients with suspected ACS admitted toa single UK Teaching Hospital. A risk scoring model wasconstructed based on tertile values for Randox Cardiac-Arraymeasurement of troponin I, H-FABP and CK-MB. These weremeasured on a blood sample taken at the time of hospital admis-sion. The lowest two lower tertiles were each given a score of 1 andthe top tertile a score of 3. The scores were then combined bysummation resulting in an overall score of between 3 and 9.Outcome measures up to 12 months were: (i) death from all causes;(ii) repeat acute coronary syndrome (ACS) (iii); readmission forheart failure; (iv) readmission for cerebrovascular event (CVA); (v)coronary revascularisation.Results The distribution of Cardio-Array scores was: 3 (n¼164;31.6%); 5 (n¼134; 25.8%); 7 (n¼110; 21.2%); 9 (n¼111; 21.4%). Thecumulative incidence of events according to the Cardiac-Array scoreis shown in Abstract 6 table 1.

Abstract 6 Table 1 The cumulative incidence of events according tothe Cardiac-Array Score

Score Death or ACS or HF or CVA or Revasc

3 0.61% 3.07% 3.11% 3.11% 4.28%

5 3.21% 5.77% 5.81% 5.81% 6.41%

7 11.11% 17.78% 19.05% 20.93% 24.44%

9 12.98% 16.23% 18.37% 18.92% 22.08%

Ratio (9/3) 21.28 5.29 5.91 6.08 5.16

p Value <0.0001 <0.0001 <0.0001 <0.0001 <0.0001

Conclusion Patients presenting with possible ACS who have aCardiac-Array biomarker score of 3 or 5, as measured on theiradmission blood sample, have a very low rate of cardiovascularevents. This tool could be used to safely triage patients towardsearly discharge and outpatient care, based upon available resources.A score of 7 or 9 would merit admission to hospital, and consid-eration of early cardiac catheterisation.

7 IN ACUTE CORONARY SYNDROMES, HEART-TYPE FATTYACID BINDING PROTEIN IS A MORE ACCURATE PREDICTOROF LONG TERM PROGNOSIS THAN TROPONIN

doi:10.1136/heartjnl-2011-300198.7

1I R Pearson, 2A S Hall, 2C P Gale, 1U M Sivananthan, 1K Viswanathan, 1N Kilcullen,2C Morrell, 1J H Barth. 1Leeds Teaching Hospitals, Leeds, UK; 2University of Leeds,Leeds, UK

Introduction We have previously shown that heart-type fatty acidbinding protein (H-FABP) has a role in predicting all-causemortality after acute coronary syndromes (ACS) and after multi-variable analysis, provides additional information to that gainedfrom the GRACE clinical risk factor score, troponin and highlysensitive CRP. H-FABP is released into the circulation duringmyocardial ischaemia and after myocardial necrosis, in contrast totroponin which is released after myocardial necrosis only. Wehave also shown that there is a group of ACS patients who are athigh risk of cardiac events and death despite normal troponinlevels on admission. This group may benefit from an early invasivestrategy.Hypothesis Plasma H-FABP level, taken between 12 and 24 h afteradmission, can identify troponin negative ACS patients who are at ahigh long term risk of death.Methods Six-year mortality data is now available for patientsenrolled in the FAB 1 study, for which 1-year mortality data waspublished in 2007. In this study, 1448 unselected patients admittedto hospital with ACS had serum H-FABP level measured in additionto usual care. Mortality was tracked by the UK Office of NationalStatistics.Results At 6 years overall all-cause mortality, available for 1421patients (98.1%), was 43.5%. If troponin �ve/H-FABP �vemortality was 20.9%; troponin �ve/H-FABP +ve 56.4%; troponin+ve/H-FABP �ve 20.2%; troponin +ve/H-FABP +ve 49.1%.Mortality rate was independent of troponin status but stronglyrelated to H-FABP status.Conclusion The current system of stratification of ACS patients forearly invasive management if troponin positive will miss a cohort ofpatients who are at high risk of death despite being troponinnegative, and who may benefit from invasive investigation.Conversely, it is likely that some ACS patients undergo angiographybased on a false positive troponin level. The addition of H-FABPmeasurement to the management of ACS could avoid this.

Abstract 7 Figure 1


BCS Abstracts 2011

8 DOES THE ADDITION OF A RADIAL ARTERY GRAFTIMPROVE SURVIVAL AFTER HIGHER RISK CORONARYSURGERY? A PROPENSITY-SCORE ANALYSIS

doi:10.1136/heartjnl-2011-300198.8

1C H Yap, 2P A Hayward, 2W Y Shi, 1D T Dinh, 1C M Reid, 3,4G C Shardey,3,4J A Smith. 1Department of Epidemiology and Preventative Medicine, MonashUniversity, Melbourne, UK; 2Department of Cardiac Surgery, Austin Hospital, Universityof Melbourne, Melbourne, UK; 3Department of Cardiothoracic Surgery and Surgery,Monash Medical Centre, Monash University, Melbourne, UK; 4Department of Surgery,Monash Medical Centre, Monash University, Melbourne, UK

Introduction The use of the radial artery as a second arterial graftduring coronary surgery has become popular due to high patency,encouraging clinical outcomes and low harvest site complicationrates. However it is not clear whether higher risk patients derivesuch benefits. We sought to assess this by examining outcomes inhigher risk subgroups.Methods A multicentre database was analysed. From 2001 to 2009,11 388 patients underwent isolated multivessel coronary surgery. Weidentified a higher risk subgroup (n¼3149) according to emergentstatus, coronary instability, low ejection fraction, aortic counter-pulsation or anticoagulant status. Among these, 2231 (71%) receivedat least 1 radial artery graft in addition to a left internal thoracicartery (LITA). The remaining 918 (29%) received LITA and veinsonly. Propensity-score matching and adjustment was performed tocorrect for group differences.Results Patientswho did not receive a radial arteryweremore likely tobe older (mean age, radial: 66610 years vs vein: 71610, p<0.0001)female (22% vs 27%, p¼0.002), have poor left ventricular function(16% vs 23%, p<0.0001), left main stenosis (35% vs 41%, p¼0.002) orbe of emergent status (11% vs 24%, p<0.0001). These patients expe-rienced higher unadjusted 30-daymortality (2.2% vs 7.1%, p<0.0001)and poorer 7-year survival (p<0.0001). Furthermore, 548 patients inthe radial groupwere propensity-scorematched to 548 receiving LITAandveins.At 30 days, therewere comparable rates ofmortality (radial:2% vs vein: 3%, p¼0.19), stroke (1% vs 1%, p¼0.51), myocardialinfarction (1% vs 1%, p¼0.77), major adverse cardiac or cere-brovascular events (MACCE) (2% vs 4%, p¼0.12), return to theatre(5% vs 7%, p¼0.19), hospital readmissions (12% vs 12%, p>0.99) andcombined anymortality/morbidity (30% vs 32%, p¼0.33). At 7 years,survival between radial and vein groups was similar (7962.5% vs8062.5%, p¼0.74). Propensity-adjusted multivariable regression didnot show radial artery to beprotective from30-daymortality (p¼0.14,OR0.67, 0.40 to1.13), 30-dayMACCE (p¼0.23,OR0.76, 0.48 to 1.20),or mid-term mortality (p¼0.79, HR 0.97, 0.78 to 1.20).Conclusions This multicentre analysis suggests that patients withthe greatest coronary instability, urgency of surgery, or impairmentof ventricular function are not disadvantaged in the early and mid-term by use of a single arterial graft. Limitations include the inabilityto correct for unquantifiable variables retrospectively. Despite this,surgeons may utilise clinical judgement to select radial or venousconduits to supplement the LITA according to other patient factorsor technical preference without prejudicing outcome.

9 EARLY HOSPITAL DISCHARGE AT 48 H FOLLOWINGPRIMARY PCI FOR MYOCARDIAL INFARCTION IS BOTHSAFE AND FEASIBLE

doi:10.1136/heartjnl-2011-300198.9

O Guttmann, D A Jones, K S Rathod, M Akhtar, A Ludman, A K Jain, C Knight,A Mathur, S Mohiddin, A Wragg, E J Smith. Barts and the London NHS trust, London,UK

Introduction Reperfusion therapy with primary PCI (PPCI) hasreduced rates of recurrent ischaemia and arrhythmia following ST

elevation myocardial infarction (STEMI), resulting in shorterhospital stays. Discharge at 72 h in selected patients has beensuggested. We investigated the feasibility and safety of very earlydischarge (<48 h) coupled with regular outpatient support for low-risk patients following PPCI.Methods 2317 patients underwent PPCI for STEMI betweenOctober 2003 and May 2010 at a regional Heart Attack Centre(HAC). Demographic and procedural data were documented at thetime of intervention. Patients with TIMI 3 flow, ST segment reso-lution, good or moderate left ventricular function, and nodysrhythmia were stratified to 48 h discharge. Remaining patientswere discharged according to physician preference. All patients werereviewed at 1, 8 and 52 weeks with a multidisciplinary teamincluding rehabilitation, heart failure, and psychology. The primaryendpoint was major adverse cardiac events (MACE) included death,myocardial infarction (MI), stroke and target vessel revascularisa-tion (TVR). All-cause mortality data were provided by the Office ofNational Statistics via the BCIS CCAD national audit. Outcomeswere compared between those discharged at#48 h, 72 h, and >72 h,out to 5 years of follow-up.Results 1079 patients (46.5%) were stratified to 48-h discharge,14% discharged at 72 h and the remainder discharged at a median of6 days (4.3e10), including those with complications. Patientsdischarged at #48 h were significantly younger and had a lowerincidence of multi-vessel disease than those discharged at 72 h(Abstract 9 table 1). Remaining baseline characteristics were similar.MACE at 3 years was similar between 48-h discharge patientsand 72- h discharge (9.1% vs 8.7%, p¼0.7). This persisted out to5 years (9.6% vs 9%, p¼0.55). As expected patients with lengthof stays >72 h had significantly worse outcomes (Abstract 9figure 1).

Abstract 9 Table 1

48 h (n[1079) 72 h (n[323) p Value

Age 60.7 64.0 0.0002

Previous MI 130 (12.0%) 35 (10.8%) 0.5569

Previous CABG 21 (1.9%) 7 (2.2%) 0.8019

Previous PCI 102 (9.4%) 29 (9.0%) 0.8007

DM 156 (14.4%) 52 (16.1%) 0.4632

HTN 455 (42.1%) 148 (45.8%) 0.2854

Hchol 403 (37.3%) 124 (38.4%) 0.7858

3 vessel disease 448 (46.4%) 156 (54.9%) 0.0112

Abstract 9 Figure 1 MACE after primary PCI.

Conclusion Early discharge at 48 h is feasible and appears to be safefor patients undergoing contemporary Primary PCI.


BCS Abstracts 2011

10 EVALUATING A NURSE LED TRIAGE PROCESS IN TREATINGPATIENTS WITH LEFT BUNDLE BRANCH BLOCK (LBBB)REFERRED FOR PRIMARY PERCUTANEOUS CORONARYINTERVENTION (PPCI)

doi:10.1136/heartjnl-2011-300198.10

1N V Joshi, 2B R Bawamia, 2S Jamieson, 2A Zaman, 2R Edwards. 1Centre forCardiovascular Science, The University of Edinburgh, Edinburgh, UK; 2The Cardio-thoracic Centre, Freeman Hospital, Newcastle Upon Tyne, UK

Background The Freeman Hospital (FRH) performs over 900 pPCI ayear. Patients with suspected Acute Myocardial Infarction (AMI) arereferred either by paramedics or networked hospitals for consid-eration of pPCI via a Telmed system, which is triaged by experiencedCCU nurses. The pPCI Pathway can be activated in patients withLBBB suspected of having an AMI. However, there remainsconsiderable variation in the clinical utility of new or presumed newLBBB as a ST-elevation myocardial infarction (STEMI)-equivalentECG diagnostic criterion. The major discriminators the triage staffuse in this population are ECG findings and symptoms suggestive ofAMI. Our aim was to evaluate outcomes in patients with LBBBaccepted to FRH or referred to local hospitals for assessment.Methods Consecutive patients referred to FRH with LBBB andsuspected AMI from 1st August 2009 to 30th November 2009 wereanalysed by recording: 1) Peak Troponin Level 2) Angiographicfindings 3) Revascularisation rates.Results 1069 patients were referred for consideration of pPCI. 177(16.6%) of patients had new or presumed new LBBB. 33 (18.6%)patients were accepted by FRH and 144 patients (81.4%) weredeclined and referred to their local hospitals for assessment. Abstract10 Table 1 Troponin levels in patients with LBBB referred forconsideration of pPCI. 26.5% of patients with LBBB referred forconsideration of pPCI had moderately to highly raised troponin. Ofthe 33 patients admitted to FRH, 13 underwent inpatient angiog-raphy and 9 patients had significant coronary disease (coronarystenosis 70%e100% in at least one coronary artery). Of those, 5 hadPCI and 1 required urgent CABG. Only one patient had a 100%coronary occlusion believed to be an acute occlusion. 4 patients hadunobstructed coronaries and were managed medically. Of the 132patients declined for pPCI only 2 (1.5%) were referred back to FRHfor PCI. Neither of these patients was found to have a 100% acuteocclusion of a coronary artery.

Abstract 10 Table 1

FRH Assessed FRH Declined

Number of patients 33 144

Number analysed 33 132 (Biochemistrydata not found for12 patients)

Troponin levels

Trop I<0.04 or Trop T<0.01 (Normal) 19 (57.6%) 84 (63.6%)

Trop I 0.04e0.1 or Trop T 0.01e0.1 (Mildly raised) 2 (6.1%) 25 (18.9%)

Trop I or Trop T 0.1e1.0 (Moderately raised) 1 (3.0%) 17 (12.9%)

Trop I or Trop T>1.0 (High) 11 (33.3%) 6 (4.5%)

Conclusion Revascularisation was performed in only 6/33 (18.2%)accepted for assessment and only 2/132 (1.5%) were referred back tothe centre for PCI. The sensitivity of the triage process in detectingpatients with LBBB requiring urgent revascularisation is 75% andthe specificity is 83%. The sensitivity of detecting patients with anacutely occluded artery diagnosed at angiography is 100% with aspecificity of 81%. In a high volume Heart Attack Centre a nurse ledtriage is effective at discriminating patients with LBBB requiringimmediate coronary intervention.

11 COPEPTIN IMPROVES EARLY RISK STRATIFICATION BYGRACE SCORE IN NON ST-ELEVATION MYOCARDIALINFARCTION; NT-PROBNP DOES NOT

doi:10.1136/heartjnl-2011-300198.11

1O S Dhillon, 1H K Narayan, 1P A Quinn, 2J Struck, 1I B Squire, 1J E Davies, 1L L Ng.1University of Leicester, Leicester, UK; 2Brahms, Hennigsdorf, UK

Background Risk stratification is vital to the optimal management ofpatients with non ST-elevation myocardial infarction (NSTEMI)however, current tools are not fully discriminatory. Copeptin, thestable 39 amino acid C-terminal portion of pro-vasopressin, is arecognised prognostic marker in ST elevation myocardial infarction(STEMI) that is also useful to excludeMI as levels rise early after onset.Copeptin has not been evaluated in a NSTEMI population to date.Aims We hypothesised that copeptin is an independent predictor ofmortality following NSTEMI and, in accordance with AHA criteriafor the evaluation of novel biomarkers, assess whether copeptin addsprognostic information to GRACE risk score (GRACE-RS). We useNT-proBNP for comparison.Methods and Results In this prospective observational study plasmacopeptin andNT-proBNPwasmeasured in 754NSTEMIpatients (519men, median age 70613 years) within 36 h of symptoms. Theprimary endpoint of all-cause mortality at 6 months was reached by56 (7.4%) patients. Median copeptin levels were 7.9 range 0.3 to523.0 pmol/l and were significantly higher in those that reached theprimary endpoint than the event free survivors; median (IQR), 32.0(12.0e88.7) vs 7.2 (4.0e16.7) respectively p<0.001. Both copeptin andNT-proBNP were predictive of the primary endpoint on univariateCox regression analysis (HR 5.98 p<0.0005 and HR 6.07 p<0.0005respectively). On adjustment for baseline clinical and biochemicalvariables copeptin remained predictive (HR 3.03 p¼0.009) but NT-proBNP did not (p¼0.70). Kaplan-Meier analysis revealed that supra-median levels of copeptin were associated with increased mortality(log rank 28.4 p<0.001). ROC curve c-statistics for GRACE-RS of0.799 increased to 0.835when combinedwith copeptin (0.785), whencombined with NT-proBNP (0.730) increased to 0.802. Re-classi-fication analysis shows that copeptin improves accuracy of riskstratification when combined with the GRACE-RS as determined bynet reclassification improvement (NRI 13.3% p¼0.008) whereas, NT-proBNP does not (NRI �4.9% p¼0.21). The relative utilities forlogistic regression models using GRACE-RS alone, GRACE-RS +copeptin and GRACE-RS + NTproBNP as covariates are shown inAbstract 11 figure 1. The relevant regionwas the region to the right ofthe sample risk for 6 months mortality, 0.074. The relative utility forGRACE-RS+ copeptinwas consistentlymore than the relative utilityfor GRACE-RS + NTproBNP across a range of risks; for example at arisk threshold of 15% the additional utility of adding copeptin to theGRACE-RS was 0.097 compared to 0.009 for NTproBNP.

Abstract 11 Figure 1


BCS Abstracts 2011

Conclusions High plasma copeptin levels indicate a worse outcomein NSTEMI patients. We have demonstrated that copeptin fulfilsAHA criteria by improving risk stratification over establishedmarkers GRACE score and NT-proBNP. Copeptin is also useful forrapid rule-out of MI and the current findings further support clinicaluptake.

12 THE RELATIONSHIP BETWEEN PSYCHOLOGICAL FACTORSAND IMPAIRED HEALTH-RELATED QUALITY OF LIFE POSTST-ELEVATION MYOCARDIAL INFARCTION

doi:10.1136/heartjnl-2011-300198.12

1L McGowan, 2H Iles-Smith, 1C Dickens, 1M Campbell, 1C Rogers, 2F Fath-Ordoubadi.1University of Manchester, Manchester, UKI; 2CMFT, Manchester, UK

Introduction Evidence suggests that psychological factors, such asdepression and anxiety, are independent risk factors for increasedmorbidity and mortality post ST-elevation myocardial infarction(STEMI). Since improved treatments have increased survival ratespost STEMI the emphasis has turned to more patient relatedoutcome measures such as health-related quality of life (HRQoL).The aim of the study was to assess the contribution of anxietyand depression to HRQoL in post STEMI patients, aftercontrolling for possible confounding factors, including type oftreatment.Methods We conducted a prospective cohort study of 385 post-STEMI patients who had undergone either lysis (183) or PPCI (202).The mean age was 60.0 years (SD 11.8) and 78% were male.Patients were assessed on a range of demographic, clinical andpsychosocial variables, including measures of cardiac risk,cardiac severity and comorbidity (Charlson Comorbidity IndexdCCI). Psychosocial assessment included anxiety and depression(Hospital Anxiety and Depression Scale), illness perceptions (briefIPQ), and health-related quality of life (SF-36). The main outcomewas the SF-36 Physical Component Score (PCS) at 6 months post-STEMI.Results Baseline results revealed a small number significant differ-ences between groups on a range of clinical variables, includinghigher GRACE scores for PPCI group (p¼0.007) but no differences inLV function. Lysis patients had more comorbid illness as measuredby the CCI (p¼0.037). Regarding psychological variables the totalHADS score was significantly higher in the PPCI vs lysis group atbaseline (means 13.2 (SD 7.9) and 11.4 (SD 8.9), p¼0.035), whileanxiety and depression almost reached significance, with raisedanxiety and depression scores in the PPCI group. In order to identifyvariables at baseline that may contribute to SF-36 PCS at 6 months,we conducted a hierarchical multiple regression with four blocks ofindependent variablesddemographic, comorbidity-related, clinicaland psychological. Factors which contributed to the final modelwere cholesterol levels (p¼0.031) and depression (p<0.001). Treat-ment group did not play a role (p¼0.199). The addition of anxietyand depression contributed significantly to the reporting of lowerphysical health-related quality of life (PCS) at 6 months (ÄR2¼0.12,p<0.001).Conclusion The findings have shown that raised levels of depres-sion and anxiety predicted impairment in health-related quality oflife at 6 months post-STEMI, regardless of mode of treatment. Theresults indicate that the assessment of psychological factors isimportant in both groups. Despite PPCI having improved clinicaloutcomes, there will always be a group of patients receivinglysis. As such it is important to assess anxiety and depression inpost STEMI patients, and to include these potentially modifiablefactors in the design of suitable interventions for this patientgroup.

13 NEUTROPHIL ACTIVATION AT THE CULPRIT LESION INACUTE ST-SEGMENT ELEVATION MYOCARDIALINFARCTION WITH MULTIPLE COMPLEX CORONARYPLAQUES

doi:10.1136/heartjnl-2011-300198.13

1C J Marshall, 1J L Mckenzie, 2T Moccata, 3M Nallaratnam, 3J Blake, 3C Frampton,3M Richards, 2A J Kettle, 3D R Mcclean. 1Sunderland Royal Hospital, Sunderland, UK;2Free Radical Research Group, University of Otago, Christchurch, New Zealand;3Department of Cardiology, Christchurch Hospital, Christchurch, New Zealand

Introduction The activation of neutrophils at the culprit coronarylesion following acute plaque disruption has not been reported. Wehypothesised that neutrophil activation occurs in ST elevationmyocardial infarction (STEMI) prior to percutaneous intervention(PCI), and that differences in activation may be detectable locally atthe culprit lesion, particularly in patients with multiple complexcoronary plaques.Methods Forty STEMI patients having primary PCI were comparedto 10 controls with chronic stable angina (CSA) undergoing elec-tive PCI. The clinical, demographic and angiographic characteristicsof patients and controls are shown in Abstract 13 table 1. Theculprit lesion was sampled after passage of a guide wire across thelesion and use of a low profile sampling catheter (Multifunctionalprobing catheter, Boston Scientific Corporation, Natick, Massa-chusetts, USA) at the site of occlusion, prior to further mechanicalintervention. Neutrophil activation was measured by flowcytometry using neutrophil intracellular myeloperoxidase content(MPO Index) and the expression of the b2- integrin CD11b, aleukocyte adhesion and activation marker at the culprit coronarylesion (CA), the aorta at the coronary ostium (Ao), the coronarysinus (CS), and femoral artery (FA) prior to primary PCI. A lowerMPO content indicates the depletion of intracellular MPO and cellactivation.

Abstract 13 Table 1

VariableSTEMI(n[40)

Elective PCI(n[10) p Value

Age (years) mean6SD 62612 6869.1 0.9

Male (%) 28 (70) 7 (70) 1

Culprit coronary artery lesion treated (%)

Left anterior descending 17 (42) 5 (50) 0.73

Diagonal 2 (5) 1 (10) 0.5

Circumflex 1 (3) 2 (20) 0.1

Right coronary 20 (50) 2 (20) 0.15

Time to presentation (mins) mean6SD 2226155 NA

Baseline TIMI flow 0e1 28 (70) 0 <0.001

Results A marked decrease in MPO content occurred at the CA, Aoand FA in STEMI compared to elective controls (p<0.01). Further-more, MPO content was lower at the CA (�23.1, (�25.6 to �17.1),n¼37) compared to Ao (�22.0, (�24.7 to �16.2), n¼37), CS (�20.6,(�24.8 to �16.9), n¼30) and FA (�20.4, (�24.4 to �13.1), n¼40), allp<0.001 (Abstract 13 figure 1). Neutrophil MPO content wascorrelated with CD11b expression only at the culprit CA in STEMI(r¼�0.4, p¼0.03, n¼31) (Abstract 13 figure 2). Neutrophil MPOcontent at the CA in patients with multiple complex plaques wassimilar to those with a single culprit however only in those withmultiple complex plaques was a correlation between MPO contentand CD11b (r¼�0.7, p¼0.02) shown. Conclusion: In acute STEMI,neutrophils are activated systemically, regionally and locally at theculprit coronary lesion. In patients with multiple complex plaques,there may be an extended local role for the activated neutrophilfollowing acute plaque destabilisation.


BCS Abstracts 2011



14 DYNAMIC CHANGES OF OEDEMA AND LATE GADOLINIUMENHANCEMENT AFTER ACUTE MYOCARDIAL INFARCTIONAND THEIR RELATIONSHIP TO FUNCTIONAL RECOVERYAND SALVAGE INDEX

doi:10.1136/heartjnl-2011-300198.14

1E Dall’Armellina, 1N Karia, 1A Lindsay, 1T D Karamitsos, 1V Ferreira, 1M D Robson,2P Kellman, 1J M Francis, 3C Forfar, 3B Prendergast, 3A P Banning, 1K Channon,3R J Kharbanda, 1S Neubauer, 1R P Choudhury. 1NIHR Biomedical Research Centre,Department of Cardiovascular Medicine, University of Oxford, Oxford, UK; 2NIH,Bethesda, USA; 3NIHR Biomedical Research Centre, Department of Cardiology, JohnRadcliffe Hospital, Oxford, UK

Introduction Changes in myocardial tissue in acute ischaemia aredynamic and complex and the characteristics of myocardial tissueon cardiovascular magnetic resonance (CMR) in the acute settingare not fully defined. We investigated changes in oedema and lategadolinium enhancement (LGE) with serial imaging early after acute

MI, relating these to global and segmental myocardial function at6 months.Methods and Results CMR scans were performed on 30 patientswith ST elevation MI (STEMI) treated by primary PCI at each of 4time points: 12e48 h (TP1); 5e7 days (TP2); 14e17 days (TP3); and6 months (TP4). All patients showed oedema at TP1. The meanvolume of oedema (% LV) was 37616 at TP1 and 39617 at TP2




BCS Abstracts 2011

with a reduction to 24613 (p<0.01) by TP 3. Myocardial segmentswith oedema also had increased signal on LGE at TP1 (k¼0.77;p<0.001). At TP1, the proportion of segments with wall motionimpairment increased in relation to the extent of both myocardialoedema (p<0.01) and LGE (p<0.01). The volume of LGE decreasedsignificantly between TP1 and TP4 (27615% vs 22612%; p¼0.002).Of segments showing LGE at 48 h, 50% showed resolution by6 months. In segments with such a reduction in LGE, 65% alsoshowed improved wall motion (p<0.0001). The area of LGE meas-ured at 6 months correlated more strongly with 48-h troponin(R2¼0.84; p<0.01) than at TP1 (R2¼0.5). The difference in LGEbetween TP1 and TP4 had profound effects on the calculation ofsalvage index (26621% at TP1 vs 42623% at TP4; p<0.02).Conclusions (1) Myocardial oedema was unchanged over the firstweek but decreased by 15 days; (2) a large majority of segments thatwere positive for oedema also showed LGE, assessed at 12e48 h; (3)In 46% of patients, LGE present on early scans had diminished insize by 6 months, (4) resolution of LGE was associated withimprovement in function; (5) the reduction in LGE at the later timehad a profound effect on the calculation of salvage index, whichvaried by up tow60%, depending on the time point used. (6) From aclinical perspective, the use of acute LGE may severely under-estimate salvaged myocardium and should not be used to predictrecovery of myocardial function.

15 INVESTIGATION OF IL-1 INHIBITION IN PATIENTSPRESENTING WITH NON-ST ELEVATION MYOCARDIALINFARCTION ACUTE CORONARY SYNDROMES (THE MRCILA HEART STUDY)

doi:10.1136/heartjnl-2011-300198.15

1A C Morton, 2C Foley, 1A Rothman, 1J Gunn, 3J P Greenwood, 3A Hall, 4K Fox,2B Lees, 2M Flather, 1D Crossman. 1NIHR Cardiovascular Biomedical Research Unit,Sheffield, UK; 2CTEU, Royal Brompton and Harefields NHS Trust, London, UK;3Academic Unit of Cardiovascular Medicine, Leeds, UK; 4Royal Infirmary of Edinburgh,Edinburgh, UK

Background Inflammatory mechanisms are involved in both coro-nary atherogenesis and its presentation as acute coronary syndromes(ACS). To date, drugs used at the time of ACS, or for primary andsecondary prevention have not primarily targeted inflammatorymechanisms. Studies with aspirin and statin drugs indicate thatanti-inflammatory properties of these compounds may contribute totheir beneficial effects. Pre-clinical studies from our group haveindicated that the pro-inflammatory cytokine IL-1 drives anumber of vascular events relevant to coronary artery disease andACS. IL-1 can be inhibited by IL-1 receptor antagonist (IL-1ra,Anakinra, Amgen) which is licensed for the treatment of rheumatoidarthritis.Aims To investigate the effects of interleukin-1 receptor antagonist(IL-1ra) on inflammatory biomarkers in patients with ACS <48 hfrom symptom onset, and to evaluate the safety and tolerability oftreatment.Design and methods The UK MRC ILA-HEART study is an inves-tigator-initiated, non-industry sponsored, phase 2, multi-centre,placebo-controlled trial, comparing the IL-1ra (100 mg) withmatching placebo given as a single, daily subcutaneous injectionover 2 weeks. The primary outcome of the study was area under thecurve (AUC) of high sensitivity CRP (hs-CRP) over the first 7 daysof treatment, and the main secondary outcomes are AUC oftroponin and safety and compliance of trial treatment. Patients wereencouraged to self-administer trial treatment, and underwent dailyassessment of hs-CRP, troponin, von Willebrand factor and otherbiomarkers up to 7 days, and again at 2 weeks and 30 days. Patientswere followed up to 12 months for safety (Abstract 15 table 1).

Abstract 15 Table 1 Log transformed values for the primary andsecondary outcomes

VariableActive Active Placebo Placebo

pn mean (SD) n mean (SD)

hsCRP AUC (days 1e7) 82 3.51 (1.42) 78 3.55 (1.46) 0.86

hsCRP at day 7 78 1.03 (1.46) 78 1.09 (1.91) 0.83

hsCRP at day 14 77 1.14 (1.54) 76 0.89 (1.56) 0.32

hsCRP at day 30 76 1.13 (0.32) 76 0.92 (1.15) 0.30

Troponin AUC (days 1e7) 82 1.35 (1.88) 78 1.50 (1.96) 0.61

Troponin at day 14 77 �3.73 (2.31) 76 �4.14 (2.35) 0.27

Troponin at day 30 76 �4.83 (2.44) 76 �4.74 (2.51) 0.82

vWF AUC (days 1e3) 89 1.33 (0.35) 84 1.31 (0.35) 0.66

vWF at day 14 77 0.36 (0.35) 84 0.28 (0.41) 0.23

vWF at day 30 78 0.37 (0.34) 75 0.29 (0.43) 0.19

IL-6 AUC (days 1e3) 86 2.62 (1.13) 83 2.56 (1.02) 0.74

IL-6 at day 14 77 1.32 (1.35) 77 0.92 (0.75) 0.028

IL-6 at day 30 73 1.22 (0.92) 75 1.08 (0.77) 0.31

Results Five UK centres randomised 182 patients with non-STelevation (NSTEMI) ACS to IL-1ra or placebo. Enrolment completedin March 2010. Mean age was 61 years, 32% female, 28% prior MI,15% diabetes, 90% were receiving a statin at the time of random-isation and 64% had early PCI or CABG. Compliance was good with85% of patients receiving daily injections during the first 7 days, and70% of patients were able to self-administer the injections. Injectionsite reactions reported as adverse events occurred in 14% of patients.There was no significant difference in area under the curve forhsCRP between active and placebo groups. The MACE and seriousadverse event rates are shown in Abstract 15 table 2.

Abstract 15 Table 2 MACE and other SAE

Event Active (n (95% CI)) Placebo (n (95% CI)) p

MACE at 30 days 6 (0.19 to 1.79) 10 (1) 0.35

MACE at 3 months 10 (0.32 to 1.84) 13 (1) 0.54

MACE at 1 year 20 (0.59 to 2.22) 18 (1) 0.69

MI (%) 3.26 7.78 0.18

Stroke (%) 1.09 0 1.00

Death (%) 4.35 4.44 1.00

CV hospitalisation (%) 19.86 25 0.31

Revascularisation (%) 69.56 60 0.8

Injection site reactions(%)

15.22 15.56 0.95

Other SAE (%) 47.83 44.44 0.65

Discussion NSTEMI ACS treated with all the current evidenced-based therapies still has significant recurrent events. MRC ILA-HEART is the first study to evaluate the effects of the anti-inflam-matory IL-1ra in ACS. The data indicates that despite encouragingpre-clinical evidence, the inflammatory driver for NSTEMI-ACS isnot IL-1 mediated.

16 ACUTE STENT THROMBOSIS RESULTING IN ST ELEVATIONMYOCARDIAL INFARCTION (STEMI) IS ASSOCIATED WITHWORSE CLINICAL OUTCOMES THAN STEMI DUE TO NATIVECORONARY THROMBOSIS

doi:10.1136/heartjnl-2011-300198.16

E C Sammut, A Graham, D A Jones, K Rathod, S May, A Jain, S Mohiddin, C Knight,A Mathur, A Wragg. The London Chest Hospital, London, UK

Background Stent thrombosis (ST) is a recognised cause of STElevation Myocardial infarction (STEMI) in patients with previous


BCS Abstracts 2011

percutaneous coronary interventions (PCI). The incidence isincreasing and to date outcomes are not well characterised, thoughthere is a suggestion that there is a worse clinical outcome.We therefore sought to compare STEMI caused by ST vs de novocoronary thrombosis to evaluate procedural risk and clinical outcome.Methods Clinical information was analysed from a prospectivedatabase on 2421 patients who underwent Primary PCI followingSTEMI between October 2003 and May 2010 at a London centre.Information was entered at the time of procedure, diagnosis of stentthrombosis made at the time by primary operator and outcomeassessed by all-cause mortality information provided by the Officeof National Statistics via the BCIS CCAD national audit.Results Stent thrombosis (ST) accounted for 7.4% (180/2421) of allSTEMIs with a frequency that has increased over time (5.4% in 2005to 9.8% 2009). ST occurred early (0e30 days) in 36% (65/180), late(30 dayse1 year) in 22% (40/180) and very late (>1 year) in 42% (75/180) of pts. Drug-eluting stents (DES) accounted for 48% of SToveralland 70% over the past 3 years. Proposed mechanisms includedpremature discontinuation of anti-platelets (11%), under-deploymentof previous stent insertion (22%) and underlying prothromboticconditions (eg, SLE) (6%). Pts with ST compared to native arteryocclusion had higher rates of previous MI (53.9% vs 11%, p<0.0001)and incidence of multi-vessel disease (59.8% vs 51.7%, p¼0.04There was no difference in age, diabetes or cardiogenic shock. SeeAbstract 16 table 1. Infarct size based on peak enzyme markers wassimilar (2.5 vs 2.2, p¼0.45). In patients with ST, angiographic success(postprocedural Thrombolysis In Myocardial Infarction grade IIIflow) was worse than in patients with de novo STEMI (87.2% vs93.7%, p¼0.02). Pts with STEMI due to ST had higher in-hospitalMACE (11% vs 3%, p¼0.0001), MACE at 30 days (19% vs 6%,p<0.0001), this persisted up to 3 years (41% vs 12%, p<0.0001). SeeAbstract 16 figure 1. MACE was driven by higher rates of MI (7% vs2%, p<0.0001), TVR (14% vs 3%, p<0.0001) and death (18% vs 6%,p¼0.0001). After adjusting for comorbidities, stent thrombosiswas anindependent predictor of long-term adverse outcome (OR¼2.1, 95%CI¼1.3 to 2.8, p<0.001).

Abstract 16 Table 1

AST (n[180) No AST (n[2241)Significance(p value)

Age 63.9618 62.964 0.406

Multi-vessel disease 101 (59.76%) 1011 (51.74%) 0.045

Previous MI 96 (53.93%) 246 (10.96%) <0.0001

Previous CABG 9 (5.06%) 60 (2.67%) 0.020

Diabetes mellitus 46 (25.84%) 407 (18.14%) 0.061

Hypertension 101 (56.74%) 975 (43.45%) 0.002

Hypercholesterolaemia 108 (60.67%) 768 (34.30%) <0.0001

Cardiogenic shock 108 (60.67%) 768 (34.30%) <0.0001


Conclusion Primary PCI for treatment of ST is less effective, andthese patients are at increased risk for in-hospital and long-termmortality compared with patients undergoing primary PCI due to denovo STEMI.

17 SUDDEN CARDIAC DEATH AND ACUTE MYOCARDIALINFARCTION: HOW HAS THE PICTURE CHANGED?

doi:10.1136/heartjnl-2011-300198.17

1G Mole, 2D Watson, 3C Davidson. 1Brighton & Sussex Medical School, Brighton, UK;2Deprtment of Informatics, University of Sussex, Brighton, UK; 3Brighton & SussexUniversity Hospital, Brighton, UK

Introduction Coronary heart disease (CHD) is a major burdenworldwide, particularly in economically developed countries such asthe UK. Between 1980 and 2000, deaths from CHD fell by over 50%,and have continued to fall. The cost of CHD manifests itself inmortality, disability and economic impact: this should be looked atin the context of a disease that is preventable.Methods Data from death certificates and studies with strict clinico-pathological criteria on mortality from CHD were accessed. Thesewere analysed in terms of hospital admissions, revascularisationrates and index of multiple deprivation (IMD). Trends in mortalityoverall and for different age groups were analysed over time todetermine how the picture has changed and predict what mayhappen going into the future. The mortality rates from the catch-ment area of the Royal Sussex County Hospital and York Hospitalwere analysed to assess reliability of official figures against strictclinico-pathological inclusion criteria. A range of statistical testsincluding, linear regression, ANOVA and JoinPoint regression wereemployed.Results Between 1993 and 2008 deaths from CHD have fallen byover 50%. The decline has been greater in older age groups partic-ularly the 65e74 age group (the oldest age group analysed).Comparison with data from studies with strict clinico-pathologicalcriteria showed this to be the age group in which official statisticswere least accurate. Regression analysis demonstrated that a higherIMD is associated with increased mortality from CHD (r2¼0.69,p<0.001). Increased admission rates was not significantly associatedwith decreased mortality from CHD (r2¼0.004, p¼0.239). Increasedrevascularisation was significantly associated with decreasedmortality from CHD (r2¼0.99, p<0.001). JoinPoint regressionanalysis shows a constant rate of decline in mortality from 1993 to2001 which then decreases faster between 2001 and 2006 before



BCS Abstracts 2011

slowing dramatically from 2006 to 2008. JoinPoint regression anal-ysis of different age groups demonstrates that the slower rate ofdecline from 2006 may be due to stubbornly high numbers of deathsin the 35e44 age group. Lastly the National figures on mortalityfrom CHD are shown to be misleading as many people are stilldying from CHD just when they have crossed the 75-year oldexclusion criteria; as a result a delay in mortality is presented asprevention of mortality from CHD.Discussion There is a danger that previous successes are being offsetby high rates in the younger cohorts, and that the overall trend maybe eventually be reversed. There is still work to be done in reducingrisk factors and also applying treatments that have had a provenpositive impact (such as revascularisation) more effectively. Stat-istically significant changes in declining CHD mortality rates.Future work This 10 000 word report formed the basis of a fundingapplication to the British Heart Foundation for a follow-up to theUnited Kingdom Heart Attack Study.


18 PATIENTS PRESENTING WITH ANAEMIA UNDERGOINGPRIMARY PCI APPEAR AT SIGNIFICANTLY HIGHER RISK OFAN ADVERSE OUTCOME

doi:10.1136/heartjnl-2011-300198.18

K R Rathod, D A Jones, B Rathod, A Graham, E Sammut, S Gallagher, J Behar,A K Jain, C Knight, A Mathur, A Wragg, R Amersey. Barts and the London NHS trust,London, UK

Background Previous studies have demonstrated a relationshipbetween pre-existing anaemia and inpatient mortality after percu-taneous coronary intervention (PCI). There is limited data lookingat the impact of baseline Haemoglobin and long term outcome afterprimary PCI.Methods Clinical information was analysed from a prospectivedatabase on 2357 STEMI patients who underwent Primary PCIbetween January 2004 and May 2010 at a London centre. Infor-mation was entered at the time of procedure and outcome assessedby all-cause mortality information provided by the Office ofNational Statistics via the BCIS/CCAD national audit. Anaemiawas defined according to WHO definition of Hb greater than orequal to 12 g/dl for females and 13 g/dl for males.Results 471 (20%) patients were anaemic at presentation. Theanaemic cohort, were older (72.2 vs 62.4, p<0.0001), had higherincidence of diabetes (27% vs 15%, p<0.0001), hypertension (42 vs35%, p¼0.01), hypercholesterolaemia (40 vs 30%, p¼0.007),previous PCI (13 vs 7%, p¼0.01), and previous MI (23% vs 12%,p<0.0001). There were similar incidences of three-vessel disease andcardiogenic shock. Over a 3-year follow-up period there was signif-icantly higher all cause mortality in the anaemic group compared tothe normal Hb group (20.4% vs 13.5%, p<0.0001). See Abstract 18

figure 1. After adjusting for comorbidities, anaemia remained anindependent predictor of long-term adverse outcome (OR¼2.4, 95%CI¼1.1 to 3.7, p<0.001). Patients with baseline anaemia whoreceived a blood transfusion were significantly more likely to sufferan adverse outcome than those that did not receive a transfusion(21% vs 6%, p<0.0001).

Abstract 18 Figure 1 All cause mortality after PCI for STEMI.

Conclusion Patients presenting with anaemia undergoing primaryPCI appear at significantly higher risk of an adverse outcome. Thisrisk increases further in population receiving RBC transfusionsduring index hospitalisation.

19 TREATMENT OF MULTIVESSEL CORONARY ARTERYDISEASE IN PRIMARY PCI FOR ST ELEVATION MYOCARDIALINFARCTION: CULPRIT ONLY REVASCULARISATION ISASSOCIATED WITH HIGHER MACE RATES

doi:10.1136/heartjnl-2011-300198.19

K S Rathod, L A McGill, E Sammut, V S Rathod, D A Jones, R Weerackody, A Jain,C Knight, A Mathur, A Wragg. London Chest Hospital NHS Trust, London, UK

Background Multi-vessel disease occurs in 40%e65% of patientsundergoing Primary PCI for STEMI and is associated with adverseprognosis. Contemporary guidelines recommend treating the infarctrelated artery alone (culprit) during the urgent procedure. There islimited data comparing outcomes of complete with infarct-relatedartery (IRA)-only revascularisation in primary PCI for STEMI withfew studies including the option of later date elective procedures forthe other lesions (staged revascularisation). We therefore sought toclarify the outcome of patients with multi-vessel disease undergoingprimary PCI dependent on management strategy.Methods Clinical information was analysed from a prospective database on 2131 STEMI patients who underwent Primary PCI betweenJanuary 2004 and May 2010 at a London centre. Patients withprevious CABG were excluded. Information was entered at the timeof procedure and outcome assessed by all-cause mortality informa-tion provided by the Office of National Statistics via the BCIS/CCAD national audit. Patients were split into three differenttreatment groups: culprit vessel angioplasty-only (COR group);staged revascularisation (SR group) and simultaneous treatment ofnon-IRA (CR group). The primary end point used was major adversecardiac events (MACE), defined as death, myocardial infarction(MI), stroke and target vessel revascularisation (TVR).Results There were 963 (45%) consecutive patients with STEMI andmultivessel CAD undergoing primary angioplasty. There weresimilar baseline characteristics between the 3 groups, aside fromcardiogenic shock, which was significantly higher in the completerevascularisation group. See Abstract 19 table 1. At 30-days offollow-up, 23/263 (9%) patients in the CR group experienced atleast one major adverse cardiac event (MACE), 1 (1%) in the SRgroup and 35 (5%) in the COR group, p¼0.01. This trend continued


BCS Abstracts 2011

up to 1-year of follow-up with the lowest rates of events in the SRgroup. However after 3 years MACE rates are significantly increasedin the COR group (24%) but were similar in the CR (18%) and SR(17%) groups. See Abstract figure 1. MACE rates were driven mainlyby death in the CR with high rates of TVR in the COR and SRgroups. See Abstract figure 2.

Abstract 19 Table 1

COR N[638 SR N[100 CR N[263 Significance

Age 64.77 61.46 64.32 0.144

Gender (female) 156 (23.7%) 13 (13.0%) 74 (27.9%) 0.0114

Ethnicity (Caucasian) 441 (67.0%) 79 (79.0%) 185 (69.8%) 0.0511

Previous MI 109 (16.6%) 11 (11.0%) 36 (13.6%) 0.2414

Previous CABG 15 (2.3%) 2 (2.0%) 3 (1.1%) 0.5231

Previous PCI 83 (12.6%) 5 (5.0%) 23 (8.7%) 0.031

Diabetes Mellitus 129 (19.6%) 16 (16.0%) 55 (20.8%) 0.5932

Hypertension 312 (48.1%) 40 (40.0%) 91 (41.2%) 0.1205

Hypercholestrolaemia 269 (41.5%) 37 (37.0%) 92 (41.6%) 0.7751

GPIIb/IIIa Inhibitor 572 (87.7%) 93 (93.0%) 231 (89.5%) 0.1724

Cardiogenic Shock 29 (4.7%) 2 (2.0%) 31 (12.26%) p<0.0001

Abstract 19 Figure 1 Comparison of MACE between multivesseldisease.

Abstract 19 Figure 2 Breakdown of MACE at 5 years.

Conclusions Culprit vessel-only angioplasty was associated with thehighest rate of long-term MACE compared with multivessel treat-ment. Patients scheduled for staged revascularisation experienced asimilar rate of MACE to patients undergoing complete simultaneoustreatment of non-IRA.

20 WHAT HAPPENS TO PLATELET FUNCTION AND VASCULARINFLAMMATION WHEN CLOPIDOGREL IS WITHDRAWN?INSIGHTS USING SHORT THROMBELASTOGRAPHY

doi:10.1136/heartjnl-2011-300198.20

1N Sambu, 2H Dent, 3T Warner, 2N Englyst, 3P Leadbeater, 1A Hobson, 1A Calver,1S Corbett, 1H Gray, 1I Simpson, 1N Curzen. 1Southampton University Hospitals NHSTrust, Southampton, UK; 2University of Southampton, School of Medicine, South-ampton, UK; 3The William Harvey Research Institute, Barts and the London School ofMedicine and Dentistry, London, UK

Introduction A clustering of adverse events, in particular stentthrombosis (ST) has been observed following clopidogrel cessation1-year after drug-eluting stenting (DES), the aetiology of which ispoorly understood. We investigated the effect of withdrawingclopidogrel in DES patients using a simple, rapid, reproducible near-patient platelet function test known as short Thrombelastography(s-TEG) that has been developed and validated by this group.Methods 33 patients on aspirin and due to stop clopidogrel at 1 yearfollowing DES were investigated. Venesection was performed at (i)4 weeks and 24 h pre clopidogrel cessation (ii) 24 h, 48 h, 1, 2 and4 weeks post clopidogrel cessation. At all time-points, plateletreactivity was determined using s-TEG and thromboxane (TX) B2,IL-6, CD40 ligand and high sensitivity CRP were measured.Results Clopidogrel cessation produced (i) a predictable increase inADP-induced platelet aggregation, and (ii) an unexpected andsignificant rise in AA-induced platelet aggregation. TXB2 wasconsistently suppressed confirming inhibition of COX by aspirin.


Conclusion We have described for the first time an aspirin-inde-pendent increase in AA-induced clotting following clopidogrelwithdrawal in DES patients. As well as potentially helping toexplain the observed clustering of ST events early after clopidogrelwithdrawal, these findings raise the question as to whether AA-induced clotting is an appropriate test of aspirin sensitivity. Ourresults also confirm s-TEG as a plausible candidate for near-patientplatelet function testing in this field.

21 INFLUENCE OF FRACTIONAL FLOW RESERVEMEASUREMENT ON TREATMENT-DECISIONS IN PATIENTSWITH RECENT ACUTE NON-ST ELEVATION MYOCARDIALINFARCTION

doi:10.1136/heartjnl-2011-300198.21

1D Carrick, 1M Behan, 1F Foo, 1J Christie, 2J Norrie, 1K Oldroyd, 1C Berry. 1Departmentof Cardiology, Golden Jubilee National Hospital, Glasgow, UK; 2Robertson Centre forBiostatistics, University of Glasgow, Glasgow, UK

Introduction Non-ST elevation acute myocardial infarction(NSTEMI) is the most common form of acute coronary syndromeand has a relatively poor prognosis. Visual interpretation of thecoronary angiogram is the standard approach to guide treatmentdecisions in patients with recent acute NSTEMI. The aim of ourstudy was to determine whether measurement of coronary pressurederived fractional flow reserve (FFR), compared to coronaryangiography alone, might influence treatment decisions.


BCS Abstracts 2011

Setting The cardiac catheterisation laboratory in a regional heartcentre in the UK.Definitions The clinical indication for FFR measurement was thepresence of an intermediate coronary stenosis (50%e75% of thereference vessel diameter) which resulted in diagnostic and treat-ment uncertainty. FFR measurement was used to provide functionalinformation on lesion severity and an FFR <0.80 was taken torepresent a flow-limiting stenosis.Methods The study involved three accredited interventional cardi-ologists and a study coordinator. The cardiologists separatelyreviewed the coronary angiograms and together with the clinicalhistory, made a decision for medical therapy, PCI, CABG/MDT, ordeferred management. The FFR results were then disclosed and theinitial management decision was reviewed in light of the FFR resultand changed as appropriate.Results Of 1621 acuteNSTEMI patients (January 2009eMarch 2010)in our hospital, 100 (6.2%) had FFR recorded. The treatment decisionsfor each cardiologist were: medical therapy 7%, 10%, 1%; PCI 64%,70%, 60%; CABG/MDT 13%, 12%, 15%; deferred management 16%,8%, 24%). The proportion of patients allocated to each treatmentgroup differed between the 2nd and 3rd Cardiologist (p¼0.02).Following FFR disclosure, each cardiologist changed his/her treatmentdecision in 58%, 50% and 62% of patients (p<0.05). Of the newdecisions made following FFR disclosure, the proportion of patientsallocated tomedical therapy increased by 26%, 19% and 29%,whereasthe proportion of patients allocated to deferredmanagement or multi-vessel PCI decreased by 16%, 8%, 24% and by 5%, 7% and 5%,respectively (all p<0.05). The number of patients in whom the treat-ment decisions made by each cardiologist independently conformed(and so represented a consensus in at least 2 of the 3 decisions)increased from 74% to 92% as a result of FFR disclosure (p<0.001).Conclusion In our hospital about 1 in 20 NSTEMI had a coronarypressure wire study because of diagnostic uncertainty based oncoronary angiography alone. In NSTEMI patients selected for FFRmeasurement, the FFR resulted in a change in management in atleast half of the patients. FFR use increased the proportion ofpatients in whom treatment decisions conformed suggesting FFRuse may also help to reduce the variation in treatment decisionsusing angiography alone. These results support further studies ofthe clinical utility and prognostic implications of FFR measurementin patients with NSTEMI.

22 COMPARISON OF 4-H HEART FATTY ACID BINDINGPROTEIN WITH 12-H TROPONIN I TO ASSESS 6-MONTHRISK FOLLOWING PERCUTANEOUS CORONARYINTERVENTION IN ACUTE CORONARY SYNDROMES

doi:10.1136/heartjnl-2011-300198.22

1I R Pearson, 1U M Sivananthan, 1J H Barth, 2C P Gale, 1A S Hall. 1Leeds TeachingHospitals, Leeds, UK; 2Division of Biostatistics, University of Leeds, Leeds, UK

Background It is known that PCI can cause myocardial injury leadingto the release of cardiac biomarkers into the circulation (proceduralMI). This occurs in approximately one third of procedures and hasbeen shown to impact negatively on prognosis. Monitoring forprocedural MI, although not yet standard practice, is increasinglyundertaken as a measure of quality control, andmay be a factor whendeciding time of discharge fromhospital following the procedure. Theuse of TnI to screen for procedural MI requires a wait of 12-h postprocedure before the blood sample may be taken, and an impact onlength of hospital stay is inevitable. Heart-type Fatty Acid BindingProtein (H-FABP) is a small protein released rapidly and in largequantities from the myocardium into the circulation, both duringischaemia and following necrosis. It allows detection of myocardialinjury associated with PCI earlier than with TnI.Hypothesis H-FABP at 4 h provides equivalent prognostic informa-tion to TnI at 12 h following PCI-induced myocardial injury.

Methods We studied 94 patients with ACS admitted to a single UKTeaching Hospital for PCI. We used the Randox Cardiac-Array tomeasure H-FABP at 4 hrs after PCI and troponin I at 12 h after PCI.Comparison of specificity and sensitivity of each biomarker foradverse cardiac events was made. Endpoint assessment consisted ofone of the following three events (i) PC-induced MI (ii) readmissionwith MI by 6 months (iii) death by 6 months.Results The area under the receiver operator curve was 0.73 forH-FABP measured at 4 h as compared to 0.72 for TnI measured as 12 h.Conclusion Early assessment of PCI-induced myocardial injury usingthe Randox Cardiac-Array to measure H-FABP is as sensitive andspecific for adverse prognosis as is TnI measurement taken at 12 hpost PCI. This approach should help to expedite early, safe hospitaldischarge following PCI.


23 SERUM NGAL IDENTIFIES CONTRAST NEPHROPATHYEARLY IN PATIENTS WITH DIABETES MELLITUS ANDCHRONIC KIDNEY DISEASE UNDERGOING CORONARYANGIOGRAPHY AND ANGIOPLASTY

doi:10.1136/heartjnl-2011-300198.23

1A C Qureshi, 2R Rampat, 3S M Harwood, 4M Roughton, 2M M Yaqoob, 1A Kapur.1The London Chest Hospital, Barts and The London NHS Trust, London, UK; 2The RoyalLondon Hospital, Barts and The London NHS Trust, London, UK; 3The William HarveyResearch Institute, London, UK; 4The Royal College of Physicians, London, UK

Background The incidence of contrast nephropathy (CIN) followingcoronary angiography or percutaneous coronary intervention (PCI)in patients with diabetes mellitus (DM) may be up to 30% and isassociated with increased long term morbidity and mortality.Methods We recruited 208 consecutive patients undergoing elective orurgent coronary angiography or PCI with known DM and chronickidney disease (CKD) (defined as eGFR<60 ml/min). CINwas definedas a post procedure rise in creatinine at day 3 of>25% from baseline oran absolute rise of 44.5 mmol/l. Severity of coronary disease wasassessed using the SYNTAX Score and risk of CIN using the Mehranrisk score. We evaluated serum and urine neutrophil gelatinase-asso-ciated lipocalin (NGAL) and albuminuria for additional informationaboutCINrisk.N-acetylcysteineand intravenoushydrationweregivento all patients with eGFR<50 ml in accordance with local guidelines.Results Baseline characteristics are summarised in table 1. 116patients underwent coronary angiography and 92 underwent PCI. 39patients (18.8%) developed CIN. Contrast dose was similar in the CINand non-CIN group (p¼0.249). The Mehran risk score was strongly


BCS Abstracts 2011

predictive of CIN development (p<0.001). The SYNTAX score did notdiffer between those who did or did not develop CIN (p¼0.188). Asignificant rise in serum NGAL was seen as early as 2 h post procedurein the CIN arm (p¼0.03) and this persisted at 4 h (p¼0.007) and12e24 h (p¼0.0015). Urine NGAL levels did not change significantlyduring the first 24 h. Neither albumin:creatinine ratio (p¼0.149) orprotein:creatinine ratio (p¼0.635) predicted development of CIN.

Abstract 23 Table 1

No CIN outcome(n[169)

CIN outcome(n[39) p Value

Age, (mean, SD) 70.8 (8.5) 71.5 (9.5) 0.64

Hypertension (%) 155 (91.7) 33 (86.8) 0.35

Hyperlipidaemia (%) 165 (97.6) 37 (97.4) 0.92

Previous MI (%) 66 (39.1) 15 (39.5) 0.96

Ex or current smoker (%) 96 (55.9) 19 (50.0) 0.75

Heart failure (%) 32 (19.1) 7 (18.4) 0.93

Valvular heart disease (%) 28 (16.6) 7 (18.4) 0.78

Family history IHD (%) 84 (49.7) 20 (52.6) 0.74

BMI, (mean, SD) 28.6 (5.4) 29.2 (6.1) 0.54

Conclusions The current gold standard for measuring CIN is a rise inserum creatinine but this is of limited value as it does not increaseuntil 48e72 h post renal injury. Neither the SYNTAX score, norurinary albuminuria or proteinuria are predictive of CIN develop-ment. A rise in serum NGAL levels within the first 12 h followingcoronary angiography or PCI appears to be a very promising markerin the early diagnosis of CIN.

24 PERCUTANEOUS MITRAL VALVE REPAIR WITH THEMITRACLIP DEVICE: A TERTIARY CARDIAC UK EXPERIENCE

doi:10.1136/heartjnl-2011-300198.24

J Dungu, C S R Baker, M F Bellamy. Hammersmith Hospital, Imperial College London,London, UK

Introduction Percutaneous mitral valve repair using the transcatheterMitraclip device is a novel therapy for patients with severe mitralregurgitation (MR) who are too high risk for conventional surgery.We report the largest UK series to date.Methods Patients were screened with transthoracic (TTE) andtransoesophageal echocardiography (TOE). Mitral regurgitation wasgraded by British Society of Echocardiography criteria. Twenty-fourpatients with $ grade 3+ symptomatic MR underwent percuta-neous mitral valve repair under general anaesthesia betweenFebruary 2009 and October 2010. The Mitraclip device was deployedunder 2- and 3-Dimensional TOE and fluoroscopic guidance. Allpatients were discussed with the manufacturing company (Evalve)and in a multidisciplinary meeting including >2 cardiologists and 2cardiothoracic surgeons with a special interest in mitral valvesurgery prior to being accepted.Results Mitraclip therapy was attempted in 24 patients aged71611 years with an average Euroscore of 16%. The indication forintervention was functional MR in 10 patients (42%), ischaemic MRin 7 patients (29%) and degenerative MR in 7 patients (29%). Twentypatients had successful deployment of the Mitraclip device (83%).Fifteen patients (75%) had 2 clips deployed. There were no vascularcomplications or strokes. We were unable to grasp the mitral valveleaflets in 2 patients due to an excessive coaptation gap. There was 1procedural death due to leaflet tear in a patient with end-stageischaemic cardiomyopathy and a grossly dilated left ventricle. Allpatients (100%) treated with the Mitraclip had severe MR (grade 3+/4+) prior to intervention. Mitral regurgitation was graded bycolour Doppler alone following intervention as standard quantitativeanalyses are not validated in the presence of a Mitraclip. At 1-month

follow-up to date, 10 patients (59%) had a reduction in MR to #

grade 2+ and 8 patients (47%) had $2 grade reduction in MR(p¼0.001). The reduction in MR grade remained significant for the 8patients with echo data at 6-month follow-up (p¼0.038). Onepatient had persistent grade 3+MR at 1-month follow-up due to latepartial detachment of one of the 2 clips deployed. NYHA classreduced significantly following intervention. Prior to Mitraclip, 63%of patients were in NYHA class III/IV. At 1-month follow-up post-Mitraclip only one patient (4%) was in NYHA class III (p¼0.042) and15 patients (83%) had at least 1 grade reduction in NYHA class. Therewas no significant change in left ventricular size following inter-vention, although there was a trend towards reduced left ventricularvolumes at 1-month follow-up (end diastolic volume 175 vs 160 ml,p¼0.102; end systolic volume 92 vs 79 ml, p¼0.076).Conclusion In selected patients Mitraclip edge-to-edge repairsuccessfully reduces the severity of mitral regurgitation andimproves symptoms. Further studies are needed to examine whetherthese results are durable and associated with improved outcome.

Abstract Figure 1 Change in MR grade post-Mitraclip.

Abstract Figure 2 Change in NYHA class post-Mitraclip.


BCS Abstracts 2011

25 TAVI OPERATOR RADIATION DOSE COMPARED TO PCI ANDICD OPERATORS: DO WE NEED ADDITIONAL RADIATIONPROTECTION FOR TRANS-CATHETER STRUCTURAL HEARTINTERVENTIONS

doi:10.1136/heartjnl-2011-300198.25

M Drury-Smith, A Maher, C Douglas-Hill, R Singh, M Bhabra, J Cotton, S Khogali.Heart and Lung Centre, New Cross Hospital, Wolverhampton, UK

Introduction Trans-catheter cardiac aortic valve implantation(TAVI), implantable cardiac defibrillators (ICD), and percutaneouscoronary intervention (PCI), are common procedures associatedwith radiation exposure to the operator and the patient. Radiationdose exposure is cumulative and if above the recommended annuallevels may have significant consequences for the operator. Theradiation dose TAVI operators are exposed to is not widely known,but it is an important consideration in view of the increasingvolume of procedures and the potential risks of over-exposure. Ouraim was to monitor and compare, radiation exposure time, dose, andindividual operator dose, in TAVI, PCI and ICD.Method TenTAVIswere performed, 6 via the trans-femoral route and4via the subclavian approach. Radiation protectionwas employed in allcases using standard lead skirts and screens. During each procedure theradiation dose exposure was monitored for each operator using Ther-moLuscent Dosemeters (TLD) on the left finger (LF), right finger (RF)and forehead. The six TAVI procedures performed via the transfemoralapproach used only two operators, while the subclavian approachinvolved three operators. The third operator was a surgeon who wasnearest to the x-ray images. Radiation exposure doses were alsocollected fromICDandPCIoperatorsduring the sameperiod, using thesame type ofTLDs onLFandRF.Operator specific radiation doseswereobtained from a central RRPPS Approved Dosimetry Service. PCI wasconsidered a standard trans-catheter procedure. TAVI and ICD oper-ator doseswere compared to themean standardised PCI operator dose.Results The mean exposure times and doses for the different typesof trans-catheter procedures performed are detailed in the tablesbelow. Despite the use of standard radiation protection measures,the mean dose to operators undertaking TAVI was 6 times higherthan the trans-femoral PCI operator (p¼0.008). The mean radiationexposure time of TAVI was seven times more than PCI. Althoughsubclavian TAVI and ICD procedures were expected to be compa-rable with respect to operator dose, subclavian TAVI operators havean unexpectedly higher dose (p¼0.03).Conclusions Overall TAVI operators are exposed to significantlyhigher radiation doses compared to PCI and ICD operators. Addi-tional radiation protection for TAVI operators is strongly advocated.We are currently evaluating the impact of using a RADPAD asadditional protection during TAVI procedures.

Abstract 25 Table 1

Variable TAVI ICD PCI p Value

Mean exposure Time (mins) 27.0* 3.26 3.825 <0.001*

Mean exposure Dose(Gy/cmq) 6 SD

196.256150.96y 11.0369.01 33.09611.5 0.008y

*Significantly increased radiation exposure time in TAVI procedures compared to ICD and PCI.ySignificantly increased radiation exposure dose in TAVI procedures compared to ICD and PCI.

Abstract 25 Table 2

Mean radiation dose(Gy/cmq) per operator ±SD p value

Trans-femoral TAVI 1.67 1.23 0.03

Subclavian TAVI 2.53 3.09 0.03

ICD 1.95 0.14 0.03

PCI 0.18 0.36

26 THE EFFECTS OF PRE-EXISTING SIGNIFICANT CORONARYARTERY DISEASE UPON OUTCOME AFTER TRANSCATHETERAORTIC VALVE IMPLANTATION USING THE EDWARDSBIOPROSTHESIS

doi:10.1136/heartjnl-2011-300198.26

1M Z Khawaja, 2H Haran, 1I Nadra, 1K Wilson, 1L Clack, 1K Macgillivray, 1J Hancock,1C Young, 1V Bapat, 1M Thomas, 1S Redwood. 1Guy’s & St. Thomas’ Hospitals NHSFoundation Trust, London, UK; 2King’s College School Of Medicine & Dentristry, London, UK

Introduction Patients undergoing surgical aortic valve replacement(sAVR) routinely undergo simultaneous coronary artery bypassgrafting (CABG) for significant coronary artery disease (CAD) due toadverse prognostic impact. While manufacturers advise percuta-neous intervention (PCI) of significant CAD prior to transcatheteraortic valve implantation (TAVI) there is considerable variationamong operators.Methods We performed a retrospective analysis of 168 patients whounderwent TAVI using the Edwards bioprosthesis fromMarch 2008 toOctober 2010 at St. Thomas Hospital, London. They were divided intotwo groups according to the results of the pre-TAVI coronary angio-gram: (Group 1) patients with $1 coronary stenosis of $70% severityand those without (Group 2). The end-point was all-cause mortality.Results In total, 70 patients (41.7%) had significant CAD prior toTAVI, with 10 (6.0%) undergoing PCI prior to their procedure. Therewere no significant differences in either the baseline characteristicsor access approach between the two groups (Abstract 26 tables 1and 2). At a mean follow-up of 3356277 days (mean6SD), theoverall mortality was 22.6%; Group 1 mortality was 30% and ingroup 2 was 17.3% (p¼0.124) (see Abstract 26 figure 1) There wasno difference seen in the length of stay in the intensive care unit(2.766.2 vs 4.1614.9 days, p¼0.462) nor in the number of days todischarge (12.6610.1 vs 12.8613, p¼0.928). Among those patientswho underwent PCI in Group 1, 8 had single vessel intervention and2 had PCI to 2 vessels. The target vessels were left main stem (LMS)(n¼2), proximal left anterior descending artery (LAD) (n¼5),circumflex (n¼1), right coronary artery (RCA) (n¼2), saphenousvein graft (SVG) to LAD (n¼1) and SVG to circumflex (n¼1).Mortality in this sub-group was not significantly different from theCAD patients who did not receive PCI (50% vs 26.7%, p¼0.272).

Abstract 26 Table 1

Group 1 SignificantCAD (n[70)

Group 2 No significantCAD (n[98) p Value

Age (years6SD) 83.767.5 81.768.5 0.112

Diabetes Mellitus 16 (22.9) 27 (27.6%) 0.492

Cerebrovascular disease 11 (15.7%) 17 (17.3%) 0.780

Peripheral vascular disease 15 (21.4%) 12 (12.2%) 0.110

Glomerular filtration rate 48.4627.9 46.8623.1 0.685

Logistic Euroscore (%6SD) 23.5612.9 21.5616.2 0.399

LV ejection fraction (%6SD) 48.8611.3 47.9612.4 0.658

Aortic valve area (cm26SD) 0.6360.20 0.6760.22 0.219

Previous CABG 18 (25.7%) 27 (27.6%) 0.791

Previous PCI 16 (22.9%) 12 (12.2%) 0.070

Abstract 26 Table 2

Group 1 SignificantCAD (n[70)

Group 2 No significantCAD (n[98) p value

Transfemoral 44 (44.9%) 29 (41.4%) 0.778

Transapical 47 (48.0%) 37 (52.9%)

Transaortic 7 (7.1%) 4 (5.7%)

Conclusion The presence of significant CAD had no significantimpact upon the all-cause mortality of patients after TAVI in our


BCS Abstracts 2011

study. As yet, the impact of PCI to significant CAD upon outcomeafter TAVI is not known and will be assessed in a prospective,randomised controlled trial currently underway.

27 PLATELET MONOCYTE AGGREGATES ARE DETERMINANTSOF MICROVASCULAR DYSFUNCTION DURINGPERCUTANEOUS CORONARY INTERVENTION FOR STABLEANGINA AND NON-ST SEGMENT ELEVATION MYOCARDIALINFARCTION

doi:10.1136/heartjnl-2011-300198.27

1C A Mavroudis, 1B Majumder, 2M Lowdell, 1R D Rakhit. 1Cardiology Department, RoyalFree Hospital, London, UK; 2Haematology Department, Royal Free Hospital, London, UK

Background Microvascular dysfunction is associated with adverseoutcome in patients with acute coronary syndrome (ACS). DuringACS platelet and monocyte derived chemokines, in conjunctionwith adhesion molecule expression, promote the inflammatoryprocess. Activated platelets express p-selectin which binds to the p-selectin glycoprotein ligand on the monocyte forming plateletmonocyte aggregates (PMA). PMA expression is a sensitive markerof platelet activation and inflammation. Although platelet mono-cyte interaction is a normal physiological process, in the presence ofplatelet activation, activated (CD62+ PMA) may be directlyinvolved in the pathophysiology of intracoronary inflammation andmicrovascular dysfunction in ACS.Aim To investigate the relationship between microvasculardysfunction and PMA expression in patients with stable angina andnon-ST elevation myocardial infarction (NSTEMI).Methods Six patients with stable angina undergoing elective PCI andsix patients with NSTEMI undergoing non-elective PCI were recruited.Microvascular dysfunction was assessed by measuring the coronarywedge pressure (CwP) and the index of Microvascular resistance (IMR)using a single pressure-temperature sensor-tipped coronary wire fromthe simultaneous measurement of distal coronary pressure and ther-modilution derived mean transit time (Tmn) of a bolus of salineinjected at room temperature into the coronary artery duringmaximum hyperaemia. Blood samples were taken from the coronaryartery (distal to the culprit lesion), aorta and the right atrium for PMAestimation. PMAs were assessed using fluorescent monoclonal anti-bodies and flow cytometry. Total PMAs were calculated and expressedas a percentage of the total monocyte count. Activated CD62+ PMAswere expressed as a percentage of total PMAs.

Results As expected CwP was higher in patients with NSTEMI(46.5 (SD) 18.8) compared with the stable angina patients (Mean(SD) 21.1 (9.3) p¼0.01). IMR was also higher in patients withNSTEMI (Mean (SD) 27.6 (12.6)) compared with patients withstable angina (Mean (SD) 20.7 (5.4) p¼0.2). Total PMAs were non-significantly higher in patients with NSTEMI (Mean (SD) 14(4.8)) compared with stable angina (Mean (SD) 10.9 (4.3) p¼0.07).CD62+ PMAs were significantly higher in patients with NSTEMI(Mean (SD) 26.9 (12.2)) compared with stable angina (Mean (SD)13.7 (5.1) p¼0.02) Abstract 27 figure 1. CwP correlated positivelywith total PMA (p¼0.01) in NSTEMI but not in stable anginapatients. However, IMR correlated positively with total PMAs in bothstable angina (p¼0.02) and NSTEMI (p¼0.08) Abstract 27 figure 2.



Conclusions PMAs are elevated in stable coronary disease and ACSwith elevated activated CD62+ PMA a hallmark of ACS. PMAscorrelate with measured microvascular dysfunction during PCI instable angina and NSTEMI. This study supports the hypothesis thatPMA formation may be important determinants of platelet activation,inflammation and microvascular dysfunction in coronary disease.

28 LOW FRAME RATE SCREENING DURING PERCUTANEOUSCORONARY ANGIOPLASTY SIGNIFICANTLY REDUCESRADIATION EXPOSURE, GIVES GOOD IMAGE QUALITYWITHOUT AFFECTING PATIENT OUTCOME

doi:10.1136/heartjnl-2011-300198.28

1S J Wilson, 1P Venables, 2O Gosling, 1V Suresh. 1South West Cardiothoracic Centre,Plymouth, UK; 2Royal Devon and Exeter Hospital, Exeter, UK

Introduction Minimisation of radiation exposure during cardiacprocedures is required by statute (IRMER 2000). During coronaryangioplasty 47% of radiation dose is related to screening at standard



BCS Abstracts 2011

frame rate (15 frames per second). Digital fluoroscopic technologyhas improved imaging making the use of lower frame ratesfeasible. This study assessed whether low frame rate screening (7.5frames per second) reduced radiation without affecting patientoutcomes.Method We prospectively collected data from consecutive coronaryangioplasty procedures performed at reduced screening frame rate(7.5 frames per second). We included elective, urgent and emergencyprocedures. Audit data from procedures performed at standard framerate with the same inclusion criteria were used as a control group.Phillips Allura flat plate XPER FD10 catheterisation equipment wasused. The frame rate could be increased at the operator ’s request,and any safety concerns were reported immediately.Data collection

Patient data:< Age< Weight (Kg)< Height (cm)

Radiation data:< Screening DAP (mGycm2)< Total DAP (mGycm2)< Total Fluoroscopy time (mm:ss)< Number of acquisition runs

Operator outcome:< Need to increase screening frame rate

Patient outcome:< 30 day incidence of major adverse cardiovascular event

(MACE): death, non-fatal myocardial infarction or need forurgent revascularisation.

Results 55 consecutive studies were examined at low-frame rateand compared with the audit control group (n¼105). Mean agewas 67 in the low screening rate group and 65 in the control group.Weight was similar in both groups (83 kg vs 82 kg). The screening

times and number of acquisition runs were similar in each group.In every case image quality was acceptable, with no requirementfor increased screening frame rate. No safety concerns werereported. 30-day incidence of major adverse cardiovascular events(MACE) was similar in both groups. In the screening group therewas 1 MACE event at 30 days (2%), with 2 MACE events (2%) inthe control group. Screening and Total DAPs (mean mGycm2) were33% and 16% lower respectively in the low frame rate group.Statistical comparison was made with the Man-WhitneyU-test. This showed a significant reduction in the Screening DAP(p#0.01) with low frame rate screening. See Abstract 28 table 1and graph.Conclusions Low frame rate screening is a practical way of reducingradiation exposure in line with the ALARA “As Low As ReasonablyAcheivable” principle. Having shown that low frame rate screeningfor coronary angiography gives good imaging quality and is safe, wenow demonstrate that low frame rate screening coronary angio-plasty is also safe. Radiation exposure from screening is significantlyreduced by 33% and total exposure is reduced by 16%. Low framerate screening should be standard practice where modern facilitiesallow. We suggest that centres currently using 15 frames per secondscreening should undertake a similar assessment in order to mini-mise radiation.

29 BIVALIRUDIN IN PATIENTS UNDERGOING PRIMARYPERCUTANEOUS CORONARY INTERVENTION FOR ACUTEST-ELEVATION MYOCARDIAL INFARCTION: OUTCOMES INA LARGE REAL-WORLD UK POPULATION

doi:10.1136/heartjnl-2011-300198.29

1C Eftychiou, 1R J Shelton, 1A Liu, 1K Somers, 1P Tooze, 2L Makri, 1D Barmby,1J M McLenachan, 1J M Blaxill, 1S B Wheatcroft, 1J P Greenwood, 1D J Blackman.1Leeds General Infirmary, Leeds, UK; 2Statistical service of Cyprus, Nicosia, Cyprus

Background The HORIZONS-AMI trial demonstrated a signifi-cantly lower early and late mortality in patients undergoing primaryPCI (PPCI) treated with bivalirudin compared to a Glycoprotein IIb/IIIa inhibitor (GPI) + heparin. However, concerns remain regardingthe increased incidence of acute stent thrombosis (ST) withbivalirudin, the apparently worse outcomes in the absence ofadditional pre-procedural heparin, and the translation of trialresults into a real-world population. We evaluated the outcomes ofpatients undergoing PPCI with bivalirudin in a large all-comers UKsetting.Methods All patients who underwent PPCI in Leeds General Infir-mary from 1 January 2009 to 31 December 2009 were prospectivelyentered into a dedicated registry. Demographic, procedural, and 30-day outcome data were obtained by abstraction from the ONSmortality database and BCIS PCI database, review of hospital notes,and telephone follow-up. Bivalirudin was administered as a bolus,high-dose intra-procedural infusion, and low-dose infusion for 4 hpost-PCI. Additional heparin was not routinely given, but wasfavoured by some operators. Bail-out GPI was administeredaccording to physician judgement. Primary endpoints were death,MACE (death, re-infarction, stroke, unplanned target vessel revas-cularisation (TVR)), and stent thrombosis (ST) (ARC definitiondefinite/probable) at 30-days follow-up.Results 968 patients (age 63.5613 years, 71.9% male, 13.2%diabetics) underwent PPCI. Bivalirudin was given in 882 patients(91.1%), and GPI + heparin in 85 (8.8%). Of bivalirudin-treatedpatients 100 (11.3%) also received heparin (29 pre-PCI and 80during) while bail-out GPI was used in 91 (10.3%). Thirty-dayoutcomes are shown in Abstract 29 table 1. All-cause mortality was5.2% in the bivalirudin treated patients. Acute SToccurred in 1.0%, amedian of 2 h post-PCI, and within 6 h in 90%. Mortality in

Abstract 28 Table 1

ScreeningDAP (mGycm2)

Total DAP(mGycm2)

Fluoro time(seconds)

Number ofacquisitions

Standard (15 fps) 28564.5 60746.9 770 26.7

Low (7.5 fps) 19248.5 50953.4 800 26.8

Mean DAP reduction �33% �16% e e

Significance p<0.01 n/s n/s n/s



BCS Abstracts 2011

patients who suffered acute ST was 20%, compared to 80%following subacute ST. There was no difference in outcomesbetween bivalirudin treated patients who also received heparincompared to those who didn9t (death 7.0% vs 5.0%, p value: 0.80;MACE 14.0% vs 10.8%, p value: 0.32; acute ST 0% vs 1.2%, p: 0.61).

Abstract 29 Table 1 Outcomes at 30 days

All patients Bivalirudin GPI + heparin p value

No. of patients 968 882 85

Death 52 (5.4%) 46 (5.2%) 6 (7.1%) 0.450

Cardiac death 45 (4.7%) 39 (4.4%) 6 (7.1%) 0.277

Re-infarction 16 (1.7%) 14 (1.6%) 2 (2.4%) 0.645

Unplanned TVR 12 (1.2%) 10 (1.1%) 2 (2.4%) 0.286

Stroke 56 (5.8%) 54 (6.1%) 2 (2.4%) 0.222

Death, re-infarction, stroke or TVR 110 (11.4%) 100 (11.3%) 10 (11.8%) 0.906

Acute stent thrombosis 10 (1.0%) 9 (1.0%) 1 (1.2%) 0.604

Subacute stent thrombosis 15 (1.6%) 13 (1.5%) 2 (2.4%) 0.386

Conclusion Routine use of bivalirudin in a large UK all-comersprimary PCI population was associated with excellent 30-dayoutcomes, including all-cause and cardiac mortality. Acute stentthrombosis was infrequent, despite the absence of routine additionalheparin.

30 COMPARISON OF BIVALIRUDIN VS ABCIXIMAB VS“UNFRACTIONATED HEPARIN ONLY” FOR PRIMARYPERCUTANEOUS CORONARY INTERVENTION IN A HIGH-VOLUME CENTRE

doi:10.1136/heartjnl-2011-300198.30

R Showkathali, J Davies, N Malik, W Taggu, J Sayer, R Aggarwal, P Kelly. The EssexCardiothoracic Centre, Basildon, UK

Introduction Primary percutaneous coronary intervention (PPCI) hasbeen established as a standard therapy for ST elevation myocardialinfarction (STEMI). In addition to thrombectomy and unfractio-nated heparin (UFH), thrombus burden in STEMI may require useof more potent antithrombotic agents. Bivalirudin is shown to besuperior to abciximab in reducing the net adverse clinical events andmajor bleeding in STEMI in the HORIZONS-AMI trial (Stone et alNEJM, 2008). We aimed to carry out a “real world” comparison ofdifferent anti-thrombotic regimes in patients undergoing PPCI inour unit.Methods Our PPCI service started in September 2009 and weincluded all patients undergoing PPCI between September 2009 andSeptember 2010. Prospectively entered data were obtained from ourdedicated cardiac service database system (Philips CVIS). Mortalitydata were obtained from the summary care record (SCR) database.We used Fisher9s exact test to compare clinical outcomes betweenthe groups.Results Of the 998 patients admitted with suspected STEMI to ourunit during the study period, 776 (77.8%) underwent PPCI. Afterexcluding patients who had both bivalirudin and abciximab duringtheir procedure (n¼15), we divided the others (n¼761) into 3 groupsaccording to the anti-thrombotic regime used (Grp 1- Abciximab+UFH, Grp 2- Bivairudin+UFH and Grp 3- “UFH only”). Patientdemographics and procedural information are given in Abstract 30table 1. Continuous data are presented as mean6 SD. Clinicaloutcomes are shown in Abstract 30 table 2. In-hospital and 30-daymortality did not differ between patients who had bivalirudin vsabciximab (5.6% vs 3.8%, p¼0.35 and 6.8% vs 5.2% p¼0.53respectively). Both acute and 30 day stent thrombosis rates werealso similar in the two groups (0.6% vs none, p¼0.3, 0.6% vs 0.9%,

p¼1.0 respectively). Even though the bleeding risk was higher in theabciximab group when compared with bivalirudin, this was notsignificant (5.8% vs 3.1%, p¼0.27). There was also no difference inthe outcomes between the bivalirudin and “UFH only” groups formortality, stent thromboses (acute and 30-day) and major bleeding.The abciximab group had significantly higher major bleeding ratesthan the “UFH only” group (5.8% vs 2.4%, p¼0.04); all otheroutcomes were similar.

Abstract 30 Table 1

Abciximab +UFH (n[346)

Bivalirudin +UFH (n[162)

UFH only(n[253)

Age in yrs (range) 64614.1(25e99)

65613.0(31e94)

67613.2(30e96)

Male (%) 77.7 72.2 66.8

Diabetes (%) 12.4 6.2 11.5

Pre-procedure cardiogenic shock (%) 7.8 6.2 4.7

Drug eluting stent (at least one) (%) 56.1 56.8 53.8

No of stents 1.460.9 1.460.8 1.460.9

Single vessel PCI (%) 91.3 87 89.3

Three vessel PCI (%) 1.4 1.9 2

Radial procedure (%) 28 26.5 31.2

Abstract 30 Table 2

%Abciximab +UFH (n[346)

Bivalirudin +UFH (n[162)

UFH only(n[253)

In-hospital Mortality (includingcardiogenic shock)

3.8 5.6 5.1

30 day Mortality (including cardiogenicshock)

5.2 6.8 7.1

30 day Mortality (excluding cardiogenicshock)

3.5 4.9 5.5

Stent Thrombosis (within 30 days) 0.9 0.6 1.2

Acute stent Thrombosis (24 h) # 0 0.6 0.4

Major bleed requiring blood transfusion(non CABG related)

5.8 3.1 2.4

Access related bleed requiringtransfusion (includes IABP related)

3.8 1.9 1.2

Conclusion These “real-world” data do not show any significantdifference in the clinical outcome for patients who had bivalirudinor abciximab. There was no advantage seen with the more expensiveagent (abciximab) in keeping with previous trial data. Thereforebivalirudin should be considered as a non-inferior alternative toabciximab. This would have considerable economic benefits in thepresent situation. The “UFH only” group had similar outcomes toboth bivalirudin and abciximab, which suggests that this may be aviable alternative in its own right. However, our study is clearlylimited by not being randomised and those patients treated withUFH alone may have been a lower risk group.

31 ASSESSMENT OF LEFT VENTRICULAR FUNCTION WITHCARDIAC MRI AFTER PERCUTANEOUS CORONARYINTERVENTION FOR CHRONIC TOTAL OCCLUSION

doi:10.1136/heartjnl-2011-300198.31

1G A Paul, 2K Connelly, 1A J Dick, 1B H Strauss, 3G A Wright. 1Sunnybrook HealthSciences Centre, Toronto, Ontorio, Canada; 2St Michaels Hospital, Toronto, Ontorio,Canada; 3University of Toronto, Toronto, Ontorio, Canada

Objective To assess the role of CMR in the treatment of true chronictotal occlusions (CTO) with percutaneous coronary intervention(PCI) and drug eluting stent implantation.


BCS Abstracts 2011

Introduction Successful PCI for CTO may confer an improvedprognosis and a reduction in major adverse cardiac events (MACE).However most trials have included occlusions of short duration (lessthan 4 weeks). In this study we assessed the impact of PCI on LVfunction in patients with true CTOs (TIMI flow grade 0 and greaterthan 12 weeks duration) using serial CMR imaging as well as thepredictive value of late gadolinium enhancement when performedprior to revascularisation.Methods Thirty patients referred for PCI to a single vessel CTOunderwent CMR examination prior to and 6 months after PCI.Technical success was defined as recanalisation of the occludedvessel and DES implantation with a final residual diameter stenosis<30%. LV function and infarct size were assessed using a 1.5T GEMRI system. Segmental wall thickening (SWT) was measuredwithin the perfusion territory of the CTO using the 16-segmentmodel and segments were dysfunctional if the SWTwas #45%. Thetransmural extent of infarction (TEI) was calculated by dividing thehyperenhanced area by the total area3100; a score of #25% wereconsidered viable.Results Technical success was achieved in 19 of the 30 patients(63%). CTO duration was greater in patients with failed revascu-larisation but other baseline demographics were well matchedbetween groups (Abstract 31 table 1). PCI-CTO success resulted in asignificant increase in LVEF when compared to both baseline (50 6

13 vs 54 6 11; p<0.01) and with PCI-CTO failure (11.8 6 19.8 vs-2.3 6 5.1, p<0.01, Abstract 31 figure 1). In dysfunctional but viablesegments only PCI success conferred a significant improvement inSWTcompared to baseline (26 6 6 vs 40 6 10; p<0.001, Abstract 31figure 2). There were no episodes of MACE in either group at21 months follow-up.

Abstract 31 Table 1

Total (n[30)CTO-PCI Success(n[19)

CTO-PCI Failure(n[11) p value

Age/ years 62.2610.2 62.469.8 61.8611.4 0.89

Male, n (%) 25 (83) 14 (74) 11 (100) 0.13

CCS Anginal Class 2.1360.68 2.2160.63 2.060.77 0.42

LVEF/ % 53.0611.6 50.3612.6 57.668.1 0.09

CTO duration, months 36.9670.8 12.6626.4 78.86101.1 0.01

Vessel, n (%)

RCA 16 (53) 9 (47) 7 (64) 0.35

LAD 11 (37) 7 (37) 4 (36)

LCx 3 (10) 3 (16) 0

Prior MI, n (%) 17 (59) 11 (58) 6 (56) 0.61

Diabetes Mellitus, n (%) 7 (23) 5 (26) 2 (18) 0.61

Hypertension 23 (77) 14 (74) 9 (82) 0.61



Conclusion PCI-CTO success of true CTOs can improve global LVfunction. The TEI, assessed with CMR, can be used to help predictimprovements in regional wall function. PCI-CTO failure was notassociated with increased MACE at medium-term follow-up.

32 DOES COMPLETE REVASCULARISATION CONFERS A LONGTERM SURVIVAL BENEFIT IN PATIENTS WITHCHRONICALLY OCCLUDED CORONARY VESSELS?

doi:10.1136/heartjnl-2011-300198.32

N H Shah, M F Khan, T Ungvari, P H Loh, L Buchanan, A Hoye, R M Oliver, S Thackray,J L Caplin, M F Alamgir. Castle Hill Hospital, Kingston upon Hull, UK

Introduction Chronic total occlusion (CTO) of coronary vessels is arelatively common finding on diagnostic angiography. There hasbeen increasing interest in this clinically important area withdevelopment of technologies resulting in improved recanalisationrates. However, long term survival data in this cohort is lacking. Inthis study we looked at survival of patients in whom complete,successful revascularisation was achieved.Methods We identified consecutive patients, found to have CTO ofat least one vessel of more than 1-month duration, on angiographyperformed between January 1999 and August 2000 in a singletertiary centre. We used a dedicated database to record data onvariables and used central National Health Service database toobtain survival data. Results were analysed using SPSS statisticsversion 17.Results We included 331 patients in the analysis. Mean age was 56.8619.8 years, 76.1% were male and 21.8% (n¼71) were diabetic.Mean duration of CTO was 29.5625.9 months and was only reli-ably estimated in 82.5% of cases. Median follow-up duration was10.0963.3 years. Complete revascularisation was successfullyachieved in 53.5% (n¼177) patients, while 46.5% (n¼154) wereeither treated medically from the outset or had failed or incompleterevascularisation. Both groups were age matched. Overall 10-yearsurvival was 66.5%; those with complete revascularisation hadsignificantly improved survival over those with incomplete revas-cularisation or medical therapy (75.1% vs 56.5%, p<0.001).Conclusion Complete revascularisation confers a significant long termsurvival in patients with CTO and underscores the importance ofimproved recanalisation rates when performing angioplasty in thispatient group. Overall survival was relatively poor and emphasises theimportance of optimal medical therapy in this cohort.


BCS Abstracts 2011


33 COMPLETENESS OF REVASCULARISATION PREDICTSMORTALITY FOLLOWING PERCUTANEOUS CORONARYINTERVENTION: UTILITY OF THE BCIS-1 JEOPARDY SCORE

doi:10.1136/heartjnl-2011-300198.33

K De Silva, G Morton, P Sicard, E Chong, A Indermeuhle, B Clapp, M Thomas,S Redwood, D Perera. St. Thomas’ Hospital, King’s College London, London, UK

Introduction Many coronary-scoring systems are complicated to useon a day-to-day basis, have varying degrees of reproducibility andexclude important subsets of patients such as those with previouscoronary artery bypass grafts (CABG) or left main stem (LMS)disease (Abstract 33 table 1). The recently described BCIS-1Myocardial Jeopardy score (BCIS-1 JS), a modification of the DukeJeopardy score to include LMS and CABG, is simple to use andovercomes many of these limitations. We assessed the prognosticrelevance of the BCIS-1 JS in patients undergoing percutaneouscoronary intervention (PCI).

Abstract 33 Table 1

Left MainStem Diseaseclassified

Patientswith CABGclassified

Easeof use

Relevance tocontemporary PCI

Prognosticvalidation

Duke JeopardyScore (Original)

x x O x O

Syntax Score O x x O O

BCIS-1 JS O O O O x

Methods Consecutive patients undergoing PCI between 2005 and2009 a single cardiac centre were screened. Patients were eligible ifthey had undergone assessment of left ventricular function beforePCI and the sample was enriched for coronary artery bypass graft(CABG) cases by using the following weightingd1 CABG: 3 non-CABG. Clinicians (who were blinded to clinical or outcome data)scored diagnostic and procedural coronary angiograms. The BCISd1JS was recorded before and after PCI (range: 0 to 12) and a Revas-cularisation Index (RI) calculated as RI¼(JSPREdJSPOST)/JSPRE.RI¼1.0 indicates full revascularisation and 0 indicates no revascu-larisation. The primary end-point was all-cause mortality. Mortalitydata was captured by tracking the database of the UK Office ofNational statistics. Predictors of outcome were assessed byunivariate and multivariate analyses.Results 660 patients were included (6869 years). 44% presented asacute coronary syndromes with 41% having left ventricular

dysfunction. Over a follow-up period of 2.661.1 years there were 42deaths. All-cause mortality was inversely related to baseline BCIS-1JS (HR 2.20 (1.34 to 3.62), p¼0.002) and to post-PCI BCIS-1 JS (HR3.98 (2.33 to 6.78), p¼0.0001). Increasing degrees of revascularisationwere associated with improved survival (Abstract 33 figure 1); arevascularisation index of $ 0.67 was associated with a survivaladvantage compared to a RI#0.66 (HR 0.39 (0.24 to 0.54), p¼0.0001)(Abstract 33 table 2). A multiple regression model, incorporating age,acuity of presentation, LV function and renal failure, demonstratedthat RI¼0.67e1 continued to be an independent predictor of survival(HR 0.51 95% CI 0.35 to 0.81, p¼0.004) (Abstract 33 figure 1).

Abstract 33 Figure 1 Cumulative survival according to Revascularisa-tion Index (RI).

Abstract 33 Table 2

VariablesUnivariate analysisHR (95% CI) p value

Multivariateanalysis HR (95% CI) p value

RevascularisationIndex (0.67e1)

0.36 (0.24 to 0.54) 0.0001 0.51 (0.33 to 0.81) 0.004

BCIS-1 JS pre PCI 1.26 (1.14 to 1.39) 0.0001 1.14 (0.65 to 2.02) 0.65

BCIS-1 JS post PCI 1.35 (1.23 to 1.48) 0.0001 1.78 (0.93 to 3.39) 0.08

LV impairment 3.76 (2.53 to 5.58) 0.0001 1.97 (1.21 to 3.20) 0.007

Age 1.04 (1.01 to 1.08) 0.01 1.04 (1.00 to 1.08) 0.05

Renal dysfunction 5.82 (2.77 to 12.24) 0.0001 3.74 (1.60 to 7.37) 0.002

Acute coronarysyndrome

2.31 (1.24 to 4.30) 0.008 1.30 (0.63 to 2.66) 0.47

Cardiogenic shock 14.56 (6.45 to 32.88) 0.0001 2.83 (0.69 to 11.54) 0.15

Previous CABG 3.35 (1.80 to 6.25) 0.0001 1.83 (0.88 to 3.82) 0.10

Conclusion The BCIS-1 Jeopardy Score predicts mortality followingPCI. Furthermore, it can be used to assess the degree of revascular-isation, with more complete revascularisation (RI$0.67) conferringa survival advantage in the medium term.

34 COMPARISON OF PCI VS CABG IN INSULIN TREATED ANDNON-INSULIN TREATED DIABETIC PATIENTS IN THECARDIA TRIAL

doi:10.1136/heartjnl-2011-300198.34

1A Baumbach, 2S Kesavan, 3K Beatt, 4E Cruddas, 4M Flather, 2G Angelini, 5R Hall,6A Kapur. 1Bristol Heart Institute, Bristol, UK; 2Bristol Heart Institute, Bristol, UK;


BCS Abstracts 2011

3Mayday University Hospital, London, UK; 4Royal Brompton, London, UK; 5ImperialCollege, London, UK; 6London Chest Hospital, London, UK

Aims The CARDia trial randomised diabetic patients to coronaryartery bypass grafting (CABG) or percutaneous coronary intervention(PCI) and concluded that PCI is a potentially safe and feasible alter-native to CABG in selected patients with diabetes mellitus (DM) andmultivessel coronary artery disease. The impact of insulin treatmenton clinical outcomes after revascularisation is unclear. The presentstudy is a sub group analysis of the CARDia trial comparing thecardiovascular outcomes at 12 months following revascularisationbetween the insulin treated (IT) and non-insulin treated (NIT) group.Methods 508 patients with an established diagnosis of DM and denovo coronary artery disease were identified and randomised toCABG or PCI. Of those, 316 patients were treated with oral anti-diabetic medication and the rest were treated with additionalsubcutaneous insulin injections. Demographics, clinical presentation,history, haemodynamic parameters, anti diabetic therapy, concom-itant medications, duration of DM and HBA1C were documented.Death, stroke andmyocardial infarctionwere classified as the primaryoutcome events. The secondary outcome events included death, MI,Stroke, repeat revascularisation and TIMI major bleed. The clinicalresults of patients in the ITand NIT groups were compared.Results There were 192 patients in the IT group (37.8%). Asianpatients constituted one fifth of the total population with a slightlyhigher representation (24.5% vs 21.6%) in the NIT. The clinicalseverity of dyspnoea, heart rate, systolic and diastolic BP, body massindex, risk factors for coronary artery disease appeared similar in the ITand NIT groups, but more patients in the IT group had a prior MI(30.7% vs 19.6%, p¼0.004) and duration of diabetes was longer in theITgroup (14vs6 yrs, p<0.001). For the comparisonofCABGvsPCI forthe primary outcome events the HR and 95% CI in the IT and NITgroups respectivelywere 1.66 (0.76 to 3.76) and 1.01 (0.51 to 2.01). Fordeath,MI, stroke, repeat revascularisation theywere2.47 (1.18 to5.20)in the ITand 1.41 (0.71 to 2.57) in the NIT group. The results suggestthat IT patients may have a worse outcome with PCI compared toCABG, whereas no difference was found for NIT patients.Conclusion Our data suggest that insulin treatment is a marker forhigher risk for PCI when compared with CABG. Treatment withinsulin rather than diabetic status alone should be considered whenchoosing the mode of revascularisation.

35 SUCCESSFUL RECANALISATION OF CHRONIC TOTALOCCLUSIONS IS ASSOCIATED WITH INCREASED LONGTERM SURVIVAL

doi:10.1136/heartjnl-2011-300198.35

1J M Behar, 1D A Jones, 1R Weerackody, 1K Rathod, 1C J Knight, 1A K Kapur, 1A Jain,1A Wragg, 2C A Thompson, 1A Mathur, 1E J Smith. 1The London Chest Hospital, Bartsand the London NHS Trust, London, UK; 2Department of Cardiology, Yale University,New Haven, Connecticut, USA

Introduction Chronic total occlusion (CTO) remains a challenginglesion subset. Despite advances in equipment and expertise, manyCTO patients may not be offered PCI as physicians perceiveprocedural success may be lower, and the anatomy is stable. The aimof this study was to investigate the impact of procedural success onmortality following CTO-PCI in a large cohort of patients in thedrug eluting stent era.Methods 6122 consecutive patients underwent elective PCI at asingle centre (October 2003eMay 2010), 836 (13.7%) for CTO.Demographic and procedural data were collected at the time ofintervention (Abstract 35 table 1). In-hospital MACE (myocardialinfarction, urgent revascularisation, stroke or death) was docu-mented at discharge. All cause mortality data was obtained from theOffice of National Statistics via the BCIS/CCAD national audit out

to 4 years (mean 2.961.6) and stratified according to successful orunsuccessful CTO recanalisation.

Abstract 35 Table 1

Successful(n[572)

Unsuccessful(n[264) p value

Age 62.460.47 63.760.69 0.1

Male 433 (75.7%) 209 (79.2%) 0.3

Diabetes 151 (26.9%) 74 (28.6%) 0.6

Hypertension 320 (63.8%) 160 (66.6%) 0.5

Hypercholesterolaemia 281 (56.0%) 147 (61.2%) 0.2

Previous MI 174 (31.7%) 94 (36.4%) 0.2

Radial access 123 (21.5%) 47 (17.8%) 0.3

Femoral access 416 (72.7%) 193 (73.1%) 0.6

Dual site access (bilateral femoralor radial + femoral)

23 (4.0%) 18 (6.8%) 0.5

Results 572 (68.4%) CTO procedures were successful. Coronarystents were implanted in 96.9% (mean 2.360.1 stents per patient,70% drug eluting). Prior revascularisation was more frequent amongpatients with unsuccessful CTO-PCI than successful; prior CABG16.5% unsuccessful vs 7.4% successful, (p<0.0001), PCI 36.0% vs21.2%, (p<0.0001). Baseline characteristics were otherwise similar(Abstract 35 table 1). Intra-procedural complications (coronarydissection, perforation, access site (dissection, haematoma) weremore frequent in unsuccessful cases (19% (52) vs 4.1% (20)(p<0.0001) but did not have an impact on in-hospital MACE (2% vs1.8%, p¼0.6). All cause mortality was 8% (21) in the unsuccessfulgroup and 3% (17) in the successful group out to 4 years, (Abstract35 figure 1). Mortality following successful CTO-PCI was similar tothat of the non-CTO elective PCI group (5.1%, p¼NS).

Abstract 35 Figure 1 All cause mortality after PCI for elective patients.

Conclusion A successful angiographic outcome following CTO-PCI isassociated with a survival advantage out to 4 years following inter-vention. These data suggest that the adoption of new techniques andtechnologies to improve procedural success may improve prognosis.

36 IN-STENT RESTENOSIS PRESENTS AS AN ACUTECORONARY SYNDROME (ACS) IN 40% OF CASES: NOTSIMPLY A BENIGN CLINICAL ENTITY

doi:10.1136/heartjnl-2011-300198.36

H Abu-Own, E Sammut, K Rathod, L A McGill, D A Jones, A Jain, C Knight, A Mathur,A Wragg. Barts and the London NHS Trust, London, UK

Background In-stent restenosis (ISR) following stent implantationmay occur in 20% to 40% of the cases according to patient and lesioncomplexity. Although in the past ISR used to be seen as a “benign”


BCS Abstracts 2011

event, more recent studies suggested that a reasonable amount ofpatients with ISRmany develop ACS as the firstmanifestation of thisadverse event. The aim of this study was to determine the differentclinical presentations of ISR in a large cohort of consecutive, non-selected patients and compare with native coronary disease.Methods 14445 consecutive patients underwent PCI at a single centre(October 2003eMay 2010), we identified 922 (6.4%) cases presentingwith restenosis after previous PCI. All patients with restenosispresented with new or recurrent symptoms. Demographic and proce-dural data were collected at the time of intervention (Abstract 36table 1). In-hospital MACE (myocardial infarction, urgent revascular-isation, stroke or death) was documented at discharge. All causemortality data was obtained from the Office of National Statistics viathe BCIS/CCAD national audit out to 3.2 years (mean 3.161.8 years).

Abstract 36 Table 1

TotalRestenosis Native disease Sign[922 n[13 523 e

Age 63.09 63.76 0.0868

Ethnicity (cau) 683 (74.2%) 9160 (97.8%) p<0.0001

Previous MI 411 (44.6%) 2160 (23.1%) p<0.0001

Previous CABG 120 (13.0%) 648 (6.9%) p<0.0001

DM 299 (32.5%) 1986 (21.2%) p<0.0001

HTN 545 (59.1%) 4170 (44.5%) p<0.0001

Hchol 544 (59.0%) 3540 (37.8%) p<0.0001

Card Shock 6 (0.7%) 100 (1.1%) 0.2339

Results Restenosis presented in 60.4% as stable angina, 30.6% asunstable angina/Non-ST elevation MI and 9% with ST-elevationMyocardial Infarction. Cardiogenic shock was reported in 6 patients(0.65%). Women had a higher incidence of unstable angina/non-STEMI comparedwithmen (32.6% vs 29.1%) but a lower incidence ofSTEMI (5%vs 10.4%). Baseline characteristics are listed inAbstract 36table 1. Mortality rate was 0.98% at 30 days, 3.9% at 1 year and 8.7%at 5 years in patientswith restenosis. Comparing the restenotic groupwith those undergoing PCI for de novo coronary artery disease, therewere similar ages and incidence of cardiogenic shock but the restenoticgroup had higher rates of baseline risk factors (diabetes, hypertension,hyerpcholesterolaemia) and higher rates of previous CABG and MI.There was also a higher proportion of South Asians in the restenoticgroup. See Abstract 36 table 1. Comparing outcome measures, therewere similar rates of inhospital MACE in the 2 groups and over a 5-year follow-up period, there was no difference in all cause mortality.There was no difference in outcome of patients with restenosis vs denovo coronary artery disease regardless of presentation (angina, UA/NSTEMI/STEMI). See Abstract 36 figures 1 and 2.

Abstract 36 Figure 1 Comparison of mortality between restenosis andno restenosis in STABLE.

Abstract 36 Figure 2 Comparison of mortality between restenosis andno restenosis in ACS.

Conclusions Clinical in-stent restenosis can frequently present as MIand such patients are more likely to have an aggressive angiographicpattern of restenosis. Drug-eluting stents with improved designs ordrug elution systems that further decrease the incidence of ISR areneeded.

37 DECREASE IN MACE RATES ASSOCIATED WITH DRUGELUTING STENT USE IN PATIENTS WITH DIABETESUNDERGOING PCI IN LARGE DIAMETER CORONARYARTERIES

doi:10.1136/heartjnl-2011-300198.37

1A Dixit, 2S Nair, 2P Williams, 2A Wiper, 2B Clarke, 2C Deaton, 2M El-Omar, 2D Fraser,2R Khattar, 2V Mahadevan, 2L Neyses, 2F Ordoubadi, 2M Mamas. 1University HospitalManchester MHC, Manchester, UK; 2Manchester Royal Infirmary, Manchester, UK

Introduction Both large multi centre trials and registry studies havedemonstrated that PCI with drug eluting stents (DES) is associatedwith reduced MACE and restenosis rates compared to bare metalstents (BMS) in native coronary vessels, although this benefit is lessevident in those patients with a larger coronary vessel diameter andMACE rates may actually paradoxically increase in this cohort asobserved in the BASKET trial. In diabetic patients, a similar or evengreater absolute reduction in MACE rates / restenosis risk is seenassociated with DES use, although it is unclear as to whether anybenefit persists in those with larger diameter native coronary vessels.Previous data derived from diabetic patients in large diameter nativecoronary vessels has come from registry studies in which numberswere either small (<200 patients) or were from highly selectedpatient sub groups excluding high-risk individuals (SCAAR registry).Methods We therefore retrospectively studied 1165 consecutivediabetic patients with target vessel diameter $3 mm admitted toour centre for PCI from 2003 to 2009, the largest series of its kind todate. Primary endpoint was defined as total mortality and secondaryendpoint was major adverse cardiac event (MACE) defined ascomposite endpoint of Death, Stroke, MI, Stent Thrombosis andTarget Lesion / Vessel Re-Vascularisation.Results Of the 1165 patients studied, 170 had BMS and 995 hadDES. Mean follow-up period was 43.3621.8 months (median41.8 months). 73.5% were male in the BMS cohort vs 73.1% in theDES cohort (p>0.05). Mean age was 62.8611.2 in BMS and 62.3610.4 years old in DES (p¼0.55). Other demographic parameters weresimilar in both groups. There were a total of 23/170 deaths in BMScohort (13.5%) and 91/995 in DES cohort (9.1%), (HR 1.38; 95% CI0.83 to 2.27, p¼0.21). A total of 42/170 (24.7%) and 163/995 (16.3%)MACE events were observed in the BMS and DES cohort respec-tively (HR 1.49; 1.02 to 2.19, p¼0.04). Multivariate analysis


BCS Abstracts 2011

illustrated that use of BMS was independently associated withincreased risk of MACE (HR 1.54; 95% CI 1.05 to 2.25, p¼0.03),driven through an increase in revascularisation.Conclusion In conclusion, in one of the largest analyses of its kind,use of DES in patients with diabetes in a real world setting under-going PCI in large diameter coronary vessels ($3 mm) is safe and isindependently associated with a reduction in MACE events. This isin contrast to that of non-diabetic patients where the benefits ofDES in large diameter coronary vessels are less evident.

38 FALSE ACTIVATION FOR PRIMARY PERCUTANEOUSCORONARY INTERVENTION IS NOT A BENIGNPHENOMENON

doi:10.1136/heartjnl-2011-300198.38

U Chaudhry, C Mavroudis, R D Rakhit. Cardiology Department, Royal Free Hospital,London, UK

Introduction Primary percutaneous coronary angioplasty (PPCI) isthe preferred reperfusion strategy following an acute ST elevationmyocardial infarction (STEMI). Since 2005 24/7 primary PCI hasbeen the first line treatment for an acute STEMI in our centre. 93%of patients are direct access admissions by London Ambulance but asignificant proportion (up to 20%) do not fulfil the diagnosticcriteria for STEMI and are termed “false activations”. Data on theoutcome of this cohort of patients is limited.Aim To review the clinical outcome of patients presenting to ourheart attack centre with false activation PPCI.Method From January 2008 until October 2010, we identified 209false PPCI activations defined as patients with incomplete diag-nostic criteria for acute STEMI: absence of chest pain and/or typicalECG features (ST elevation or new LBBB). Data was collected via a“false activation” database together with retrospective review of caserecords.Results Complete data was available in 165 cases. 71% were maleand 29% were female (mean age 67). The mean length of stay was4 days (range 1e33). 71% presented with chest pains and 29% hadno chest pains, but presented with breathlessness, palpitations orsyncope. The ECG abnormality was non-specific ST-T changes in22%, LBBB in 19%, left ventricular hypertrophy in 15%, fixed STelevation or Q waves in 14%, early repolarisation changes in 10%,RBBB in 8% and other ECG abnormalities in 12%. The final diag-nosis was non-STelevation acute coronary syndrome (NSTEACS) in19%, sepsis in 19% and congestive heart failure (CHF) in 15%. Stableangina was observed in 8% and syncope in 7%. Musculoskeletal ornon-cardiac chest pains were noted in 8% and 7% of the patientsrespectively. 2% of the patients had pulmonary embolism and in 5%,a gastric cause for presentation was diagnosed. 14% had othercardiac problems, including arrhythmia, dilated cardiomyopathy,hypertension, pericarditis, pericardial effusion and late presentationSTEMI. 15% had other diagnoses. The mean follow-up period was18.7 months, during which 21.5% of false PPCI activation admis-sions died (n¼45). 25% (n¼11) died during the index admission and33% (n¼15) died within 30 days of admission. The overall 30-daymortality for false activations was 7.2%, which is higher than theoverall PPCI mortality of 6.0% (including cardiogenic shock)(p¼0.008) and 3.3% (excluding shock) (p<0.0001) in our centre. 49%of deaths were cardiac (NSTEACS and CHF), 29% sepsis and 22%other causes. The mean age for this cohort was 83.Conclusion Patients presenting with false PPCI activation have ahigh observed mortality. This is probably due to significant asso-ciated comorbidities, including occult cardiac disease. Thus, falsePPCI activation is not a benign phenomenon and masks underlyingsignificant disease. Robust pathways are required to minimise delayin further investigations and a need for risk stratification for asignificant proportion who present with NSTEACS.

39 A RANDOMISED CONTROLLED TRIAL COMPARINGCONVENTIONAL CORONARY ARTERY BYPASS GRAFTSURGERY WITH A COMPOSITE ARTERIAL GRAFTTECHNIQUE

doi:10.1136/heartjnl-2011-300198.39

1A Alahmar, 2R A Perry, 2R H Stables. 1Leicester University Hospital Glenfield,Leicester, UK; 2Liverpool Heart and Chest Hospital, Liverpool, UK

Background Composite (Y/T) coronary artery bypass graft surgery(CABG) confers full arterial revascularisation, and “hands off ” aortacompared to conventional bypass graft surgery. However, thecomposite surgical configuration could lead to preferential blood flowdown one arm than the other (left internal mammary artery LIMA orradial artery RA) with its potential impact on graft patency.Aim To investigate the impact of bypass graft configuration onshort-term grafts patency and cardiac related quality of life.Methods and Results This is a single centre randomised, controlled trialBetween March 2006 and July 2007, 322 patients undergoing isolatedbypass graft surgery at our institutionwere screened and 89 (27%)metthe inclusion criteria and were randomised. Patients were allocated toconventional (conv n¼46) or composite (comp n¼43). The twoprimary end points were graft patency defined as (Thrombolysis InMyocardial Infarction) TIMI III flow in distal anastomosis at angiog-raphy 12e24 months after surgery, and cardiac-related health statusassesses bySeattle angina questionnaire (SAQ). Baseline characteristicswere similar between the two groups apart from diabetes where therewere more diabetic patients in the composite arm than the conven-tional one (15(35%) vs 5(11%) p<0.01 respectively). Trial was stoppedprematurely following 18 months interim analysis which showedsignificant graft failure in the composite arm (40%). Final Analysiswasperformed on intention to treat basis. Sixty-five (73%) had follow-upangiography (34conv, 31 comp),with total of 116graft in conventionalarm and 100 grafts in composite arm. All patients in both groups hadLIMAgraft to left anterior descending artery (LAD).Graft patency ratewas significantly higher in the conventional compared to compositearm (95(82%) vs 59(59%) p<0.001 respectively). Three main domainsof the SAQs there was significant improvement between before and6 months after surgery in both groups. There were no significantdifferences between the two groups in the percentage of improvementin these four domains (Physical limitation, Angina stability, Anginafrequency, Quality of life).Conclusions In our randomised trial, composite bypass graft surgerywas associated with higher graft failure rate at 12e24 months aftersurgery compared to conventional type. This difference may be dueto the composite conduit configuration. Further blood flow charac-teristics study in this configuration can help understand such animportant finding and its implication on our clinical practice.Despite the difference in graft patency there were no differences inphysical limitation, angina stability, angina frequency, or quality oflife between the two groups.

40 PATIENT VS PHYSICIAN REPORTED ANGINA BEFORE ANDAFTER REVASCULARISATION OF CORONARY ARTERYDISEASE: EVIDENCE FROM A LARGE RANDOMISEDCONTROLLED TRIAL (THE SOS TRIAL)

doi:10.1136/heartjnl-2011-300198.40

C Appleby, I Kemp, R H Stables. Liverpool Heart and Chest Hospital, Liverpool, UK

Introduction The success of revascularisation therapies for coronaryartery disease (CAD) must be measured by both an improvement inhard clinical endpointsdmortality, repeat revascularisation proce-dures and myocardial infarction, the traditional focus of clinicaltrialsdand, critically for patients, the relief of angina symptoms.Interest in patient reported outcomes (PROMs) has increased,although their use in cardiovascular trials is far from universal. In


BCS Abstracts 2011

particular the differences between physician and patient reportedoutcomes has not been analysed. High quality data from the Stentor Surgery (SOS) trial allows such an analysis.Methods The SoS trial was a large RCT (n¼988) comparing stent-assisted percutaneous coronary intervention (PCI) with coronaryartery bypass grafting (CABG) in patients with multivessel CAD.Participation in the SoS trial included an appraisal of angina symptomsby both patient and physician according to the Canadian Cardiovas-cular Society (CCS) Classification System prior to, and subsequentlyat 6 and 12 months following coronary intervention. In this studypatient and doctor reported outcomes were compared systematically.Results Paired CCS scores at baseline, 6 months and 12 months wereavailable for 919, 886 and 888 cases respectively. At baseline the overalllevel of agreementwas goodwith>75%paireddata setsdemonstratinga difference of#61CCS class. Patterns of discordance change howeverbetweenbaseline and follow-up timepoints.Abstract 40figure1 showsthe paired scores at baseline, charting the patient score and, for eachCCS grade, the observed differenceddoctor (D) minus patient (P).Doctors are reluctant to record scores of 0 or 4, preferring CCS grades 2and 3. Thus there is little overall difference inmean CCS score (P 2.2 vsD 2.5, p<0.001). Yet at follow-up, doctors record freedom from angina(CCS¼0) in a more substantial proportion of the population, consid-erablymore so than patients self-report (p<0.0001) (Abstract 40 figure2). The published results of the SOS trial used doctor gradings to reportfreedom from angina at 1 year in 79% of CABG patients vs 66% of PCIpatients (p<0.0001). If patient gradings are used instead these figuresare reduced to 57% in CABG and 44 % in PCI (p<0.0001), renderingboth treatment strategies significantly less effective at relieving anginafrom a patients perspective (p<0.0001), Abstract 40 table 1.

Abstract 40 Figure 1 Difference between doctor and patientclassification of Angina before revascularisation.

Abstract 40 Figure 2 Difference between doctor and patientclassification of Angina at 12 m Fup.

Abstract 40 Table 1

PCI (%) CABG (%) p

Doctor scoring CCS class 0 66 79 <0.0001

Patient scoring CCS class 0 44 57 <0.0001

Conclusions This is the first randomised study to compare theimprovement in angina status reported by patients and cliniciansfollowing revascularisation therapy for coronary artery disease. Theobserved trend for doctors to insist that all patients must have somesymptoms at baseline, and more importantly, to suggest that agreater proportion of patients have been rendered symptom free atfollow-up (than is suggested by self-reported estimates) has impor-tant implications and may call into question our current under-standing of the impact of revascularisation.

41 REDUCED ARTERIAL WAVE REFLECTION AND ENHANCEDLV RELAXATION CONTRIBUTE TO WARM-UP ANGINA

doi:10.1136/heartjnl-2011-300198.41

1T P E Lockie, 2A Guilcher, 3C Rolandi, 1D Perera, 1K De Silva, 1R Williams, 3M Siebes,2P Chowienczyk, 1S Redwood, 1M Marber. 1Rayne Institute, St Thomas Hospital, KCL,

Abstract 41 Figure 1 (A) shows aortic pressure traces taken at peakexertion with a reduction in pressure augmentation during Ex2; (B)shows WIA with an increase in the backward expansion, or “sucking”wave originating from the microvasculature.


BCS Abstracts 2011

London, UK; 2Clinical Pharmacology, St Thomas Hospital, KCL, London, UK; 3Depart-ment of Bio-Engineering, University of Amsterdam, AMC, Amsterdam, The Netherlands

Background The mechanisms of the clinically observed phenom-enon of reduced angina on second exertion, or warm-up angina, arepoorly understood. This study compared changes in centralhaemodynamics, peripheral wave reflection and patterns of coro-nary blood flow during serial exercise that may contribute.Methods and Results 16 patients (15male, 6164.3 yrs)with a positiveexercise stress test and exertional angina completed the protocol.During cardiac catheterisation via radial access they performed 2consecutive exertions (Ex1, Ex2) using a supine cycle ergometer.Throughout exertions, distal coronary pressure (Pd) and flow velocity(V) were recorded in the culprit vessel using a dual sensor coronaryguide wire while aortic pressure was recorded using a second wire.Time to 1 mm ST depression was longer in Ex2 (p¼0.003) and ratepressure product (RPP) was higher (p¼0.025) confirming warm-up. A33% decline in aortic wave reflection (p<0.0001) in Ex2 (see Abstract41 figure 1A) coincided with a reduction in both tension time indexand diastolic time index (p<0.0001).However, the latterwas offset byreduced microvascular resistance (Pd/V), p¼0.0002, and enhanced leftventricular relaxation during Ex2 as suggested by a larger backward-travelling suctionwave (p¼0.01) onwave intensity analysis (WIA) ofthe intra-coronary signals. See Abstract 41 figure 1B. The energy of theforward compression wave and overall coronary blood flow, asmeasured by the velocity time integral, did not change.Conclusions In patients with warm-up angina, exercise induceschanges in the aortic pressure waveform, microvascular functionand LV relaxation. These combine to reduce afterload withoutcompromising myocardial diastolic blood flow thereby likelyenabling improved performance on second exercise.

42 RETROSPECTIVE CALCULATION OF SYNTAX SCORE IN 200PATIENTS UNDERGOING ELECTIVE CORONARY ARTERYBYPASS GRAFTING (CABG) AND PERCUTANEOUSCORONARY INTERVENTION (PCI); ARE WE FOLLOWINGBEST PRACTICE?

doi:10.1136/heartjnl-2011-300198.42

L J Mullen, R Edwards, R Taylor, B Nyawo. Freeman Hospital, Newcastle upon Tyne

Introduction Cardiologists are generally the gatekeepers of coronaryartery disease and have been much criticised for not discussing allpatients being considered for revascularisation therapy at an MDT(multi-disciplinary team) meeting or not referring patients with tradi-tional “surgical disease” for CABG. At the Freeman Hospital (FRH), alarge cardiothoracic unit in the North of England, patients are typicallyreferred forPCIorCABGbycardiologistsworkingwithin theNewcastleupon Tyne Trust or from district general hospitals within the network.Patients are not routinely discussed at MDT but can be brought to theweekly meeting at the discretion of the referrer. The recently reportedSYNTAX study allows objective quantification of the degree of coro-nary disease and facilitates an evidence based decision between CABGand PCI. This gives us the opportunity to examine whether electiverevascularisation is being performed appropriately at our institution.Methods We performed a retrospective analysis at the FreemanHospital. 200 patients who had elective revascularisation betweenApril 2009 and April 2010 were selected. This included 100 cases ofCABG and 100 of PCI.Half of eachwere referrals fromother hospitals.Patients’ SYNTAX scores were calculated using pre-procedure angio-grams. MDT meeting records and patients’ notes were reviewed.Results The average SYNTAX score for patients undergoing electivePCI was 15, compared to 29 for those undergoing CABG. 84% ofpatients undergoing elective PCI had SYNTAX scores less than 22.35% of all patients referred for elective CABG had scores greaterthan 33. The average SYNTAX score for CABG referrals fromoutside the trust was lower (25) than from within the trust (31).

Discussion The majority of patients undergoing PCI at the FRH haveSYNTAX scores in the lowest tertile. There is no difference in theSYNTAX scores in patients having PCI from referral bases within thecentre or from outside. In total almost one quarter of all patientsundergoing CABGhave a SYNTAX score in the lowest tertile. And thisrises to almost one third in those patients referred from district generalhospitals. Only a small number of these patients have an additionalclear indication for CABG over PCI. Furthermore we found that asignificantproportionof these donotgo throughMDTplanning.Theseresults may indicate that cardiologists are more likely to bring patientsto MDT meetings than surgeons and, according to SYNTAX scoring,more patients are inappropriately having CABG than are inappropri-ately having PCI. Based on this data in our institution discussing allpatients at anMDTand the use of SYNTAX scoring at point of referralwould be more likely to increase PCI revascularisation rates.

Abstract 42 Table 1

AverageSYNTAXscore

% Patients withSYNTAX scorein lower third(0e22)

% Patients withSYNTAX scorein middle third(23e32)

% Patients withSYNTAX score inhigher third (>33)

All Referrals for PCI 14.9 84 11 5

Referrals for PCIfrom within trust

15.0 84 12 4

Referrals for PCIfrom outside trust

14.8 84 10 6

Abstract 42 Table 2

AverageSYNTAXscore

% Patients withSYNTAX scorein lower third(0e22)

% Patients withSYNTAX scorein middle third(23e32)

% Patients withSYNTAX score inhigher third (>33)

All Referrals for CABG 28.8 24 40 35

Referrals for CABGfrom within trust

31.0 19 34 47

Referrals for CABGfrom outside trust

25.0 32 50 18

43 PROGNOSIS AFTER PRIMARY PERCUTANEOUS CORONARYINTERVENTION FOR STEMI: CAN THE SYNTAX SCORE HELP?

doi:10.1136/heartjnl-2011-300198.43

A J Brown, L M McCormick, N E J West. Papworth Hospital, Cambridge, UK

Background Factors affecting prognosis after primary percutaneouscoronary intervention (PPCI) for ST-elevation myocardial infarction(STEMI) include age at presentation, the presence of diabetesmellitus,left ventricular function and/or cardiogenic shock. Although thedebate continues over a strategy of complete revascularisation(immediate or staged) vs culprit-only, little is known about the impactof the extent of coronary disease at presentation on prognosis afterPPCI.TheSYNTAXscore, designed to stratifyoutcomes inmultivesselPCI and CABG, has been validated in unselected populations under-going elective PCI; to date, no studies have assessed its utility in PPCI.Methods Consecutive patients attending a single UK tertiary centrefor PPCI between September 2008 and June 2010 (n¼695) wereincluded. SYNTAX scoring was performed by a single trainedoperator blinded to patient details and outcome. Scoring was vali-dated by analysis of 3 separate cohorts by 2 other experiencedoperators. Patients were split into 3 subgroups as in the SYNTAXtrial (score #22 (low, L), 22.5e32 (intermediate, IM) and $32.5(high, H)), and patient data and outcome measures obtained byinterrogation of local and national databases.Results 671 of 695 patients were included in the analysis with 24being excluded owing to inability to score (previous CABG, images


BCS Abstracts 2011

unavailable). The ability to allocate a SYNTAX tertile was reprodu-cible between observers (r¼0.94). Median scores in the 3 groups were:L 14, IM 26, H 36 (Abstract 43 figure 1A). Although there was nocorrelation between SYNTAX score and patient sex or diabetic status,there was a linear relationship with patient age (r2¼0.03; p<0.0001).1-year absolute survival (Abstract 43 figure 1B) followed SYNTAXscore groups: L 94.7%, IM 88.7%, H 82.1% (p¼0.0002). Similar resultswere obtained for freedom from death or unplanned revascularisation(p<0.0001) and death or any revascularisation (p<0.0001).


Conclusions The SYNTAX score, when applied to an unselectedpopulation of patients undergoing PPCI for STEMI, providesimportant prognostic information regarding 1-year survival fromdeath and revascularisation. These findings may provide supportingevidence towards routine complete revascularisation of obstructivecoronary artery disease after PPCI.

44 PRIMARY PERCUTANEOUS INTERVENTION: HAVE WETAKEN OUR EYE OFF THE MEDICINE BALL?

doi:10.1136/heartjnl-2011-300198.44

1J D Jones, 2E Damm, 2M Nijjar, 2S Pettit, 2N M Hawkins, 2R Perry. 1UniversityHospital Aintree NHS Foundation Trust, Liverpool, UK; 2Liverpool Heart and ChestHospital, Liverpool, UK

Introduction Primary percutaneous intervention (PPCI) improvessurvival in patients with ST elevation myocardial infarction(STEMI). Significant resources have been directed to achieving timelyreperfusion throughout the UK. However, intensive medical therapyis of equal importance, with landmark clinical trials demonstratingunequivocal morbidity and mortality benefits from b-blockers,angiotensin-converting enzyme inhibitors (ACEI), and angiotensin IIreceptor blockers (ARB). All trials employed rigorous titration to

maximum clinically tolerated doses. We examined whether medicaltherapy is being applied appropriately in patients referred for PPCI.Methods Consecutive patients with STEMI referred for PPCI to alarge tertiary centre between 1st March and 1st August 2009 wereincluded (n¼167). The case records of all patients were reviewed.Myocardial infarction was diagnosed according to standard criteria.Medications and doses on admission, discharge and follow-up wererecorded. Contraindications and limits to dose escalation were noted(symptoms, systolic blood pressure <90 mm Hg, heart rate<50 bpm, serum creatinine and potassium).Results Mean agewas 62.0611.9 years, 72%weremale.Ondischarge,100% of patients were prescribed clopidogrel, 95.8% aspirin, 98.8%statin, 88.6% b-blockers, and 91.0% ACEI/ARB. However, the inpa-tient dose of b-blocker or ACEI/ARB was maximum or clinicallylimited in only 13% and 15% of patients respectively (Abstract 44figure 1). Outpatient follow-up at a mean of 5.0 months was equallyconcerning. The majority of patients (83%) were neither receivingmaximum tolerated doses of b-blocker or ACEI/ARB, nor receivedinstructions to escalate the dose (Abstract 44 figure 2).

Abstract 44 Figure 1 Inpatient titration of b-blocker and ACEI/ARB.

Abstract 44 Figure 2 Outpatient titration of b-blocker and ACEI/ARB.

Conclusion The national service framework and target driveninitiatives such as advancing quality promote “tick box” medicine.Quantitative prescribing of secondary prevention is excellent.Qualitative follow-up and titration is not. Whether suboptimaldoses convey the mortality benefits observed in landmark clinicaltrials is unknown. Frameworks to deliver titration of medicaltherapy must be explored. Options include nurse or pharmacy ledservices and expansion of cardiac rehabilitation. Reorientation isneeded to focus on both quantity and quality.

45 SHOULD PRIMARY PERCUTANEOUS CORONARYINTERVENTION BE THE ROUTINE REPERFUSION STRATEGYIN OCTOGENARIANS AND NON-AGENARIANS PRESENTINGWITH ST ELEVATION MYOCARDIAL INFARCTION?

doi:10.1136/heartjnl-2011-300198.45

R Showkathali, E Boston-Griffiths, J Davies, G Clesham, J Sayer, P Kelly, R Aggarwal.The Essex Cardiothoracic Centre, Basildon, UK

Introduction Primary percutaneous coronary intervention (PPCI) hasbeen established as standard therapy for ST elevation myocardialinfarction (STEMI). Very few trials have looked at the outcome ofPPCI in elderly patients. Even in trials which claimed to have looked


BCS Abstracts 2011

at PPCI in elderly patients such as SENIOR PAMI (Grines, 2005) andTRIANA (Bueno, 2009) the minimum age for inclusion was 70 yrsand 75 yrs respectively. With an ageing population in the westernworld, about 20% of patients admitted for suspected STEMI are$80 yrs. We evaluated the outcome of PPCI in patients $80 yrswho were admitted to our unit with STEMI.Methods Our PPCI service was started in September 2009 and weanalysed all the patients who were $80 yrs presenting to the PPCIservice between September 2009 and September 2010 (13 months).Prospectively entered data were obtained from our dedicated cardiacservice database system (Philips CVIS). Mortality data wereobtained from the summary care record (SCR) database. Follow-updata were obtained from patients’ respective district general hospi-tals and general practitioners medical records.Results Of the 998 patients who were admitted to our unit forprimary PCI for suspected STEMI during the study period, 183(18.3%) were $80 yrs of age. After excluding 51 patients (27.9%)who did not undergo PPCI, we included 132 (70.1%) patients foranalysis. Of those who were included in the study (n¼132, 63female), the mean age was 8563.95 yrs (range 80e99 yrs, median85 yrs). There were 20 diabetics (15.2%) and 39 (29.5%) hadprevious myocardial infarction. Ten patients (7.6%) were in cardio-genic shock on arrival of which 9 (90%) had an Intra aortic balloonpump (IABP). The infarct related vessel was the right coronary in42.4% and left anterior descending in 37.1%. Drug eluting stentswere used in 40.2% of patients. In-hospital and 30-day mortalitywas 14.4% and 19.7% respectively. There was a significant differencein the mortality between patients age <80 yrs and those $80 yrs(Abstract 45 figure 1). In patients $80 yrs, mortality and bleedingrisk increased markedly with advancing age (Abstract 45 table 1).


Abstract 45 Table 1

%80e84 yrs(N[62)

85e89 yrs(N[51)

‡90 yrs(N[19)

Inhospital mortality 9.7 15.7 26.3

30-day mortality 14.5 23.5 26.3

30-day MI 3.2 2.0 0

30-day CVA 1.6 0 0

Major bleeding requiring blood transfusion 1.6 3.9 10.5

Conclusion This study clearly demonstrates a significant mortalitydifference between patients aged <80 yrs and those $80 yrs treatedwith PPCI. Our 30-day mortality outcome in patients $80 yrs(19.7%) was similar to the subgroup analysis of the PPCI arm insimilar SENIOR-PAMI patients (19%). In the same analysis, thethrombolytic group had a lower (16%) mortality. Further studies arerequired to determine whether PPCI should be routinely used in veryelderly patients presenting with STEMI.

46 PROGNOSTIC VALUE OF BASELINE RENAL FUNCTION ONLONG TERM OUTCOME IN PATIENTS UNDERGOINGPRIMARY PERCUTANEOUS CORONARY INTERVENTION FORST-ELEVATION MYOCARDIAL INFARCTION

doi:10.1136/heartjnl-2011-300198.46

O P Guttmann, K Rathod, B Rathod, E Wicks, S Gallagher, D A Jones, A Jain, C Knight,A Mathur, A Kapur, A Wragg. Barts and the London NHS Trust, The London ChestHospital, London, UK

Background Renal impairment is associated with increased cardio-vascular mortality following acute coronary syndromes (ACS),however there is limited data assessing this relationship in thecontext of primary PCI and whether it exists with other majoradverse cardiovascular events.Methods Clinical information was analysed from a prospective database on 2310 STEMI patients who underwent primary PCI betweenJanuary 2004 and May 2010 at a London centre. Information wasentered at the time of procedure and outcome assessed by all-causemortality information provided by the Office of National Statisticsvia the BCIS/CCAD national audit. Estimated glomerular filtrationrate (eGFR) was calculated using the modified diet in renal diseaseequation and patients were divided into groups based on eGFR (<40,40e50, 50e60, >60 ml/min/1.73 m2). 3-year composite of MACE(death, reinfarction, stroke and target vessel revascularisation) werecompared between groups.Results The average eGFR in all patients was 73.40623.37 (95% CI72.25 to 74.56) ml/min/1.73 m2. The prevalence of coexisting riskfactors (hypertension, diabetes mellitus, hypercholesterolaemia),previous MI, previous CABG and cardiogenic shock were higher amongpatients with reduced eGFR. There was a progressive increase in MACEwith declining eGFR (OR¼4.84, 95% CI 2.94 to 7.96, for comparisonbetween the highest and lowest eGFR groups). See Abstract 46 figure 1.After adjustment for baseline characteristics including age, diabetes andcardiogenic shock renal function based on the GFR at admissionremained a strong independent predictor of outcome.

Abstract 46 Figure 1 All MACE after PCI for STEMI.

Conclusion Baseline renal dysfunction in patients undergoingprimary PCI is associated with an increased risk for combined death,re-infarction and recurrent angina. This risk increases linearly withdeclining eGFR.

47 CARDIOVASCULAR EVALUATION OF ENGLISH PREMIERSHIPRUGBY PLAYERS

doi:10.1136/heartjnl-2011-300198.47

1S Ghani, 1H Raju, 1A Zaidi, 1N Sheikh, 1S Gati, 2J Somauroo, 3S Kemp, 1S Sharma.1St Georges University London, London, UK; 2Countess of Chester Hospital; 3RugbyFootball Union (RFU)

Introduction Recent experience of pre-participation cardiovascularevaluation (PPCE) in Italian athletes demonstrates a significant


BCS Abstracts 2011

reduction in mortality from cardiomyopathies and cardiac conduc-tion disorders. Although PPCE is endorsed by large medical andsporting organisations, including the European Society of SportsCardiology, the International Olympic Committee and FIFA, thestate health system in the UK (and many other Western countries)does not support cardiovascular evaluation of athletes. Certain elitesporting organisations in the UK mandate PPCE in all athletes priorto competition. In 2010 the English Premier Rugby league intro-duced formal PPCE in all competing players.Methods Athletes participating in the English Premiership Rugbyunderwent PPCE with a structured clinical questionnaire and 12-lead ECG. Trans-thoracic echocardiogram (TTE) and additionalinvestigations were performed where indicated.Results A total of 606 players were assessed (mean age 22.9 years;range 15e37). Of these, 45 (7.4%) required TTE (35 (5.7%) due toECG abnormalities; 5 (0.08%) due to family history of sudden death;5 (0.08%) due to symptoms). ECG abnormalities warranting TTEincluded right axis deviation (n¼4), left axis deviation (n¼17), rightbundle branch block (n¼3), right ventricular hypertrophy (n¼1),abnormal Twave inversion (n¼5) and prolonged QT (n¼1). Six ofthe 45 subjects demonstrated mild changes on TTE (markedlydilated LV cavity (n¼3), mitral regurgitation (n¼1), pulmonarystenosis (n¼1), dilated aortic root (n¼1)), requiring serial surveil-lance. Five demonstrated abnormalities on TTE and/or ECG thatwarranted referral for further evaluation including exercise stresstest (n¼5), 24-h ECG (n¼5) and cardiac MRI (n¼3). The reasons forthese tests included possible arrhythmogenic right ventricularcardiomyopathy (n¼3), suspicion of hypertrophic cardiomyopathy(n¼1) and QT prolongation on ECG (n¼1). None of the playersexhibited a cardiac disorder that warranted disqualification fromsport. Overall 7.4% of athletes required further investigationfollowing initial ECG, and 1.8% required further tests followingTTE. False positive rate was 5.6%.Conclusion Cardiovascular evaluation of British rugby players with astructured questionnaire and ECG resulted in clearance of 92.6%following initial tests, and 5.6% were reassured after TTE. Only 1%players required surveillance echocardiograms and 0.8% werereferred for further diagnostic evaluation. False positive rate was5.6%. The results indicate that PPCE carried out in an expert settingresults in a relatively small number of athletes requiring furthertests, and a low false positive rate.

48 DARBEPOETIN ENHANCES ENDOTHELIAL-DEPENDENTVASOMOTOR FUNCTION IN PATIENTS WITH STABLECORONARY ARTERY DISEASE ONLY AFTER PRECEDINGISCHAEMIA-REPERFUSION

doi:10.1136/heartjnl-2011-300198.48

1,2L M Tilling, 1,2J Hunt, 1,2A Donald, 1,2B Clapp, 1,2P Chowienczyk. 1British Heart FoundationCentre, King’s College London, St Thomas’, London, UK; 2Department of ClinicalPharmacology, Cardiovascular Division, King’s College London, St Thomas’, London, UK

Background Vasoprotective effects of erythropoietin in animalmodels are mediated by endothelium-derived nitric oxide (NO) and/or mobilisation of endothelial progenitor cells (EPC) and may be

enhanced by ischaemia; whether they are present in humans isunknown. We examined whether the erythropoietin analoguedarbepoetin improves flow mediated dilatation (FMD), a measure ofendothelium-derived NO, and whether this is influenced bypreceding ischaemia-reperfusion.Methods 36 patients (50e75 years) with stable coronary arterydisease were randomised to receive a single dose of darbepoetin 300 mgor saline placebo. Immunoreactive erythropoietin was measured byan enzyme linked immunospecific assay. FMD was measured at thebrachial artery using high resolution ultrasound. CD34+/VEGFR2+/133+ circulating EPC were enumerated by flow cytometry.Measurements were made immediately before darbepoetin/placeboand at 24 h, 72 h and 7 days. At 24 h FMD was repeated after 20 minischaemia-reperfusion of the upper limb. A further group of 11patients were studied according to the same protocol, all receivingdarbepoetin, with omission of forearm ischaemia-reperfusion at 24 h.Results Immunoreactive erythropoietin peaked at 24 h in thedarbepoetin group (median value of 724 U/l (IQR 576e733 U/l),compared to 12 U/l (IQR 9e21 U/l) in the placebo group) andremained elevated at approximately 500 fold baseline at 72 h. FMDdid not differ significantly between groups at 24 h (before ischaemia-reperfusion). At 72 h, (48 h after ischaemia-reperfusion) FMDincreased from baseline in the darbepoetin group but not in theplacebo group so that FMD (and change in FMD from baseline) wassignificantly greater in the darbepoetin group (change from baseline1.760.3% and �0.660.4% in darbepoetin and placebo groupsrespectively, p<0.001).The increase in FMD at 72 h after darbe-poetin and ischaemia-reperfusion at 24 h was significantly greaterthan that without preceding ischaemia-reperfusion (p<0.01). Aw20% increase in CD133+/VEGFR2+ cells after darbepoetin wastemporally dissociated with the increase in FMD.Conclusions Preceding ischaemia-reperfusion is required for darbe-poetin to enhance endothelial function, possibly by increasingexpression of the erythropoietin receptor and by a mechanism likelyto involve Akt/NO rather than circulating EPC.

Abstract 48 Figure 1 Change from baseline in FMD at 72 h, 48 h afterischaemia-reperfusion (+IR), after placebo and darbepoetin (study 1) andafter darbepoetin without preceding ischaemia-reperfusion (�IR, study 2).

Abstract 48 Table 1 Endothelial function and EPC

Placebo Placebo Placebo Placebo Darbepoetin Darbepoetin Darbepoetin DarbepoetinEndothelial function Baseline 24 h 72 h 7d Baseline 24 h 72 h 7d

FMD% 3.560.80 3.460.73 2.960.63 3.460.75 3.560.92 4.460.97 5.260.9*** 3.760.61

Progenitor cells

CD133+/VEGFR2+ 110617.6 117619 101619 123621 146613 180613* 180611* 182616*

CD34+/VEGFR2+ 4.160.6 3.960.5 3.160.6 4.760.8 8.763.07 11.163.9 7.161.4 13.364.5

CD133+/CD34+ 17.362.9 17.061.9 17.763.6 20.463.2 23.662.4 29.364* 31.664.6* 29.964.8*

Values are means6SE.* p<0.05, ***p<0.001


BCS Abstracts 2011

49 ETHNIC VARIATION IN QT INTERVAL AMONG HIGHLYTRAINED ATHLETES

doi:10.1136/heartjnl-2011-300198.49

1H Raju, 1M Papadakis, 2V Panoulas, 2J Rawlins, 2S Basavarajaiah, 1N Chandra,1E R Behr, 1S Sharma. 1St George’s University of London, London, UK; 2UniversityHospital Lewisham, London, UK

Background Studies in Caucasian (white) athletes indicate that asignificant proportion exhibit an isolated prolonged corrected QTinterval (QTc), raising concerns for potentially false diagnoses anddisqualification from competitive sport. The prevalence of prolongedQTc interval in athletes of African/Afro-Caribbean (black) descent isunknown. However, this ethnic group generally exhibits a highproportion of ECG repolarisation changes and increased leftventricular wall thickness, that may impact on QTc.Aim We aimed to assess the impact of ethnicity on QTc in youngelite athletes.Methods We assessed 3035 elite athletes, aged 14e35 years, whowere participating at national and international level in a variety ofsporting disciplines. Athletes were evaluated with ECG and 2Dechocardiography. Athletes diagnosed with structural heart diseaseor hypertension were excluded from analysis.Results Demographic and cardiological results are summarised inAbstract 49 table 1. Black male athletes exhibited shorter QTc thanwhite male athletes, but QTc was similar among black and whitefemale athletes. Bivariate analysis revealed that none of T waveinversions, ST segment elevation, or left ventricular wall thicknesswere associated with QTc. No ethnic difference was observed inprevalence of QT prolongation, as defined by ESC Sports Consensuscriteria (male >440 ms; female >460 ms).

Abstract 49 Table 1 Characteristics of athletes evaluated

Black Male(n[901)

White Male(n[1652)

Black Female(n[122)

White Female(n[360)

Mean Age, years 2265 1764 2165 1864

Mean Heart Rate, bpm 61612 56610 63610 5969

Mean QRS duration, ms 88614 96610 84610 8869

Mean LV wall thickness, mm 10.661.6 9.4*61.2 9.261.2 7.9*62.9

ST segment elevation, n (%) 570 (63.3%) 406 (24.6%) 20 (16.3%) 64 (17.8%)

T wave inversions, n (%) 204 (22.6%) 66* (4.0%) 18 (14.6%) 15* (4.2%)

Mean QTc (Bazett’s), ms 393626 404*620 407625 412627

QTc >440 ms, n (%) 20 (2.2%) 49 (3.0%) 13 (10.6%) 39 (10.9%)

QTc >460 ms, n (%) 4 (0.4%) 7 (0.4%) 1 (0.8%) 5 (1.4%)

Means presented as mean 6 SD.*p<0.001 white vs black athletes.

Conclusion Despite demonstrating a higher prevalence of repolar-isation changes and morphological left ventricular hypertrophy,black athletes do not exhibit a longer QTc than white counterparts.Based on ESC Sports Consensus criteria, prevalence of a long QTc inblack and white athletes is similar, obviating the need for ethnicityspecific criteria for defining a long QTc.

50 DIAGNOSTIC ROLE OF EXERCISE TOLERANCE TESTING INFAMILIAL PREMATURE SUDDEN CARDIAC DEATH

doi:10.1136/heartjnl-2011-300198.50

H Raju, M Papadakis, R Bastiaenen, A Zaidi, N Chandra, M Muggenthaler, N Spath,S Sharma, E R Behr. St George’s University of London, London, UK

Background Investigation of blood relatives for evidence of aninherited cardiac condition is advocated following an unexplainedsudden cardiac death (SCD).

Aim We determined the diagnostic yield of exercise tolerance testing(ETT) in investigation of inherited cardiac conditions followingfamilial premature SCD.Methods Between 2006 and 2010, we evaluated 308 blood relativesof 148 SCD victims, who completed at least 3 min of the Bruceprotocol. ETTs were analysed for: QT prolongation; Brugada type 1pattern; ST depression: blood pressure (BP) response; multipleventricular ectopics or arrhythmia. Individual pathological pheno-types were determined by a combination of 12-lead ECG, echo-cardiogram, 24-h holter monitor, with additional MRI, CTcoronaryangiography and genetic mutation analysis, as appropriate.Results Thirty (9.8%) patients had an abnormality during ETT,details of which are summarised in Abstract 50 figure 1. All ETTswith abnormal QT prolongation and dynamic Brugada pattern wereassociated with diagnoses of long QT syndrome and Brugadasyndrome respectively. An example of dynamic Brugada phenotypeis given in Abstract 50 figure 2. Ventricular ectopy was seen in 15patients, of whom 5 demonstrated phenotypic cardiomyopathy orchannelopathy on further investigations. No patients with signifi-cant ST depression had evidence of coronary abnormalities onimaging. No hypotensive BP response was seen, but exertionalhypertension was associated with systemic hypertension.

Abstract 50 Figure 1 ETT abnormalities and associated diagnoses atfamilial evaluation.

Abstract 50 Figure 2 Exercise tolerance test demonstrating dynamicBrugada ECG pattern. Stage 1 of Bruce protocol exercise (left) and post-exercise recovery (right).

Conclusion The ETT is a useful diagnostic adjunct when evaluatingrelatives of victims of premature SCD. Reliable diagnostic indicatorsinclude inappropriate QT prolongation and dynamic Brugada


BCS Abstracts 2011

pattern. Ventricular ectopy is non-specific, but is associated withboth cardiomyopathic and channelopathic processes in a significantminority. ST segment depression, however, is unhelpful and shouldbe viewed in the context of the patient’s cardiovascular risk profile.

51 LOW-DOSE SODIUM NITRITE RELIEVES MYOCARDIALISCHAEMIA IN PATIENTS WITH CORONARY ARTERYDISEASE: A TARGETED NO-DONOR EFFECT

doi:10.1136/heartjnl-2011-300198.51

1T E Ingram, 2R A Bleasdale, 2C Templeton, 2C Williams, 1A Margulescu, 1A G Fraser,1P E James. 1Cardiff University, Cardiff, UK; 2Royal Glamorgan Hospital, Llantrisant, UK

Introduction Sodium nitrite (NaNO2) became a popular means oftreating angina in the 19th century, as its stable chemical structureallowed for cheap preparation and easy storage. However, the effectswere slow and unpredictable and so it fell out of favour as morepotent and faster-acting agents became available, (eg, organicnitrates). Recent in vitro evidence shows that nitrite (NO2

-) exhibitsan enhanced vasodilator effect in hypoxia; an environmentalmodification which encourages its reduction to nitric oxide (NO).Therefore NaNO2 could potentially be an anti-ischaemic agent atmuch lower doses than those used historically, and be without theadverse side effects associated with organic nitrates (eg, systemichypotension and tachyphylaxis).Method A double-blind, placebo-controlled, cross-over study wasperformed in 10 subjects with proven myocardial ischaemia docu-mented by exercise tolerance testing and coronary angiography. Twodobutamine stress echocardiography (DSE) studies were performedon each subject: one with 0.9% saline and one with NaNO2,1.5 mmol/min for 20 min. This dose of NaNO2 has previously beenshown to be inert in normoxia but to vasodilate hypoxic tissue.Myocardial ischaemia was identified by the peak systolic velocity(PSV) response during DSE in a six basal-wall segment model of theleft ventricle. Using placebo study data-set, walls were classified intotertiles: the lowest tertile of responders of PSV to an increase inheart rate (DHR) labelled ischaemia (n¼18) and the upper tertilecontrol (n¼18). Data was divided into four groups according to thestudy-infusion received and the myocardial-wall examined: saline/ischaemia, NO2

-/ischaemia, saline/control and NO2-/control.

Results Data from each stage of each DSE was plotted on a scatterplot graph with change in (DHR) on the x-axis and correspondingchange in PSV (DPSV) on the y-axis (increase in both valuescompared to baseline), see Abstract 51 figures 1 and 2. Linearregression analysis of the saline/ischaemia group was lower than theNO2

-/ischaemia group, with no overlap in their 95% CI, see Abstract51 figure 1. In addition, the linear regression gradient of the NO2

-/ischaemia group was similar to the saline/control and the NO2

-/controlgradient, see Abstract 51 figure 2. The peak-dose dobutamine values

of DPSV/DHR were different in the saline/ischaemia group comparedto the three other groups (ie, saline/ischaemia ¼3.760.6 cm/s/s,NO2

-/ischaemia ¼8.261.0 cm/s/s, saline/control ¼10.561.1 cm/s/s,NO2

-/control ¼8.460.7 cm/s/s; p<0.01, repeated-measures ANOVAwith Bonferroni post-test). No difference was present between thethree other groups.


Conclusions Low-dose NaNO2 delivers a therapeutic effect toischaemic myocardial tissue in the absence of a vasodilator effect onnormoxic tissue. This is the first study in patients to demonstrate atargeted vasodilator effect of NO2

- to tissues in need only.

52 BRAIN NATRIURETIC PEPTIDE PREDICTS ALL CAUSEMORTALITY IN PATIENTS WITH TYPE 2 DIABETES ANDNORMAL EJECTION FRACTIONS

doi:10.1136/heartjnl-2011-300198.52

1B R Szwejkowski, 1D H J Elder, 2A Dawson, 1A D Struthers. 1University of Dundee,Dundee, UK; 2Department of Cardiology, NHS Tayside, Dundee, UK

Introduction The use of brain naturetic peptide (BNP) to predictoutcome in patients with normal ejection fractions (EF) and type 2diabetes (T2DM) is understudied. Only three previous studies havespecifically addressed the question as to whether BNP adds prognosticinformation in T2DM. There appears to be a link between survivaland BNP in T2DM, however these studies included small numbers ofpatients and did not fully exclude left ventricular systolic dysfunction(LVSD). We therefore studied the 5-year survival in a cohort of 500T2DM patients prospectively phenotyped with echocardiography.Methods 500 patients with T2DM where studied with echocardiog-raphy between April 2002 and October 2003. Patients were recruitedfrom the diabetes clinics at Ninewells Hospital, Dundee. Transthoracicechocardiography was performed by one trained operator and leftventricular (LV) assessment was performed using modified biplaneSimpson’s method over three cycles. We excluded individuals with EF of<55%. Follow-up data was linked via the Health Informatics Centre(HIC), to mortality data, laboratory test data, hospitalisation, andprescribing via the community health index (CHI) number. Coxproportional hazards model was used to examine the effects of BNP(bedside stick measurement) measure on all-cause mortality using age,sex, smoking status, hypertension, IHD, duration of diabetes, and diabeticdrug prescription as co-variants. Outcome was all cause mortality.Results In total we followed 316 patients over 8 years. 56 patientsdied over this time. After adjusting for confounding factors wehave shown that for every 10-unit increase in BNP there is a 6%increased risk of death. HR 1.06 (95% CI 1.02 to 1.10) (p¼<0.01).Conclusions In patients with normal EF, BNP is an independentpredictor of death in a cohort of T2DM patients. Although moreresearch is needed, BNP may become an important tool in riskstratifying T2DM patients in the future.Abstract 51 Figure 1


BCS Abstracts 2011

53 B-TYPE NATRIURETIC PEPTIDE PERFORMS BETTER THANCURRENT CARDIOVASCULAR RISK SCORES IN IDENTIFYINGSILENT “PANCARDIAC” TARGET ORGAN DAMAGE INALREADY TREATED PRIMARY PREVENTION PATIENTS

doi:10.1136/heartjnl-2011-300198.53

1A Nadir, 1S Rekhraj, 2J Davidson, 1T M MacDonald, 1C C Lang, 1A D Struthers.1University of Dundee, Dundee, UK; 2Ninewells Hospital, Dundee

Background Primary prevention needs to be improved because up to70%ofcardiovascular (CV) eventsoccuroutwith those classifiedashighrisk by CV risk scores currently used in clinical practice (eg,Framingham). One possible way to improve primary prevention of CVdisease is to identify those patients who may already harbour silentpancardiac target organ damage in the form of left ventricular hyper-trophy (LVH), systolic dysfunction (LVSD), diastolic dysfunction(LVDD), left atrial enlargement (LAE) or silent myocardial ischaemia.This could be achieved by reapplying traditional CV risk scores toprimarypreventionpatients after theyhavebeen treatedorby screeningwith a simple biomarker like B-type natriuretic peptide (BNP).Methods We prospectively recruited 300 asymptomatic individualswithout known cardiovascular disease already on primary preventiontherapy. Patients with valvular heart disease, atrial fibrillation andrenal impairmentwere excluded.Wemeasured BNPand calculated 10-year global CV risk scores (based on Framingham, QRISK andASSIGN) in each participant. Transthoracic echocardiography wasused to assess LV mass, LV systolic and diastolic function, and leftatrial volume while the presence of inducible ischaemia was assessedby dobutamine stress echocardiography or dipyridamole myocardialperfusion imaging. Patients were divided into low, intermediate andhigh risk groups based on 10-year global CV risk. The prevalence ofvarious cardiac TOD in each group was compared and ROC curveswere constructed for BNP and for 10-year global CV risk scores toassess their ability to detect presence of silent cardiac TOD.Results One hundred and two (34%) patients (Mean age6466.0 years, 58% males) had evidence of silent cardiac TOD(29.7% LVH, 18% LAE, 17.3% LVDD, 7.3% LVSD and 6.3%Ischaemia). The prevalence of cardiac TOD ranged from 19 to 28%in the low risk, 26%e33% in the intermediate risk and 36%e41% inthe high risk groups based on three commonly used CV risk equa-tions. BNP levels were significantly higher (median (IQR); 21.6(13.6e40.0) vs 11.4 (6.3e20.0) pg/ml, p<0.0001) in those withcardiac TOD compared to those without. The AUC for BNP toidentify any form of cardiac TOD was 0.77 (p<0.0001) overall and0.83 (p<0.0001) in males. However, discrimination power of CV riskscores was poor with area under curve of only 0.58 (p¼0.02) forQRISK, 0.62 (p¼0.001) for Framingham and 0.62 (p¼0.001) forASSIGN to detect presence of any form of TOD.Conclusion Silent cardiac TOD is highly prevalent (34%) in alreadytreated primary prevention population but current CV risk estima-tion alone performs poorly in the detection of these silent cardiacabnormalities. However, a raised BNP is able to identify existingsilent cardiac TOD of various subtypes particularly in males. UsingBNP to identify silent cardiac TOD could, in the future, become anew way to improve the primary prevention of CV events.

54 CAN MICROALBUMINURIA IDENTIFY SILENT“PANCARDIAC” TARGET ORGAN DAMAGE IN ANON-DIABETIC PRIMARY PREVENTION POPULATION?

doi:10.1136/heartjnl-2011-300198.54

1A Nadir, 1S Rekhraj, 2J Davidson, 1T M MacDonald, 1C C Lang, 1A D Struthers.1University of Dundee, Dundee, UK; 2Ninewells Hospital, Dundee, UK

Background Microalbuminuria is associated with increased cardio-vascular mortality and is a marker of generalised vascular dysfunc-

tion. We sought to investigate whether micoalbuminuria identifiedthe presence of silent “pancardiac” target organ damage (TOD) ie, leftventricular hypertrophy (LVH), systolic dysfunction (LVSD), diastolicdysfunction (LVDD), left atrial enlargement (LAE) or silent myocar-dial ischaemia in a non-diabetic primary prevention population.Methods Two hundred and sixty-three asymptomatic individualswithout diabetes or previous cardiovascular disease on primarypreventive therapy were prospectively recruited. Each participantunderwent a comprehensive echocardiographic examination for theassessment of LV mass, LV systolic and diastolic function, and leftatrial volume while the presence of inducible ischaemia was assessedby dobutamine stress echocardiography or dipyridamole myocardialperfusion imaging. A spot urine sample was analysed for micro-albuminuria and urinary creatinine by a trained laboratory techni-cian blinded to clinical or echocardiographic data. Microalbuminuriawas defined as urinary albumin to creatinine ratio of$2.5 mg/mmolin males and $3.5 mg/mmol in females.Results Out of 263 participants (Mean age 6466.3 years, 57%males) 89 (33.8%) had evidence of silent cardiac TOD (29.7% LVH,16.7% LAE, 17.1% LVDD, 6.8% LVSD and 6.1% Ischaemia). Theprevalence of cardiac TOD was significantly higher (53% vs 29%,p¼0.002) among those with microalbuminuria than those without.In multivariate analysis adjusted for age, gender, hypertension anddyslipidemia, presence of microalbuminuria was an independentpredictor of cardiac TOD with an adjusted HR of 2.54 (95% CI, 1.2to 4.4, p¼0.005). The utility of UACR in discriminating betweenthose with or without cardiac TOD was assessed by receiver oper-ating characteristic analysis but the area under curve was only0.6160.04, p¼0.003.Conclusion Microalbuminuria is an independent predictor of silent“pancardiac” target organ damage in a non-diabetic primary preven-tion population. Presence of microalbuminuria may help to identifythose primary prevention patients who are at a particularly higher risk.

55 GENE-GENE INTERACTIONS IN CORONARY ARTERYDISEASE

doi:10.1136/heartjnl-2011-300198.55

1M D Musameh, 2W Y S Wang, 3A J Balmforth, 4S G Ball, 5A S Hall, 1M Tomaszewski,1N J Samani. 1Department of Cardiovascular Sciences, University of Leicester, Glen-field Hospital; Leicester BRU, Leicester, UK; 2School of Medical Sciences and Institutefor Biomedical Research, University of Sydney, Sydney, Australia; 3Division of Cardi-ovascular and Diabetes Research, University of Leeds, Leeds, UK; 4LIGHT ResearchInstitute, Faculty of Medicine and Health, University of Leeds, Leeds, UK; 5Division ofCardiovascular and Neuronal Remodelling, University of Leeds, Leeds, UK

Background Only a small fraction of the heritability of coronaryartery disease (CAD) has been explained by common variantsidentified by genome-wide association studies. Among the stones tobe turned in the hunt for the missing heritability of CAD are gene-gene interactions. We investigated whether interactions betweencommon alleles in genes and pathways of known importance tocardiovascular regulation may contribute to the heritability of CAD.Methods 2101 CAD cases and 2426 controls of Caucasian originrecruited into Wellcome Trust Case Control Consortium weregenotyped using 50 K IBC gene-centric array containing 45 707 singlenucleotide polymorphisms (SNPs) of the highest biological relevanceto cardiovascular system. After applying appropriate quality controlfilters, 11 332 common (minor allele frequency >10%), independent(r2 linkage disequilibrium coefficient of #0.5) were included inpair-wise SNP-SNP interaction analysis using two complementarystatistical approaches: logistic regression (PLINK and INTERSNPsoftware packages) and Bayesian model (BEAM software).Results None of the analysed SNP-SNP interactions was statisti-cally significant after correction for multiple testing (p¼7.8310-10).The most significant interaction identified in this analysis was


BCS Abstracts 2011

between rs727139 (KCNH8) on chromosome 3 and rs11167496(PDGFRB) on chromosome 5 (p¼2.45310�8). Analysis of subsetsof SNPs pre-selected based on their nominal association with CAD(p<0.05) or molecular functionality (non-synonymous SNPs) didnot contribute more significant findings than investigation ofrandom set of SNPs.Conclusion Our analysis suggests that common SNP-SNP inter-actions are unlikely to account for a large proportion of the missingheritability of CAD.

56 CLINICAL AND FINANCIAL REPERCUSSIONS OF THEMARCH 2010 NATIONAL INSTITUTE FOR HEALTH ANDCLINICAL EXCELLENCE (NICE) GUIDELINE “CHEST PAIN OFRECENT ONSET” ON THE RAPID ACCESS CHEST PAINCLINIC (RACPC)

doi:10.1136/heartjnl-2011-300198.56

T Rogers, S Claridge, K Al Fakih. University Hospital Lewisham, London, UK

Background The RACPC is a well-established “one-stop” service,with goal to identify patients with stable chest pain due to coronaryartery disease (CAD) and quickly reassure those with non-cardiacpain. In March 2010, NICE published a new guideline, whichadvocates assessing likelihood of CAD based on age, gender, historyand risk factors (RF). If estimated likelihood is >60%, invasivecoronary angiography (ICA) is recommended as the first-line diag-nostic investigation. If estimated likelihood is 30%e60%, functionalimaging is recommended. If estimated likelihood is <30%, CTcalcium scoring CTcoronary angiography (CTCA) is recommended.Significantly, the guideline discourages the use of ETT to diagnose orexclude stable angina in patients without known CAD.Methods 167 consecutive patients referred to RACPC betweenOctober 2009 and March 2010 were retrospectively assessed forlikelihood of CAD according to the new NICE guideline. Choice ofinvestigations and eventual outcome (confirmed CAD vs noevidence of CAD) were compared between subgroups defined byestimated likelihood of CAD. An economic analysis of cost ofinvestigation per patient was undertaken using current Payment byResults national tariffs.Results Our patient population had a high prevalence of RF with38.1% having a total of three or more RF. Consequently 23.2% ofpatients had an estimated likelihood of CAD of <30%. 27.4% had anestimated likelihood of 30%e60%. 49.4% had an estimated like-lihood of >60%. 7.2% of patients were lost to follow-up. 14.4% ofpatients were ultimately confirmed to have CAD on ICA, whichcorrelated with pre-test estimated likelihood. 6% of patients withlikelihood <30%, 8.7% of those with likelihood 30%e60% and23.2% of those with likelihood >60% were confirmed to have CAD.Average cost of investigation per patient was £528. A negative ETTresulted in average cost per patient of £347. An inconclusive ETTresulted in higher cost (£728) as did inability to exercise (£435) dueto the need for further investigations. A positive ETT resulted inaverage cost of £1174 due to the high cost of ICA. Were the NICEguideline strictly applied to our patient population, average cost perpatient would have been substantially higher at £838 (£362 perpatient if likelihood <30%, £566 if likelihood 30%e60% and £1218 iflikelihood >60%). Overall this corresponds to a 60% increase in cost.Conclusion The 2010 NICE guideline appears to significantly over-estimate the true risk in our patient population. Were the guidelinestrictly applied, almost half of our patients would proceed to ICA asa first-line investigation, but many of them would be found to haveunobstructed coronary arteries. As ICA is an expensive inves-tigation, this would inevitably result in a significant increase inaverage cost per patient. Relatively few patients would be eligiblefor CTCA, which is an excellent non-invasive “rule-out” test forCAD and relatively inexpensive compared with other investigations.


Abstract 56 Table 1

All Men Women

Numbers (%) 167 86 (51.5) 51 (48.5)

Age (mean 6 SD) 56.0612.6 55.2613.3 57.0611.8

Diabetes (%) 29 (17.4) 15 (17.4) 14 (17.3)

Hypertension (%) 72 (43.1) 40 (46.5) 32 (39.5)

Hypercholesterolaemia (%) 100 (59.9) 53 (61.6) 47 (58.0)

Family history (%) 58 (34.7) 28 (32.6) 30 (37.0)

Smoking (%) 87 (52.1) 54 (62.8) 33 (40.7)

Systolic BP (mean 6 SD) /Diastolic BP (mean 6 SD)

134621 /79611

131618 /80611

137623 /78610

Fasting glucose (mean 6 SD) 5.261.0 5.761.0 5.361.1

Total cholesterol (mean 6 SD) 5.1761.04 5.0661.06 5.2961.01

LDL (mean 6 SD) / HDL(mean 6 SD)

3.1960.90 /1.3060.43

3.1460.90 /1.2360.46

3.3360.86 /1.3860.38

BMI (mean 6 SD) 30.166.0 29.464.9 30.966.9

Abstract 56 Table 2

Average cost perpatient prior to NICEguideline implementation

Average cost perpatient were NICEguideline strictlyimplemented

Predicted likelihood <30% £324 £362

Predicted likelihood 30%e60% £467 £566

Predicted likelihood >60% £661 £1218

OVERALL AVERAGE COST PERPATIENT

£528 £838

57 THE IMPACT OF PREOPERATIVE RENAL DYSFUNCTION ANDTHERAPY TYPE IN PATIENTS WITH TYPE 2 DIABETESUNDERGOING CORONARY ARTERY BYPASS SURGERY

doi:10.1136/heartjnl-2011-300198.57

A Menon, J Hodson, D Pagano, J Mascaro, I C Wilson, S J Rooney, T R Graham,R S Bonser. University Hospital Birmingham, Birmingham, UK

Introduction There is limited data addressing the impact of preoper-ative renal dysfunction in type 2 diabetics (T2DM) undergoing first-time coronary artery bypass surgery (CABG); specifically exploringthe influence of diabetic management (oral hypoglycaemic (OH) andinsulin therapy (IN)). We assessed the impact of preoperative renalstatus and diabetic management on the post operative renal status,morbidity, 30-day and long-term survival in T2DM-CABG.Methods We reviewed prospectively accrued data from 1/1/1999 to31/12/2009. Pre and 4 to 5-day postoperative creatinine clearance


BCS Abstracts 2011

(CrCl) was calculated using Cockcroft-Gault formula. Patients weresubgrouped into 5 grades based on preoperative CrCl; Group ICrCl$90 ml/min; II 60e89; III 30e59; IV 15e29; V <15 orhaemodialysis. Late KaplaneMeier survival data (compared by logrank method), censored at 1/10/2009 were obtained from the UKCCAD. Surgical morbidity outcomes included re-exploration forbleeding, stroke (type 1 deficit) and low cardiac output state (LCOS)requiring inotropes 6 intra-aortic balloon counterpulsation werecompared using Fisher ’s Exact tests.Results 1215 patients (921 males) with a mean age of 64 years(31e89 years) underwent CABG; 742 on OH and 472 on IN. Preop-erative renal status in the groups were Group I -209(17%), II-584(48%), III-387(32%), IV-26(2%) and V (8(1%). Similar percentages ineach group had $1 grade deterioration of renal function post-operatively 19%, 18%, 16% and 23% (grades IeIV respectively;p¼0.470). When examined as a continuous variable, higher preoper-ative CrCl correlated with a better postoperative improvement inCrCl (r¼0.073, p¼0.012 Spearman Rank). Overall 30-day mortalitywas 3.33% (CI 2.32 to 4.34%) and was not different by group I-3.37%(CI 0.92 to 5.82), II-2.09% (CI 0.92 to 3.26%), III 4.92% (CI 2.76 to7.08%), IV 8% (CI 0 to 18.6%) and Stage V 0% (CI 0 to 0.4%; p¼0.101)or by therapy type; (p¼0.411). IN patients had similar preoperativerenal function (median CrCl 66.8 vs 68.6; p¼0.828) but a higher rateof postoperative renal deterioration (53.3 vs 46.7%, p<0.001). Stroke(p¼1.000), bleeding (p¼0.755) and LCOS (p¼0.335) incidence werenot different between therapy type. Overall mean survival was9 years (CI 8.7 to 9.2 years) and was not different by renal functiongrade (p¼0.612). However, IN patients had shorter mean survival 8.7(8.3 to 9.0) vs OH 9.1(8.8 to 9.4) years; p¼0.03.Conclusions InT2DM-CABG,36%ofpatientshaveCrCl<60 ml/min.Higher CrCl protects against postoperative renal deterioration. Renaldysfunction does not appear to affect hospital outcome or survival.However, preoperative IN requirement increases the risk of renaldysfunction and is associated with worse longer-term survival.

58 TEMPORAL EVALUATION OF REFERRAL FOR AND LONG-TERM SURVIVAL FROM CARDIAC REHABILITATION FORACUTE MYOCARDIAL INFARCTION

doi:10.1136/heartjnl-2011-300198.58

1C L Lewinter, 1M B Bland, 2P D Doherty, 1B L Lewin, 3A S H Hall, 4C P G Gale.1University of York, York, UK; 2York St John University, York, UK; 3Yorkshire HeartCentre, Leeds, UK; 4University of Leeds, Leeds, UK

Background Cardiac rehabilitation (CR) is a cost-effective, evidence-based approach to managing heart disease. Rates of uptake have andcontinue to vary despite recommendations from the NSF for CHDand NICE. The Evaluation of the Management and Methods ofAcute Coronary Events (EMMACE) 1 and 2 studies are 2 largeprospective multi-centre registries of care of acute coronarysyndromes (ACS) in Yorkshire undertaken in 1995 and 2003 inrespectively. We studied the temporal changes in referral for andlong-term survival from CR in patients who were admitted tohospital with an acute myocardial infarction (AMI).Methods Baseline characteristics were described as numbers (%) or asmeans with IQRs. For Continuous variables, the Kruskal Wallis testwas used for comparisons. Discrete variables were assessed by the c2

test. Unadjusted relative risk ratios (RRR) were calculated to assessmortality after referral for CR. KaplaneMeier (KM) curves comparedunadjusted survival stratified byCR referral and EMMACE study. Logrank tests compared the survival estimates. Sex, age, STEMI, heartfailure, diabetes, COPD and mini-GRACE score, revascularisation,reperfusion, ACE-inhibitors, a-blockers, statins, anti-platelet agentsand admitting cardiologist were regressed (backward logistic, p<0.10and goodness of fit with a group of 10) on CR referral and representedas 95% CI OR. A Cox proportional model (Model 1: mini-GRACEscore, Model 2: sex, age, STEMI, heart failure, diabetes, COPD,

EMMACE risk score, revascularisation, reperfusion, ACE-inhibitor, a-blocker, statins, anti-platelet agent, admitting cardiologist) was usedto compare the temporal long-term survival estimates (all causemortality) by CR referral.Results 4341 had AMI. CR referral was 44% in 1995 and 59 % in 2003(p<0.001). CR referral was associated with reduced mortality in 2003(RRR, 95%CI: 0.54; 0.50 to 0.60), but was not in 1995 (1.02; 0.96 to1.09). Unadjusted survival for patients not referral for CR in 1995wassimilar to that for patients referred for CR in 1995; (Abstract 58figure.1). For those referred for CR, the mean mini-GRACE score forCR referrals was lower in 2003 than 1995; 0.53 and 0.72, p<0.001.After adjustment using the min-GRACE score (Model 1), the impact(HR, 95% CI) of CR referral was 0.63, 0.55 to 0.73 in 2003 and 1.07,0.92 to 1.3 in 1995. After adjustment using Model 2, the impact(HR, 95% CI) of CR referral was 0.57, 0.48 to 0.66 in 2003 and 1.31,1.04 to 1.60 in 1995.Conclusion Between 1995 and 2003, referral for CR increased andbecame a significantly important factor contributing to reducedmortality rates post-AMI. This is despite the differences in patientand treatment factors between the 2 studies periods. Even so, rate ofreferral for CR remain sub-optimal.

Abstract 58 Figure 1 KaplaneMeier survival estimates.

59 SHORT TERM ELEVATION OF CHOLESTEROL LEVEL INNEONATAL LIFE AND LONG TERM CHANGES IN AORTICSTIFFNESS: INSIGHTS FROM USE OF INTRAVENOUS LIPIDS

doi:10.1136/heartjnl-2011-300198.59

1A J Lewandowski, 1M Lazdam, 1E Davis, 1R Poole, 1J Diesch, 1J Francis,1S Neubauer, 2A Lucas, 2A Singhal, 1B Kelly, 1P Leeson. 1Cardiovascular Medicine,University of Oxford, Oxford, UK; 2Institute of Child Health, University College London,London, UK

Introduction Offspring born to hypercholesterolaemic mothers haveincreased fatty streak formation in the fetal aorta, which persists intoadolescence. To understand whether exposure to elevated cholesterolin early life, independent of a maternal history of hyper-cholesterolaemia, also has a long-term impact on the cardiovascularsystem we studied the vascular phenotype of adults in whom choles-terol levels were artificially elevated for a short period postnatally.Methods We prospectively followed-up 102 subjects born prematurenow aged 23 to 28 years. Individuals exposed to maternal hyper-cholesterolaemia were excluded. 18 received intravenous (IV) lipidsduring the first nine weeks of life and were matched 2:1 for preg-nancy and early life complications, age, sex, birthweight and gesta-tional age with controls that did not receive IV lipids. Aortic pulsewave velocity (aPWV), regional aortic distensibility, left ventricularmass and ejection fraction were determined by cardiovascular


BCS Abstracts 2011

magnetic resonance. Detailed lifestyle information and anthro-pometric measurements were collected during childhood andadolescence. Metabolic parameters were measured multiple timesper week for the first 9 weeks of life and again at follow-up visits.Results Individuals that received IV lipids achieved significantlyhigher maximum cholesterol levels during the first 9 weeks of lifethan those that did not (mean6SD¼4.3861.65 vs 3.1260.78 mmol/l, p¼0.006). Dose given and number of days on IV lipids also asso-ciated with maximum cholesterol level during this period (r¼0.557,p<0.001 and r¼0.567, p<0.001, respectively). There was a gradedrelation between the maximum elevation in circulating cholesterolpostnatally and aortic stiffness (aPWV) in young adulthood(r¼0.596, p<0.001). The greatest increase in stiffness was seen inthe abdominal aorta, where distensibility was significantly reducedin the group that received IV lipids (mean6SD¼9.7464.27 vs12.9164.11/mm Hg3103, p¼0.012). There were no differencesbetween the groups in other vascular or left ventricular measures. Ina stepwise regression model, maximum cholesterol level achievedin the first few weeks of life was an independent predictor of aPWVin young adulthood (b¼0.596, p<0.001) and accounted for 30.9% ofthe variance in hierarchical multiple regression (b¼0.584, p<0.001).Conclusions Brief artificial elevation of cholesterol level in immediatepostnatal life is associated with long term changes in aortic functionindependent of later cholesterol levels. The association is gradeddepending on the degree of elevation of circulating cholesterol. Highcholesterol exposure during sensitive periods of early postnatal lifemay have long term impacts on the cardiovascular system.

60 ETHNIC DIFFERENCES IN REPOLARISATION PATTERNS ANDLEFT VENTRICULAR REMODELLING IN HIGHLY TRAINEDMALE ADOLESCENT (14e18 YEARS) ATHLETES

doi:10.1136/heartjnl-2011-300198.60

1N Sheikh, 1M Papadakis, 2F Carre, 2G Kervio, 1J Rawlins, 3V Panoulas, 1N Chandra,1H Raju, 1R Bastiaenen, 1E Behr, 1S Sharma. 1St. George’s University of London,London, UK; 2French Institute of Health and Medical Research, University of Rennes,Rennes, France; 3University Hospital Lewisham, London, UK

Purpose Studies in adult, black athletes (BA) demonstrate a highprevalence of ECG repolarisation changes and echocardiographic leftventricular hypertrophy (LVH) that may overlap with hypertrophiccardiomyopathy (HCM). The prevalence of ECG repolarisationchanges and echocardiographic LVH in adolescent BA, the groupmostvulnerable to exercise-related sudden death from HCM, is unknown.Methods This study evaluated 219 male adolescent BA (14e18 years,inclusive) with 12-lead ECG and 2-D echocardiography. Results werecompared with 1440 male adolescent WA. Athletes with T waveinversions and morphological LVH were invited for further inves-tigation with exercise stress test, 24 h Holter and CMR.Results STsegment elevation was common in both groups but morefrequent in BA (63.5% vs 14.9%, p<0.001), while ST segmentdepression was exceedingly rare. Both Twave inversions (21.5% vs2.9%, p<0.001) and deep Twave inversions (11% vs 0.3%, p<0.001)were commoner in BA. Black athletes demonstrated greater leftventricular wall thickness (10.461.6 vs 9.461.2 mm, p<0.001)compared to WA. Twenty-three (10.5%) BA exhibited a leftventricular wall thickness >12 mm vs only 6 (0.4%) WA (p<0.001).None of the athletes exhibited the broader phenotype of HCM onfurther investigation. In multivariable analysis black ethnicity wasthe strongest independent predictor for the presence of T waveinversions (OR 3.56, 95% CI 1.56 to 8.13, p¼0.003) and LVH (OR3.17, 95% CI 1.77 to 5.71, p<0.001).Conclusions As with adult athletes, Twave inversions and LVH weremore prevalent in adolescent BA compared to WA. These findingshave important implications in the pre-participation screening era,particularly in countries with a high proportion of BA competing at

elite level, since extrapolation of ECG and echocardiographic criteria,solely derived from Caucasian cohorts, would result in 25.6% of BArequiring further investigations for cardiac pathology.

Abstract 60 Figure 1 Bar chart depicting the distribution of leftventricular wall thickness in black and white adolescent athletes.

61 FIVE-MIN HEART RATE VARIABILITY CAN PREDICTOBSTRUCTIVE ANGIOGRAPHIC CORONARY DISEASE

doi:10.1136/heartjnl-2011-300198.61

1D Kotecha, 2G New, 1M Flather, 3D Eccleston, 3H Krum. 1Royal Brompton Hospital,London, UK; 2Box Hill Hospital, Melbourne, UK; 3Monash University, Melbourne, UK

Background Obstructive coronary artery disease (CAD) is evident inonly half of patients referred for diagnostic coronary angiography.Five-minute heart rate variability (HRV) is a marker for autonomiccontrol of the vasculature, which we hypothesised could be used torisk-stratify cardiac patients (the Alternative Risk Markers inCoronary Artery Disease (ARM-CAD) study.Methods Resting HRV prior to elective coronary angiography wasanalysed in 470 participants with predominantly normal cardiacrhythm, regardless of comorbidity. The presence of obstructive CAD($50% stenosis) was regressed in a multivariate model including riskfactors, ECG variables and medications.Results Mean age was 65 years (SD 11), 67% were male, 21% haddiabetes, mean blood pressure was 144/79 mm Hg (SD 21/10) and16% had impaired left-ventricular function. Patients with obstruc-tive CAD had significantly reduced HRV, particularly in the lowfrequency (LF) range (median 180 vs 267 ms2 without CAD;p<0.001). There was a linear trend according to the severity ofCAD; the median LF power (IQR) in patients with normal coro-naries was 275 (612), with minor coronary irregularities 255 (400),



BCS Abstracts 2011

single-vessel CAD 212 (396) and more severe disease 170 (327) ms2;p-value for trend¼0.003. There was a similar reduction in LF powerregardless of the anatomical location of coronary stenoses (seeAbstract 61 figure 1). Comparing patients with LF<250 and$250 ms2, the OR for obstructive CAD was 2.42, 95% CI 1.33 to4.38 (p¼0.004) after adjusting for risk factors, medications and heartrate. No interactions were noted in sub-group analysis of age,gender, diabetes, prior cardiovascular disease or high-sensitivity CRP.In addition, HRV added to risk prediction irrespective of baselineFramingham risk (p<0.0001).Conclusion Low HRV is strongly predictive of angiographic coronarydisease regardless of other comorbidities and is clinically useful as arisk predictor in patients with sinus rhythm.

62 WILL THE NEW EUROPEAN AF GUIDELINES LEAD TO MOREPATIENTS RECEIVING ORAL ANTICOAGULATION THERAPY?

doi:10.1136/heartjnl-2011-300198.62

T Kaier, R Williams, S Khan. West Middlesex University Hospital, London, UK

Atrial fibrillation (AF) confers a 5-fold risk of stroke and the risk ofdeath from AF-related stroke is doubled; it is the most commonsustained cardiac arrhythmia, occurring in 1%e2% of the generalpopulation. Meta-analysis of anti-platelet therapy demonstrate anon-significant 19% reduction in the incidence of stroke, provingoral anticoagulation (OAC; such as warfarin) to be far superior(64% relative risk reduction) in stroke prevention. Recent guidelineshave been published by the European Society of Cardiology (ESC)which focus on the most effective antithrombotic therapy in AFand propose a new risk scoring system, the CHA2DS2-VASc score.In our institution the prescription and documentation of antith-rombotic therapy in AF has been the focus of a previous audit thatdemonstrated poor compliance with the guidelines and docu-mentation of decision making. The focus of this audit was twofold:first to determine whether compliance with the guidelines anddocumentation had improved and second determine the effect ofthe new risk scoring system on prescription of OAC. A random10% of cases of patients discharged with a coding diagnosis of AFwere selected (125 cases). They were risk assessed using the NICE2006 stroke risk stratification, CHADS2 and CHA2DS2-VASc score.In all cases the agent used for thromboprophylaxis was reviewed asto whether NICE recommendations had been met. The scoringsystems were compared to identify patients in whom the ESCguidelines would change treatment―ie, OAC instead of Aspirin oras first line antithrombotic therapy. Of the 125 selected case notes114 arrived in time for analysis; out of these 8 were excluded due toerroneous coding as AF. 106 patients were risk stratified, of whom68.22% (73) were high risk, 28.04% (30) moderate and 2.80% (3)low risk according to NICE guidelines. 74.77% (80) scored 2 ormore points on the CHADS2 risk assessment―this numberincreased to 93.46% (100) if CHA2DS2-VASc was applied, forwhom OAC would be the recommended antithrombotic therapy(see Abstract 62 figure 1). Only 57.50% or 61 patients were on theappropriate choice of thromboprophylaxis if the NICE guidelinewas used as risk assessment. 73.77% (45) of patients had no formaldocumentation why NICE guideline had not been followed; thismostly comprised patients who were on Aspirin but, correctly riskassessed, were candidates for OAC. We extrapolate our findings tosuggest that in an average sized DGH, 200 patients more per yearwould be considered high thromboembolic risk and henceappropriate for OAC. This audit shows that the rate of appropriateanti-coagulation among patients with AF is still low and couldbe improved further. The new ESC guidelines add to thischallenge as significantly more patients will be considered forOAC therapy.

Abstract 62 Figure 1 Comparison CHADS2 & CHA2DS2-VASc scores.

63 IMPLICATIONS OF A LIKELIHOOD BASED APPROACH TODIAGNOSTIC TESTING IN CORONARY ARTERY DISEASE:IMPACT OF THE NEW NICE GUIDELINES

doi:10.1136/heartjnl-2011-300198.63

I U Haq, J S Skinner, P C Adams. Royal Victoria Infirmary, Newcastle upon Tyne, UK

Background The NICE clinical guideline for chest pain of recentonset (NCG 95) published in March 2010 recommends diagnosingangina based on clinical assessment and likelihood of coronaryartery disease (CAD). If the estimated likelihood of CAD is 61%e90%, coronary angiography should be offered. If 30%e60%, func-tional imaging should be offered, and if 10%e29%, CT calciumscoring should be offered. We determined the number and types ofdifferent investigations to diagnose coronary artery disease inpatients referred with suspected cardiac pain before the publicationof NCG 95 and compared this with the predicted investigationsafter the application of the guidelines.Methods Data was collected prospectively in a bespoke databasefor patients referred to the Rapid Access Chest Pain Clinic,Newcastle upon Tyne, UK. Patients with chest pain of suspectedcardiac origin were referred from primary care between February2002 and March 2010. Patients with previous MI, PCI or CABGwere excluded. The analysis comprised 5598 men and women withno past history of coronary disease. Likelihood of CAD wascalculated by the Pryor equation using the variables age, sex, typeof chest pain (typical or atypical), ECG Q waves, smoking,hyperlipidaemia, diabetes and ST/T changes on ECG. The mainoutcome measures were actual and predicted future frequency ofexercise tests, CT coronary angiograms, functional imaging testsand invasive coronary angiograms by pretest likelihood of coronaryartery disease.Results The proportion of the study population before and after theguidelines undergoing exercise testing was 50.1% vs 0.0%; forcalcium score/CT coronary angiography 0.0% vs 14.7%; for func-tional imaging 25.6% vs 13.4%; and for invasive coronary angiog-raphy 15.3% vs 25.8%. The proportion not requiring further testingwas unchanged (30.0% vs 31.0%).Conclusions Application of NICE CG95 will change the inves-tigation of patients with chest pain substantially. A significantreallocation of resources will be required. Exercise testing will bereplaced by anatomic or functional imaging. CT coronary angiog-raphy will play an important role and replace functional imaging insome patients. Invasive angiography will take on a more importantrole in the diagnosis of coronary artery disease. It will, however,empower us to reassure almost a third of referrals that they do nothave angina on clinical assessment alone.


BCS Abstracts 2011

64 DIAGNOSTIC ACCURACY OF EXERCISE STRESS TESTING ININDIVIDUALS WITHOUT KNOWN CORONARY ARTERYDISEASE: A SYSTEMATIC REVIEW AND META-ANALYSIS

doi:10.1136/heartjnl-2011-300198.64

1A Banerjee, 2D Newman, 2A Van den Bruel, 2C Heneghan. 1West Midlands Deanery,Birmingham, UK; 2Department of Primary Care, University of Oxford, Oxford, UK

Background Exercise stress testing offers a non-invasive, lessexpensive way of risk stratification prior to coronary angiography,and a negative stress test may actually avoid angiography.However, previous meta-analyses have not included all exercisetest modalities, or patients without known coronary arterydisease.Objectives To systematically review the literature to determine thediagnostic accuracy of exercise stress testing for coronary arterydisease on angiography.Search methods MEDLINE (January 1966eNovember 2009) andEMBASE (1980e2009) databases were searched for articles ondiagnostic accuracy of exercise stress testing.Selection criteria We included prospective studies comparing exer-cise stress testing with a reference standard of coronary angiographyin patients without known coronary artery disease. Results From6055 records, we included 34 studies with 3352 participants.Overall, we found published studies regarding five different exercisetesting modalities: treadmill ECG, treadmill echo, bicycle ECG,bicycle echo and myocardial perfusion imaging. The prevalence ofCAD ranged from 12% to 83%. Positive and negative likelihoodratios of stress testing increased in low prevalence settings. Tread-mill echo testing (LR+ ¼7.94) performed better than treadmill ECGtesting (LR+ 3.57) for ruling in CAD and ruling out CAD (echo LR�¼0.19 vs ECG LR� ¼0.38). Bicycle echo testing (LR+ ¼11.34)performed better than treadmill echo testing (LR+ ¼7.94), whichoutperformed both treadmill ECG and bicycle ECG. A positiveexercise test is more helpful in younger patients (LR+ ¼4.74) thanin older patients (LR+ ¼2.8).Conclusions The diagnostic accuracy of exercise testing varies,depending upon the age, sex and clinical characteristics of thepatient, prevalence of CAD, and modality of test used. Exercisetesting, whether by echocardiography or ECG, is more usefulat excluding CAD than confirming it. Clinicians have concen-trated on individualising the treatment of CAD, but there is greatscope for individualising the diagnosis of CAD using exercisetesting.


Abstract 64 Figure 2 Probability of coronary artery disease.


65 OUTCOMES AFTER CARDIAC SURGERY: ARE WOMEN OFSOUTH ASIAN ORIGIN AT INCREASED RISK?

doi:10.1136/heartjnl-2011-300198.65

1D A George, 1D Morrice, 2A M Nevill, 1M Bhabra. 1New Cross Hospital,Wolverhampton, UK; 2University of Wolverhampton, Wolverhampton, UK

Objectives The population served by our centre has a relatively highproportion of people originating from the Indian subcontinent(“South Asians”) compared to the national average (14.3% vs 4.6%).We observed that the mortality rate in South Asian women under-going cardiac surgery in our unit appeared to be relatively high. Weinvestigated this observation further to determine whether ethnicorigin was an independent risk factor for postoperative death infemales.Methods Data for all patients undergoing cardiac surgery werecollected prospectively in a registry. Retrospective analysis wascarried using SPSS on data for 4901 patients operated on in the 6-year period April 2004 to March 2010. Categorical data associatedwith mortality were analysed using c2 tests. Risk factors for in-hospital mortality were subjected to univariate analysis, and thosefound to be significant were tested for independence using multi-variate logistic regression.Results During the study period, 1160 female patients underwentsurgery with a mortality rate of 4.7%. Mortality in 113 South Asianswas 8.9% vs 4.3% in non-Asians (p¼0.03). Of 20 risk factors testedwith univariate analysis, 16 were significantly associated withmortality. Logistic regression showed the following to be inde-pendent predictors of postoperative mortality: urgency of operation(OR 32.0; p<0.001), older age (OR 24.2; p<0.001), preoperativerenal dysfunction (OR 15.8; p<0.001), diabetes (OR 7.8; p¼0.005),South Asian ethnicity (OR 7.3; p¼0.007), extra-cardiac arteriopathy(OR 4.8; p¼0.028), and an operation other than isolated CABG (OR5.8; p¼0.016).Conclusions In our population, South Asian ethnicity appears to bean independent risk factor for mortality in females undergoingcardiac surgery. Studies in larger populations are warranted.


BCS Abstracts 2011

66 ETHNIC DIFFERENCES IN CAROTID INTIMAL MEDIALTHICKNESS AND CAROTID-FEMORAL PULSE WAVEVELOCITY ARE PRESENT IN UK CHILDREN

doi:10.1136/heartjnl-2011-300198.66

1P H Whincup, 1C M Nightingale, 2A Rapala, 2D Joysurry, 2M Prescott, 2A E Donald,2E Ellins, 1A Donin, 2S Masi, 1C G Owen, 1A R Rudnicka, 1D G Cook, 2J E Deanfield.1Division of Population Health Sciences, St George’s, University of London, London, UK;2Vascular Physiology Unit, Institute of Child Health, UCL, London, UK

Introduction There are marked ethnic differences in cardiovasculardisease risks in UK adults; South Asians have high risks of coronaryheart disease and stroke while black African-Caribbeans have highrisks of stroke and slightly low risks of coronary heart disease whencompared with white Europeans. Ethnic differences in cardiovas-cular risk factors are apparent in childhood, but little is known abutethnic differences in vascular structure and function during child-hood. We set out to measure two vascular markers of cardiovascularrisk, common carotid intimal-medial thickness (cIMT) and carotid-femoral pulse wave velocity (PWV) in UK children from differentethnic groups.Methods We conducted a school-based study examining the cardi-ovascular risk profiles of 9e10 year-old UK children, includingsimilar numbers of South Asian, black African-Caribbean and whiteEuropean participants. Following a baseline cardiovascular risksurvey with measurements of body build, blood pressure, fastingblood lipids, insulin and HbA1c, 1400 children were invited to havemeasurements of cIMT (bilateral measurements were made with aZonare ultrasound scanner). A subgroup of these children (n¼900)was also invited for PWV measurements, made with a Vicorderdevice. All analyses were adjusted for age, gender and allowed forclustering at school level.Results In all, 939 children (67% response) had measurements ofcIMT and 631 children (70% response) had measurements of PWV.Mean cIMTwas 0.475 mm (SD 0.035 mm); mean PWV was 5.2 m/s(SD 0.7 m/s). Compared with white European children, blackAfrican-Caribbeans had higher cIMT (mean difference 0.014 mm,95%CI 0.008 to 0.021 mm) and PWV (% difference 3.3, 95%CI 0.4 to6.2); South Asian children had similar cIMT to white Europeans butslightly higher PWV (% difference 2.7, 95%CI �0.1 to 5.5%). cIMTwas positively associated with systolic and diastolic blood pressurebut not with other cardiovascular risk markers. In contrast, PWVwas positively associated with adiposity, diastolic blood pressureand insulin resistance. Black African-Caribbean children had lowerLDL-cholesterol levels and higher insulin and HbA1c levels thanwhite Europeans; South Asian children had higher insulin, HbA1cand triglyceride levels. However, adjustment for these risk factorshad little effect on the ethnic differences in cIMTand PWVobserved.Conclusions Ethnic differences in cIMT and PWV, markers of long-term cardiovascular risk, are apparent in childhood. These differencesare not fully explained by the ethnic differences in established cardi-ovascular riskmarkers observed. The results suggest that theremay beimportant opportunities for prevention of cardiovascular diseasebefore adult life, particularly in high-risk ethnic minority groups.

67 SPONTANEOUS CARDIAC HYPERTROPHY AND ADVERSE LVREMODELLING IN A NOVEL HUMAN RELEVANT MOUSEMODEL OF DIABETES; A MECHANISTIC INSIGHT

doi:10.1136/heartjnl-2011-300198.67

1S M Gibbons, 1Z Hegab, 1M Zi, 1S Prehar, 1T M A Mohammed, 2R D Cox,1E J Cartwright, 1L Neyses, 1M A Mamas. 1University of Manchester, Manchester, UK;2MRC mammalian genetics unit, oxford, UK

Heart failure (HF) is one of the commonest cardiovascular compli-cations of Diabetes Mellitus (DM) with the prevalence of DM

reported at around 30% in many pivotal heart failure studies. DM isan independent predictor of mortality in patients with HF, howevermolecular mechanisms that contribute to HF development in thediabetic population are poorly understood. Using a novel humanrelevant mouse model of DM (GENA348), identified through theMRC mouse mutagenesis programme with a point mutation in thepancreatic glucokinase (GLK) gene we investigate the molecularmechanisms that contribute to the HF phenotype in DM. GLK isthe glucose sensor which regulates insulin secretion and GLKactivity is reduced by 90% by the GENA348 point mutationresulting in severe hyperglycaemia. Similar mutations underlieMaturity Onset Diabetes of the Young Type 2 (MODY 2) inhumans. Mean random blood glucose was found to be increased inthe GENA348 mutant (HO) mice compared to wild type (WT)littermates (WT 6.960.3 mmol/l vs HO 20.660.8 mmol/l,p<0.001). Serial echocardiography was performed, at 3, 6 and12 months. No significant changes in echocardiographic parameterswere observed at 3 months, although by 6 months development ofsignificant cardiac hypertrophy in HO mice was observed. At12 months of age left ventricular dilatation was evident, charac-terised by an 8% increase in diastolic diameter (WT 4.0860.10 vsHO 4.4160.12, p<0.05). Systolic function was preserved althoughsignificant diastolic dysfunction was evident at 6 and 12 monthswith a 31% reduction in the E:A ratio. Histological staining illus-trated significant cellular hypertrophy with real time PCR datademonstrating a relative 150% increase in the hypertrophic markerBNP. Hypertrophic pathways were examined through western blotanalysis revealing an age dependant increase in Akt phosphorylation(3 months- no increase, 6 months-140%, 12 months-460%). Serumlevels of advanced glycation end products (AGE) were also elevatedby 86% (WT 2163.5 ng/ml vs HO 3968.3 ng/ml, p<0.05) as wasthe protein expression level of the receptor for AGEs (RAGE). Invitro cellular experiments also revealed AGEs directly activate Aktthrough phosphorylation and increase levels of the receptor RAGE.AGE induced phosphorylation of Akt is inhibited in the presence ofwortmannin, suggesting a PI3K dependent signalling mechanism.This was further confirmed in vivo where a bolus injection ofwortmannin in 6-month old mutant mice returned Akt phosphor-ylation levels to those seen in WTmice. In conclusion, using the firsthuman relevant mouse model of diabetes, GENA348 we demon-strate the development of a progressive cardiac phenotype includingcardiac hypertrophy, LV dilatation and diastolic dysfunction similarto the clinical manifestations of diabetic cardiomyopathy. Wepropose that the RAGE/PI3K/Akt pathway contributes to themolecular mechanisms associated with the cardiac phenotype.

68 RARE ALLELES IN GENETIC PREDISPOSITION TO CORONARYARTERY DISEASE: INSIGHTS FROM THE NOVEL ANALYSISOF GENE-CENTRIC ARRAY

doi:10.1136/heartjnl-2011-300198.68

1P Christofidou, 1R Debiec, 1C P Nelson, 1P S Braund, 1L D S Bloomer, 2S G Ball,2A J Balmforth, 2A S Hall, 1M Tomaszewski, 1N J Samani. 1University of Leicester,Leicester, UK; 2University of Leeds, Leeds, UK

Background Genome-wide association studies have been successfulin identifying association between several common variants andcoronary artery disease (CAD). However, collectively these variantsexplain only a small proportion of CAD heritability. It is becomingincreasingly clear that the remainder of the “missing CAD herit-ability” could be explained by low frequency/rare alleles. Because ofthe small number of observations for any given rare allele, the powerto detect its association with a phenotype is a major limiting factorin genetic analysis. In this study we have undertaken a novel stat-istical approach that combines information from all low frequency


BCS Abstracts 2011

(MAF<5%) SNPs at one locus in gene-centric analysis of CAD. Wehypothesised that patients with CAD will show over-representationof rare alleles compared to controls.Methods To examine associations between rare alleles and CAD, wehave used data from 2119 CAD cases and 2440 healthy controlsrecruited to the Welcome Trust Case-Control Consortium(WTCCC) Study. DNA from each subject was genotyped forapproximately 45 000 SNPs in more than 2000 genes/loci using 50KIBC array (version 1). Association analysis was based on theCCRaVAT (Case-Control Rare Variant Analysis Tool) algorithm thatmaximises statistical power by combining all rare alleles withindefined regions into a single “super locus”. Differences in theproportion of cases and controls carrying rare “super loci” weretested by Pearson’s or Fisher ’s exact test. Empirical p values weregenerated by permuting case-control status a predefined number oftimes and repeating the analysis for each replicate.Results 5 candidate regions (MMP23B, VEGFA, DVL1, RIPK1,LPAL2) showed an over-accumulation of rare alleles in patients withCAD when compared to controls (FDR<50%). The number ofanalysed rare alleles at each of these loci ranged from 4 to 42. Themost significant over-representation of rare variants were identifiedat MMP23B (matrix metallopeptidase 23B gene; p¼1.3/104), a genepreviously unsuspected to play a major role in CAD and VEGFA(vascular endothelial growth factor A; p¼2.6310�4). Only one ofthe identified genes (LPAL2; p¼1.7310�3) lies within the locus thatwas previously shown to harbour rare variants associated withsusceptibility to CAD.Conclusions Rare alleles are associated with predisposition to CADand this gene-centric analysis combining information from low-frequency variants of the same locus has a potential to uncover, atleast a proportion of, the “missing heritability” of CAD.

69 GENOME WIDE METHYLATION ANALYSIS IN CORONARYARTERY DISEASE

doi:10.1136/heartjnl-2011-300198.69

K J Dick, C P Nelson, P S Braund, A H Goodall, N J Samani. University of Leicester,Leicester, UK

Background Using genome-wide association studies several geneshave been identified that affect the risk of CAD. However, thesegenes only explain part of the heritability. There is increasingevidence of the role of epigenetic regulation in complex diseases thatmay explain part of the missing heritability. DNA methylation is animportant epigenetic change that regulates gene expression. Anyrole of methylation in CAD is poorly understood. Therefore weundertook an exploratory genome-wide screen to identify genesdifferentially methylated in CAD cases and controls.Methods We characterised DNA methylation in 24 CAD patientswith a documented history of MI and 24 matched controls from theCardiogenics case-control cohort. All subjects were male, ranging inage from 40 to 57 years. For each subject, genomic DNA, isolatedfrom whole blood, was bisulphite converted and run on IlluminaHumanMethylation27 bead chips. The HumanMethylation27 chipsinterrogate 27 578 CpG sites spanning 14 495 genes with an averageof 2 CpG sites per gene.Results Global DNA methylation level was significantly higher incases compared to controls (p¼9.0310�4). Furthermore, 686 indi-vidual CpG sites, spanning 633 genes showed statistically significantdifferences in methylation levels between cases and controls.Significant signals after Bonferroni correction for multiple compar-isons included GNAS (p¼7.94310�5), which is involved in receptor-mediated signal transduction, PCMT1 (p¼7.94310�5), ACD(p¼3.48310�4 part of the telosome/shelterin complex), ATXN2 andAPOA1 (p¼5.6310�3 and p¼0.01). To explore the potential func-

tional importance of differences in methylation level in cases andcontrols for individual genes, we examined the relationship ofmethylation level to transcript level in monocytes and macrophageson a gene by gene basis and identified several genes including GNASand PCMT1 that showed significant correlations between geneexpression and methylation. Pathway enrichment analysis of thedifferentially methylated genes using the DAVID bioinformaticsresource identified a number of pathways that showed significantenrichment including the calcium signalling pathway(p¼3.85310�7).Conclusions This pilot study has shown several significant differ-ences in gene methylation patterns between CAD cases andcontrols. We also found a correlation between methylation level andgene expression for a number of these genes. Genes differentiallymethylated in CAD are significantly enriched for a number ofpathways including the calcium signalling pathway. While thesefindings require further validation they suggest that epigeneticchanges may play an important role in the pathogenesis of CAD.

70 GENE EXPRESSION AT THE 9p21 LOCUS AND CAD RISK

doi:10.1136/heartjnl-2011-300198.70

1C P Nelson, 2P Lundmark, 1V Codd, 1A H Goodall, 2A C Syvanen, 1N J Samani.1Department of Cardiovascular Sciences, University of Leicester, Glenfield Hospital,Leicester, UK; 2Department of Medical Sciences, Molecular Medicine, Science for LifeLaboratory, Uppsala University, Uppsala, Sweden

Background Human chromosome 9p21 harbours a locus that affectsrisk of coronary artery disease (CAD) through an unknown mech-anism. The variants at the locus most strongly associated with CADlie in non-coding regions suggesting that the affect on CAD risk maybe mediated through regulation of gene expression. We investigatedthe association of single nucleotide polymorphisms (SNPs) acrossthe locus with expression of genes in the locus and compared thiswith association of the same SNPs with CAD risk.Methods We quantified transcript levels for CDKN2A, CDKN2B,ARF and MTAP in circulating monocytes from 422 healthy blooddonors and 386 CAD cases and obtained genotypes for SNPs in the9p21 region in the same subjects using genome-wide platforms. Wealso quantified allelic expression (AE) for these genes and for ANRILin 186 of the healthy blood donors. We compared expressionquantitative trait loci (eQTL) associations for the genes with asso-ciation findings for the same SNPs for CAD in the Wellcome TrustCase Control Consortium study.Results In the global gene expression analysis, we found strong ciseQTLs for both CDKN2B (p¼1.3310�38) and MTAP(p¼6.6310�23), explaining 17.0% and 8.0% of the expression ofthese genes. AE analysis confirmed these findings (CDKN2B,p¼6.0310�64; MTAP, p¼1.4310�38) and also showed a significantcis-eQTL effect on ANRIL expression (p¼3.5310�28). Interestingly,the SNPs associated with CDKN2B and ANRIL expression were thesame. However, the SNPs showing e-QTL effects were distinct fromSNPs that showed an association with CAD risk (p¼2.2310�12).Even in the region with a physical overlap of variants affectingexpression of CDKN2B/ANRIL and CAD risk, the effects of therespective variants were independent of each other. Expression ofCDKN2A and ARF was low but did not show any obviouseQTL effect, or differences according to genotype at CAD-associatedSNPs.Conclusions Our findings in monocytes do not support thehypothesis that the chromosome 9p21 locus mediates CAD risk byaffecting expression of the genes at the locus. The mechanism bywhich the chromosome 9p21 locus affects CAD risk requires furtherelucidation.


BCS Abstracts 2011

71 A GENOME-WIDE ASSOCIATION STUDY IN INDIAN ASIANSIDENTIFIES FOUR SUSCEPTIBILITY LOCI FOR TYPE-2DIABETES

doi:10.1136/heartjnl-2011-300198.71

1,2J Sehmi, 3D Salaheen, 4Y Yeo, 5W Zhang, 1,2D Das, 6M I McCarthy, 4E S Tai,3J Danesh, 1,2J Kooner, 2,7J Chambers. 1National Heart and Lung Institute, ImperialCollege, London, UK; 2Ealing Hospital NHS Trust, London, UK; 3Department of PublicHealth and Primary care, Cambridge University, Cambridge, UK; 4Department ofMedicine, National University of Singapore, Singapore; 5Department of Epidemiologyand biostatistics, Imperial College London, London, UK; 6Wellcome Trust for humanGenetics, Oxford University, Oxford, UK; 7Department of Epidemiology and biostatistics,Imperial College London, London, UK

Background Type-2 diabetes (T2D) is a major risk factor for cardi-ovascular disease, and a leading causing of mortality worldwide.T2D is 2e4 fold more common among Indian Asians than Euro-peans, and contributes to higher cardiovascular disease mortality inAsians. Little is known of the genetic basis of T2D in Indian Asians.Methods We carried out a genome-wide association (GWA) study ofT2D in 5561 Indian Asian cases and 14 458 controls from LOLIPOP,PROMIS and SINDI cohorts. Whole genome scans were performedusing the Illumina 317 k or 610 k arrays. Further testing of sugges-tive SNPs was carried out in independent cohorts of Indian Asian(12 K T2D cases and 25 K controls) and European ancestry(DIAGRAM+, 8 K T2D cases and 39 K controls).Results There were two novel loci associated with T2D at p<10�6,and an additional 57 loci associated with T2D at p<10�4 in theGWA study. We used results from DIAGRAM+ to prioritise 19 locifor further testing in Indian Asians. In combined analysis of resultsfrom GWA and further testing, four loci now reached genome-widesignificance (p<5310�8) among Indian Asians. Coding variant andeQTL studies at these loci identify genes closely involved in insulinsecretion and signalling.Conclusion We identify four novel genetic loci associated with T2Din Indian Asians. Our observations provide new insights into thebiological mechanisms underlying T2D, a major risk factor forcardiovascular disease.

72 ANGIOGENESIS IN RESPONSE TO UPREGULATED HYPOXICSIGNALLING IS DEPENDENT ON HAEMODYNAMIC FLOW

doi:10.1136/heartjnl-2011-300198.72

1O J Watson, 1F J van Eeden, 1C Gray, 2T J A Chico. 1MRC Centre for DevelopmentalBiology and Genetics, University of Sheffield, Sheffield, UK; 2NIHR CardiovascularBiomedical Research Unit, Sheffield Teaching Hospitals NHS Foundation Trust,Sheffield, UK

Introduction Hypoxia drives angiogenesis in a range of pathologies.Mutations in von hippel lindau protein (vhl) lead to excessiveangiogenesis via upregulation of hypoxic signalling, due to impairedHIF-1a degradation. Physical forces exerted by blood flow have beenshown to contribute to vascular remodelling. We therefore used vhlmutant zebrafish to observe the interplay between hypoxic signal-ling, haemodynamic flow and vascular development. Since NO hasbeen shown to be both pro-angiogenic and released in response tohaemodynamic force, we assessed whether NO contributed toangiogenesis in this model.Methods Vhl mutant zebrafish were crossed with a fli1; GFPtransgenic that expresses Green Fluorescent Protein (GFP) in theendothelium. Embryonic vascular development was observed inmutants and wild type siblings by confocal microscopy. To deter-mine the role of blood flow in the angiogenic response, cardiactroponin t2 was knocked down by morpholino antisense injection.To assess the contribution of nitric oxide, embryos were treatedwith either L-NAME (nitric oxide synthase inhibitor) (1mM) or

sodium nitroprusside (NO donor) (100mM) from 24-h post fertil-isation (hpf) until imaging at 4dpf.Results Imaging of the developing trunk vasculature revealed thatvhl mutant embryos display excessive and aberrant angiogenesisfrom 3dpf (Abstract 72 figure 1A, B). Cardiac troponin T2 knock-down prevented any cardiac contraction, but embryos developnormally due to passive oxygen diffusion. Loss of blood flow did notalter normal intersegmental vessel patterning in either controls(Abstract 72 figure 1C) or vhl mutants (Abstract 72 figure 1D).However, loss of blood flow completely prevented excessive angio-genesis in vhl mutants (Abstract 72 figures 1D and 2), implying thatboth blood flow and hypoxic signalling are required for “patho-logical” angiogenesis but not developmental angiogenesis (vasculo-genesis). NO synthase inhibition with L-NAME had no effect,suggesting that the contribution of flow to excessive angiogenesis inresponse to upregulated hypoxic signalling is NO independent.


Abstract 72 Figure 2 Effect of troponin T knockdown on total vessellength.

Conclusion Angiogenesis in response to hypoxic signalling is crit-ically dependent upon haemodynamic force, compared with devel-opmental vasculogenesis that can proceed in the absence of anyblood flow. This indicates a different mechanism of development forhypoxia driven angiogenesis and vasculogenesis which may haveimportant therapeutic implications.

73 HERITABILITY OF CORONARY FLOW RESERVE

doi:10.1136/heartjnl-2011-300198.73

R Ahmed, P Muckett, S Cook. Clinical Sciences Centre, Imperial College, London, UK

Introduction Coronary flow reserve (CRF) is the ratio of peak coro-nary flow during maximal coronary artery dilatation to basal


BCS Abstracts 2011

coronary flow and is an important predictor of coronary micro-vascular function. A variety of environmental stimuli have beenshown to affect CFR but little is known about the geneticcomponent of CFR. To characterise the genetics of CFR we initiallymeasured in vivo blood pressure (BP) and ex vivo cardiac pheno-types including CFR in two inbred rat strains, Brown Norway(BN) and Spontaneously Hypertensive Rat (SHR) which is agenetic model for hypertension and microvascular dysfunction. Wethen studied BP and coronary flow (CF) phenotypes in F1 and F2crosses derived from BN and SHR to estimate the heritability ofCFR and its relationship with BP.Methods Animals were anaesthetized using amixture of Oxygen andIsoflurane. BP was measured invasively by cannulation of carotidartery. Following BP measurement hearts were excised and rapidlytransferred to the ex vivo perfusion apparatus where retrogradeperfusion was established using the Langendorff technique. Heartswere perfused with Carbogen buffered Kreb9s solution and pacedconstantly at 360 bpm. A fluid filled balloon was placed in the leftventricular (LV) cavity to measure the pressure indices. CF, LVdeveloped pressure, myocardial contractility (LV dP/dtmax) andmyocardial relaxation (LV dP/dtmin) were recorded at baseline, duringpeak hyperaemia, regional ischaemia (induced by ligation of theproximal left anterior descending artery) and reperfusion.Results 1) CFR differs significantly between the two inbred parentalrat strains. (BN¼2.1 6 0.32, SHR¼1.5 6 0.18, p¼2.6310�7, n¼16each). 2) Heritability of CFR: Broad sense heritability (the propor-tion of total phenotypic variance attributable to total geneticvariance) for CFR is 62% indicating a large and previously unrec-ognised genetic component of CFR. 3) Relationship between CFRand BP: We did not find statistically significant correlation betweenCFR and BP in the F2 intercross (r¼0.11, p¼0.11, n¼176). 4) Rela-tionship between CF and myocardial relaxation (LV dP/dtmin): LVdP/dtmin correlated strongly with CF during all stages of theexperiment (baseline CF, r¼�0.36, p<0.0001, reperfusion CF,r¼�0.40, p<0.0001).Conclusions Our results demonstrate that CFR has a significantgenetic component and is largely independent of BP effects.Furthermore we demonstrate a very significant relationship betweenCF and LV dP/dtmin indicating a link between LV diastolicdysfunction and impaired CF. Using 768 SNP genotyping assay forlinkage mapping and gene expression analysis with Affymetrix ratgene chip, we will determine the quantitative trait loci and tran-scripts associated with CFR to improve our understanding of thegenomic architecture of CFR.

Abstract 73 Figure 1 Coronary flow reserve.

Abstract 73 Figure 2 Correlation between BP and CFR.

74 MECHANISTIC STUDY FOR THE ROLE OF ADVANCEDGLYCATION END PRODUCTS IN THE DEVELOPMENT OFDIABETIC HEART FAILURE

doi:10.1136/heartjnl-2011-300198.74

Z Hegab, T M A Mohammed, L Neyses, M Mamas. Manchester University,Manchester, UK

Advanced glycation end products (AGEs) are thought to play acrucial role in the development of diabetic complications includingheart failure, a leading cause of morbidity and mortality in diabeticpatients. However, the molecular mechanisms that underlie thepathophysiological contribution of AGEs to heart failure develop-ment are not yet fully understood. We therefore investigated theeffects and mechanisms of action of AGEs on isolated neonatal ratcardiomyocytes (NRCM). Standard molecular techniques wereapplied. Western blot showed that RAGE receptor is expressed inNRCM and adult mouse cardiomyocytes. Incubation of NRCM for24 h with AGEs showed a dose dependant reduction of calciumtransient amplitude with a maximum of 52% at 1 g/l (p<0.01)accompanied with 32% reduction in SR calcium content with nosignificant changes in the protein expression of calcium handlingproteins. We demonstrated a 24% increase (p<0.01) in the produc-tion of reactive oxygen species ROS in AGE treated cardiomyocytesmediated through increased NADPH oxidase activity (p<0.05).Subsequent translocation of NF-KB, a transcriptional factor fromthe cytoplasm to the nucleus together with increased NF-KBactivity resulted in a 56% increase in iNOS gene protein expression(p<0.01), a downstream target of NF-KB. The latter was associatedwith 10% increase in NO production (p<0.05) with subsequentnitrosylation of the Ryanodine receptor shown through immuno-fluoresence. Changes in calcium transient were completely inhibitedwhen we incubated the cardiomyocytes with inhibitors of NADPHoxidase, NOS or NF-KB prior to their incubation with AGEs. Inconclusion, AGEs directly decline cardiomyocytes function throughbinding to their RAGE receptor leading to calcium handlingimpairment through increased ROS production inducing activationand translocation of NF-KB to the nucleus. The latter increasedtranscription of iNOS with increased NO production. Coexistence


BCS Abstracts 2011

of ROS and NO favours the production of peroxynitrite that iscapable of nitrosylation of key cellular proteins such as theRyanodine receptor that has a crucial role in cardiac excitation-contraction coupling. This study provides novel insights into themechanisms of cardiac damage in diabetes that occur independent ofvascular disease through AGEs.

75 OPTIMISATION OF MEDICAL THERAPY AFTER CARDIACRESYNCHRONISATION: A NURSING OPPORTUNITY NOT TOBE MISSED

doi:10.1136/heartjnl-2011-300198.75

1S J Russell, 2J Bell, 3L Edmunds, 4J Davies, 3H Rose, 3Z R Yousef. 1Wales HeartResearch Institute, Cardiff, UK; 2Cardiff University, Cardiff, UK; 3University Hospital ofWales, Cardiff, UK; 4University Hospital Llandough, Cardiff, UK

Introduction Cardiac resynchronisation therapy (CRT) is indicated inpatients with left ventricular dysfunction (EF#35%), electro-mechanical dyssynchrony, and limiting heart failure (HF) symptomsdespite optimal medical therapy. In many cases target doses of HFmedications prior to CRTare not achieved due to bradycardia and/orlimiting hypotension. CRT however provides bradycardia backupand improved haemodynamics, thus providing an opportunity tofurther optimise HF medical therapies known to confer substantialmorbidity and mortality benefits. We conducted the present studyto evaluate the potential to further optimise medical treatments inpatients receiving CRTwithin the framework of nurse-led pre andpost CRT clinics.Methods Our unit operates an integrated CRT service with pre-assessment, implantation, and follow-up components. Pre-assess-ment and follow-up incorporate dedicated HF nurse clinics tosupport protocol-driven optimisation of medical therapies. Wetherefore conducted a retrospective analysis of our CRT databaseover a 9-month period to quantify the frequency of use, and dose ofHF medications (bblockers; bB, angiotensin converting enzymeinhibitors: ACE-I or angiotensin receptor blockers: ARB, aldosteroneantagonists, digoxin, and loop diuretics) before and 6 months afterCRT. Total daily dose equivalences within each class of medication(bisoprolol for bB, lisinopril for ACE-I/ARB, spironolactone foraldosterone antagonists, and frusemide for loop diuretics) andtitration protocols were based on National Institute of ClinicalExcellence guidelines for HF (guideline 5).Results Between October 2009 and Jun 2010, 74 patients (age:67611 yrs, 86% male) underwent implantation of a CRT device. All

patients attended the pre and post CRT nurse clinic to optimisemedical therapies and provide adjunctive HF support. Abstract 75table 1 describes the frequency of use and daily dose equivalent ofeach class of medication used in the patients prior to and 6 monthsafter device implantation. The frequency of bB and digoxin useincreased by 10% and 5% respectively. In addition, the dose of bB,ACE-I/ARB, and digoxin significantly increased, while the dose ofloop diuretics significantly reduced in the 6 months after CRTimplantation.Conclusions The beneficial haemodynamic and pacing profilesprovided by CRT offer important opportunities to further optimiseheart failure medications after device implantation. In a dedicatednurse-led CRT follow-up clinic, we successfully initiated b blockersand digoxin in previously naive patients, and significantly uptitratedthe doses of b blockers, ACE-I/ARB, and digoxin, whilesignificantly reducing loop diuretic use in the 6 months after deviceimplantation.

76 EXPANDING THE ROLE OF CARDIAC CARE UNIT NURSES TOREDUCE TIME TO TREATMENT FOR PATIENTS REQUIRINGPRIMARY ANGIOPLASTY

doi:10.1136/heartjnl-2011-300198.76

S Young, G Pretsell, A Gibbins, G Dixon, A de Belder. Royal Sussex County Hospital,Brighton, UK

Introduction In Brighton, UK, 24-h Primary Angioplasty has beenused for the treatment of ST segment elevation myocardial infarc-tion (STEMI) since October 2008, with local patients beingadmitted via the Accident and Emergency (A&E) department. Withthe publication of the National Infarct Angioplasty project report(DH 2008) it was evident that direct admission into the cardiaccatheter lab from the ambulance could further reduce time totreatment. Call to Balloon time (CTBT) <150 mins is a nationallyrecognised indicator measuring the time the patient first calls forprofessional help (usually the ambulance) to the opening of thecoronary artery on the catheter lab Abstract 76 table 1.

Abstract 76 Table 1

% CTBT £150 mins Median CTBT

Financial Year 2009e2010 59/78 76%* 125 mins

Quarter 1 (AprileJune 2010) 32/36 89% 111 mins

Quarter 2 (JulyeSeptember 2010) 44/45 98%* 99 mins

*p¼0.0013.

Methods The on-call cardiology team are non-resident out of hours.It was therefore agreed the point of contact and immediate decisionmaking would lie with the Cardiac Care Unit (CCU) nurses. Apathway was developed following consultation with the multi-disciplinary team at an educational and mapping day, and risks wereaddressed. It was agreed that the ambulance crew would telephonethe CCU nurse who would review the clinical history and thetelemetry ECG. They would then make the decision to activate thecatheter lab team. Patient Group Directions for the administrationof GTN, diamorphine, metoclopramide and clopidogrel were devel-oped so that immediate treatment could be delivered by the CCUnurse without medical prescription before the cardiac catheter labteam arrived, if required. The nurses were trained in their use andassessed as competent. Nurses were already competent in ECGinterpretation, defibrillation, cannulation and venepuncture.Nursing documentation was developed to prioritise the patient9semergency care. CTBTwere monitored.

Abstract 75 Table 1 Heart Failure nurse supervised use of medicationsbefore and 6 months After CRT

Pre-CRT Post-CRT p Value

b-Blocker: exemplar bisoprolol

Frequency of use 78% 88% <0.01

Daily dose equivalent 5.363.1 mg/day 6.963.2 mg/day

ACE-I/ARB: exemplar lisinopril

Frequency of use 93% 93% 0.02


Aldosterone Antagonist: exemplar spironolactone

Frequency of use 32% 28% 0.32


Cardiac Glycoside: exemplar digoxin



Loop Diuretic: exemplar frusemide




BCS Abstracts 2011

Results Following the implementation of the direct entry pathwayin May 2010 the CTBT for all patients admitted direct to ourhospital have reduced. This is statistically significant when lookingat Quarter 2 results from baseline. Patient safety has not beencompromised. Patients who were admitted directly have been askedabout their experience and if anything could be done differentlyfrom their perspective. They have said:< The process is quick which is good from their perspective< They are fully informed< The ambulance crews deal with them competently< The lab staff are waiting for their arrival.Conclusions The CCU nurses have embraced this development andexpansion of their nursing practice, allowing major changes to bemade to the Primary Angioplasty pathway within the existinginfrastructure, despite the challenges of working within thecomplex nature of traditional geographical referral patterns. Alongwith the work of all members of the multi disciplinary team this hassignificantly reduced times to treatment for patients.

77 SCREENING FIRST DEGREE RELATIVES FOR HYPERTROPHICCARDIOMYOPATHY: 12-MONTH EXPERIENCE OF A CARDIO-GENETICS NURSE SERVICE

doi:10.1136/heartjnl-2011-300198.77

1S Finch, 2S Russell, 1D Kumar, 1Z R Yousef. 1University Hospital of Wales, Cardiff, UK;2Wales Heart Research Institute, Cardiff, UK

Introduction Hypertrophic cardiomyopathy (HCM) is an autoso-mally transmitted cardiomyopathy with an estimated gene preva-lence of 1:500, and an important cause of sudden cardiac death.Screening to identify at risk first degree relatives is thereforerecommended. The British Heart Foundation (BHF) recently fundednine Nationwide cardio-genetic nurses to support local initiatives.Our application for a nurse was successful and we present our12-month experience of HCM screening.Methods We mapped the course of patients with suspected HCMreferred to our tertiary heart muscle clinic which serves a populationof 1.4million. Following phenotype confirmation, a family tree andcontact details from the index case were recorded by the cardio-genetic nurse. The index case was given literature to pass onto atrisk relatives. The information pack included an open invitation(referral via primary care) to attend for screening. For relativesresiding outside our catchment area screening was arranged via linkswith the BHF cardio-genetic network and other health careproviders. Relatives domiciled outside UK were given our detailswith offers to support screening. Throughout, strict adherence topatient confidentiality was maintained.Results Over 12 months, 64 index HCM cases presented to ourheart muscle clinic. Pedigree analysis identified 221 first degreerelatives at risk of carrying the HCM gene; mean index-to-at RR:1-to-3.4 (range 0e14 subjects). Of the 221 at risk subjects, 71(19 through paediatrics) have undergone screening throughclinical assessment at our unit with plans for long-term 2e5 yearlyfollow-up in view of variable gene penetrance. Of the 71 screenedsubjects, 15 were newly diagnosed with HCM. Newly diagnosedHCM patients underwent further risk stratification for suddencardiac death; where we identified 3 patients at high risk ($2conventional high sudden death risk factors). After appropriatecounselling, these 3 patients have received primary preventiondefibrillators. Despite our approach, 52 subjects remain unscreened(Abstract 77 table 1), either due to complex family relationships(n¼14), personal preference (n¼28) and/or geographical/logisticalreasons (n¼10).

Abstract 77 Table 1 Screening outcomes of 221 at risk subjectsidentified from 64 index cases of hypertrophic cardiomyopathy

Number of Patients

New screening initiated (local heart muscle clinic) 52

New screening initiated (local paediatric clinic) 19

New screening initiated (out of area service) 6

Pre-existing screening in place 63

Personal preference (declined screening) 28

Awaiting response from subject (literature delivered) 19

Complex family relationships (unable to deliver literature) 14

Geographical/Logistical constraints 10

Subject deceased (non-hypertrophic cardiomyopathy) 3

Subject deceased (hypertrophic cardiomyopathy) 7

Conclusions Proactive screening for HCM can be effectively facili-tated by cardio-genetic nurse services. Each new index case generates3e4 at risk relatives who require long-term surveillance. Of 71asymptomatic at risk subjects screened in our unit, we diagnosed 15new cases of HCM, and 3 patients at high risk of sudden cardiacdeath who subsequently received primary prevention defibrillatorimplantation.


78 FIRST YEAR EXPERIENCE OF A DEDICATED “RADIALLOUNGE” FOR PATIENTS UNDERGOING ELECTIVEPERCUTANEOUS CORONARY PROCEDURES

doi:10.1136/heartjnl-2011-300198.78

S Brewster, R Weerackody, K Khimdas, A Little, N Cleary, A Penswick, M Rothman,A Archbold. London Chest Hospital, London, UK

Introduction The potential to achieve safe early mobilisation andsame day discharge on a consistent basis after radial artery accesshas provided us with the opportunity to make a step change in theway we deliver elective care to patients undergoing percutaneouscoronary procedures. We designed a dedicated “radial lounge” toaccommodate patients before and after their procedure with the aimof minimising the feeling of “hospitalisation” that accompaniesmost encounters with health services. The lounge is a day case unitthat has no beds, only chairs, and televisions but no cardiac moni-tors. Patients remain in their clothes throughout their hospital visit.Here we report our first year9s experience of this facility. Methods:The study population comprised all patients who attended theradial lounge between July 2009eJune 2010 for coronary angiog-raphy or percutaneous coronary intervention (PCI). Patients weresuitable for the radial lounge if they were elective cases who had asatisfactory radial pulse and no pre-procedure contraindication to


BCS Abstracts 2011

same day discharge. Patients were excluded if they had any of thefollowing: an unsuitable radial pulse, planned femoral access, priorcoronary artery bypass surgery, or the requirement for an overnighthospital stay for planned complex/high risk PCI, renal impairment,or social reasons. The final decision regarding route of arterial accesswas left to the operator.Results In the one year study period, 1548 patients were managed inthe radial lounge. 1109 patients underwent coronary angiography,114 (10.2%) of whom also had a pressure wire or intravascularultrasound, and 439 underwent PCI. This represented approx-imately 88% of our unit9s elective angiograms and 60% of ourelective PCIs. Among the patients who underwent angiography, 938(84.5%) were performed radially and 1076 (97.0%) were dischargedfrom the radial lounge on the same day as their procedure. Amongthe PCI patients, 359 (81.8%) were performed radially and 372(84.7%) were discharged the same day. The PCI group included 326(74.3%) patients who had a single vessel treated, 105 (23.9%) whohad two vessels or a bifurcation with a significant side branchtreated, and 8 (1.8%) patients who had three vessels treated. Therewere no deaths or arrhythmias in the radial lounge. Requirement forovernight admission was significantly more common after femoralaccess compared with radial access for both angiography (4.1% vs2.8%; p<0.05) and PCI (21.3% vs 14.2%; p<0.01).Conclusions A dedicated radial lounge free of cardiac monitors is asafe environment in which to manage most patients before and afterelective coronary angiography and PCI. The lack of monitoringnecessitates patient selection but this does not prevent the loungebeing suitable for the majority of elective patients. Femoral access isassociated with a significantly greater requirement for overnightadmission.

79 REMAINING CLOTHED FOR RADIAL DIAGNOSTICCORONARY ANGIOGRAPHY: AN IMPROVEMENT IN THEPATIENT JOURNEY

doi:10.1136/heartjnl-2011-300198.79

S Eve, M Sinha, T A Wells. Salisbury District Hospital, Salisbury, UK

Background Patients undergoing invasive diagnostic coronaryangiography (DCA) for the first time often display high levels ofanxiety at the time of their procedure as they are unfamiliar withthe cardiac catheter laboratory set up. It is therefore part of thecardiac catheter laboratory staff9s role to reduce patient fears andhence improve their journey through the cardiac catheter laboratory.Several Cardiac centres have recently introduced radial loungeswhereby patients feel less “hospitalised” by not needing to undressfor their procedure.Methods Following infection control approval, between mid-August2010 and the end of October 2010, patients undergoing radial DCAat Salisbury District Hospital were offered the option of remainingclothed for their procedure. Each patient was given an informationleaflet included in which was explained possible downsides to beingdressed including if CPR were needed then clothes would be cut,failure of radial access and the subsequent need for femoral access,and the possibility of soiling the clothes with either blood or iodine.The only caveat stated was that female patients were not allowed towear an underwire bra. Following their DCA, patients were thenasked to fill in an anonymous questionnaire in which they wereasked about their experience and whether not having to undressmade them feel more relaxed.Results 57 consecutive patients underwent (DCA) during this timeperiod (100% uptake) with an average age of 68.1 6 9.6 years. 71%were male and 21% (12/57) had undergone a DCA previously. Of

these 12 patients 92% (11/12) stated that not having to undress wasa good idea while an identical number felt much more relaxed thantheir previous DCA experience. Of the 45 patients that had not hada previous DCA, 96% (43/45) stated that not having to undress wasa good idea while 96% (43/45) felt that this had made them feel veryrelaxed during their pathway. The other two patients felt that itmade no difference. No patients required cross-over to femoralaccess and there were no blood or iodine stains on any clothes.Having patients remain dressed did not reduce fluoroscopic imagequality and there were no issues with infection.Conclusion Offering patients the option of having their radial DCAdone without undressing is safe and helps to improve the patientjourney through the cardiac catheter laboratory by making themfeel more relaxed and less hospitalised. This is now standard at ourInstitution.

80 PPCI: IS THERE A ROLE FOR THE ACS ANP?

doi:10.1136/heartjnl-2011-300198.80

V Oriolo, J Tagney. Bristol Heart Institute University Hospital Bristol NHS FT, Bristol, UK

Introduction Primary Percutaneous Coronary Intervention (PPCI) isnow considered the treatment of choice for patients experiencing STElevation Myocardial Infarction (STEMI) (European Society ofCardiology/European Association for Cardio-Thoracic Surgery2010). One of the many benefits claimed is reduced length of staydue to decreased morbidity (Zahn et al 2000, Kalla et al 2006). Toassess performance of one English PPCI 24/7 provider organisationagainst the national average length of stay (LOS) for patients postPPCI, a retrospective baseline audit was conducted. This demon-strated an average LOS of 4.4 days which is above the NationalInfarct Angioplasty Project 2008 average LOS of 3 days. A 48 h nurseled discharge (NLD) protocol was therefore developed and intro-duced by the acute coronary syndromes advanced nurse practitionerto streamline the patient journey. This instructed the nurse and/orphysician to ensure appropriate investigation and documentationwas carried out in a timely manner to avoid unnecessary delays inpatient discharge.Method Suitability criteria for the 48-h NLD protocol were estab-lished, which included: absence of acute complications (eg,:bleeding, haemodynamic instability, ongoing chest pain, ejectionfraction <40%, respiratory compromise); appropriate support atdischarge. Following the baseline audit, data were electronicallycollected prospectively for 5 months, measuring date of admissionto date of discharge to the usual place of residence. After 5 monthsthe audit was repeated to assess the average length of stay forpatients presenting with STEMI.Result Between 1st April 2010 and 31st August 2010, a total of 274patients were admitted with STEMI. Of these, 122 (45%) met theNLD criteria and were discharged by the ACS ANP. The remaining152 (55%) were discharged by the medical physicians. It was notedthat introduction of the protocol also facilitated a structuredapproach to discharge for the medical team. The average LOS for allPPCI patients (n¼274) decreased from 4.4 days to 3 days (30%). Forpatients that were seen and discharged solely by the advanced nursepractitioner (n¼122), the average LOS decreased from 4.4 days to2.0 days (55%).Conclusions/Implications In the current financial climate, a decreasein LOS can have a significant impact on any organisational resourcesthus increasing efficiency saving and patient throughput. Thisdemonstrates part of the added value the advanced nurse practi-tioner brings to patient care and to tertiary centres that provide a24/7 PPCI service.


BCS Abstracts 2011

PPCI -REPERFUSION TIME 0

CCU

12HRS PATIENT STABLE*

(SYSTOLIC BP> 100 mmHg - NO NEW ECG CHANGES - PAIN FREE - SaO2>93%)

REG/SENIOR NURSE TO BOOK ECHO & TRANSFER TO WARD (IF APPLICABLE)

24 HRS PATIENT STABLE * (AS ABOVE)

DEMONITOR + ECHOCARDIOGRAM+

REHAB REFERRAL

TTA’S, D/C LETTER, MINAP, REHAB TO SEE

36 HRS PATIENT STABLE* (AS ABOVE)

EF≥ 40% & KILLIP I

48 HRS PATIENT STABLE* (AS ABOVE)

BLOODS, ECG, COPY OF PROCEDURE,

VENFLON OUT, MINAP

HOME

TO BE SEEN BY

ACS NURSE/DR

ECHO PERFORMED?

Y N

ARRANGE

ASAP

EF≤ 39%

DR R/V SCREEN FOR

ICD

IF PATIENT UNSTABLE AT ANY STAGE PLEASE D/W DR/ACS NURSE

PPCI NURSE LED DISCHARGE PROTOCOL

Abstract 80 Figure 1 PPCI nurse LED discharge protocol.

81 DYSSYNCHRONOUS THREE PLANE MOTION AND IMPAIREDLEFT VENTRICULAR TWIST IN PATIENTS WITH HEARTFAILURE AND NORMAL EJECTION FRACTION

doi:10.1136/heartjnl-2011-300198.81

1Y T Tan, 2F W G Wenzelburger, 3F Leyva, 3J E Sanderson. 1Department of Cardio-vascular Medicine, Birmingham, UK; 2Institute for Science and Technology in Medicine,Keele University, Stoke on Trent, Keele, UK; 3Department of Cardiovascular Medicine,University of Birmingham, Birmingham, UK

Background The pathophysiology of heart failure with normalejection fraction (HFNEF) is complex and not fully understood.Recent publications showed a loss of apical rotation and longi-tudinal function particularly on exercise in these patients. Whethera deterioration of basal rotation and a dyssynchrony of differentthree plane motions on exercise might contribute to symptoms inthese patients is not known.Method 72 Patients (age 7367 years, 48 female) with breathlessnesson exertion and normal EF (6067%) underwent cardiopulmonaryexercise test to rule out alternative clinical reasons (VO2max18.464.9 ml/min/kg). Data were compared to 38 age-matchedcontrol subjects (age 7167 years, 29 female, EF 6367%) with anormal exercise tolerance (VO2max 28.665.1 ml/min/kg). Allunderwent full Doppler 2D-echocardiography at rest and on supineexercise. Echo images were analysed off-line. Apical and basal rota-tion, longitudinal and radial displacement were measured by speckletracking. Speckle tracking pictures and colour TDI curves wereloaded into custom made software. The software interpolated allcurves and calculated twist as the difference of rotation at apex andat base. The software offered timing information to calculate SDand time delays for different motions.

Results As previously described apical rotation was reduced at restand on exercise Basal rotation was comparable at rest but signifi-cantly reduced on exercise in patients. The SD for four differentsystolic peak motions (basal and apical rotation, longitudinal andradial displacement) was comparable at rest but on exercise controlsshowed a significantly reduced SD compared to patients showing agreater ability to synchronise motions. Furthermore a ratio ofuntwist during IVRT and longitudinal extension (Ratio Untwist/Extension in IVRT) showed a significant deeper slope on exercisefor patients indicating a loss of synchrony in diastole, too. All resultsare presented in Abstract 81 table 1.

Abstract 81 Table 1

PatientsRest

ControlsRest p value

PatientsExercise

ControlsExercise p value

Apical Rotation (8) 9.964.4 13.464.0 <0.001 12.564.7 16.663.9 <0.001

Basal Rotation (8) �8.363.3 �8.063.4 0.676 �7.763.2 �9.763.0 0.011

Twist (8) 18.065.7 21.064.9 0.01 19.565.9 25.966.0 <0.001

SD Systolic Motions (ms) 48.6632.9 43.1625.3 0.38 40.1627.1 25.9615.5 0.01

Ratio Untwist/Extensionin IVRT (8/mm)

25.3651.4 7.1610.7 0.059 9.6614.7 3.363.8 0.034

Conclusion Patients with HFNEF show a deterioration of basalrotation and a systolic and diastolic three plane dyssynchronyparticularly on exercise. This might further contribute to the dete-rioration of early diastolic suction and therefore decrease strokevolume on exercise. This might be a major contribution to theirsymptoms.

82 MANAGEMENT OF ADVANCED HEART FAILURE IN THE UK:TRENDS IN HEART TRANSPLANTATION AND MECHANICALCIRCULATORY SUPPORT

doi:10.1136/heartjnl-2011-300198.82

1A Emin, 2C A Rogers, 3H L Thomas, 4S Tsui, 5S Schueler, 5G MacGowan, 6A Simon,7R S Bonser, 4J Parameshwar, 8N R Banner. 1Clinical Effectiveness Unit, The RoyalCollege of Surgeons of England, London, UK; 2Clinical Trials and Evaluation Unit,University of Bristol, Bristol, UK; 3NHS Blood and Transplant, Bristol, UK; 4Cardio-pulmonary Transplantation, Papworth Hospital NHS Foundation Trust, Cambridge, UK;5Cardiopulmonary Transplantation, Freeman Hospital, Newcastle, UK; 6Heart and LungTransplantation, Royal Brompton and Harefield NHS Trust, Middlesex, UK; 7Cardio-pulmonary Transplantation, Queen Elizabeth Hospital, University of Birmingham,Birmingham, UK; 8Royal Brompton and Harefield NHS Trust - on behalf of the UK VADForum and UKCTA Steering Group, Middlesex, UK

Introduction Patients with advanced heart failure due to systolicventricular dysfunction require “pump replacement” therapy.Previously, heart transplantation (HTx) met this need but waitingtimes have increased due to shortage of donor hearts. Consequently,more patients require a ventricular assist device (VAD) as a bridge totransplant (BTT). We report UK activity, trends and outcome forHTx and BTT VAD.Methods Data were acquired from a comprehensive national data-base using 3 eras for analysis: E1: 5/2002e12/2004, E2: 1/2005e12/2007 & E3: 1/2008e6/2010. Paediatric and multi-organ transplantswere excluded from the transplant cohort. Patients who receivedprior short-term support (bridge to bridge) were excluded from theVAD group.Results 1278 patients were listed for HTx over the 3 eras: E1 155 peryear, E2 165 per year, E3 148 per year. The number of adult HTx fellfrom 132 per year in E1 to 94 per year in E3. The median waitingtime for non-urgent HTx increased from 87 days in E1 (95%CI 55 to119) to 321 days in E3 (95%CI 203 to 439) (p<0.001). 239 patientsneeded left VAD support as BTT; 75 (31%) also received a right VAD.Activity rose from 26 per year in E1 to 41 per year in E3. Device


BCS Abstracts 2011

choice has changed in favour of rotary pumps; 19%, 69% and 96%for E1, E2 and E3 respectively. Median duration of VAD supportincreased from 84 days (IQR 20e209) in E1 to 280 days (IQR86e661) in E3 (p<0.01). Overall survival to 1 year after VADimplant rose from 52.9% (95%CI 40 to 64) in E1 to 65.6% (95%CI 54to 75) in E3 (p¼0.10). Of the 239 patients implanted, 83 (35%) haveundergone HTx, 52 (22%) are alive on VAD support & 84 (35%) diedon support. Twenty were explanted following myocardial recovery;18 of these remain alive & 2 died. Survival after HTx for patientswith or without a pre-HTx VAD was 81.4% (95%CI 71 to 88) &90.3% (95%CI 88 to 92) respectively at 30-days (p<0.01) and 80.0%(95%CI 63 to 82) & 84.3% (95%CI 82 to 87) respectively at 1-year(p<0.01). 1-year survival conditional on 30-day survival was similarwith & without a pre-HTx VAD (93% vs 91%, p¼0.48).Conclusion Heart transplant activity has declined and waiting timeshave become prolonged leading to an increased need for bridging totransplantation. There has been a shift from volume displacementVADs to rotary blood pumps and the duration of support hasincreased. Post VAD survival has improved. While bridging appears toincrease mortality early after HTx, longer term survival is unaffected.

83 CLINICAL AND HAEMODYNAMIC STATUS BEYOND3 MONTHS OF MECHANICAL SUPPORT WITH THEHEARTWARE VENTRICULAR ASSIST DEVICE

doi:10.1136/heartjnl-2011-300198.83

B Gordon, A McDiarmid, N Robinson, N Wrightson, G Parry, S Schueler, G MacGowan.Freeman hospital, Newcastle upon Tyne, UK

Introduction Limited data exist on the longer term clinical andhaemodynamic impact of the HeartWare left ventricular assistdevice (HVAD�) when used as a bridge to heart transplantation.Patients who had a device longer than 3 months were reviewed.Methods 26 patients had a HVAD implanted from 07/2009 to 07/2010 (mean age 46.8 years, 18 male, 5174 total days of support).Baseline and follow-up NYHA functional class, peak VO2 (bicycleexercise), right heart haemodynamics, biochemistry and mortalityoutcome were compared using paired t test. Results: 22/26 (85%)patients survived beyond 3 months. 4 patients died before (meansurvival 40 days, 2 stroke and 2 multi-organ failure) and 2 died after(mean survival 173 days, 1 stroke, 1 right heart failure) dischargefrom hospital. 2 patients were transplanted (at 3 and 241 days afterimplant) and 1 had recovery of LV function. Follow-up data isavailable for 14/20 survivors (mean 197 days from implant).Significant results are shown in the Abstract 83 table 1. There wasno significant change in peak VO2 (9.961.8 to 12.963.8, p¼0.08),haemoglobin (12.761.7 to 12.161.2, p¼0.3) or creatinine (122641to 105638, p¼0.19).

Abstract 83 Table 1

Parameter Baseline Follow-up p Value

NYHA functional class 3.660.4 2.160.6 <0.001

Mean PA pressure (mm Hg) 3869 2168 <0.001

Mean PW pressure (mm Hg) 2565 1065 <0.001

Transpulmonary gradient (mm Hg) 1265 963 0.02

Right atrial pressure (mm Hg) 1166 564 0.006

PA oxygen saturation (%) 5168 6667 0.003

Cardiac Output (l/min) 2.960.8 4.360.9 <0.001

Sodium (mmol/l) 13464 13963 0.002

Conclusions The HVAD� results in significant improvement infunctional class, right heart haemodynamics, cardiac output andsodium levels beyond 3 months of therapy. Ongoing randomisedclinical trials will establish the long-term outcome of this device.

84 TREATMENT OF REFRACTORY RIGHT HEART FAILURE AFTERIMPLANTATION OF A LEFT VENTRICULAR ASSIST DEVICE.IS THE LEVITRONIX CENTRIMAG RIGHT HEART SUPPORT ASOLUTION?

doi:10.1136/heartjnl-2011-300198.84

B Zych, A F Popov, A Barsan, M Hedger, R Hards, N R Banner, A R Simon. RoyalBrompton&Harefield NHS Foundation Trust, Harefield Hospital, Harefield, UK

Introduction Right heart failure after left Ventricular Assist Device(LVAD) implantation is a severe complication, in extreme casesnecessitating additional mechanical assist. We present our institu-tional experience with the Levitronix CentriMag used for rightventricular support commencing LVAD implantation with refrac-tory right ventricular failure.Material and Methods Between March 2001 and November 2010 109patients underwent implantation of long term, total implantable,continuous flow LVADs: 60 HeartMate II, 25 Jarvik 2000 and 24HeartWare. All patients requiring right ventricular support wereincluded (n¼24), for which the Levitronix CentriMag continuousflow, paracorporeal device was used. The analysis included patientdemographics as well as overall duration of support and outcomeparameters, including survival at 30, 90 days and 1 year.Results 24 pts. underwent implantation, age 37.9613.7 years,gender: M/F-15/9, underlying disease: dilated cardiomyopathy 22(92%), peripartum cardiomyopathy 1(4%), viral myocarditis 1(4%).Median duration of support: 28 days (5e146). 3(12.5%) pts. under-went heart transplantation (HTx) on RV support, 14(58.5%)underwent RVAD explantation. Of these, 3 underwent successfulHTx, 4 recovered LV function and underwent successful LVADexplantation, 3 remain on continuing LVAD support, 4 patients diedafter RVAD explantation (post explantation day 1, months 3 and 4and at 2 years), 7(29%) patients died during RV support. MedianITU/hospital stay: 19.5 days (6e145)/78.5 days (10e219). 30-day/90-day/1-year survival: 79%/71%/60%. 15(62.5%) patients weredischarged from hospital after treatment. Median survival afterprocedure: 473.5 days (10e1917).Conclusion Levitronix CentriMag right ventricular support is anexcellent option for post LVAD implantation treatment of refractoryRV failure. It allows either bridging to transplantation or RV func-tion improvement and provides an acceptable rate of survival.

85 PREDICTION OF RESPONSE TO BIVENTRICULAR PACINGFROM DYSSYNCHRONY INDICES: THE ABSOLUTE LIMIT ONPREDICTABILITY, AND ITS CLINICAL IMPLICATIONS

doi:10.1136/heartjnl-2011-300198.85

1S S Nijjer, 2P Pabari, 3B Stegemann, 4V Palmieri, 5N Freemantle, 2A Hughes, 2D P Francis.1Imperial College Healthcare NHS Trust, London, UK; 2Imperial College London, London,UK; 3Medtronic Bakken Research Center, Maastricht, The Netherlands; 4Ospedale deiPellegrini, Naples, Italy; 5University of Birmingham, Birmingham, UK

Background It may be incorrect to believe that, with a good echo-cardiographic marker of mechanical dyssynchrony, response tobiventricular pacing (BVP) should be predictable with a high r2

value. Variability between repeat echocardiographic measurements,and between successive dyssynchrony measurements, may reducer2. Both will mandatorily limit the achievable r2; we determine this“contraction factor”.Method and Results We compared correlation coefficients of dyssyn-chrony indices with response markers, in externally monitoredrandomised controlled trials (EMRCTs) and highly skilled singlecentre studies (HSSCSs). DLVEF in CRT recipients comprises trueCRT effect plus unpredictable spontaneous variability present incontrol patients (Abstract 85 figure 1, upper panel). The resultantdepression in r2 is calculated. HSSCSs overstate r2 between


BCS Abstracts 2011

dyssynchrony and remodelling response in contrast to EMRCTs(p<0.0000000001),whether response is LVEF (0.40 vs 0.01), ESV (0.26vs 0.01); EDV (0.53 v 0.01). An “averaged” reported r2 betweendiffering dyssynchronymarkers to commonly used echocardiographicresponse markers is shown in Abstract 85 figure 1, lower panel.


EMRCT data shows maximal r2 between dyssynchrony andDLVEF is 0.57 (DESV, 0.54; DEDV, 0.50). Dyssynchrony indices’ ownvariability further contracts observable r2 values (by x0.68). Theoverall ceiling to r2 is between dyssynchrony and DLVEF is 0.39(DESV, 0.37; DEDV, 0.34). All EMRCTr2 values obey these statisticallimits; 29% of HSSCSs results do not.Conclusions HSSCSs suggest dyssynchrony markers strongly predictresponse to BVP but EMRCTs cannot confirm this. Natural varia-bility forces observed correlation coefficients between dyssynchronyand response to be low. EMRCTs, being less susceptible to publicationbias, reflect this reliably. Frequent citation (without verification inindependent cohorts) of the most exuberant values, from HSSCSscreates mathematically unviable, unrealistic, expectations. Simplysearching for progressively more extreme correlations is thereforemisguided. Rationally, we should concentrate on improving test-retest reproducibility of markers of dyssynchrony and of response.

86 HOW OFTEN IS IMPORTANT ADJUSTMENT OF PACINGINTERVALS REQUIRED FOR OPTIMAL RESPONSEFOLLOWING CRT?

doi:10.1136/heartjnl-2011-300198.86

1V Nayar, 1F Z Khan, 1A Rawling, 1L Ayers, 2M S Virdee, 2D Begley, 1D P Dutka,1P J Pugh. 1Addenbrooke’s Hospital, Cambridge, UK; 2Papworth Hospital, Cambridge,UK

Introduction A significant minority of patients do not experienceclinical benefit following cardiac resynchronisation therapy (CRT).Haemodynamically-guided adjustment of the intervals betweenchambers paced (“optimisation” of atrio-ventricular (AV) and left-

right ventricular (VV) delays) may be undertaken to improve thechance of response to CRT. However, data to support this approachas standard management are lacking and many institutionsprogramme CRT devices to deliver “out-of-the-box” intervals, onlyundertaking optimisation when clinical response is lacking. Wesought to determine how often the “out-of-the-box” settings areoptimal or acceptable and how often CRT optimisation results insignificant alteration of the pre-programmed pacing intervals.Methods Data were collected from 180 consecutive patients whounderwent CRT followed by optimisation within 24 h. Opti-misation was performed with serial adjustment of AV and VVintervals. Haemodynamic assessment was undertaken using eitherechocardiography or Non-Invasive Cardiac Output Measurement.The optimal pacing intervals were considered to be those whichresulted in greatest acute augmentation of cardiac output and thedevice was programmed accordingly. The final settings werecompared with the pre-programmed settings for that device and thedifference (AV or VV Adjustment) derived, taking into account thepreset paced or sensed AV delay. An AV or VV Adjustment of morethan 40 ms was considered to be clinically significant. Data arepresented as mean (SD).Results Optimal AV delay ranged from 60 to 200 ms (mean 124 ms(30)), VV delay ranged from 0 to 100 ms (mean 23 ms (19)). Withthe pre-set pacing parameters, cardiac output was acutelyaugmented by 13.1 (34)%. Optimised CRT produced furtherimprovement of cardiac output, to 24.9 (32)% augmentation. “Out-of-the-box” settings were found to be optimal in 11 (6.1%), orrequiring only minor alteration in 120 (66.7%). A clinically signifi-cant alteration in AV delay was made in 40 (22.2%), in VV delay in12 (6.7%) or in either parameter in 49 (27.2%).Conclusions Significant adjustment of AV or VV delay is required inover a quarter of patients receiving CRT. Optimisation of pacingintervals provides augmentation of cardiac output over and abovethe “out-of-the-box” settings. The findings suggest that optimisationis an important component of resynchronisation therapy.

Abstract 86 Table 1 Adjustment of pacing intervals followingoptimisation of CRT

0 1e20 mS 21e40 mS 41e60 mS 61e80 mS 81e100 mS

AV Adjustment 29 (16.1) 89 (49.4) 22 (12.2) 32 (17.8) 7 (3.9) 1 (0.6)

VV Adjustment 50 (27.6) 65 (35.9) 53 (29.3) 11 (6.1) 0 1 (0.6)

Data as N (%).

87 OPTIMISATION OF VV DELAY OF CRT IS MOREREPRODUCIBLE USING PEAK VELOCITIES THAN USINGVELOCITY TIME INTEGRAL, AS WELL AS BEING QUICKER

doi:10.1136/heartjnl-2011-300198.87

P A Pabari, A Kyriacou, M Moraldo, C Manisty, A D Hughes, J Mayet, D P Francis.Imperial College London, London, UK

Background It is not obvious which is a better echocardiographicmarker for optimisation of AVor VV delay: stroke distance (VTI) orpeak velocity. The biggest problem is genuine physiological varia-bility between beats. Because optimisation of VV delay requiresdetection of persistent changes in cardiac function (“signal”), whichmay be small in relation to beat-to-beat variability (“noise”), weshould choose measurements with the best signal-to-noise ratio andreproducibility. The standard echocardiographic method of choicefor VV delay optimisation is to maximise left ventricular outflowtract velocity time integral (LVOT VTI). An alternative is peakvelocity instead of VTI as the parameter to be measured. But surelyVTI, which is encompassing and cumulating more data, is moreimmune to disruption by spontaneous variability between beats,


BCS Abstracts 2011

and therefore simply using peak velocity might give a less reliableoptimum? Surely the time saved by using peak would have a priceto pay in poorer reproducibility of the optimum? In this study, weevaluate whether peak velocity is a suitable alternative to VTI,having regard to both time consumed and reproducibility. We alsoexamine whether averaging multiple replicate measurementsimproves optimisation.Methods & Results VV optimisation was performed on 40 subjectswith biventricular pacemakers using LVOT velocity (VTI or peak) asthe echocardiographic marker being maximised. Importantly, 6successive replicate optimisations were performed per patient at asingle session. Scatter of apparent VV optimum between repeatoptimisationswas threefold smaller for peak thanVTI (p<0.03), witha single measurement for each. Peak velocity had a higher intraclasscorrelation coefficient (ICC) thanVTI (0.66 vs 0.53, p¼0.003). Scatterbetween replicate optimisations is reduced if, instead of singlemeasurements, we use pairs, or triplicates (ANOVA p<0.0001). Thisbenefit occurs with both peak and VTI (p<0.001 among each). Timetaken for acquisition and analysis of a single optimisation (6 settings)was 17.5 s for peak and 57.5 s for VTI (p<0.0001).Conclusions Doppler optimisation of VV delay using peak velocityrather than VTI is (as expected) quicker but (surprisingly) moreaccurate. Making replicate measurements further improves repro-ducibility. Perhaps guidelines should favour peak over VTI andmandate multi-replicate averaging? These data suggest a rareopportunity to reduce labour while increasing reliability of opti-misation. Indeed, triplicate peak velocity assessment takes the sameamount of time as a single VTI, and identifies the VV optimum 3times more confidently. While VTI measurement remains essentialfor assessing stroke volume and cardiac output, for optimisationpurposes it comparison of peak velocity between different settings isboth faster and more reliable.


88 EVALUATION OF THE IMPACT OF AV DELAY VARIATION ONTHE ACUTE MECHANOENERGETIC EFFICIENCY OF CARDIACRESYNCHRONISATION THERAPY AND ASSESSMENT OFPERFORMANCE OF NON-INVASIVE VS INVASIVEHAEMODYNAMIC OPTIMISATION

doi:10.1136/heartjnl-2011-300198.88

A Kyriacou, P Pabari, K Willson, R Baruah, S Sayan, D W Davies, J Mayet, N S Peters,P Kanagaratnam, Z Whinnett, D P Francis. International Centre for Circulatory Health,London, UK

Background The impact of varying AV delay on the acute mechanoe-nergetic efficiency of cardiac resynchronisation therapy (CRT) is notknown; nor is known if non-invasive haemodynamic optimisation byblood pressure agrees with invasive haemodynamic measures duringoptimisation. We studied these invasively, in contemporary patients.Methods Eleven patients with heart failure (EF 2968%) and leftbundle branch block (LBBB, QRS 154626 ms) underwent measure-

ments of left ventricular (LV) pulse pressure (systolic minus diastolic),aortic flow velocity and myocardial oxygen consumption (MVO2) atfour settings: 3 AV delays during biventricular (BiV) pacing (referenceBiV-AV120 ms; BiV-AV40 ms; individualised haemodynamic BiV-AVoptimum), and at intrinsic ventricular conduction (LBBB). Atrialpacing at 100 bpm ensured a fixed heart rate.Results LV pulse pressure rose from LBBB to BiV-AV120 ms by1062% (p<0.001) and 261% more (p<0.05) at the haemodynamicBiV-AVoptimum. At BiV-AV40 ms, pressure was 1062% worse thanBiV-AV120 ms (p<0.001), no different to LBBB (D¼0.860.4%,p¼ns). Invasive aortic flow velocity, measured at a fixed positionthroughout each individual’s study (ie, cardiac output index), roseby 962% (p<0.01) from LBBB to BiV-AV120 ms, rising a further361% (p<0.01) at BiV-AVoptimum. At BiV-AV40 ms, aortic flowwas, no different to LBBB (p¼NS). MVO2 increased from LBBB toBiV-AV120 ms by 964% (p¼0.035) and to BiV-AVoptimum by1263% (p¼0.002). MVO2 at At BiV-AV40 ms and LBBB was notsignificantly different (D463%, p¼ns), The 4 pacing states lay ona straight line: for Dpressure against Dflow, r¼0.99 (p<0.01),Abstract 88 figure 1. Dexternal work (Dpressure 3Dflow) correlatedwith D MVO2, r¼0.99 (p<0.01), with slope 1.6160.17, significantlygreater than 1.00 (p<0.05), Abstract 88 figure 2.

Abstract 88 Figure 1 The correlation of LV pulse pressure and aorticflow velocity during acute biventricular pacing, (at three AV delays) andduring LBBB, at a fixed heart rate.

Abstract 88 Figure 2 The correlation of cardiac work and myocardialoxygen consumption during acute biventricular pacing, (at three AVdelays) and during LBBB, at a fixed heart rate.


BCS Abstracts 2011

The correlations of optimal AV delays by non-invasive (Finom-eter) systolic blood pressure (SBP) vs invasive measures were asfollows; aortic SBP, r2¼0.96, p<0.01; aortic flow velocity, r2¼0.81,p<0.01; LV dP/dtmax, r2¼0.68, p<0.01.Conclusions During acute biventricular pacing, at a fixed heart rate,changing the AV delay affects the cardiac mechanoenergetics. Whenan AV delay improves external cardiac work, compared to LBBB or aphysiologically too short AV delay (eg, AV 40 ms), it also increasesthe myocardial oxygen consumption. However, only 1% moreenergy is consumed per 1.6% more external work (pressure3flow)done; as a result cardiac efficiency improves. Haemodynamic opti-misation of AV delay can be achieved with high precision using noninvasive beat-to-beat pressure measurements. This should enableroutine haemodynamic optimisation (easily automated) of CRTdevices in clinical practice.

89 ELECTROMECHANICAL INTERACTION IN PATIENTSUNDERGOING CARDIAC RESYNCHRONISATION THERAPY:COMPARISON OF INTRACARDIAC ACTIVATION MAPS ANDEARLY SEPTAL CONTRACTION IN LEFT BUNDLE BRANCHBLOCK

doi:10.1136/heartjnl-2011-300198.89

1S G Duckett, 2O Camara, 1M Ginks, 3J Bostock, 1P Chinchapatnam, 1M Sermesant,2A Pashaei, 3J S Gill, 3G Carr-White, 2A F Frangi, 1R S Razavi, 2B H Bijnens,3C A Rinaldi. 1Kings College London, London, UK; 2UPF, Barcelona, Spain; 3Guy’s andSt Thomas’ Hospital, London, UK

Introduction Early inward motion and thickening/thinning of theventricular septum associated with left bundle branch block (LBBB)is known as the septal flash (SF). Correction of SF corresponds withCRT response. We hypothesise that electromechanical interactionsinduced by SF are associated with functional changes in conduc-tivity and a “U-shaped” activation pattern. Characterising thespatio-temporal relationship between electrical and mechanicalevents may explain why patients with a SF respond to CRT.Methods 13 patients (63610 years, 10 men) with severe heartfailure (EF 22.865.8%) undergoing CRT underwent echocardiog-raphy and non-contact mapping (NCM) pre-implant. Presence andextent of a SF was defined visually and with M-mode and fusedwith NCM bull9s eye plots of endocardial activation patterns. LVdP/dtmax was measured during different pacing modes.Results Five patients had a large SF, four small SF and four no SF.Patients with large SF had areas of conduction block in non-infarcted regions whereas those with small or no SF did not(Abstract 89 figure 1). Patients with large SF had greater acuteresponse to left ventricular (LV) and biventricular (BIV) pacing vsthose with small/no SF (% increase dP/dt 28614% vs 11619% forLV pacing and 42628% vs 22621% for BIV pacing). The lines ofconduction block disappeared after LV and BIV pacing, whileremaining present with RV pacing (Abstract 89 figure 2). Abstract89 figure 1 Patient with a large SF. Unipolar isochronal map withNCM electrograms showing fragmented signals (development ofsplit potentials) indicating a reduction of conduction and inabilityto cross throughout the inferior region. The NCM mapping elec-trograms show the criteria used by Auricchio et al to define block,with the emergence of R-wave, smallest and earliest at the superiorpart of the block (where area of block begins) with largest negativepeak. Bold white arrows on the electrogram indicate how the elec-trical activation spreads superiorly in a U-shape pattern leading tothe development of split potentials. Abstract 89 figure 2 Activationmaps of patient with a large SF. Row A, baseline with area of blockand late anterior breakthrough. Row B, RV pacing showing the areaof anterior block remains. Row C, BIV pacing. Functional conduc-tion block has disappeared.



Conclusion A strong interaction exists between electrical activationand mechanical deformation of the septum. Correction of bothmechanical synchrony and the functional conduction block by CRTmay explain the large positive response in patients with a SF.

90 INVASIVE ACUTE HAEMODYNAMIC RESPONSE TO GUIDELV LEAD IMPLANTATION PREDICTS CHRONICREMODELLING IN PATIENTS UNDERGOING CARDIACRESYNCHRONISATION THERAPY

doi:10.1136/heartjnl-2011-300198.90

1S G Duckett, 1M Ginks, 1A Shetty, 2J Bostock, 2J S Gill, 2S G Hamid, 2S Kapetanakis,2E Cunliffe, 1R S Razavi, 2G Carr-White, 2C A Rinaldi. 1Kings College London, London,UK; 2Guy’s and St Thomas’ Hospital, London, UK

Introduction Cardiac resynchronisation therapy (CRT) reducesmortality and morbidity in heart failure patients, however up to 30%of patients do not derive symptomatic benefit. Higher proportions donot remodel. Multi-centre trials have shown echocardiographictechniques are poor at improving response rates. We hypothesisedthat the degree of acute haemodynamic response (AHR) at implantcan predict which patients remodel. We evaluated the relationshipbetween AHR and reverse remodelling (RR) in CRT. Methods 33patients undergoing CRT (21 dilated & 12 ischaemic cardiomyop-athy) were studied. Left ventricular (LV) volumes were assessed preand post CRT. AHR (LV-dP/dtmax) was assessed at implant using apressure wire in the LV cavity. The LV lead was placed in potentialtarget veins and the largest percentage rise in LV-dP/dtmax frombaseline (AAI or RV pacing with atrial fibrillation) to DDDLV wasused to determine optimal LV lead position. RR was defined asreduction in LV end systolic volume (ESV) $15% at 6 months.Results LV-dP/dtmax increased significantly from baseline(8016194 mm Hg/s to 9246203 mm Hg/s (p<0.001)) withDDDLV pacing for the optimal LV lead position. There was asignificant difference in the percentage rise in LV-dP/dtmax betweenthe best and worst LV lead position (Abstract 90 figure 1). LVESV


BCS Abstracts 2011

decreased from 186668 ml to 157668 ml (p<0.001). 18 (56%)patients exhibited RR. There was a significant relationship betweenpercentage rise in LV-dP/dtmax and RR for DDDLV pacing(p<0.001) (Abstract 90 figure 2). A similar relationship for AHR andRR in DCM and ICM (p¼0.01 & p¼0.006) was seen.



Conclusions Acute haemodynamic response to LV pacing is usefulfor predicting which patients are likely to remodel in response toCRT both for DCM and ICM. There is much variation in the rise inLV-dP/dtmax depending on LV lead position. Using acute haemo-dynamic response measured with a pressure wire during CRTimplant has the potential to guide LV lead positioning and improveresponse rates in the future.

91 RIGHT VENTRICULAR DYSFUNCTION IDENTIFIES CLINICALOUTCOMES FOLLOWING CARDIAC RESYNCHRONISATIONTHERAPY

doi:10.1136/heartjnl-2011-300198.91

1K Guha, 2F Alpendurada, 2S Prasad, 2T McDonagh, 1M R Cowie, 2R Sharma. 1RoyalBrompton Hospital, Imperial College, London, UK; 2Royal Brompton Hospital, London,UK

Background Cardiac resynchronisation therapy (CRT) is an estab-lished treatment for patients with advanced heart failure (HF).However, a proportion of patients do not derive benefit postimplantation of CRT. Despite an established predictive role in HF,the significance of RV dysfunction in gauging clinical benefit from

CRT has not been investigated. Cardiovascular magnetic resonance(CMR) is an important tool in the assessment of HF and isconsidered the gold-standard in estimating RV function. We usedthis technique to assess the impact of RV dysfunction on clinicaloutcomes following CRT implantation.Methods We evaluated 48 consecutive patients attending a heartfailure pacing clinic who had a CMR study within 6 months prior toCRT implantation. Clinical, biochemical, ECG and imaging datawere collected. Biventricular function and myocardial scar wereassessed by CMR including gadolinium enhancement. The primaryend-point was a composite of all cause mortality (ACM) orunplanned cardiovascular hospitalisation.Results The mean age was 64.5612.7 years. HF was ischaemic in42% of patients, and 85% were in NYHA class III/IV at the time ofimplantation. Atrial fibrillation/flutter was found in 27% ofpatients. The mean LVEF estimated by CMR was 2768%, while themedian RVEF was 52% (IQR 35%e63%). The mean tricuspidannular plane systolic excursion (TAPSE) was 14.066.0 mm, andthe mean pulmonary artery pressure (on echocardiography) was 37610 mm Hg. Ten patients (21%) met the primary end-point over amean follow-up of 28.6 months. On time-to-event analysis, onlyatrial fibrillation (HR 4.8, p¼0.02) and RV dysfunction, ie, reducedRVEF (HR 0.96 per 1% EF, p¼0.01) or TAPSE (HR 0.80 per mm,p<0.01) were independent predictors of the primary end-point.Atrial fibrillation and low RVEF were the only independent predic-tors of mortality (p¼0.03 and 0.04, respectively).Conclusions RV dysfunction is an independent predictor of adverseclinical outcomes following CRT. The assessment of RV functionmay be considered in patient selection for CRT implantation.

92 IDENTIFYING PATIENTS WITH CHRONIC HEART FAILURE: ACOMPARISON OF THE GOLD STANDARDS FRAMEWORKWITH A CLINICAL PROGNOSTIC MODEL

doi:10.1136/heartjnl-2011-300198.92

1K K Haga, 2S A Murray, 3J Reid, 3A Ness, 3M O’Donnell, 4D Yellowlees, 5M A Denvir.1University of Edinburgh, School of Medicine and Veterinary Medicine, Edinburgh, UK;2Primary Palliative Care Research Group, University of Edinburgh, Edinburgh, UK;3Heart Failure Nursing Service, NHS Lothian, Edinburgh, UK; 4University of Edinburgh,Edinburgh, UK; 5Department of Cardiology, University of Edinburgh, School of Medicineand Veterinary Medicine, Edinburgh, UK

Introduction Heart failure has a worse survival rate than manycommon cancers, yet few patients receive any palliative care inputduring the course of their illness. One of the main difficulties inproviding palliative care for heart failure patients is the uncertaintyaround the course of the disease and the patient9s life expectancy.The aim of this study was to compare the “Gold StandardsFramework” (GSF) criteria, which were developed to determine theneed for palliative care in non-cancer patients, with the “SeattleHeart Failure (SHF) Model”, which provides a method of calculatinga patient9s predicted mean life expectancy using physiological vari-ables.Methods Chronic heart failure patients, in NYHA class III or IV,who were being managed in the specialist, heart failure nursingservice, were identified from a clinical heart failure database. GSFcriteria were assessed by interviewing the specialist nurse respon-sible for each patient9s care. Clinical data required for the SHF modelwere obtained from two, online databases and were used to estimatemean life expectancy and predicted mortality at 1 year. Patientswere then followed up, at 1 year, to evaluate; 1) all cause mortality,2) place of death, and 3) the sensitivity and specificity of the GSFand SHF to predict death at 1 year.Results 138 NYHA III-IV patients were identified from a total of368 patients currently managed within the specialist nurse service;66% were male, and the mean age was 77 years. GSF criteria,


BCS Abstracts 2011

identified 119/138 (86%) patients that met the minimum require-ment for palliative care input. However, the SHF model predictedthat only 6/138 patients (4.3%) had a predicted life expectancy ofless than 1 year. Patients who met GSF criteria for palliative care hadsignificantly more hospital admissions (p¼0.001) and had signifi-cantly lower predicted survival rates at 1 year (p¼0.038) than thosepatients that did not meet GSF criteria. At follow-up, 43/138patients had died (31%). Of these, 58% (25/43) died in hospital,following an acute admission. The sensitivity and specificity for theGSF and SHF model were 22%/83% and 98%/12% respectively.Overall, the patients renal function (eGFR<35 ml/min) was the bestpredictor of mortality, (sensitivity/specificity¼82%/56%).Discussion Neither the GSF nor the SHF were very accurate inpredicting which patients were in the last year of life, in thisselected sample. Despite the increasing drive towards palliation inheart failure, clinicians are still faced with a substantial prognosticbarrier. Therefore, the progress of palliative care in heart failurepatients may require a shift away from the traditional “end of life”model developed in cancer treatment, and focus instead on apatient9s increasing needs coupled with an understanding thatdeath, itself, may remain unpredictable.

93 OPTIMAL MEDICAL THERAPY IN HEART FAILURE: IS THERESPACE FOR ADDITIONAL HEART RATE CONTROL?

doi:10.1136/heartjnl-2011-300198.93

1S Russell, 2M Oliver, 3H Rose, 2J Davies, 4H Llewellyn-Griffiths, 4A Raybould, 2V Sim,3Z R Yousef. 1Wales Heart Research Institute, Cardiff, UK; 2University Hospital Llan-dough, Cardiff, UK; 3University Hospital of Wales, Cardiff, UK; 4Hywel Dda HealthBoard, Camarthen, UK

Introduction Current evidence suggests that heart rate (HR) mayserve both as a modifiable risk factor, and as a disease modifyingvariable in patients with impaired left ventricular (LV) systolicfunction. The systolic heart failure (HF) treatment with If inhibitorivabradine trial (SHIFT) for example recently demonstrated signifi-cantly improved outcomes in otherwise optimally treated HFpatients following additional HR reduction with ivabradine. Wetherefore estimated the number of patients who after optimisationof conventional HF medications may be suitable for additional HRreduction.Methods We performed a retrospective analysis from two HF clinicswhere patients are referred for nurse lead, protocol-guided opti-misation of conventional HF therapies. Data on patient demo-graphics and classification of HF including; severity (ejectionfraction>35% vs ejection fraction#35%), functional limitation(New York Heart Association; NYHA class), and cause (ischaemic vsnon-ischaemic) were recorded. In addition, we collected data onpatient’s resting pulse (absolute value and rhythm: sinus vs atrialfibrillation), and blood pressure at the first and last clinic visits.Between the two clinic visits, patients underwent protocol-guidedforced up-titration of standard neurohormonal HF therapies. Wealso collected data on the maximal tolerated doses of beta blocker(bB), ACE inhibitor (ACE-I) or angiotensin receptor blocker (ARB),and the reasons for the inability to achieve target doses of bB.Results Of 172 consecutive patients referred for optimisation of HFtherapies (age 71613 yrs, 67% male), 71 (41%) were in atrial fibril-lation. Of the patients in sinus rhythm, 78% had severe LV systolicdysfunction (Abstract 93 figure 1). Overall, 145 of 172 patients(83%) tolerated bB therapy; of these, 39% achieved the targetmaximal dose, 57% at least half target dose, and 4% less than half ofthe target dose of bB. Reasons for failure to initiate bB (n¼27, 17%)included; severe and limiting hypotension (48%), intractable leth-argy (26%), and hospitalisation with worsening airways disease(26%). ACE-I/ARB, aldosterone antagonists, and digoxin were

tolerated in 92%, 30%, and 18% of patients respectively (Abstract 93table 1). Resting heart rate and blood pressure before and afteroptimisation of medical therapy are shown in Abstract 93 table 2.

Abstract 93 Figure 1 Heart Failure Patients Potentially Suitable forAdditional Heart Rate Reduction After Optimisation of Standard MedicalTherapy.

Abstract 93 Table 2 Haemodynamic profiles before and afteroptimisation of medication

First Clinic Visit(pre-optimisation)

Final Clinic Visit(post-optimisation) p value

Resting Heart Rate (beats/min) 73.8614.8 67.369.5 <0.001

Systolic Blood Pressure (mm Hg) 120619.6 115.1618.0 <0.001

Diastolic Blood Pressure (mm Hg) 71.7611.6 67.2610.4 <0.001

Conclusions Of 172 unselected patients attending HF clinics foroptimisation of medical therapy,w50% are in sinus rhythm with anejection fraction #35%. Despite forced optimisation of medicaltherapy, half of these patients have a resting heart rate $70 beats/minute. Overall, w1 in every 3 patients attending a heart failureclinic may be suitable for additional heart rate control.

Abstract 93 Table 1 Patient Characteristics (n¼172)

NYHA Class (%)

I 10.5

II 62.2

III 26.2

IV 1.1

HF aetiology (%)

Ischaemic 57

Non-ischaemic 43

LV function (%)

Ejection Fraction #35% 92.4

Ejection Fraction >35% 7.6

Cardiac rhythm (%)

Sinus 58.7

Atrial Fibrillation 41.3

Medication use (%)

b-blockers 83

ACE-I/ARBs 92

Aldosterone antagonists 30

Digoxin 18


BCS Abstracts 2011

94 A COMPARISON OF FUNCTIONAL ANDECHOCARDIOGRAPHIC OUTCOMES IN NICE COMPLIANTAND NON-COMPLIANT PATIENTS UNDERGOING CRT IN THEREAL WORLD

doi:10.1136/heartjnl-2011-300198.94

1S J Russell, 1I Rees, 2P O’Callaghan, 2Z R Yousef. 1Wales Heart Research Institute,Cardiff, UK; 2University Hospital of Wales, Cardiff, UK

Introduction The National Institute for Clinical Excellence (NICE)define a population of patients that are most likely to respond tocardiac resynchronisation therapy (CRT) and have a favourablehealth economic profile. Current NICE criteria (technologyappraisal; TA120) for CRT include: NYHA class III or IV symptomsdespite optimal medical therapy, sinus rhythm, ejection fraction#35%, and either QRS duration >150 mS alone or 120e149 mStogether with echocardiographic (echo) evidence of mechanicaldyssynchrony. Several randomised clinical trials however haveconsistently reported beneficial effects of CRT in patients outsidecurrent NICE guidelines. In our centre, potential CRT patients arediscussed at a multi-disciplinary team (MDT) meeting attended by aheart failure specialist, electrophysiologist, interventional cardiologist,cardiac surgeon and hospital manager. CRT is offered where there isconsensus agreement that the individual patient is likely to benefit.This individualised and evidence based approach provides for acomparison of outcomes in NICE compliant (NICE:+ve) and NICE:�ve patients (patients with a clinical need and evidence basesupporting CRT, but who do not meet NICE criteria).Methods Our unit operates an integrated CRT service with pre-assessment, implantation, and follow-up components. Pre-assess-ment includes clinical evaluation and baseline echo (EF: ejectionfraction, and ESV: left ventricular end-systolic volume) and func-

tional characterisation: a) Minnesota quality of life score (QoL), b)6 min walk test (6MWT), and c) peak oxygen consumption oncardiopulmonary exercise test (VO2). Follow-up at 3 and 6 monthspost CRT includes clinical evaluation, device/medical optimisation,and reassessment of echo and functional outcomes. This studyinvolves a retrospective analysis of our CRT database and comparesoutcomes in NICE:+ve and NICE: �ve patients.Results Between January 2007 and December 2009, 253 patientsreceived CRT. Complete paired data comparing baseline and6 month functional and echo data are available for 139 patients; 89NICE:+ve and 50 NICE: �ve (Abstract 94 table 1). Exclusions forthe NICE: �ve patients included: atrial fibrillation (n¼19), QRS120e149 mS without mechanical dyssynchrony (n¼12); QRS<120 mS (n¼5); pacemaker upgrades (n¼9). An additional 5patients with right bundle branch block and otherwise NICE CRTcompliance are analysed as NICE: �ve in this study. Compared tobaseline, 6-month outcomes were similar and significantly improvedin both NICE:+ve and NICE: �ve groups (Abstract 94 table 2).Conclusions We observed significantly favourable and similar func-tional and echocardiographic responses to CRT in patients meetingand not meeting current NICE criteria for CRT. Guidelines shouldguide therapy but ultimately each therapy should be individualisedand evidence based.

95 IMPAIRED CARDIAC ENERGETICS IN DILATEDCARDIOMYOPATHY: MAGNETIC RESONANCESPECTROSCOPY AT 3T

doi:10.1136/heartjnl-2011-300198.95

1R M Beadle, 2L K Williams, 2M Kuehl, 2S Bowater, 2K Abozguia, 2F Leyva Leon,1M P Frenneaux. 1University of Aberdeen, Aberdeen, UK; 2University of Birmingham,Birmingham, UK

Introduction The aim was to measure the cardiac phosphocreatine toATP (PCr/ATP) ratio non-invasively in patients with dilated cardi-omyopathy and normal controls, and to correlate the patient’sresults to symptom status, ejection fraction (EF) and quality of lifescores. Dilated cardiomyopathy is known to be associated withcardiac energy deficiency. Magnetic resonance spectroscopy (MRS)has been proposed as a non-invasive method of assessing cardiacenergetics and as a method of measuring response to therapy.Interrogation at high field strength improves signal to noise ratio.Methods 32 patients and 22 control subjects were studied usingphosphorus-31 (31P) MRS and patients were classified to NYHAsymptom class. In vivo energetics were measured using a commer-cially available Philips Achieva 3 Tesla scanner and dedicated 31P coilwith ISIS volume selection. Java Magnetic Resonance User Interface(jMRUI) was used for analysis. Furthermore, all patients completeda Minnesota Living with Heart Failure (MLWHF) score and under-went echocardiography. LVEF was measured using biplane Simp-son’s method.Results The PCr/ATP ratio was significantly reduced in patients(1.3560.31) compared with control subjects (1.9060.40; p<0.005).The PCr/ATP ratio was correlated with NHYA class (r¼�0.68,n¼32, p<0.0005). No correlation was found with LVEF or MLWHFscore.Conclusions This study confirms the presence of energy deficiency indilated cardiomyopathy as measured by MRS at 3T. The energystatus correlates strongly with symptom status but not with ejec-tion fraction nor quality of life score. Cardiac energetic status isdirectly proportional to symptoms status and therefore any treat-ments targeted to improve cardiac energetics may improve patientsymptoms in dilated cardiomyopathy.

Abstract 94 Table 1 Characteristics of NICE compliant and non-compliant patients

NICE:+ve NICE:Lve

Age: years (SD) 65 (11) 66 (11)

Male: % 83 90

Ejection fraction: % (SD) 22 (7.1) 24 (7.2)

QRS duration: mS (SD) 164 (26) 158 (37)

CRT-Defibrillator: % 57 46

Abstract 94 Table 2 Outcomes in NICE compliant and noncompliantpatients

NICE:+ve(n[89)

p value(baseline v6 months)

NICE:Lve(n[50)

p value(baseline v6 months)

QOL (score): baseline (SD) 58.2 (23.6) 63.5 (31.7)

QOL (score): 6 months (SD) 40.1 (25.0) <0.001 38.9 (25.6) <0.001

6MWT (m): baseline (SD) 216.4 (118.3) 208.5 (131.8)

6MWT (m): 6 months (SD) 324.5 (131.2) <0.001 291.5 (127.5) <0.01

V02 (ml/kg/min): baseline (SD) 13.2 (5.6) 12.3 (4.0)

V02 (ml/kg/min): 6 months (SD) 14.4 (3.3) 0.24 12.8 (4.2) 0.64

Ejection fraction- baseline (SD) 21.5 (7.1) 24.3 (7.2)

Ejection fraction- 6 months (SD) 30.7 (10.4) <0.001 31.4 (6.6) <0.001

End-systolic volume-baseline /ml (SD)

185.1 (79.3) 177.2 (67.2)

End-systolic volume-6 months /ml (SD)

144.6 (73.5) 0.02 135.2 (57.2) 0.01


BCS Abstracts 2011



96 A TEST TO CONFIRM MAXIMAL OXYGEN UPTAKE INCHRONIC HEART FAILURE PATIENTS WITHOUT THE NEEDFOR SECONDARY CRITERIA

doi:10.1136/heartjnl-2011-300198.96

T S Bowen, D T Cannon, G Begg, V Baliga, K K Witte, H B Rossiter. University ofLeeds, Leeds, UK

Cardiopulmonary exercise testing for peak oxygen uptake (VO2peak)is widely used to evaluate severity, pathophysiology and prognosis inpatients with chronic heart failure (CHF). AVO2peak#14 (or 12 withb-blocker) ml/kg/min is associated with increased mortality and is akey criterion for cardiac transplant listing. A symptom-limitedexercise test, however, may elicit a VO2peak lower than the maximumphysiological limit (VO2max); the latter commonly “confirmed” usingthe secondary criterion of respiratory exchange ratio (RER) >1.05.RER, however, is sensitive to the test format. We, therefore, deter-mined if a ramp-incremental (RI) step-exercise (SE) (or RISE) testcould determine VO2max in CHF patients without using RER, bysatisfying the criterion that two different work rates are terminatedat the same VO2peak. Twenty-one male CHF patients (NYHA class I:n¼3, II: n¼16, and III: n¼1) initially performed a modified Brucetreadmill test. Patients then completed a symptom-limited RISE95cycle ergometer test in the format: RI (4e18 W/min; w10 min);5-min recovery (10 W); SE (95% of peak RI work rate). Thirteen ofthese patients also performed RISE95 tests using slow (RI 3e8 W/min; w15 min) and fast (RI 10e30 W/min; w6 min) ramp rates.VO2 and RER were measured breath-by-breath by a mass spec-trometer and turbine (MSX, NSpire, UK). Peak VO2 and RER werecompared within-subjects, between RI and SE, by unpaired t test ofthe final 12 breaths of exercise. This approach allowed VO2max and itsassociated 95% confidence limits to be estimated. VO2peak wassimilar (p>0.05) in treadmill and cycle exercise (mean6SD: 16.262.7vs 15.063.2 ml/kg/min, n¼20, respectively), despite RER beinggreater in cycling (1.0860.12 vs 1.1560.09; p<0.05). As a group,VO2peak was similar (p>0.05) between RI and SE (mean6SD:14.663.2 vs 14.963.2 ml/kg/min, n¼21). A within-subject compar-ison, however, revealed that the VO2max criterion was met in 14 of 21patients (measurement sensitivity range 0.6e3.8 ml/kg/min),despite RER being >1.05 in the remaining 7 (1.1660.09). There wasno effect of ramp rate on VO2peak (p>0.05), however RERwas greater(p<0.05) in the fast ramp (1.2460.09) compared to the slow

(1.1260.06). The single-visit RISE95 test incorporating incremental-and step- exercise phases, each to the volitional limit, was welltolerated by CHF patients: The SE phase was contraindicated in only3 of the 47 tests. The RISE95 detected VO2max in 14 of 21 patientswith a sensitivity of w10% (ie, similar to healthy subjects), andwithout the need for secondary criteria or incidence of false-positive.In contrast, the end-exercise RER was sensitive to both modality andramp rate and provided a false-positive for VO2max attainment inevery incidence. Therefore, the RISE95 protocol provides a robustmeasure of VO2max in CHF patients, to within an individually-defined CI without dependence on secondary criteria.

97 INCREASING SKELETAL MUSCLE OXYGENATION BY PRIORMODERATE-INTENSITY EXERCISE INCREASES AEROBICENERGY PROVISION IN CHRONIC HEART FAILURE

doi:10.1136/heartjnl-2011-300198.97

T S Bowen, D T Cannon, S R Murgatroyd, K K Witte, H B Rossiter. University of Leeds,Leeds, UK

Rapid adaptation of pulmonary oxygen uptake (VO2) at exerciseonset reduces the reliance on limited anaerobic energy stores and isassociated with increased exercise tolerance. These VO2 kinetics,however, are slow in patients with chronic heart failure (CHF). Thiscould be due to limitations in the control of muscle O2 consumptionand/or O2 delivery. Recent evidence in CHF of a transient overshootin microvascular deoxygenation at exercise onset supports the latter.As prior exercise is known to increase muscle blood flow in healthyindividuals, we examined whether it could attenuate the fall inmicrovascular deoxygenation and speed VO2 kinetics on transition tomoderate exercise in CHF patients. Thirteen CHF patients (NYHAclass I: n¼3, II: n¼9, and III: n¼1) performed a ramp test on a cycleergometer for estimation of lactate threshold (LT) and VO2max.Patients subsequently repeated two 6-min moderate-intensity exercisetransitions (bout 1, bout 2) from rest to 90%LT, separated by 6-min ofrest. Measurements included breath-by-breath VO2 using a turbineand mass spectrometer (MSX, NSpire, UK), and tissue oxygenationindex (TOI) of the vastus lateralis by spatially resolved near-infraredspectroscopy (NIRO200, Hamamatsu, Japan). The exponential time-constant (s) for TOI and phase II VO2 were estimated using non-linear least-squares regression. The sVO2/sTOI, or “kinetic index”,was taken to reflect the relative matching of muscle oxygenation toits instantaneous requirement. LT and VO2max were 9.961.7(mean6SD) and 15.063.2 ml/kg/min, respectively. Prior exerciseincreased resting TOI by 1063% (p<0.05), attenuated the transientovershoot in muscle deoxygenation by w50% (p<0.05) and slowedthe rate of deoxygenation in the transient (sTOI: 1061 vs 21613 s;p<0.05). Both sVO2 (46620 vs 39618 s; p<0.05) and the kineticindex (4.561.8 vs 2.260.9; p<0.05) were reduced following priorexercise. sVO2 was well correlated to the kinetic index (R2¼0.92) inbout 1. However, although a lower sVO2 was typically reflected in areduced kinetic index in bout 2, VO2 kinetics remained slowed in 4patients. These patients had a higher NYHA class (2.360.5 vs1.660.5; p¼0.06) and greater initial sVO2 (62617 vs 3369 s; p<0.05)than the others. In CHF prior moderate-intensity exercise improvedthe dynamic matching of muscle oxygenation to its instantaneousrequirement and speeded VO2 kinetics in all patients. This suggeststhat slow VO2 kinetics in CHF are due, at least in part, to a dynamiclimitation in O2 delivery. However, this approach revealed an appa-rent limitation in the control of muscle O2 consumption in the mostsevere patients, which was only partly ameliorated by improving O2

delivery. Nevertheless, these findings suggest that an acute inter-vention to improve muscle oxygenation can increase aerobic energyprovision on transition to exercise in CHF patients.


BCS Abstracts 2011

98 HIGH PREVALENCE OF UNDIAGNOSED CARDIACDYSFUNCTION IN THE OLDEST OLD: FINDINGS FROM THENEWCASTLE 85+ STUDY

doi:10.1136/heartjnl-2011-300198.98

1F Yousaf, 1J Collerton, 2A Kenny, 1T Kirkwood, 1C Jagger, 1A Kingston, 3B Keavney.1Institute of Ageing and Health, Newcastle University, Newcastle upon Tyne, UK;2Freeman Hospital, Newcastle upon Tyne, UK; 3Institute of Human Genetics,Newcastle University, Newcastle upon Tyne, UK

Background Heart failure prevalence increases sharply at older ages.The section of the population aged 85 and over represents the mostrapidly increasing demographic worldwide. Previous epidemiologicalstudies of ventricular dysfunction and heart failure have includedonly small numbers of the “oldest old”, and have generally beenconducted in hospital-based settings, potentially introducing ascer-tainment biases. We conducted a community-based study of theoldest old using domiciliary echocardiography to estimate theprevalence of cardiac dysfunction and heart failure. Since in elderlypeople with multiple comorbidities, heart failure may morefrequently be incorrectly diagnosed, we cross-referenced our findingsto preceding diagnoses present in general practice records.Methods Four hundred and twenty-seven individuals aged86e89 years (mean age 87.9 years; 39.1% (n¼167) men, 60.9%(n¼260) women) were visited in their usual place of residence. A fullcardiovascular and medical history, including current medication,was taken; symptoms were graded using the NYHA classification.Previous diagnoses of heart failure (HF) were abstracted from the GPrecord. Participants underwent 2-D and Doppler echocardiography,including tissue Doppler measurements of LV long axis velocities,using a portable instrument (Vivid-I, GE Healthcare). LV systolicand diastolic dysfunction were graded according to AmericanSociety of Echocardiography guidelines.Results LV systolic function could be quantified in 93.2% (n¼398)participants anddiastolic function (classifiedasnormal,mild,moderateor severe dysfunction) in 88.1% (n¼376). 37.2% of participants(n¼140/376) had normal cardiac function or isolated mild diastolicdysfunction; 19.6% (n¼78/398) had moderate or severe LV systolicdysfunction, which was commoner in men (27.4%) than women(14.5%); and 14.4% (n¼54/376) had isolated moderate or severediastolic dysfunction. 65.1% (278/427) ofparticipantshadvaliddataonprevious diagnosis of HF, NYHA class and echocardiographic assess-ment of cardiac dysfunction. Of these, 37.4% (104/278) had clinicalevidence of HF, which was defined as NYHA class II, III, or IV symp-tomswithunderlying systolic dysfunction (29.5% (82/278)) or isolatedmoderate or severe diastolic dysfunction (7.9% (22/278)) on echo.Only7.6% (21/278)hadapreviousdiagnosis ofHF. 33.1%(n¼92/278) hadnoprevious diagnosis of HF but had clinical evidence of HF and an addi-tional 21.6% (n¼60/278) had no previous diagnosis but evidence ofpre-clinical HF (NYHA class I with systolic or moderate/severediastolic dysfunction). Of thosewith a previous diagnosis ofHF, 23.5%(n¼5/21) had no echocardiographic evidence of cardiac dysfunction.Conclusions Systolic and diastolic dysfunction and HF werecommoner in our population than previous studies in the “youngerold” have suggested. There are significant levels of both undiagnosedand misdiagnosed HF in this age group.

99 IS VO2MAX/KG A RELIABLE INDICATOR OF CARDIACDYSFUNCTION IN OVERWEIGHT HEART FAILUREPATIENTS?

doi:10.1136/heartjnl-2011-300198.99

S Chinnappa, N Lewis, D Barker, L B Tan. Leeds Teaching Hospitals NHS Trust, Leeds,UK

Background Peak O2 consumption (Vo2max) of #14 ml/kg/min hasbeen widely accepted as being indicative of poor cardiac function

warranting consideration for transplantation (Circulation 2010;122:173). We examined whether this variable is a good indicator ofcardiac function in overweight heart failure (HF) patients.Methods We compared the cardiopulmonary exercise performanceand non-invasive haemodynamics of overweight (BMI >34 kg/m2)and non-overweight (BMI #30) male heart failure patients inNYHA Classes II and III, with those of healthy male volunteerswith no known cardiovascular diseases (n¼101, age 43.2618.1(SD) years, BMI 26.063.1) as controls. Their physical and cardiacfunctional reserves were measured during treadmill exercise testingwith standard respiratory gas analyses and rebreathing method ofnon-invasively measuring cardiac outputs during peak exercise.Results Consecutive overweightHFwere screenedand24patients (age4968(SD) years, BMI 44.966.8, NYHA 2.5060.50) managed toexercise to acceptable cardiopulmonary limits (peak RER¼1.0160.12),and achieved Vo2max of 16.864.6 mls/kg/min which was significantlylower than controls (37.0610.7 mls/kg/min, p<10�6) and also lowerthan those of 30 non-overweight HF counterparts (20.063.7 mls/kg/min, p¼0.0019, age 49615 years, BMI 25.062.9, NYHA 2.4860.51).As shown in Abstract 99 figure 1, the overweight HF patients hadVo2max values which distributed around the 14 mls/kg/min cut-offvalue, and 9 of whomwere indeed below this cut-off value. However,the uncorrected Vo2max were higher than those of non-overweightcounterparts (Overweight: 25756748vs15946325mls/min,p<10�6),and its range of 1485e4210 mls/min overlapped with the range of1244e5774 mls/min in controls. The peak cardiac power output(CPOmax, 4.561.6 W, minimum 2.7 W) of overweight HF patientswere clearly above those of non-overweight (2.460.6 W, p<10�6,Abstract 99figure 2) and all above the transplant cut-off value of 1.5 W.




BCS Abstracts 2011

Conclusion These results indicate that in principle Vo2max in ml/kg/minas an indirect indicator of cardiac function or for cardiac trans-plantation selection is unreliable when applied to overweight heartfailure patients. Extending this concept to the entire spectrum ofbody weights, the practice of correcting Vo2max by body weight incardiological practice would also require urgent reconsideration.

100 PRESSURE VS FLOW AS A GUIDE FOR PACEMAKEROPTIMISATION? THE ACUTE HAEMODYNAMIC EFFECTS OFCHANGES TO ATRIOVENTRICULAR DELAY

doi:10.1136/heartjnl-2011-300198.100

C H Manisty, B Unsworth, R Baruah, P Pabari, Z I Whinnett, J Mayet, D P Francis.Imperial College, St. Marys Hospital, London, UK

Background Non-invasive blood pressure monitoring by continuousfinger photoplethysmography (Finometer) may have value in pacemaker

optimisation. However, the immediate increment in blood pressureseems to diminish somewhat in the initial minute: it is unclearwhether this is due to an (undesirable) fall in stroke volume or a(desirable) compensatory reduction in peripheral resistance.Methods and Results We studied this question by measuring beat-by-beat stroke volume (flow) using Doppler echocardiography, andblood pressure using continuous finger photoplethysmography,during and after atrioventricular delay adjustment from 40 to120 ms in 19 subjects with cardiac pacemakers. Quintuplicateexperimental runs were performed. Blood pressure and strokevolume (flow) both increased immediately (p<0.00001 within oneheartbeat). The immediate pressure increment correlated stronglywith the immediate flow increment (r¼0.74, p¼0.0001). Pressureshowed a partial decline a few seconds later (average rate 0.65 mmHg/beat, r¼e0.98, p<0.0001), in contrast, flow did not decline(p¼NS), Abstract 100 figure 1. Signal-to-noise ratio was significantlybetter for pressure than flow (6.363.6 vs 2.161.4, p<0.0001),Abstract 100 figure 2.Conclusions Improving atrioventricular delay immediately increasesblood pressure; however this effect decays slightly over the subse-quent minute. This is due to compensatory vasodilatation ratherthan a reduction in cardiac function. Pressure changes are simpler tomeasure and easier to distinguish from random variation thanDoppler measurements of flow, but are best measured immediately,before the vascular compensation.

101 WHAT DEGREE OF PULMONARY HYPERTENSIONPREDICTS POOR OUTCOME IN PATIENTS WITH LEFTVENTRICULAR SYSTOLIC DYSFUNCTION? A 10-YEARFOLLOW-UP STUDY

doi:10.1136/heartjnl-2011-300198.101

1B R Szwejkowski, 1D H J Elder, 1A M J Choy, 2S D Pringle, 1A D Struthers, 1C C Lang.1University of Dundee, Dundee, UK; 2Department of Cardiology, NHS Tayside, Dundee

Introduction The presence of pulmonary hypertension in leftventricular systolic dysfunction (LVSD) is an ominous sign. Itremains unclear the level at which pulmonary hypertension conveysa mortality risk in patients with LVSD.Methods We performed a record-linkage study in Tayside, UK(population approximately 400 000) utilising the Tayside echo-cardiogram database (>100 000 echo’s) maintained by the HealthInformatics Centre (HIC). Datasets from HIC include mortalitydata and other health care activities linked anonymously by thecommunity health index (CHI) number. Patients were included inthe analysis if they had LVSD and had a right ventricular systolicpressure (RVSP) measurement. Cox proportional hazards regressionanalysis was used to examine the effects of different ranges of RVSPmeasures on all cause mortality.Results 2910 patients (mean age, 74.5611.4 years; 43 % male) metentry criteria. Mean RVSP was 43.3 6 12.7 mm Hg and medianfollow was 362 days (IQR 129e850 days). There was a significantcorrelation between RVSP and survival (p<0.0001). In quartiles ofRVSP, the HR after adjustment for confounding factors includingLVSD and the presence of chronic obstructive pulmonary disease(COPD) were: RVSP 35e41 mm Hg, HR 1.12 (95% CI 0.95 to 1.32,p¼0.175), RVSP 42e50 mm Hg, 1.27 (1.07 to 1.49, p<0.001) andRVSP 51e106 mm Hg 1.62 (1.38 to 1.1, p<0.001). For each 5 mmHg stepwise increase in RVSP the HR for all cause mortality was1.07 (1.04 to 1.09, p<0.001). Abstract 101 figure 1 shows theKaplan-Meier survival curves for all cause mortality for all patientsexpressed as different RVSP quartiles.




BCS Abstracts 2011

Abstract 101 Figure 1 Survival of RVSP quartile.

Conclusion An RVSP of greater than 42 mm Hg is predictive ofincreased mortality in heart failure. This is finding is independent ofLVSD and COPD.

102 ETHNIC DIFFERENCES IN ENDOTHELIAL FUNCTION INCHRONIC HEART FAILURE

doi:10.1136/heartjnl-2011-300198.102

1E Shantsila, 2P S Gill, 3G Y H Lip. 1University of Birmingham Centre for CardiovascularSciences, City Hospital, Birmingham, UK; 2University of Birmingham, Primary Care andPopulational Sciences, Birmingham, UK; 3University of Birmingham Centre for Cardi-ovascular Science, Birmingham, UK

Background Endothelial dysfunction is characteristic of patientswith heart failure (HF) and is associated with an increased risk offuture cardiovascular events. However, data on ethnic differences inendothelial function in HF are scarce. In this study we aimed tocompare parameters of macro- and micro-vascular endothelialfunction and arterial elasticity in HF age- and sex-matched patientsof different ethnic origin: (i) white European, (ii) south Asian and(iii) African-Caribbean. Additionally, SA patients with systolic HFwere compared to two matched control groups: (i) south Asianpatients with coronary artery disease without HF(disease controls)and (ii) south Asian “healthy controls”.Methods We recruited 186 age/sex-matched patients with HF (ejectionfraction <40%) of SA (n¼43, age 66.5611.1 years), white (n¼44, age68.469.4 years) and African-Caribbean (n¼21, age 69.2610.3 years)origin; as well as 36 disease controls (age 64.0610.6 years) and40 healthy controls (n¼40, age 63.369.24 years). Macrovascularendothelial function was assessed as brachial artery flow mediateddilation in response to hyperaemia (FMD) and glyceryltrinitrate wereassessed by vascular ultrasonography (iE33, Philips, USA). Micro-vascular endothelial function was evaluated by laser Doppler flow-metry of forearm skin (DRT4, Moor Instruments, UK) afteriontophoresis of acetylcholine and sodium nitroprusside. Arterialstiffness was quantified by pulse wave velocity and augmentationindex using (Sphygmocor, Australia).Results Compared to disease controls and healthy controls southAsian patients with HF had impaired microvascular response toacetylcholine (3906302%, 5496264%, and 123695.5%, respectively,p<0.05) and reduced FMD (7.1263.64%, 11.864.66%, and4.8664.88%, respectively). HF patients of south Asian origin hadimpaired microvascular endothelial function (response toacetylcholine123695.5%) compared to white (258615.6%) andAfrican-Caribbean (286617.3%) groups (p>0.05). HF patients ofwhite origin had higher FMD than south Asian (4.8664.88%) andAfrican-Caribbean (5.3663.24%) patients (p<0.05). No difference in

glyceryltrinitrate- and sodium nitroprusside-mediated (endothelial-independent) response was observed between study groups. Insouth Asian subjects, parameters of pulse wave velocity andaugmentation index did not differ between those with HF and thosein control groups. No ethnic differences were detected in pulse wavevelocity. Conclusion: South Asian patients with HF have impairedmicro- and macro-vascular endothelial function, but preservedarterial elastic properties. Significant ethnic differences in endothe-lial function are present in patients with HF.

103 SENILE SYSTEMIC AMYLOIDOSIS: A COMMON CAUSE OFHEART FAILURE IN THE ELDERLY?

doi:10.1136/heartjnl-2011-300198.103

1J H Pinney, 2H J Lachmann, 2J D Gillmore, 2A Wechalekar, 3S D J Gibbs,3P Sattianayagam, 4,5S M Banypersad, 6,7J Dungu, 3N Wassef, 3C A McCarthy,3P N Hawkins, 3C J Whelan. 1National Amyloidosis Centre and UCL Centre forNephrology, UCL Division of Medicine, Royal Free Hospital, London, UK; 2NationalAmyloidosis Centre, UCL Division of Medicine, Royal Free Hospital, London, UK;3National Amyloidosis Centre, UCL Medical School, Royal Free Hospital, London, UK;4National Amyloidosis Centre, London, UK; 5The Heart Hospital, UCL Medical School,London, UK; 6National Amyloidosis Centre, UCL Medical School, University of London,London, UK; 7St George’s Hospital, University of London, London, UK

Senile systemic amyloidosis (SSA) is a rare cause of heart failure dueto the deposition of wildtype transthyretin. The clinical featuresand outcome are ill defined; our aim was to evaluate the naturalhistory of the disease in the UK in a group of thoroughly charac-terised patients. The series included all cases of biopsy proventransthyretin (TTR) amyloidosis with wildtype TTR genesequencing who were prospectively followed up between January2001 and May 2010. Clinical, biochemical, ECG and echocardio-graphic evaluation were performed at presentation to our centre.Patient survival was estimated using KaplaneMeier analysis. 55patients with histologically proven SSA; 36 (65.5%) from cardiac, 14(25.4%) from GI tract, 3 (5.5%) from bladder, 1 (1.8%) from fat and1 (1.8%) from carpal tunnel tissue were identified. 49 (89%) weremale. The median age at diagnosis and death were 74 (range 66e89)and 79 (range 69e84) years respectively. Survival from symptomonset and diagnosis was 7.04 (range 0.54e8.41) and 4.58 (range0.07e5.41) years respectively. In recent years more patients havebeen diagnosed with 2 (3.6%), 14 (25.5%) and 39 (70.9%) patientsbetween 2001e2003, 2004e2006 and 2007e2009 respectively. Themost common presentation was with breathlessness in 28 patients(51%). Twenty-four patients (43.6%) had prior carpal tunnel oper-ations. Twelve (21.8%) patients had a history of ischaemic heartdisease. Fifteen had had a coronary angiogram; 8 were reportedlynormal and 7 required intervention. Arrhythmias were common, 20patients (36.3%) had a history of atrial fibrillation and 6 (10.9) hadpacemakers in situ. ECG findings were; 24 (43.6%) in AF, 6 (10.9%)first degree block, 10 (18.2%) left bundle and 6 (10.9%) right bundlebranch block, 27 (49%) Twave changes, 11 (20%) <5 mm complexesin all inferior leads. Echocardiographic findings revealed the medianIVSd was 1.7 (range 1.1e2.5) cm, median E/A ratio was 2.7 (range0.79e5.4), E/E9 15.81 (range 7.5e41.1) and ejection fraction was45.5 (range 13e83)%. Blood results showed; the median baselineNT-proBNP was 356.1 (range 5e2611) and troponin T 0.03 (range0.01e0.28). Twenty-five patients had a troponin T >0.03 (45%). Tenpatients (18%) had a detectable paraprotein and 2 (3.6%) had bencejones proteins. SSA is present in >25% of the very elderly at postmortem but was rarely diagnosed during life. It is becoming morefrequently recognised perhaps due to widespread use of cardiac MRI.Most patients are male but women can be affected. A history ofcarpal tunnel syndrome is common. The diagnosis is often madeafter the onset of breathlessness. Systolic and diastolic dysfunction


BCS Abstracts 2011

can be seen on echocardiogram. A positive troponin is a commonfinding with a subsequent normal coronary angiogram. Incidentalparaproteins are prevalent in up to 8% of this population and it isimportant to obtain a tissue diagnosis to rule out AL amyloidosis.With supportive management medium term outcomes are good.

104 PROGNOSTIC UTILITY OF CALCULATED PLASMA VOLUMESTATUS IN CHRONIC HEART FAILURE

doi:10.1136/heartjnl-2011-300198.104

1H Z Ling, 1N Aung, 1,2J Flint, 1,2S Aggarwal, 1,2S Weissert, 1,2A Cheng, 3D P Francis,3J Mayet, 1,2M Thomas, 1,2S Woldman, 1,2,4D O Okonko. 1University College LondonHospital, London, UK; 2The Heart Hospital, London, UK; 3International Center forCirculatory Health, NHLI, Imperial College London, London, UK; 4NHLI Imperial CollegeLondon, London, UK

Background Plasma volume (PV) expansion is a hallmark feature ofworsening heart failure that is notoriously underestimated by clin-ical examination. While radioisotope assays optimally quantify PVstatus, numerous haemodialysis-based equations also exist for itsestimation. The prognostic utility of such formulas in chronic heartfailure (CHF) is unknown.Methods We analysed the relation between estimated PV status andmortality in 246outpatientswithCHF(mean (6SD)age 67613 years,NYHA class 261, LVEF 2868%). PV status was calculated (HakimRM, et al) by subtracting the patients actual PV ((1-haematocrit) 3(a + (b 3 weight)); a and b are gender-specific constants) from theirideal PV ((c3 weight); c¼gender-specific constant).Results Median (6IQR) PV status wasd2616550 ml with 78% and21% of patients having PV contraction and expansion, respectively.Patients with PVexcess had significantly higher creatinine and loweralbumin levels. Over a median follow-up of 13616 months, 36 (15%)patients died. PV status predicted mortality (HR 1.001, 95% CI 1.001to 1.002, p¼0.001) in a graded fashion (Abstract 104 figure 1A) and didso independently of NYHA class, LVEF, weight, haematocrit andcreatinine. A PV status#�178 ml optimally predicted survival (ROCAUC 0.68, p¼0.0007) and conferred a 75% reduced hazard for death(HR 0.16, 95% CI 0.07 to 0.37, p<0.0001, Abstract 104 figure 1B).


Conclusions Calculating plasma volume status in CHF patientsappears prognostically useful and suggests that dehydration is bettertolerated than volume excess in these individuals and that targetingtherapy to achieve a plasma volume status #178 ml might incre-ment survival.

105 CLINICAL AND ECHOCARDIOGRAPHIC DETERMINANTS OFN-TERMINAL PRO B-TYPE NATRIURETIC PEPTIDE LEVEL INPATIENTS WITH STABLE CHRONIC OBSTRUCTIVE AIRWAYSDISEASE: A PROSPECTIVE OBSERVATIONAL STUDY OF 140PATIENTS

doi:10.1136/heartjnl-2011-300198.105

1C P Gale, 2J White, 2A Hunter, 3J Owen, 4J Watson, 5I R Pearson, 4I Holbrook,6N Durham, 6M Pye. 1Division of Biostatistics, University of Leeds, Leeds, UK;

2Department of Respiratory Medicine, York Hospitals NHS Foundation Trust, York, UK;3Department of Echocardiography, York Hospitals NHS Foundation Trust, York, UK;4Department of Biochemistry, York Hospitals NHS Foundation Trust, York, UK; 5LeedsTeaching Hospitals, Leeds, UK; 6Department of Cardiology, York Hospitals NHSFoundation Trust, York, UK

Background Brain natriuretic peptides have been shown to bereliable indicators of left ventricular failure and markers of risk incardiac disease. However, patients with chronic obstructivepulmonary disease (COPD) are also known to have elevatedconcentrations of brain natriuretic peptides in the absence of overtcardiac disease, likely due to right ventricular strain. This has beenshown to have prognostic value and has a potential role in themanagement of the condition; for example, it has been suggestedthat it could be used to guide the initiation of non-invasiveventilation. The aim of this study was to identify clinical andechocardiographic determinants of the polypeptide N-terminal pro-Brain Natriuretic Peptide (NT pro-BNP) in patients with stableCOPD.Method Arterial blood gases, plasma NT pro-BNP and transthoracicechocardiographic parameters were studied in 140 patients withstable COPD attending a respiratory outpatient clinic.Results Of the 140 patients, 65 (46%) were male, 26 (19%) receivedhome oxygen therapy, 115 (82%) were current smokers, 38 (27%)were prescribed diuretics and 15 (11%) had a left ventricular ejectionfraction <45%. Patients with cor pulmonale (n¼6) were more likelyto have left ventricular systolic dysfunction (p<0.001), reducedtricuspid annular plane systolic excursion (p¼0.017) and higherpulmonary artery systolic pressures (p¼0.01). The median (IQR)NT pro-BNP concentration was 16.2 (25.4) pmol/l. Concentrationswere significantly higher in those with a dilated left atrium, aorticstenosis, left ventricular systolic dysfunction, right ventricularimpairment, atrial fibrillation and those prescribed diuretics andACE inhibitors. Significant predictors of NT pro-BNP were a dilatedleft atrium, aortic stenosis and left ventricular systolic dysfunction.NT Pro-BNP was an excellent discriminator of RV impairment (Cstatistic¼0.90).Conclusions NT pro-BNP readily identifies patients with stableCOPD who have right ventricular dysfunction. However, severalother clinical variables also associated with increased NT pro-BNPconcentrations are prevalent in this population. This is likely toconfound clinical decision making.

106 CHF PATIENTS ARE VITAMIN D DEFICIENT ANDHYPERPARATHYROID, WITH LEVELS OF EACH RELATED TOMARKERS OF SEVERITY

doi:10.1136/heartjnl-2011-300198.106

1G A Begg, 1L Kearney, 2A C Wheatcroft, 1R Byrom, 1S Barnes, 1J Gierula, 2J Barth,2R Cubbon, 2M T Kearney, 2K K Witte. 1Leeds General Infirmary, Leeds, UK; 2Universityof Leeds, Leeds, UK

Background The vitamin D-parathyroid (PTH) axis is increasinglyrecognised as potentially being involved with many of the featuresof the syndrome of CHF. We wanted to explore the relationshipbetween vitamin D and PTH levels in a group of CHF patients andrelate these to markers of severity.Methods We analysed serum 25(OH) vitamin D3 levels in 406consecutive attendees of the Leeds Advanced Heart Failure clinic(310 men) and correlated these to clinical markers of severity.Results Mean age (SE) was 69 (3) years, mean left ventricular ejec-tion fraction (LVEF) 31 (2)%, mean serum creatinine 117 mmol/l(2.4), median vitamin D levels (IQR) 30 (20e43) nmol/l (normal forskeletal health>75 nmol/l) and median parathyroid levels 8.8(6.2e13.5) pmol/l (normal<6.5 pmol/l). Aetiology was ischaemic


BCS Abstracts 2011

heart disease in 63% and 23% had diabetes mellitus. Patients wereoptimally treated (84% on b-blockers, 88% on ACE inhibitors, and46% on spironolactone). The mean daily dose of furosemide was 60(3) mg. Very few patients (5%) were sufficient in vitamin D. Patientswith worse symptoms as measured by NYHA status had lowervitamin D levels and higher PTH levels (Abstract 106 figures 1 and2). There was also a negative relationship between furosemide doseand vitamin D (Abstract 106 figure 3) and, in an unselected subset of160 patients (mean peak oxygen uptake (pVo2) 16.6 (0.5) ml/kg/min), there was a positive relationship between pVo2 and vitamin D(Abstract 106 figure 4). Patients with diabetes had lower vitamin Dlevels than non-diabetics (p<0.001) and there was a negativecorrelation between vitamin D and fasting glucose levels (r¼0.13;p¼0.02). There was no relationship between vitamin D levelsand age, calcium, creatinine or CRP, and no differences betweenthose patients taking and those not taking b-blockers and ACEinhibitors. In 8 unselected patients we found a negativerelationship between tumour necrosis factor-alpha (TNF-a) levelsand vitamin D (r¼0.62; p¼0.05). Although there was no relation-ship between vitamin D levels and baseline LVEF, in a subgroup of150 patients followed up one year after titration to optimal CHFtherapy, there was a significant positive relationship betweenchange in LV dimensions and vitamin D levels at the time of thebaseline scan (p<0.05).Conclusions The vitamin D-PTH axis is abnormal in CHF, related tothe severity of the condition. Our data suggest that reverseremodelling in response to optimal drug titration is greater in thosewith higher vitamin D levels. Whether vitamin D deficiency iscausally related to CHF remains unknown and requires a long-term,randomised, placebo-controlled study in CHF patients with efficacyand mechanistic outcomes, using a dose of vitamin D capable ofnormalising both vitamin D and PTH levels.





107 EXPANSION OF THE RED CELL DISTRIBUTION WIDTH ANDEVOLVING IRON DEFICIENCY AS PREDICTORS OF POOROUTCOME IN CHRONIC HEART FAILURE

doi:10.1136/heartjnl-2011-300198.107

1N Aung, 1H Z Ling, 1,2S Aggarwal, 1,2J Flint, 1,2S Weissert, 1,2A Cheng, 1T Richards,3D P Francis, 3J Mayet, 1,2M Thomas, 1,2,4D O Okonko. 1University College LondonHospital, London, UK; 2The Heart Hospital, London, UK; 3International Center forCirculatory Health, NHLI Imperial College London, London, UK; 4NHLI Imperial College,London, UK

Background Red cell distribution width (RDW) is a surrogate ofmany aberrations (inflammation, malnutrition, iron deficiency (ID))that may drive chronic heart failure (CHF) progression. While anelevated RDW and iron deficiency at baseline predict mortality inCHF, little is known about the prognostic implications of theirtemporal trends.Methods We analysed the relation of red cell indices on firstconsultation and over time with mortality in 274 outpatients withCHF (mean (6SD) age 70614 years, LVEF 2868%, NYHA class261, 54% ischaemic). The combination of a rising RDW and afalling mean cell volume (MCV) identified evolving ID.Results On initial consultation, an RDW >15%, Hb<12.5 g/dl, andMCV <80 fl were evident in 41%, 46%, and 8% of patients. Over amedian (6IQR) follow-up of 15617 months, 60 (22%) patientsdied. On Cox proportional hazards analyses, a higher RDW inde-pendently predicted increased mortality (HR 1.21, p<0.0001). Overtime, 51%, 58%, 40%, and 23% of patients had a rise in RDW, a fallin Hb, a fall in MCV, and evolving ID, respectively. A rising RDWpredicted death (HR 1.18, p¼0.002) independently of baselineRDWs and changes in Hb, with an absolute increase >1% conferringa twofold escalated risk of mortality (Abstract 107 figure 1A).Evolving ID was also associated with poorer survival (HR 2.89,p<0.0001, Abstract 107 figure 1B).


BCS Abstracts 2011


Conclusions An expanding RDW and evolving iron deficiency overtime predict an amplified risk of death in CHF and could be utilisedfor risk stratification or therapeutically targeted to improveoutcomes.

108 4D-FLOW CMR DEMONSTRATES THE REGIONALDISTRIBUTION OF AORTIC FLOW DISTURBANCE INMARFAN SYNDROME

doi:10.1136/heartjnl-2011-300198.108

1A Pitcher, 1T E Cassar, 1J Suttie, 1J M Francis, 2P Leeson, 3E Blair, 4B P Wordsworth,5J C Forfar, 6M Markl, 1S N Neubauer, 7S E Petersen. 1Oxford Centre for ClinicalMagnetic Resonance Imaging, Oxford, UK; 2Department of Cardiovascular Medicine,University of Oxford, Oxford, UK; 3Department of Clinical Genetics, Churchill Hospital,Oxford, UK; 4Nuffield Department of Medicine, University of Oxford, Oxford, UK; 5JohnRadcliffe Hospital, Oxford, UK; 6University Hospital, Freiburg, Germany; 7Centre forAdvanced Cardiovascular Imaging, William Harvey Research Institute, London, UK

Background Marfan syndrome (MFS) commonly leads to progres-sive aortic dilation, aneurysm formation and aortic dissection,particularly at the aortic sinuses (w60% of dissections), anddescending thoracic aorta (w30% of dissections). Abnormal aorticblood flow patterns may contribute to the enlargement anddissection of an inherently weak aorta, or to late complications afteraortic dissection.Methods 18 patients with MFS (3 with a prior history of aorticdissection and aortic root surgery, 15 with no such history) and 18healthy volunteers matched for age, sex and height underwent CMRat 3T, using a time-resolved 3-dimensional flow technique. The aortawas segmented into regions on the basis of anatomic features(Abstract 108 figure 1A). Each segment was visualised usingstreamlines (Abstract 108 figure 1B) and particle traces, and wasrated as normal or abnormal, (defined as the presence of turbulentflow vortices) and, where abnormal, extent of abnormality wasclassified on a 4-point scale determined by the extent of radialinvolvement of the aortic lumen. Wall shear stress (WSS) quantifi-cation was undertaken at predefined aortic locations (Abstract 108figure 1A).

Abstract 108 Figure 1 A. Planes for aortic segmentation and WSSquantification. B. Flow visualisation in a healthy volunteer. C. Flowvisualisation in a patient with prior aortic dissection fulfilling the GhentCriteria for Marfan syndrome.

Results Significant vortical flow in any segment (defined as flowdisturbance occupying more than one half of the aortic lumen) waspresent in all patients with MFS, but in only 7/18 controls(p<0.0005). The severity of flow disturbance was greater in MFSpatients than controls (median severity score 3 for Marfan patients,1 for controls, p<0.0005). There was marked regional variation inthe prevalence of major flow disturbance (Abstract 108 figure 2),with the sinuses of Valsalva and proximal descending aorta beingmost frequently affected. Prior repaired aortic dissection was asso-ciated with marked abnormalities of blood flow (Abstract 108 figure1C), with corresponding increases in axial WSS within the truelumen of the dissected aorta (typical axial WSS in the dissectedascending aorta was +0.9 N/m2, compared to +0.54 N/m2 inhealthy controls). Aortic flow disturbance in MFS was of one ofthree types: Type A: flow disturbance confined to the sinuses ofValsalva, Type B: flow disturbance confined to the proximaldescending aorta, Type C: flow disturbance in both the sinuses ofValsalva and the proximal descending aorta.

Abstract 108 Figure 2 Prevalence of vortical flow disturbanceoccupying >50% luminal diameter for each aortic region for Marfanpatients and controls.

Conclusion Patients with MFS commonly show aortic flowdisturbance. The sinuses of Valsalva and proximal descending aortaare most frequently affected. Flow disturbance can be categorisedinto one of three categories, and we anticipate that flow abnor-malities within a segment will predict progressive aortic dilation anddissection in an ongoing follow-up study.

109 3T MRI OF ACUTE ATHEROSCLEROTIC PLAQUE RUPTUREAND DOWNSTREAM EMBOLIC INJURY

doi:10.1136/heartjnl-2011-300198.109

A C Lindsay, L Biasiolli, J M Lee, I Kylintireas, H Watt, W Kuker, A Handa,M D Robson, S Neubauer, J Kennedy, R P Choudhury. University of Oxford, Oxford, UK

Introduction Luminal stenosis is a poor predictor of the risk posed byany given atherosclerotic plaque, therefore current angiographicimaging techniques cannot reliably determine which patients aremost likely to suffer future ischaemic events. However, MRI may beable to detect features of atherosclerotic plaque rupture that havebeen associated with an increased risk of recurrent athero-thrombosis.


BCS Abstracts 2011

Hypothesis 3T MRI of the carotid artery can identify atheroscleroticplaque rupture in patients presenting with TIA or minor stroke.Methods 81 patients with carotid artery disease were recruited; 41presented acutely with TIA or minor stroke and 40 asymptomaticpatients acted as the control group. Median time from symptomonset to MRI in the symptomatic group was 2.1 days (range0.17e7.0). All patients underwent T1, T2 and proton density-weighted turbo spin echo MRI to 10 mm either side of the carotid.As part of a combined scan protocol, study participants thenunderwent diffusion-weighted imaging (DWI) and Fluid-AttenuatedInversion Recovery (FLAIR) imaging of the brain to assess acute andchronic injury, respectively. If physically able, patients underwentfollow-up scanning a minimum of six weeks later. Plaques weregraded according to the MRI modified American Heart Association(AHA) system by two independent reviewers blinded to the clinicalstatus of the patient. Statistical analysis was performed using theWilcoxon sign rank test and Fisher9s exact test to compare plaques,in addition to the Mann Whitney U test to compare cerebral injury.Results AHAtypeVI (ruptured) plaquewas seen in 22/41(54%) in thesymptomatic group vs 8/41(20%) in the asymptomatic group(p<0.05), either due to intra-plaque haemorrhage (34% vs 18%,p¼0.08; Abstract 109 figure 1A), surface rupture (24% vs 5%, p¼0.03;Abstract 109 figure 1B), or luminal thrombus (7% vs 0%, p¼0.24;Abstract 109 figure 1C). Of particular note, 17/30 (57%) cases of AHAVI (ruptured) plaque were seen to cause <70% stenosis―the currentcut-off for surgical treatment. At follow-up scanning a minimum of6 weeks later, only two cases of AHA VI plaque showed evidence offull healing. Of the 41 patients in the acute group, evidence of cerebralinjury on DWI imaging was seen in 32/41 patients; the mediannumber of lesions per patient was 7 and the median total lesionvolume was 10.62 ml (range 0e522 ml). No significant associationswere noted between AHA plaque type and downstream cerebralinjury, however the presence of plaque surface rupture independentlypredicted a higher number of DWI lesions, a higher total DWI burdenat presentation, and higher total cerebral FLAIR signal at follow-upwhen compared to all other plaque types (p<0.05).


Conclusion Acute atherosclerotic plaque rupture can be visualisedusing 3T MRI. In particular, MRI can provide detailed informationon plaque morphology that can predict downstream embolic injury,independent of the degree of luminal stenosis caused.

110 MYOCARDIAL SYSTOLIC STRAIN AND SUBCLINICALATHEROSCLEROSIS IN YOUNG ADULT LIFE

doi:10.1136/heartjnl-2011-300198.110

A J Lewandowski, M Lazdam, E Davis, R Poole, J Diesch, J Francis, D Augustine,R Banerjee, J Suttie, S Neubauer, P Leeson. Cardiovascular Medicine, University ofOxford, Oxford, UK

Background In the elderly, reduced left ventricular function is relatedto elevated carotid intima media thickness (IMT), a well-establishedsubclinical marker of atherosclerosis. Cardiovascular magneticresonance (CMR) allows for precise quantification of changes inmyocardial structure and function. We therefore sought to deter-mine if in young adults, without overt cardiovascular risk factors,

there was already evidence of early changes in systolic functionrelated to subclinical atherosclerosis.Methods We studied 81 individuals (44 females, 37 males) withoutcardiovascular risk factors and with a mean age of 28.4265.36 years(mean6SD). Peak mid-ventricular myocardial circumferentialsystolic strain and left ventricular mass adjusted for body surfacearea (LVM) were assessed by CMR. Carotid IMTwas measured as amarker of subclinical atherosclerosis using ultrasound. Demographicand anthropometric characteristics were measured as well asmetabolic parameters and peripheral and central blood pressure.Results Individuals with reduced peak myocardial circumferentialsystolic strain had higher carotid IMT (r¼0.392, p<0.001). Totalcholesterol level and waist to hip ratio were both significantly asso-ciated with reduced myocardial strain. Increased LVM, central andperipheral systolic blood pressure, peripheral pulse pressure, glucose,triglycerides, age, body mass index and waist to hip ratio, as well asreduced high-density lipoprotein, were all significantly associatedwith increased carotid IMT (p<0.01). Males also had higher carotidIMT than females (mean6SD¼ 0.5460.068 mmvs 0.4760.042 mm,p<0.001). The association between carotid IMTand peak myocardialcircumferential systolic strain was independent of gender, smoking,LVM as well as peripheral and central blood pressure measures.Conclusions We have shown for the first time that subclinicalchanges in cardiac function and subclinical atherosclerosis areclosely interrelated in young adults, with associations that extend tothose in the normal range of cardiovascular risk. This study furtherestablishes the ability of CMR to detect early changes in cardio-vascular disease development.

111 SINGLE CENTRE PROSPECTIVE CARDIAC CT STUDY TODETERMINE THE PREVALENCE OF PATIENTS WITHCORONARY ARTERY DISEASE WITH A ZERO CORONARYARTERY CALCIUM SCORE AND ASSOCIATED NON-CARDIACINCIDENTAL FINDINGS

doi:10.1136/heartjnl-2011-300198.111

A J Shah, D R Obaid, D Gopalan, J Babar, J H F Rudd. Addenbrooke’s Hospital,Cambridge, UK

Introduction Cardiac CT, incorporating coronary artery calcium(CAC) scoring and angiography, is being increasingly used to eval-uate patients with chest pain and exclude coronary artery disease(CAD), as recommended in the recent NICE guidelines. Calcificationof the coronary arteries is an excellent marker of underlying athe-rosclerosis, but a zero CAC score does not completely exclude thediagnosis as potentially significant non-calcified plaques will not bedetected by CAC scoring. CT imaging may also identify non-cardiacincidental findings that can lead to further downstream testing andmedical expense.Objectives (1) To evaluate the probability of CAD in patients with aCAC score of zero. (2) To determine the incidence of non-cardiac inci-dental findings on cardiac CTand to quantify resulting investigations.Methods 116 symptomatic patients undergoing cardiacCT to excludeCAD from November 2009 to October 2010 were prospectivelyenrolled. Patients underwent CAC scoring and had contrast-enhanced, 128-slice, dual source CT coronary angiography (CTCA―Siemens Flash). Scanswere dual-reported by a cardiac radiologist and acardiologist. Statistical analysis was performed using GraphPadPrism.Results 62/116 patients had a CAC score of zero. Of these, 57 (91.9%)patients had normal coronary arteries, 4 (6.5%) patients had non-obstructive CAD (stenosis <50%), and 1 patient (1.6%) had signifi-cant obstructive CAD (stenosis>50%). This patient with obstructiveCAD had a high grade lesion in the proximal left anterior descendingartery that required intervention. 54/116 had non-zero CAC scores.Of these, 13 (24%) had obstructive CAD and 41 (76%) non-obstructive CAD. 42/116 (36%) patients had incidental findings on


BCS Abstracts 2011

cardiac CT that are summarised in Abstract 111 table 1. These inci-dental findings resulted in further investigations, documented inAbstract 111 table 2. The mean radiation dose (6 SEM) for CACscoring was 0.6160.03 mSv. The mean radiation dose (6 SEM) forsubsequent CTCA was 2.66 6 0.32 mSv in high pitch “flash” mode(n¼27), 5.8660.50 mSv in prospective mode (n¼64) and17.1561.68 mSv in the retrospective mode (n¼25).

Abstract 111 Table 1 Incidental findings on cardiac CT

Area Structure Incidental Finding n

Chest (n¼27) Lung parenchyma Nodule <1 cm 5

Emphysema 3

Atelectasis 6

Fibrosis 4

Tumour recurrence 1

Bronchiectasis 2

Pleura Effusion 2

Calcification 2

Lymph node Adenopathy 2

Abdomen (n¼7) Liver Cyst/Nodules 6

Adernal Adenoma/metastasis 1

Diaphragm (n¼5) Hiatus Hernia 5

Vasculature (n¼11) Aorta Dilatation 8

Aneurysm 1

Renal Stenosis 1

Coeliac Stenosis 1

Abstract 111 Table 2 Further investigation of incidental findings onCardiac CT

Investigation n

Bone scintigraphy 1

Chest clinic referral 2

CT chest 4

DMSA 1

MR adrenals 1

MRA renal 1

Nephrology clinic referral 1

Pleural fluid aspiration 1

Ultrasound kidneys 1

Ultrasound liver 3

Abstract 111 Table 3 Investigations and referrals generated byincidental findings

Investigations or referrals Number

Bone scintigraphy 1

Chest clinic referral 2

CT chest 4

DMSA 1

MR adrenals 1

MR cardiac 2

MRA renal 1

Nephrology clinic referral 1

Pleural fluid aspiration 1

Ultrasound kidneys 1

Ultrasound liver 3

Conclusions Despite 62patientshaving a reassuringCACscore of zero,8%of this grouphad evidence of non-calcified plaque,with one patienthaving obstructive CAD that required intervention. We conclude that

if strong clinical suspicion remains in patientswith aCACscore of zerofurther coronary investigation may be warranted. Incidental findingsare common, and can result in multiple further investigations forpatients. Further research is needed to evaluate the added cost, clinicalbenefits and radiation exposure created by investigation of such inci-dental findings in the context of cardiac CT.

112 COMPUTED TOMOGRAPHIC CORONARY ANGIOGRAPHY TOSCREEN FOR ALLOGRAFT VASCULOPATHY AFTER HEARTTRANSPLANTATION

doi:10.1136/heartjnl-2011-300198.112

M G Panicker, A G Mitchell, N R Banner, T K Mittal. Harefield Hospital, Harefield, UK

Objective To evaluate ComputedTomographic Coronary Angiog-raphy (CTA) as an alternative to Invasive Coronary Angiography(ICA) for the detection of Cardiac Allograft Vasculopathy (CAV).Background CAV is an important cause of late mortality after hearttransplantation (HT). Because patients are often asymptomatic,surveillance ICA is performed in our institution. CTA is effective forthe diagnosis of coronary disease in non-transplant patients, but fewstudies have been done after HT.Methods 117 HT patients, 1 to 24 years post transplant (mean¼12years SD6 6) underwent CTcoronary artery calcification (CTCAC)followed by retrospective ECG gated coronary angiogram on a 64-slice scanner without the use of any b-blockers. Majority (89%) ofpatients had CTA within 24 h before ICA. The Agatston calciumscore (CS) was calculated for all patients. The CTA images weresystematically analysed for image quality and the presence of CAV(graded as significant if >50% luminal stenosis) using a fifteencoronary segments model by an independent investigator blinded tothe results of ICA.Results CS ranged from 0 to 1681 (Mean¼91.76275). Out of 77patients with absent CS, 3 had significant CAV on ICA. Despite amean resting heart rate of 82 bpm SD613 and body mass index of27 kg/m2 SD65, 81% of the CTA images were graded as excellent orsatisfactory. For all the 1755 segments assessed by CTA irrespectiveof the image quality, CTA had sensitivity, specificity, positive andnegative predictive values of 71%, 79%, 72% and 78% respectively forthe detection of any CAV found by ICA. On a patient basis, CTA bestperformed in diagnosing CAV of more than 25% with sensitivity,specificity, positive and negative predictive values of 74%, 94%, 79%,and 92% respectively. None of the 61 patients with completelynormal CTA had CAV on ICA. 83 (92%) out of 90 patients whoresponded to a patient survey preferred CTA to ICA as a screeningtest for CAV. Non-coronary cardiac and non-cardiac abnormalitieswere identified in 18% and 14% patients respectively.Conclusion The study shows that CTA compares favourably withICA in detecting CAV in heart transplant recipients, and may be apreferable screening technique because of its non-invasive nature,patient preference and yield of additional information. One has toexercise caution in just using CS in these patients as significant CAVcan be missed out.

113 DUAL ENERGY CT IMPROVES DIFFERENTIATION OFCORONARY ATHEROSCLEROTIC PLAQUE COMPONENTSCOMPARED TO CONVENTIONAL SINGLE ENERGY CT

doi:10.1136/heartjnl-2011-300198.113

1D R Obaid, 1P A Calvert, 1J H F Rudd, 2D Gopalan, 1M R Bennett. 1University ofCambridge, Cambridge, UK; 2Papworth Hospital NHS Trust, Cambridge, UK

Introduction Vulnerable plaques have a relatively high necrotic corearea and low fibrous tissue content. Although CT can identifyplaque components on the basis of their x-ray attenuation, there is


BCS Abstracts 2011

significant overlap between their attenuation ranges, most cruciallybetween necrotic core and fibrous plaque. Recently introduced dualenergy CT (DECT) permits acquisition of 2 different energy datasets simultaneously, with the change in attenuation of plaquecomponents to different energies depending upon their materialcomposition. We therefore examined whether DECT was betterthan single energy CT in determining plaque components defined byvirtual histology IVUS.Methods 20 patients underwent DECT and 3-vessel VH-IVUS. CTdata was obtained at peak voltages of 100 kVand 140 kV. 52 plaqueswere chosen with either homogenous fibrous plaque or confluentareas of calcified plaque or necrotic core as defined by VH-IVUS. VH-IVUS images were co-registered and orientated with the corre-sponding CT images using distance from coronary ostia and fidu-ciary markers (Abstract 113 figure 1). Multiple regions of interest(ROI) were placed within the plaque components or in lumen oncross sectional CT images pre-classified by VH-IVUS (Abstract 113figure 1). ROI densities were measured (in Hounsfield Units) andassigned to the plaque component. A dual energy index (DEI) wascreated for each component, defined as the ratio of the difference inattenuation at 2 different energies / sum of attenuation with 1000added to each attenuation value to avoid negatives.

Abstract 113 Figure 1 Demonstration of plaque co-registrationbetween VH-IUS and 140kV/100kV CT data sets. Calcified plaque isidentified 5mm from side branch adjacent to characteristic calcification(yellow line). Cross section taken through this plaque (blue arrow) andfollowing orientation with VH-IVUS cross section HU region of interestsampling is performed in calcified plaque.

Results Attenuation values for 1088 ROIs were measured from 70paired data sets at 100 kV and 140 kV creating 70 DEIs (12 necroticcore, 11 fibrous plaque, 29 calcified plaques and 18 lumen). Valuesobtained using a single energy data set showed good differentiationbetween calcified plaque and all others (p<0.05), but considerableoverlap between necrotic core and fibrous plaque (p¼ns) (Abstract113 figure 2A) (Abstract 113 table 1). In DECT, lumen (iodinatedcontrast) showed the greatest change in attenuation and hence hadthe highest DEI. Necrotic core had the lowest DEI and could bedistinguished from all other components (p<000.1) Importantly, incontrast to the single energy data, DEI derived from both energy

data sets permitted resolution of necrotic core and fibrous plaquewithout overlap (Abstract 113 figure 2B).

Abstract 113 Figure 2 (A) Defined CT attenuation spectra ofplaque components using a single energy (140kV), calcified plaqueis distinguishable from all others but necrotic core and fibrousplaque overlap. (B) The use of dual energy index from the attenuationdata at 2 energies (100/140kV) allows significant separation ofnecrotic core and fibrous plaque (p<0.05) (Tukeys multiplecomparison test).

Abstract 113 Table 1

PlaqueComponent

100 kV meanHU (SD)

140 kV meanHU (SD)

Mean Difference(100e140 kV)

Dual EnergyIndex (mean)

Necrotic Core 57.26 (42.20) 42.69 (31.51) 14.57 0.0071

Fibrous Plaque 148.30 (49.47) 84.60 (30.34) 63.69 0.0283

Calcified Plaque 733.10 (226.7) 582.20 (194.9) 150.9 0.0450

Lumen 411.5 (82.27) 282.90 (55.93) 128.6 0.0483

Conclusions The additional attenuation data provided by DECTimproves the differentiation of plaque components when comparedto conventional single energy CT. In particular, DECT may allowbetter differentiation of necrotic core and fibrous plaque, a weaknessof conventional cardiac CT, allowing for more accurate non-invasiveidentification of vulnerable plaques.

114 RADIATION DOSES TRENDS FROM CARDIAC CT USING ACARDIAC SPECIFIC CONVERSION FACTOR: SYSTEMUNDERSTANDING & AN OPTIMISATION STRATEGYSIGNIFICANTLY REDUCES THE DOSE TO THE PATIENTS IN ACLINICAL SERVICE

doi:10.1136/heartjnl-2011-300198.114

1O E Gosling, 1S Iyengar, 1R Loader, 1G Morgan-Hughes, 2W D Strain, 3C Roobottom.1Plymouth Hospitals NHS Trust, Plymouth, UK; 2Peninsula College of Medicine andDentistry, Exeter, UK; 3Peninsula College of Medicine and Dentistry, Plymouth, UK

Background CT coronary angiography CTCA now has an estab-lished role in the investigation of patients with chest pain. Underthe IRMER regulations radiation doses to patients should be kept aslow as reasonably practical (ALARP). Previous publications haveused a chest conversion factor to calculate the effective dose (mSv)from CTCA. We have previously demonstrated that chest conver-sion factors significantly under-estimate the effective dose to thepatient when applied to CTCA and have calculated a cardiac specificconversion factor of 0.028 mSv (mGy.cm)-1. Our departmentfollows the ALARP ethos and has implemented new technologiestogether with physician training to reduce the radiation dose fromCTCA. We aimed to investigate what impact the implementation ofnew technologies has had on the radiation dose of CTCA.Method All patients who were coded as attending for a cardiac CTscan on the PACS and CRIS systems were included in the analysis.Scan indication included: rule out coronary artery disease, CABGassessment, pre-EP studies and problem solving. CT scanning


BCS Abstracts 2011

between September 2007 and August 2010 was included; the totaldose for the whole examination is used including the scout and non-enhanced scan (calcium score). Scans were performed on a Light-speed VCT or HD750 (GE Healthcare). To calculate the effectivedose a conversion factor was applied to the dose length product ofeach examination. The DLP is the radiation dose in one CT slicemultiplied by the length of the scan. A cardiac specific conversionfactor was used rather than a chest conversion factor (0.014) whichsignificantly underestimates the effective dose from CTCA. Datawas transformed and expressed as a geometric mean with 99% CI.For each analysis period all scans were included; retrospective,prospective, low kV and zero padding.Results In the 3-year period 1736 scans were performed. The meanradiation dose in the first 6 months of the study (retrospectivegating) was 29.6 mSv; using the accepted conversion factor at thetime the mean dose was 14.9 mSv. In March 2008 prospective ECGgating was installed; this resulted in a halving of the mean radiationdose to 13.6 mSv. In March 2009 the scanner parameters was set tozero padding and 100 KV reducing the dose to 7.4 mSv. For the final6 months the mean radiation dose for a cardiac scan was 5.9 mSv;this Abstract 114 figure 1 incorporates scans performed withstandard filtered back projection, iterative reconstruction, highdefinition scanning and retrospective ECG gating for a variety ofdiffering clinical scenarios.

Abstract 114 Figure 1 Effective dose (mSv) by protocol period.

Conclusion The introduction of dose saving strategies and appro-priate physician training has lead to a significant reduction in theradiation dose from cardiac CT. As CTCA programmes becomeestablished in hospitals around the UK it is important that clinicianshave the appropriate training and experience to keep the radiationdose to the patients as low as reasonably practical.


Scanning protocol

RetrospectiveGatingddosemodulation

Prospectivegating

Zero paddingd100 kV

Final 6months

Number of Patients 150 489 636 461

Mean EffectiveDose (mSv)

29.6 13.6 7.4 5.9

CIs (99%) (mSv) 33 14.9 8 6.5

26.6 12.5 6.8 5.3

115 ATRIAL HIGH RATE EPISODES AND ATRIAL FIBRILLATIONBURDEN: DO THEY HAVE SIMILAR ASSOCIATION WITHCARDIAC REMODELLING?

doi:10.1136/heartjnl-2011-300198.115

C W Khoo, S Krishnamoorthy, G Dwivedi, B Balakrishnan, H S Lim, G Y H Lip.University Department of Medicine Centre for Cardiovascular Sciences, City Hospital,Birmingham, UK

Background and Objectives Contemporary pacemaker devices allowquantification of atrial high-rate episodes (AHREs) and atrial fibril-lation burden (AFB) accurately. Cumulative ventricular pacing (Vp)is associated with development of atrial fibrillation, but it is notclear if AHREs and AFB share similar pathophysiologic associationswith left atrium (LA) and ventricle (LV) function and remodelling.Methods In total, 87 patients with dual-chamber pacemakerunderwent two-dimension (2D) and tissue Doppler imaging (TDI)echocardiography. LA volume (LAV) was evaluated by area-lengthmethod and indexed to body surface area. Septal A9 was used tomeasured regional LA function. LV systolic and diastolic parameterswere evaluated by mitral inflow velocity (E, A, E/A), LV ejectionfraction (biplane Simpson’s) and septal TDI velocity. The presenceof AHREs (defined by atrial-rate $220 beats/min and $5 minutes)and AFB were derived from pacemaker diagnostics. Plasma markersof remodelling, matrix metalloproteinases-1 (MMP1) and tissueinhibitors of metalloproteinases-1 (TIMP1), were analysed by ELISA.Results Baseline characteristics and comorbidities were comparablebetween groups (Abstract 115 table 1). Patients with AHREs hadsignificantly larger indexed LAV (p¼0.011) and higher cumulativeVp (p¼0.012), but this was not associated with elevation of MMP1and TIMP1. Plasma markers, LV systolic and diastolic parameterswere comparable between groups. In patients with AHREs, the AFBranged from 0 to 99% and correlated with E/A (r¼0.966, p<0.001),and inversely correlated with late acceleration velocity (A)(r¼�0.612, p¼0.009). On linear regression analysis, A, E/A, septal A9

were independently associated with AFB (all p<0.01).Conclusion Cumulative Vp and increased LAV are associated withthe development of AHRE, but AFB is independently associatedwith changes in LA function and LV diastolic function. This studysuggests AHREs and AFB have dissimilar pathophysiologic associa-tions with left atrium and ventricle remodelling.


No AHRE (n[70) AHRE (n[17) p Value

Age (years) 71.0611.6 75.468.8 0.1

Body mass index (kg/m2) 26.464.4 27.664.7 0.38

Indexed LA volume (ml/m2) 27.467.9 34.869.4 0.01

LV ejection fraction (%) 52.8611.9 55.169.2 0.4

E/A 0.860.2 1.060.6 0.23

Septal A’ (cm/s) 8.962.2 7.962.6 0.16

Septal S’ (cm/s) 6.661.8 6.561.4 0.71

Septal E/E’ 13.766.2 14.163.5 0.74

Percentage Vp 21.9 (1.8e99.0) 98.6 (41.0e99.9) 0.01

116 CRT OPTIMISATION: IMPROVING ECHOCARDIOGRAPHICTECHNIQUES BY ACCOMMODATING BIOLOGICALVARIABILITY WITHIN DIFFERENT ECHOCARDIOGRAPHICPARAMETERS

doi:10.1136/heartjnl-2011-300198.116

P A Pabari, A Kyriacou, M Moraldo, B Unsworth, N Sutaria, J Mayet, A D Hughes,D P Francis. Imperial College London, London, UK

Background In optimisation of CRT (and even selection for implan-tation) we may have underestimated the impact of beat-to-beat


BCS Abstracts 2011

variation on echocardiographic measurements. This can be quanti-fied most clearly in the optimisation process, in which genuine smallchanges in cardiac function (signal) must be detected amongpotentially large beat-to-beat variation (noise).Methods and Results In this large study of biological variability, weperformed over 2000 echocardigraphic measurements in 12 patients.We performed separate, replicate measurements at a series of inter-ventricular delays of each potential optimisation modality at rest.This included (i) 3D systolic dyssynchrony index, (ii) aortic pre-ejection time, (iii) interventricular mechanical delay, (iv) LVOT VTIand (v) QRS width. The equivalent of 31 optimisations per patientwere performed. For single measurements at each setting, agreementbetween successive optimisations was low, at 39% for SDI, 41% foraortic pre-ejection time, 32% for IVMD, 54% for LVOT VTI and58% for QRS width. Agreement between one method and another,using single replicates, was similarly low, with the average agree-ment between optima by two methods being only 18% similar topure guesswork. The intraclass correlation coefficient was low for allmethods at 0.11, 0.51, 0.30, 0.50 and 0.55 respectively. The intraclasscorrelation coefficients improved to 0.19, 0.63, 0.42, 0.54 and 0.66(p¼0.001) when averages of paired measurements were used. Tooptimise within 20 ms or 10 ms of the true optimum, requires agreater number of measurements, as seen in Abstract 116 figure 1,dependant on the intraclass correlation coefficient. The scatter ofoptima obtained reduced (improved) significantly when usingaveraged pairs of measurements compared to single measurementsfrom 23 ms to 18 ms (3D SDI), 14 ms to 10 ms (aortic pre-ejectiontime), 28 ms to 22 ms (IVMD), 21 ms to 16 ms (LVOT VTI) and14 ms to 10 ms QRS duration (p¼0.0002).


Conclusions Because of beat-to-beat variability, VV delay opti-misation by any of the echocardiographic techniques is not realisticunless multiple replicates are performed and averaged. Smoothingbiological variation by averaging multiple measurements allows thefull potential of echocardiographic optimisation to be achieved andimproves the consistency of optimisation. Trying to save time byperforming inadequate numbers of replicates is a false economy andleads to optimisation being a form of randomisation. These obser-vations may also cast light on to why attempts to identify futureresponders from CRT has not e when tested in externally moni-tored randomised trialsdbeen fruitful: dyssynchrony assessment toselect patients for implantation may need averaging too, and of farmore replicate measurements than is current practice. Integration ofthis biological insight into technological achievements of clinicalimaging is necessary, if reliable predictors of which patients willbenefit from CRT, are to be developed.

117 TRICUSPID VALVE ANNULAR DYNAMICS IN NORMAL VSDILATED RIGHT HEARTS; A 3D TOE STUDY

doi:10.1136/heartjnl-2011-300198.117

L Ring, B Rana, R A Rusk. Papworth Hospital NHS Foundation Trust, Cambridge, UK

Background The tricuspid valve annulus (TVA) is a complex threedimensional structure that is non-planar, and is incompletelyunderstood. The dynamics of the normal TVA has not beendescribed in any significant detail, nor has the impact of abnormalright hearts on the TVA been described. This study was designed toassess the feasibility of assessing the TVA throughout the cardiaccycle using 3D transoesophageal echo (TOE).Methods 20 patients were included, divided into 2 groups: normalright hearts (n¼10), and dilated right hearts (n¼10). 3D zoomimages of the TVA were acquired using an iE33 imaging platformand X7-2t transducer (Phillips, Andover, Massachusetts, USA).Antero-posterior (AP) diameter, septo-lateral (SL) diameter, area andheight were measured at 6 points of the cardiac cycle adaptingcommercially available software designed for assessing the mitralvalve (MVQ, Phillips). The eccentricity ratio was calculated asAP/SL.Results TVA area decreases during systole in both groups, and isgreatest in mid-diastole. The area is significantly larger in theabnormal group (mean 1795 mm2 abnormal vs 1204 mm2 normal;p<0.01). The SL diameter increased more in the abnormal group,resulting in a circular orifice and lower eccentricity ratio throughoutthe cycle (mean 0.91 abnormal v 1.22 normal; p<0.01, see graph).Annular height is similar in both groups but has an upward trend insystole in normals and reduces in abnormals, reaching significance atend systole (6.7 mm vs 4.9 mm; p¼0.046).Conclusions In patients with abnormal right hearts, the TVA dilatesin a septo-lateral direction, resulting in a more circular orifice. Thedynamic changes of the TVA are similar in dilated vs normal righthearts, with the exception of annular height. This pilot studysuggests that 3D TOE provides insight into understanding tricuspidannular dynamics.

Abstract 117 Figure 1 Eccentricity ratio of the tricuspid valve annulusduring the cardiac cycle: normal vs dilated rated hearts.


BCS Abstracts 2011

118 HIGH-RESOLUTION CARDIAC MAGNETIC RESONANCEPERFUSION IMAGING VS POSITRON EMISSIONTOMOGRAPHY FOR THE DETECTION AND LOCALISATION OFCORONARY ARTERY DISEASE

doi:10.1136/heartjnl-2011-300198.118

G D J Morton, M Ishida, A Chiribiri, A Schuster, S Baker, S Hussain, D Perera,M O’Doherty, S Barrington, E Nagel. King’s College London, London, UK

Background Non-invasive imaging has a key role in the detection ofcoronary artery disease (CAD). Its importance has been affirmed byrecent National Institute of Clinical Excellence (NICE) guidelines.Localisation of ischaemia to a coronary territory is also important inpatient management. Cardiac Magnetic Resonance (CMR) perfu-sion imaging is a well-established and radiation-free test for thesepurposes. However, there are few data comparing perfusion CMRwith Positron Emission Tomography (PET), which is widelyregarded as the non-invasive gold standard. Furthermore novel CMRmethods, including those based on k-t acceleration techniques, allowmyocardial perfusion imaging with unprecedented spatial reso-lution.Methods 31 patients with known or suspected CAD referred fordiagnostic x-ray coronary angiography (XCA) underwent both CMRand PET examinations. Both PET and CMR protocols includedadenosine stress and rest perfusion imaging. CMR perfusionimaging was performed at 1.5T with a k-t-accelerated steady-statefree-precession sequence. PET imaging was performed with 13N-Ammonia. The Abstract 118 figure 1 shows an example. Expertsblinded to the clinical data analysed the imaging data and expertsblinded to the imaging results visually analysed the XCA data. Asignificant coronary artery stenosis was defined as $70% reductionin diameter or a fractional flow reserve <0.8 where available.Sensitivity and specificity for PETand CMR vs invasive angiographywere calculated. Localisation of ischaemia was assessed in patientswith CAD by classifying myocardial territories as either supplied by,or remote from, a stenotic artery.


Results Patient characteristics are shown in the Abstract 118 table 1.Mean age 6 SD was 6469 years. One CMR examination was non-diagnostic. The interval between PETand CMR was 266 days (77%same day), between PET and XCA 22628 days and between CMRand XCA 22629 days. The prevalence of CAD was 81%. For thedetection of CAD PET sensitivity was 80% (95% CI 59% to 92%)and specificity was 67% (24% to 94%). CMR sensitivity was 83%(95% CI 62 to 95%) and specificity was also 83% (36% to 99%). Inpatients with CAD ischaemia was localised to 63% of the territories

supplied by stenotic arteries by PET and 76% by CMR. Remoteischaemia was detected in 24% of territories by PET and 16% byCMR.


CharacteristicNumber (percentage)of affected patients

Male 25 (81%)

Diabetes 12 (39%)

Previous PCI 10 (32%)

Hypertension 22 (71%)

Conclusions CMR is at least as accurate as PET for the diagnosis ofCAD and also for the localisation of ischaemia to coronary terri-tories. Relatively low numbers mean that CIs are wide and furtherwork is required. Using an anatomic test as the reference-standardfor functional tests has well-described limitations. Remoteischaemia is likely to occur for several reasons including under-estimation of disease severity at XCA, microvascular disease and alsofalse positive results.

119 CARDIOVASCULAR MAGNETIC RESONANCE IMAGING(CMR) DETECTS SUBCLINICAL CARDIOMYOPATHY INASYMPTOMATIC PATIENTS WITH LEFT BUNDLE BRANCHBLOCK (LBBB) AND NORMAL ECHOCARDIOGRAPHY

doi:10.1136/heartjnl-2011-300198.119

M Mahmod, T D Karamitsos, J J Suttie, S G Myerson, S Neubauer, J M Francis.University of Oxford Centre for Clinical Magnetic Resonance Research (OCMR), Oxford, UK

Introduction Asymptomatic left bundle branch block (LBBB) is acommon indication for referral for cardiovascular magnetic reso-nance (CMR) imaging. However, it is not known whether referralfor LBBB returns a high diagnostic yield. We evaluated the diagnosticvalue of CMR in these patients.Methods All clinical CMR referrals for LBBB from January 2005 toNovember 2010 were reviewed by two independent investigators.Only patients with asymptomatic LBBB and normal echocardio-grams (echos) who underwent complete CMR evaluation wereincluded in the study. Patients were excluded if they had cardiacsymptoms or known coronary artery disease. Anthropometric data,pre-existing conditions, medications, smoking status, family historyand echocardiographic data were recorded.Results From January 2005 to November 2010, 63 asymptomaticpatients with LBBB were referred to our institution for CMR from atotal of 3596 CMR referrals. Of these, 34 had normal echos; 20subjects who had abnormal echos and 9 who had no echos atpresentation were excluded from further analysis. Mean age of the34 patients with normal echos was 5469 years, and 19 (56%) weremen. Demographic data and left ventricular (LV) measurements arepresented in the Abstract 119 table 1. The most common associatedmedical conditions were hypertension (11 patientsd33%) andhyperlipidaemia (8 patientsd24%). Ten subjects (30%) had a familyhistory of heart disease. Nine (27%) patients underwent coronaryangiography which was normal. Of the 34 patients, 14 (41%) werefound to have pathological findings on CMR. The commonestabnormalities were dilated cardiomyopathy (DCM) (23%), followedby LV hypertrophy (LVHddefined as LV wall thickness >13 mm)(9%), arrhythmogenic right ventricular cardiomyopathy (ARVC)(6%) and Ebstein anomaly (3%). Two patients (6%) had mid walllate gadolinium enhancement. In the remaining 20 (59%) patients,no abnormalities on CMR were detected.


BCS Abstracts 2011


All patients(n[34)

Normal CMR(n[20)

Abnormal CMR(n[14) p value

Age (years (median, IQR)) 54.368.9* 57.5 (19.7) 48.5 (17.0) 0.6

Male gender (no, %) 19 (55.8%) 11 (55.0%) 8 (57.1%) 0.59

BMI (mean, kg/m2) 28.365.6 27.664.9 29.366.5 0.37

LVEDV (ml (median, IQR)) 155.0 (58.0) 133.0 (41.5) 182.5 (60.5) 0.012

LVESV (ml (median, IQR)) 51.0 (26.0) 48.0 (12.5) 71.5 (39.5) 0.005

LVEF (ml (mean, SD)) 60.6613.9 66.165.5 55.7613.6 0.004

LV thickness (mm (median, IQR)) 11.0 (7.4) 9.0 (6.1) 12.5 (9.4) 0.059

LVMI (g/m2 (median, IQR)) 72.5618.1* 64.0 (15.0) 83.0 (14.5) 0.001

*mean, SD. IQR.

Conclusions There is a high rate of sub-clinical cardiomyopathy(41%) detected by CMR in asymptomatic patients with LBBB despitenormal echocardiograms. These findings support the claim that CMRis a valuable adjunct to conventional investigations in asymptomaticLBBB. Further studies are needed to evaluate the prognostic impli-cations of CMR abnormalities in this cohort of patients.

Abstract 119 Figure 1 CMR findings in asymptomatic patients withLBBB and normal echocardiogram.

120 COMPARISON AND REPRODUCIBILITY OF STANDARD ANDHIGH TEMPORAL RESOLUTION MYOCARDIAL TISSUETAGGING IN PATIENTS WITH SEVERE AORTIC STENOSIS

doi:10.1136/heartjnl-2011-300198.120

1C D Steadman, 2N A Razvi, 1K I E Snell, 3J P A Kuijer, 3A C van Rossum,4G P McCann. 1Leicester Cardiovascular Biomedical Research Unit, Leicester, UK;2Department of Cardiovascular Sciences, University Hospitals of Leicester, Leicester,UK; 3Department of Physics and Medical Technology, ICaR-VU, VU University MedicalCenter, Amsterdam, The Netherlands; 4University Hospitals of Leicester, Leicester, UK

Objectives The aim of this study was to compare and assess thereproducibility of left ventricular (LV) circumferential peak systolicstrain (PeakEcc) and strain rate (SR) measurements using standardand high temporal resolution myocardial tissue tagging in patientswith severe aortic stenosis (AS).Background Myocardial tissue tagging with cardiac magnetic reso-nance (CMR) can be used to quantify strain and SR, however, thereare little data on the reproducibility. Diastolic SR may be ofparticular interest as it may be the most sensitive marker of diastolicdysfunction often occurring early in the course of disease.Methods Eight patients with isolated severe AS without obstructivecoronary artery disease were prospectively enrolled. They under-went CMR in a 1.5T scanner (Siemens Avanto) on two separateoccasions, median interval 12 days. Complementary tagged(CSPAMM) images were acquired with both a single breath-hold(SBH: temporal resolution 42 ms), and a multiple brief expirationbreath-hold (MBH: high temporal resolution 17 ms) sequence. Mid-wall PeakEcc was measured in the LV at mid-ventricular level withHARP Version 2.7 (Diagnosoft, USA). SR was calculated from thestrain data; SR¼Ecc2-Ecc1/Time2-Time1. PeakEcc, peak systolic anddiastolic SR were read from curves of strain and SR against time.

The MBH SR curves were filtered with a moving average (MA) toreduce noise sensitivity, results from a sample width of three andfive were examined. Differences between SBH and MBH wereassessed using Wilcoxon signed-rank test as not all measures werenormally distributed. Reproducibility assessments were carried outon all techniques.Results PeakEcc was significantly higher with MBH vs SBH, butreproducibility was slightly worse. Results are summarised inAbstract 120 table 1. Systolic SR was approximately equal with alltechniques although MBH using MA of five led to a borderlinesignificant reduction. Diastolic SR was higher when measured withMBH although only significant using MA of three. Systolic anddiastolic SR measures were more reproducible with MBH comparedwith SBH, except for the diastolic SR using MA of three, which wassubstantially worse. Strain and SR curves for the same patient areshown in Abstract 120 figure 1.


Peak systolicstrain (%)

Peak systolicstrain rate (1/s)

Peak diastolicstrain rate (1/s)

SBH e13.762.4 e0.7460.15 0.7560.27

MBH (MA of three) e15.163.1(p¼0.023 vs SBH)

e0.7360.11(p¼0.877 vs SBH)

1.1260.54(p¼0.017 vs SBH)

MBH (MA of five) e15.163.1(p¼0.023 vs SBH)

e0.6960.10(p¼0.049 vs SBH)

0.9160.36(p¼0.535 vs SBH)

SBH reproducibility(MD6SD; CoV; B-A)

0.5061.52; 11.1%;e2.5 to 3.5

e0.0160.13; 18.1%;e0.26 to 0.28

e0.0460.16; 21.0%;e0.36 to 0.27

MBH reproducibility(MA of three)(MD6SD; CoV; B-A)

1.1362.23; 14.7%;e3.3 to 5.6

0.0660.04; 5.3%;e0.02 to 0.14

e0.1360.44; 39.0%;e1.00 to 0.75

MBH reproducibility(MA of five)(MD6SD; CoV; B-A)

1.1362.23; 14.7%;e3.3 to 5.6

0.0460.05; 7.8%;e0.07 to 0.15

0.0960.15; 16.9%;e0.39 to 0.22

MD6SD¼meandifference 6 SD

CoV¼coefficientof variation

BeA¼BlandeAltman95% limits of agreement


Conclusions It is likely than SBH may be adequate or even superior toMBH for assessment of PeakEcc. The increased temporal resolution ofMBH may be advantageous for examining systolic and diastolic SR; aMA of five for diastolic SR may be the preferred method for quan-tification given the improved reproducibility of this measure.


BCS Abstracts 2011

121 INCIDENTAL EXTRA-CARDIAC FINDINGS ON CLINICALCMR; A COMPARISON OF 3 HASTE TECHNIQUES

doi:10.1136/heartjnl-2011-300198.121

1R B Irwin, 2T Newton, 3C Peebles, 4A Borg, 5D Clark, 4C Miller, 6N Abidin,4M Greaves, 4M Schmitt. 1Wythenshawe Hospital, Manchester, UK; 2Royal BlackburnInfirmary, Blackburn, UK; 3Southampton General Hospital, Southampton, UK;4Wythenshawe Hospital, University Hospitals of South Manchester NHS Trust,Manchester, UK; 5Alliance Medical, Wythenshawe Hospital CME unit, Manchester,UK; 6Salford Royal Hospital, Salford, UK

Introduction Cardiac magnetic resonance (CMR) is an increasinglyimportant imaging modality, which by necessity incorporates alarge field of view. Both “localiser” and multiple slice half-fourierspin echo (eg, HASTE) sequences provide coverage of the thorax andupper abdomen. Such imaging may reveal hitherto unexpectedincidental extra-cardiac findings (IEF). First we sought to assess thefrequency of IEF found on clinically indicated CMR scans. Secondwe compared the 3 clinically used HASTE acquisition protocols inthis context. Lastly we determined the impact of the 3 differentprotocols on acquisition time and image quality.Methods Three subsequent groups of 238 patients (714 patients intotal), all referred for clinically indicated CMR, were scanned witheither breath-hold (BH) HASTE (Group 1), free breathing (FB)HASTE (Group 2) or diaphragmatic navigated (NAV) HASTE(Group 3). Additionally “localiser” sequences performed in 3 orthog-onal planes were analysed. All 714 clinical reports were reviewedregarding the presence of IEF. These were categorised as either minor,or major if recommendations for further investigation, follow-up,and/or clinical correlation were made. Finally, to determine theimpact of each HASTE protocol on acquisition time and imagequality, an additional cohort of 15 patients underwent 3 protocolsback to back in a random fashion. The length of each acquisition wastimed and image quality was reviewed and scored externally.Results A total of 180 IEF were found in 162 (22.7%) out of 714patients. There was no significant difference in frequency of IEFbetween the 3 HASTE groups. Out of 180 IEF, 88 were consideredminor and 92 major findings. Of the latter, 8 (1.1%) were consideredhighly significant. These included one bronchoalveolar carcinomastage 1B requiring lobectomy, 2 cases of florid sarcoidosis in patientspresenting with VT and “structurally normal hearts” on echo-cardiography, one case of pulmonary aspergillosis, 2 cases ofadvanced pulmonary fibrosis, one ascending thoracic aorticaneurysm and a case of iatrogenic liver haemorrhage followingplacement of a pericardial drain. FB HASTE acquisition (6962.5 s)was significantly faster than BH (10563.8 s) and NAV (12162.7 s),p<0.001, but also produced the lowest image quality on a 5 pointscale; 3.5 (FB) vs 3.9 (BH) vs 3.8 (NAV), p¼0.08.Conclusion Overall, IEF are common and lead to follow on inves-tigations in a substantial minority of cases. However, the overallincidence of highly significant findings in the current study was low(w1%). There was no difference in the frequency of incidental extra-cardiac findings between the 3 HASTE protocols. While the freebreathing HASTE technique is statistically significantly faster thanbreath hold and navigated HASTE, the absolute time saving is smalland probably out-weighted by lesser image quality.

122 OBESITY AND PERIVASCULAR ADIPOSITY INATHEROSCLEROSIS

doi:10.1136/heartjnl-2011-300198.122

I Kylintireas, C Shirodaria, O Rider, J M Lee, I Bechar, J Digby, M D Robson,S Neubuer, R P Choudhury. University of Oxford, Oxford, UK

Introduction It has been proposed that perivascular adipose tissue(PVAT) contributes to inflammation and advancement of athero-

sclerosis via a direct paracrine or vasocrine route. Excess adiposetissue accumulation leads to adipose tissue dysfunction charac-terised by a pro-inflammatory and potentially pro-atherogenicpattern of adipokine secretion. We used MRI for PVAT imaging andquantification and evaluated the effects of obesity and increasedperivascular adiposity on the relationship of PVATwith the functionand structure of the underlying vessels.Methods We measured peri-aortic fat, aortic stiffness and atheromaburden by MRI in 128 cardiovascular patients and in 18 healthy leansubjects at baseline and in 22 healthy obese subjects (before andafter weight loss intervention (diet or bariatric surgery). Fat aroundthe brachial artery and FMD of the brachial artery was measuredamong 75 cardiovascular patients.Results There was good inter-observer and intra-observer reprodu-cibility (coefficient of variance (CV) <6% and<5%) and inter-scanrepeatability (CV<8%) of the measurement of PVAT. After adjust-ment for anthropometric indices, demographics and cardiovascularrisk factors as appropriate: I) A positive independent associationbetween PVAT and aortic atheroma was detected among obeseparticipants (BMI$30 kg/m2)(p<0.005) but not among individualswith intermediate (BMI<30 and $26) and low BMI (BMI<26). II)Perivascular fat was independently, inversely associated with aorticstiffness among lean patients (p<0.0005) while the association wasindependent and positive for obese participants (p<0.05). III) Anindependent, negative linear correlation between peri-brachial fat andFMD was noted among overweight and obese subjects (BMI$26)(p<0.001), but not among normal weight participants (BMI<26).IV) PVATwas an independent negative predictor of aortic elasticityamong healthy obese individuals (BMI$30) (p<0.01) while it waspositively and independently associated with aortic elasticity amonglean healthy controls (BMI#18) (p<0.05). V) Following weight lossintervention, PVAT reduction was an independent predictor of aorticelasticity improvement in the obese group (p<0.05).Conclusions Our results suggest an influence of both generalised andregional excess adiposity on the functional state and the effects ofperivascular adipose tissue on dysfunction and remodelling of theunderlying vessels.

123 CARDIOVASCULAR RISK IN ASYMPTOMATIC POTENTIALSIMULTANEOUS PANCREAS-KIDNEY TRANSPLANTRECIPIENTS IS DETERMINED BY MYOCARDIAL PERFUSIONSCINTIGRAPHY

doi:10.1136/heartjnl-2011-300198.123

V M S Stoll, N S Sabharwal, O O Ormerod. The John Radcliffe Hospital, Oxford, UK

Introduction More than 50% of renal transplant recipients will die asa consequence of cardiovascular disease (CVD). Type I diabeticsundergoing simultaneous pancreas-kidney transplantation (SPK) areat an even greater risk of CVD. Optimising a patient9s cardiovas-cular status is necessary before SPK transplant surgery. Patients canremain on transplant waiting lists for years. There is little evidenceas to how frequently repeat cardiovascular risk assessments arerequired in asymptomatic patients. Myocardial perfusion scintig-raphy is used in SPK patients to detect any asymptomatic myocar-dial ischaemia or abnormal left ventricular function. This studyanalyses data from a SPK transplant centre with an annual surveil-lance programme to aim to establish the suitable frequency of MPS.Methods Potential SPK transplant recipients who had undergonetwo perfusion scans were included for analysis. An abnormal MPSwas defined as either showing a regional wall motion abnormality,inducible ischaemia, or impaired left ventricular function. The scanresults were both documented and compared. Angiography resultsfrom the study period were also recorded.Results 99 out of 130 patients on the SPK waiting list in November2009 had undergone two perfusion scans as part of their


BCS Abstracts 2011

pre-transplant assessment. The median age was 45 yrs (range26e63), with 41% female and a median time between scans of1.4 yrs (range 0.6e3.0). 59 patients (60%) had two consecutivenormal scans. The remaining 40 patients had at least one abnormalscan. 16% of patients with a normal 1st scan developed an abnormal2nd scan within a median period of 1.4 years. 28 (70%) of thepatients with an abnormal MPS underwent angiography, of these 12required revascularisation (either PCI or CABG). Of the remaining16 patients; 1 died before angiography and the other 15 patientswere treated with medical therapy. Of the 59 patients with twonormal scans; 3 underwent angiography during the study period (fornew symptoms), 1 of these patients required revascularisation afterpresenting with an ACS. 2 had minor plaque disease.Conclusions 40% of SPK patients on the waiting list have anabnormal MPS. Of the patients with normal scans 5% required anangiogram because of new symptoms with only 2% requiringrevascularisation. Of the patients undergoing angiography driven byMPS 43% subsequently underwent revascularisation. The currentscreening interval is successfully monitoring changes in the patients’cardiovascular status with only one patient requiring an inter-vention which was not predicted by MPS. Therefore a near annualMPS is a useful, non-invasive means by which to monitor patientsat very high risk of asymptomatic cardiovascular disease whileawaiting a SPK transplant.

124 VALIDATION OF THE BCIS-1 MYOCARDIAL JEOPARDYSCORE USING CARDIAC MRI

doi:10.1136/heartjnl-2011-300198.124

1G D J Morton, 2K De Silva, 1M Ishida, 1A Chiribiri, 1A Indermuhle, 1A Schuster,2S Redwood, 1E Nagel, 1D Perera. 1King’s College London, London, UK; 2Guy’s and StThomas’ NHS Foundation Trust, London, UK

Introduction The recently described angiographic BCIS-1 MyocardialJeopardy Score (BCIS JS) was designed to classify the extent ofcoronary artery disease (CAD). It provides a semi quantitativeestimate of the amount of myocardium at risk as a result of severecoronary stenoses (0¼no jeopardy; 12¼maximum jeopardy).Advantages include ease of use and universal applicability includingclassification of left main stem disease and CABG. Howeveranatomic tests, including the BCIS JS, do not incorporate myocar-dial ischaemia and scar, which are important for management andprognosis. Cardiac magnetic resonance (CMR) imaging allows reli-able assessments of myocardial ischaemia and scar in a singleexamination and was used to examine the functional relevance ofthe BCIS JS.Methods 60 consecutive patients with angina and known orsuspected CAD referred for diagnostic x-ray coronary angiographyunderwent CMR examination at a single UK centre. CMR includedstandard functional and scar imaging and also high-resolution k-taccelerated adenosine stress and rest perfusion imaging at 1.5T (40patients) or 3T (20 patients). Expert observers blinded to the clinicaldata analysed the angiographic and CMR data. The BCIS JS wascalculated from visual analysis of the coronary angiogram. CMRperfusion and scar data were segmented according to the standard17-segment model excluding the apex. Segments were subdividedinto equal endo- and epicardial sub-segments, each assigned 3% ofthe total myocardial volume and classified as normal, ischaemia orscar. Myocardial ischaemia and scar burden were calculated andcorrelated with the BCIS JS individually and as a combined score(Abstract 124 figure 1).


Results Patient characteristics are summarised in the Abstract 124table 1. 2 patients were excluded (1 claustrophobia; 1 incompleteimaging data). Mean interval 6 SD between CMR and coronaryangiography was 40647 days. 13 patients (22%) with no history ofmyocardial infarction had CMR evidence of prior infarction. Therewas a strong correlation between the BCIS JS and myocardialischaemic burden: Pearson’s r¼0.75, p<0.00001 (Abstract 124 figure2). The BCIS JS was also correlated with the combined burden ofscar and ischaemia: r ¼ 0.77, p<0.00001. There was no differencebetween 3T and 1.5T CMR imaging. Area under the receiver-oper-ating characteristic curve for BCIS JS to detect $10% myocardialischaemic burden was 0.87 (95% CI 0.77 to 0.97). BCIS JS $6predicted $10% myocardial ischaemic burden with sensitivity 68%and specificity 90%.

Abstract 124 Figure 2 Correlation between myocardial ischemicburden and BCIS JS.

Conclusions The BCIS JS correlated well with ischaemic burden onCMR. A BCIS JS $6 predicts the prognostically importantischaemic threshold of 10% with high specificity. As expected, thecorrelation is imperfect which is likely to be a result of difficultypredicting haemodynamic effects of angiographically moderatedisease, microvascular disease and limitations of CMR imaging.


BCS Abstracts 2011


Characteristic Number of patients

Age (mean6SD) 65610

Left ventricular ejection fraction (mean6standard deviation) 59614%

Male 48 (83%)

Diabetes 17 (29%)

Previous CABG 13 (22%)

Previous percutaneous coronary intervention 22 (38%)

Previous MI 10 (17%)

Hypertension 38 (66%)

125 ASSESSING PATIENT BENEFIT FROM THEREVASCULARISATION OF CHRONICALLY OCCLUDEDCORONARY ARTERIES BY ADVANCED CARDIOVASCULARMRI TECHNIQUES

doi:10.1136/heartjnl-2011-300198.125

1N J Artis, 2A Crean, 1A Zaman, 1S Sorbron, 1A N Mather, 1S G Ball, 1S Plein,1J P Greenwood. 1University of Leeds, Leeds, UK; 2Toronto General Hospital, Toronto,Canada

Background Cardiovascular magnetic resonance (CMR) imaging canprovide an array of information about cardiac function and anatomy.The utility of CMR in the setting of coronary artery chronic totalocclusion (CTO) has not been fully investigated. We set out toexamine the ability of CMR to show regional improvements in leftventricular (LV) function and perfusion and to investigate if anyfeatureswere able to predict those that benefit from revascularisation.Methods Twenty-seven patients with single vessel CTO wererecruited from clinical waiting lists and underwent a comprehensiveCMR assessment prior to and 6 months following attempted CTOrevascularisation. A multi-parametric CMR protocol was performedwhich included cine imaging to assess regional wall thickness/thickening and global LV function, rest and adenosine stress perfu-sion imaging (Fermi model), low dose dobutamine stress to assessinotropic reserve, and late gadolinium enhancement (LGE) imagingto determine scar location and extent. Using the AHA 16 segmentmodel only segments supplied by the CTO artery were studied forfunctional improvement. Data are presented as mean (SD).Results Procedural success in terms of revascularisation of theoccluded artery was achieved in 23 of the 27 patients (85%, 20 with

PCI and 3 with CABG). In those with successful revascularisationby PCI LV volumes reduced (EDV 185 (54) vs 174 (50) p<0.05; ESV85(60) vs 77(58) p<0.001) and the left ventricular ejection fractionimproved (56.5(12)% vs 58.9(12)% p¼0.01). During adenosine stressimaging there was a significant improvement in absolute myocardialblood flow in the revascularised segments (from 1.87(0.51) to 3.77(0.67) ml/g/min p<0.001) but not in the remote regions (from 3.76(0.52) to 3.95(0.58) ml/g/min p¼ns). LGE was only present in 25(20%) revascularised segments. In these segments there was a stronginverse correlation between the extent of scar and improvement insegmental systolic thickening (r¼�0.736, p<0.001). There was aweaker association between the segmental response to low dosedobutamine and the degree of functional improvement followingsuccessful revascularisation (Pearson r¼0.249, p<0.01).Conclusion Following revascularisation of CTO, myocardial perfu-sion increases and both regional and global systolic functionimproves. While the majority of subjects in this study had no scaron LGE imaging, when segments are scarred there is a negativecorrelation with improvement in regional systolic thickening.

Abstract 125 Figure 2 Relationship of improvement in segmentalsystolic thickening against segmental scar (top panel) and change inthickening with low dose dobutamine (bottom panel).

126 THE IMPACT OF NICE GUIDELINES FOR THE INVESTIGATIONOF CHEST PAIN ON OUTPATIENT CARDIOLOGY SERVICES INTHE UK

doi:10.1136/heartjnl-2011-300198.126

1C Patterson, 2E Nicol, 3L Bryan, 4T Woodcock, 1S Padley, 1D Bell. 1Imperial College,London, UK; 2Royal Brompton Hospital, London, UK; 3Chelsea and WestminsterHospital, London, UK; 4NIHR CLAHRC for Northwest London, London, UK

Abstract 125 Figure 1 Changes observed in the CTO and a remoteterritory following attempted revascularisation. (non-revasculariseddblack lines). Adenosine stress increases perfusion only in the CTOterritory with no change in resting perfusion.


BCS Abstracts 2011

Introduction The National Institute for Health and Clinical Excellence(NICE) have released guidelines for the investigation of chest pain ofrecent onset (1). There is concern that the guidelines will increasethe burden on cardiac imaging, requiring service reconfiguration andinvestment (2, 3). This study was performed to assess the impact ofthe guidelines on outpatient cardiology services in the UK.Methods 595 consecutive patients attending chest pain clinics attwo hospitals over six months preceding release of the NICEguidelines (51% male; median age 55 yrs (range 22e94 yrs)) wererisk stratified using NICE criteria. Preliminary cardiac investigationsrecommended by NICE were compared with existing clinical prac-tice and the relative costs calculated.Results NICE would have recommended 443 patients (74%) fordischarge without cardiac investigation, 10 (2%) for cardiaccomputed tomography (CCT), 69 (12%) for functional cardiacimaging and 73 (12%) for invasive coronary angiography (ICA).Relative to existing practice there would have been a trend towardsreduced functional cardiac imaging (�24%; p¼0.06) and increasedCCT (+43%; p¼0.436) but a significant increase in ICA (+508%;p<0.001). The cost of investigations recommended by NICE wouldhave been £15 881 greater than existing practice.Conclusions This study suggests implementation of the NICE guide-lines will require investment in cardiology services, particularly ICA. Itwill be necessary to establish and maintain CCT for relatively fewpatients; also to establish and maintain functional cardiac imagingeven though referrals are likely to decline. Individual hospitals shouldassess their local populations prior to service reconfiguration.

Abstract 126 Table 1 Preliminary cardiac investigations undertaken(pre-NICE) compared with those recommended by NICE (N¼595)

Pre-NICE NICE % change

No investigation 33 443 +1242% (p<0.001)

Cardiac CT 7 10 +43% (p 0.436)

Functional cardiac assessment 91 69 �24% (p 0.06)

Invasive angiography 12 73 +508 (p<0.001)

127 TIMING OF CARDIOVASCULAR MRI AFTER ACUTEMYOCARDIAL INFARCTION: EFFECT ON ESTIMATES OFINFARCT CHARACTERISTICS AND PREDICTION OF LATEVENTRICULAR REMODELLING

doi:10.1136/heartjnl-2011-300198.127

A N Mather, T A Fairbairn, N J Artis, J P Greenwood, S Plein. University of Leeds,Leeds, UK

Background The pathophysiological remodelling processes asso-ciated with acute myocardial infarction (AMI) evolve over time andthe optimal acute imaging time point to predict medium-termsurrogates for outcome has not been established. This study aimed todefine the evolution of infarct characteristics by cardiovascularmagnetic resonance (CMR), and to assess whether CMR dataacquired at “day 2” or at “1 week” post-AMI are stronger predictors ofinfarct size and left ventricular (LV) function measured at 3 months.Methods Fifty-seven patients were recruited with first presentationST elevation AMI treated successfully with primary percutaneouscoronary intervention. Cine, T2- weighted and late gadoliniumenhancement CMR imaging were performed at days 2, 7, 30 and 90after index presentation.Results Infarct size and extent of myocardial oedema decreasedsignificantly between “day 2” and “1 week” (mean %LV-scar (SD)27.2 (13.9) vs 21.6 (14.1), p<0.001 and %LV-AAR (Area At Risk)(SD), 37.9 (15.2) vs 32.3 (14.3), p¼0.003). These changes wereaccompanied by a significant improvement in LV ejection fraction(%LVEF (SD), 41.7 (9.6) vs 44.6 (10.1), p<0.001). CMR data

acquired at “1 week” were better predictors of LVEF and infarct sizeat “3 months” than data collected at “day 2”.Conclusions The extent of myocardial oedema and infarct sizedecrease significantly during the first week after reperfusion for AMIand these changes are associated with a significant improvement inLVEF over the same interval. These findings have implications forthe timing of CMR studies in the early post-infarct period. Wefound that the percentage myocardial salvage index did not changesignificantly between “day 2” and “1 week”. Therefore, accurateassessment of the efficacy of reperfusion therapy can be made up toone week after revascularization. In addition, CMR data acquired at“1 week” were better predictors of CMR endpoints measured at“3 months”. Thus, we conclude that the optimal time point toimage patients post-reperfusion therapy for AMI is at 1 week.

128 BRIGHT BLOOD T2 WEIGHTED MRI HAS HIGHERDIAGNOSTIC PRECISION AND ACCURACY THAN DARKBLOOD STIR MRI FOR ASSESSMENT OF THE ISCHAEMICAREA-AT-RISK AND MYOCARDIAL SALVAGE IN ACUTEMYOCARDIAL INFARCTION

doi:10.1136/heartjnl-2011-300198.128

1A R Payne, 1M Casey, 1J McClure, 2R McGeoch, 2A Murphy, 2R Woodward, 2A Saul,2J Gilchrist, 2C Clark, 2K G Oldroyd, 1N Tzemos, 1C Berry. 1University of Glasgow,Glasgow, UK; 2Golden Jubilee National Hospital, Glasgow, UK

Background T2-weightedMRI revealsmyocardial oedemaand enablesestimation of the ischaemic area-at-risk and myocardial salvage inpatients with acute myocardial infarction (MI). We compared thediagnostic accuracy of a new bright blood T2-weighted with astandard black blood T2-weighted MRI in patients with acute MI.Methods A breath hold bright blood T2-weighted ACUTE pulsesequence with normalisation for coil sensitivity and a breath hold T2dark blood short s inversion recovery (STIR) sequence were used todepict the area-at-risk in 54 consecutive acute MI patients. Infarctsize was measured on gadolinium late contrast enhancement images.Results Compared with dark blood T2-weighted MRI, consensusagreements between independent observers for identification ofmyocardial oedemawere higher with bright blood T2 -weightedMRIwhen evaluated per patient (p<0.001) and per segment of leftventricle (p<0.001). Compared to bright blood T2-weighted MRI,dark blood T2-weighted MRI under-estimated the area-at-riskcompared to infarct size (p<0.001). The 95% limits of agreement forinter-observer agreements for the ischaemic area-at-risk and myocar-dial salvage were wider with dark blood T2-weighted MRI than withbright blood T2-weighted MRI. Bright blood enabled more accurateidentification of the culprit coronary arterywith correct identificationin 94% of cases compared to 61% for dark blood (p<0.001).Conclusion Bright blood T2-weighted MRI has higher diagnosticaccuracy than dark blood T2-weighted MRI. Additionally, darkblood T2-weighted MRI may underestimate area-at-risk andmyocardial salvage.

129 MYOCARDIAL SALVAGE DURING PRIMARY PCI CAN BEPREDICTED IN THE CATH LAB

doi:10.1136/heartjnl-2011-300198.129

1A R Payne, 1C Berry, 1O Doolin, 2M B McEntegart, 2R Woodward, 2A Saul, 2S D Robb,2M C Petrie, 1I Ford, 2K G Oldroyd. 1University of Glasgow, Glasgow, UK; 2GoldenJubilee National Hospital, Glasgow, UK

Objectives This study investigated the relationship between theindex of microcirculatory resistance (IMR) and myocardial salvageas determined by T2-weighted and contrast-enhanced cardiacmagnetic resonance (CMR) imaging in patients undergoing primarypercutaneous coronary intervention (pPCI) for ST elevationmyocardial infarction (STEMI).


BCS Abstracts 2011

Background IMR is a simple invasive measure of microvascularfunction available at the time of pPCI. T2-weighted non-contrastCMR can reveal myocardial oedema, and in the post-infarct popu-lation this represents the ischaemic area at risk (AAR). Contrast-enhanced CMR delineates the area of myocardial infarction. Thevolume of myocardium within the AAR, but not contained withinthe infarct area is salvaged myocardium.Methods 108 patients with STEMI underwent invasive coronaryphysiology measurements during pPCI and had a subsequent CMRscan at a median of 19 h post pPCI. Short axis non-contrast T2-weighted images were acquired and delayed enhancement imagingwas performed following administration of intravenous gadolinium(0.1 mmol/kg). AAR was determined and myocardial salvage wascalculated as AARdinfarct area.Results IMR was 29 (21), AAR 32% (13%) and myocardial salvage6% (9%)dall mean (SD). Spearman rank correlation between IMRand AAR was 0.27 (p 0.02) and between IMR and salvage was �0.31(p 0.01). IMR was also a multivariate predictor of AAR (p 0.01) anda negative multivariate predictor of myocardial salvage (p 0.02).Conclusions IMR measured acutely correlates with AAR and corre-lates negatively with myocardial salvage as determined by MRI.

130 COMPARISON OF HARMONIC PHASE IMAGING WITHLOCAL SINE WAVE MODELLING FOR THE ASSESSMENT OFCIRCUMFERENTIAL MYOCARDIAL STRAIN USING TAGGEDCARDIOVASCULAR MAGNETIC RESONANCE IMAGES

doi:10.1136/heartjnl-2011-300198.130

1A N Borg, 1C A Miller, 2C D Steadman, 2G P McCann, 1M Schmitt. 1UniversityHospital of South Manchester, Manchester; 2NIHR Leicester Cardiovascular BiomedicalResearch Unit, Leicester, UK

Introduction Assessment of myocardial strain promises to become animportant quantitative tool in early diagnosis of cardiac disease andtreatment monitoring. Advances in image processing software havefacilitated rapid and clinically feasible analysis of strain from taggedcardiac magnetic resonance (CMR) images. Harmonic Phase Anal-ysis (HARP) or Local Sine Wave Modelling (SinMod) can be used forautomated derivation of strain. We obtained tagged CMR images tocompare measurements of left ventricular (LV) circumferentialstrain obtained using a HARP with a SinMod method.Methods Ten normal controls, 10 hypertrophic and 10 dilatedcardiomyopathy patients (mean age 46.6614.8 years) wereincluded. Spatial modulation of magnetisation using short-axis LVslices at mid-ventricular level, with a temporal resolution of30e50 mS, were obtained using a 1.5 Tesla scanner (SiemensAvanto) with a 32-channel coil. Global and segmental transmuralpeak circumferential strains (ecc) were measured using HARP(Diagnosoft, USA, version 2.7) and SinMod (InTag, University ofLyons, France, version 3.6.1). Prior to running the algorithm, bothmethods involve manual tracing of the endocardial and epicardialborders, and localisation of right ventricle-to-septum insertionpoints, in one frame. Agreement between HARP and SinMod wasassessed by Spearman’s correlation co-efficient R and Bland Altmanmethods. Repeated measurements were carried out on 10 randomlyselected scans to assess reproducibility.Results There was a high level of agreement between HARP andSinMod for global ecc (HARPdSinMod mean difference: �0.12%,95% limits of agreement: �5.69% to 5.45%, R¼0.83, p<0.001)(Abstract 130 figure 1). Agreement was much lower for segmentalecc, ranging from very poor in lateral segments to modest in infer-oseptal segments (Abstract 130 table 1). Analysis time using SinModwas significantly shorter than for HARP (84642 vs 2016120 S,p¼0.02). Inter- and intra-observer reproducibility were extremelyhigh for SinMod measurements of global ecc (inter-observer R¼0.99,

repeatability co-efficient (RC) 2.14; intra-observer R¼0.99, RC 1.49).Reproducibility of global ecc measurements by HARP was some-what lower, but still high (inter-observer R¼0.89, RC 4.80; intra-observer R¼0.98, RC 2.73). There was much greater variability insegmental ecc measurements using both methods, particularly withHARP (Abstract 130 figure 2).



Analysed segmentfor circumferentialstrain

HARP vs SinModMean Difference± SD (%)

HARP vs SinMod95% Limits ofagreement (%)

HARP vsSinModCorrelationCoefficient

p-value forcorrelation

Anterior �1.6866.38 �14.18 to 10.82 0.59 0.001

Anterolateral �3.1868.07 �18.99 to 12.64 0.22 0.25

Inferolateral 1.4868.24 �14.67 to 17.62 0.24 0.21

Inferior 1.3366.17 �10.76 to 13.42 0.48 0.008

Inferoseptal �1.6665.63 �12.68 to 9.37 0.59 0.001

Anteroseptal �2.4766.77 �15.73 to 10.79 0.52 0.007

All 6 segmentspooled

�0.9967.11 �14.92 to 12.95 0.43 <0.001

Abstract 130 Figure 2 Inter- and intra observer variability for HARPlocal sine wave modelling: repeatability co-efficients.


BCS Abstracts 2011

Conclusions HARP and SinMod methods show a high level ofagreement for assessment of global mid-ventricular transmuralcircumferential strain, with good reproducibility for both individualmethods. Agreement is much lower for segmental measurements;poor reproducibility for segmental measurements using both tech-niques probably reflect user variability in identification of rightventricular septal insertion points and contour tracing.

131 AETIOLOGICAL ROLE OF FOLATE DEFICIENCY INCONGENITAL HEART DISEASE: EVIDENCE FROMMENDELIAN RANDOMISATION AND META-ANALYSIS

doi:10.1136/heartjnl-2011-300198.131

V Mamasoula, T Pierscionek, D Hall, J Palomino-Doza, A Topf, T Rahman, J Goodship,B Keavney. Institute of Human Genetics, Newcastle upon Tyne, UK

Background The existence of a causal relationship between lowerlevels of plasma folate and congenital heart disease (CHD) remainscontentious. Randomised trials of this question are not possible, inview of the known protective effect of folate against neural tubedefects (NTDs). Folate fortification of flour is known to reduce theincidence of NTDs, but it is not known whether there is any effecton CHD. Clarity regarding the relationship between folate andCHD could potentially inform the practice of folate fortification. Wepresent a genetic approach using “Mendelian randomisation” todetermining the causality of folate in CHD risk.Methods We compared genotype frequencies at the methylenetetrahydrofolate reductase (MTHFR) C677T single nucleotidepolymorphism (SNP) in 1186 CHD cases and 4168 controls. The TTgenotype at MTHFR C677T is known to be associated with loweractivity of MTHFR and plasma folate, and higher levels of plasmahomocysteine. The effect of TT genotype on plasma folate levels isgreater in conditions of folate deficiency. Thus, if lower plasmafolate had a causal effect on CHD risk, a higher frequency of TTgenotype among CHD cases than among healthy controls wouldbe anticipated, and this would be expected to be more markedin conditions of folate deficiency. We placed our results in thecontext of a meta-analysis of all previously published studies of thisquestion (to September 2010), which together included 1883 casesand 3069 controls in 25 studies. Thus, the combined analysesincluded 3069 CHD cases and 7271 controls. We used random-effects models to combine the data. We conducted sensitivityanalyses to examine folate fortification of flour as a potential sourceof heterogeneity.Results The primary genotyping data in 1186 cases and 4168controls revealed a trend towards increased risk with the TTgenotype, but this did not reach statistical significance (OR 1.15(95% CI 0.94 to 1.40)). Combination of our primary data withprevious studies, however, revealed association in the larger dataset(OR 1.45 (95% CI 1.12 to 1.89); p¼0.005). The population attrib-utable fraction for the TT genotype was 3% of CHD. There was noevidence of publication bias among the contributing studies. Wediscovered folate fortification status to be a significant source ofheterogeneity. Studies conducted in countries with mandatoryfolate fortification showed no effect of C677T genotype on CHDrisk (OR 0.96 (95% CI 0.64 to 1.44)), whereas studies conducted incountries without mandatory fortification showed a significanteffect of genotype (OR 1.63 (95% CI 1.19 to 2.25)). These ORs weresignificantly different from each other (p¼0.032).Conclusions We demonstrate genetic evidence in favour of a causalrelationship between plasma folate and CHD. The absence of agenetic association in countries practicing folate fortificationsuggests that fortification largely abrogates the risk of CHDattributable to folate deficiency.

132 NON-SYNONYMOUS SMAD6 MUTATIONS IMPAIREDINHIBITION OF BMP SIGNALLING IN PATIENTS WITHCONGENITAL CARDIOVASCULAR MALFORMATION

doi:10.1136/heartjnl-2011-300198.132

1H L Tan, 1E A Glen, 1A L Topf, 1D H Hall, 2J J O’Sullivan, 2L Sneddon, 2C Wren,3P Avery, 4R J Lewis, 5P ten Dijke, 1H M Arthur, 1J A Goodship, 1B D Keavney.1Institute of Human Genetics, Newcastle University, Newcastle upon Tyne, UK;2Freeman Hospital, Newcastle upon Tyne, UK; 3School of Mathematics & Statistics,Newcastle University, Newcastle upon Tyne, UK; 4Institute for Cell and MolecularBiosciences, Newcastle University, Newcastle upon Tyne, UK; 5Department ofMolecular Cell Biology and Center for Biomedical Genetics, Leiden University, Leiden, UK

Introduction Congenital cardiovascular malformation (CVM)exhibits familial predisposition but the specific genetic factorsinvolved are unknown. Bone morphogenetic proteins (BMPs) regu-late many processes during development, including cardiac devel-opment. Five genes of the BMP signalling were surveyed for novelvariants predisposing to CVM risk. One of the genes, SMAD6,functions as an inhibitory SMAD which preferentially inhibits BMPsignalling. The SMAD6 knockout mouse is characterised by cardiacvalve and outflow tract defects, including aortic ossification. Wehypothesised that rare functional variation in SMAD6 couldpredispose to congenital cardiovascular malformation (CVM).Methods The coding regions of BMP2, BMP4, BMPR1A, BMPR2 andSMAD6 were sequenced in 90 unrelated Caucasian cases of CVM.The MH2 domain of SMAD6 were further sequenced in additional346 CVM patients. Functional effects of the wild-type and variantSMAD6 proteins were expressed in C2C12 cells and their capacity toinhibit ALK3 activated expression of a BMP-responsive reporter, orto inhibit osteogenic differentiation (using an alkaline phosphataseassay) was assessed.Results We identified two novel non-synonymous variants, P415Land C484F, that were not present in 1000 ethnically-matchedcontrols. P415L was identified in a patient with congenital aorticstenosis and C484F was identified in a patient with coarctation andcalcification of the aorta. Both mutations are in evolutionarilyconserved amino acid residues and are predicted to be damaging byin silico analysis. This was confirmed in functional assays as bothSMAD6 variants failed to inhibit BMP signalling compared withwild-type SMAD6. The P415L mutant appeared to be hypomorphicwhereas C484F appeared to be a null allele in the luciferase assay.The C484F mutant had a significantly (p<0.05) lower capacity toinhibit alkaline phosphatase generation in response to BMP signalling.Conclusions This is the first time that functional mutations inSMAD6 have been described in patients with CVM, specificallythose with calcific aortic malformations. Our data suggest thatinadequate inhibition of BMP signalling pathway due to geneticvariation in SMAD6 may be an important factor in CVM.

133 ACTIVITY AND PSYCHOSOCIAL HEALTH IN ADOLESCENTSWITH CONGENITAL HEART DISEASE (CHD)

doi:10.1136/heartjnl-2011-300198.133

1M L Morrison, 1A J Sands, 1,2C G McCusker, 2P P McKeown, 1M McMahon,1J Gordon, 1B G Craig, 1,2F A Casey. 1Royal Belfast Hospital for Sick Children, Belfast,UK; 2The Queen’s University of Belfast, Belfast, UK

Many patients with CHD are now adolescents. Like other patientswith chronic illnesses they may be at higher risk of psychological/emotional problems. Ability to exercise is an important quality oflife measure and indicator of physical health. We aimed to ascertainif activity and psychosocial health were reduced in adolescents withmajor CHD compared to those with a minor diagnosis. Patientsaged 12e20 years were identified using the Northern Irelandregional database (HeartSuite). Participants were categorised ashaving major or minor CHD and divided into four diagnostic


BCS Abstracts 2011

subgroups. Participants completed validated, age-appropriate ques-tionnaires examining standard psychological parameters. Partic-ipants also underwent an evaluation of exercise, including formalexercise stress testing and measurement of free-living activity usingan ActiGraph accelerometer. Results were analysed using parametricmethods. 143 patients (mean age 15.6 years) consented to partic-ipate, 86 were male (60%) and 105 had major CHD (73%). Diag-nostic subgroups included 39 acyanotic (27.3%), 61 acyanoticcorrected (42.7%), 30 cyanotic corrected (21.0%) and 13 (9%)cyanotic palliated patients. Beck Youth Inventory demonstratedthat individuals with major CHD, particularly cyanotic palliatedpatients, had higher anxiety scores (p value 0.01 (�8.42, �1.13)).There were no significant differences across study groups for self-esteem or other psychological parameters. 134 participants (93.7%)took part in regular exercise each week. There was no significantdifference in activity score between study groups. On formal exer-cise testing, more complex patients performed worse at peak exer-cise. Exercise time for acyanotic group 11.73 mins (sd 3.74)compared to 8.26 mins (sd 4.08) in cyanotic palliated group, p value0.002 (1.32, 5.61)). However, patients with major CHD had signif-icantly higher activity counts. Correlation analysis showed that self-esteem and health locus of control were important predictor varia-bles for activity. Self-esteem and mood seem well preserved inadolescents with CHD as a whole. The majority of young peoplewith CHD, in this group, take part in regular exercise. Surprisingly,complex patients rate themselves to be as active as those with minorCHD. While accelerometer data indicate that the group may bemore active day to day, they are limited in terms of peak exerciseduration. The experience of growing up with a chronic conditionmay therefore have a positive effect on psychological health andinterventions targeted around this area may influence activity.

134 MUTATIONS IN THE SARCOMERE PROTEIN GENE MYH7 INEBSTEIN’S ANOMALY

doi:10.1136/heartjnl-2011-300198.134

1T Rahman, 1J Goodship, 2A Postma, 2K Engelen, 2B Mulder, 3S Klaassen, 4B Keavney.1Institute of Human Genetics, Newcastle upon Tyne, UK; 2Academic Medical Centre,Amsterdam, The Netherlands; 3Max-Delbrueck-Center for Molecular Medicine, Berlin,Germany; 4Institute of Human Genetics, Newcastle upon Tyne, UK

Background Ebstein’s anomaly is a rare congenital heart malforma-tion characterised by adherence of the septal and posterior leaflets ofthe tricuspid valve to the underlying myocardium. As there havebeen reports of abnormal left ventricular morphology and functionin patients with Ebstein’s anomaly we hypothesised that mutationsin the b-myosin heavy chain (MYH7) may be associated withEbstein’s anomaly.Methods MYH7 mutation analysis was undertaken in 141 unrelatedaffected individuals with Ebstein’s anomaly using next-generationsequencing on the 454 platform. 64 probands had no associatedcardiac anomalies. The most common associated cardiac malfor-mation were atrial septal defect (48 probands) and left ventricularnon-compaction (LVNC) (7 probands). Where mutations werediscovered, family studies were undertaken and the segregation ofthe mutation with disease was investigated.Results Heterozygous mutations were identified in eight of theprobands including six of the seven with LVNC. Two patients hadthe same mutation; of the seven distinct mutations, five were novel(four missense changes and an in-frame deletion) and two have beenpreviously reported in patients with hypertrophic cardiomyopathy.Family studies revealed additional members with LVNC for three ofthe probands, one of whom also had a relative with Ebstein’s anomaly.In these three pedigrees the mutation segregated with disease.Conclusions Mutations in MYH7 occur relatively frequently inEbstein’s anomaly accompanied by LVNC. This study is another

example of mutations in a sarcomere protein causing congenitalheart malformation.

135 GENE SCREENING OF THE SECONDARY HEART FIELDNETWORK IN TETRALOGY OF FALLOT PATIENTS

doi:10.1136/heartjnl-2011-300198.135

A Topf, H R Griffin, D H Hall, E Glen, B D Keavney, J A Goodship; The Change StudyCollaborators. Institute of Human Genetics, Newcastle upon Tyne, UK

Background Tetralogy of Fallot (TOF) is the most common cyanoticheart defect, affecting 3e6 infants for every 10 000 births. TOF isphenotypically well defined; it consists of four heart abnormalities: aVSD, an over-riding aorta, a narrowed pulmonary valve and rightventricular hypertrophy. During heart development two heart fieldscan be distinguished. The first one gives origin to the left ventricleand contributes to the right and left atria. The secondary heart fieldgives origin to the right ventricle and the outflow tract. Each ofthese fields can be identified by the expression of specific markers. AsTOF is a malformation of the outflow tract, we hypothesised genesinvolved in the regulatory network of the secondary heart field wereparticularly good candidates for TOF susceptibility.Methods We examined by standard Sanger method the full exonicand intron boundary regions of 14 secondary heart field genes,namely NKX2-5, GATA4, TBX20, MEF2C, BOP, HAND2, FOXC1,FOXC2, TBX1, FOXA2, FGF10, FGF8, ISL1 and FOXH1, in a panel of93 TOF patients. All newly discovered rare variants were checked ina panel of 1000 control chromosomes by multiplex Sequenomassays. When available, parents of cases were screened to assessinheritance of the rare variant.Results We re-sequenced a total of 80 exons and w30 Kb. Amongthe 14 genes studied we found a total of 50 new variants, of which23 were exclusive to the patient population, ie, were absent from1000 normal chromosomes. Nine of these variants cause change inthe aminoacid sequence. We found a functional 19aa deletion of ahighly conserved region of TBX1. In FOXC1 we found a contractionof both alanine and glycine tracts. An alanine expansion, usuallyknown to be deleterious, was found in HAND2. Four non-synon-ymous changes were found in FOXA2. Most patients presented justone variant, however 3 patients presented two, and one patientpresented up to 3 variants. All patients were heterozygotes for thevariants, and had inherited them from one of their phenotypicallynormal parents (when parental information was available). Inaddition, 75% of the variants were inherited from the mother.Conclusions Although genes of the secondary heart field seemedgood candidates for TOF susceptibility, thus far we have not foundany strong indication of unique causal effect, as all variation foundin probands was also present in their unaffected parents. However,the presence of multiple variants in the same proband may result inthe disruption of gene-gene interactions in the secondary heart fieldpathway, which in turn may lead to outflow tract defects. Based onour results, it would seem more likely that susceptibility to TOF bedetermined by a larger number of small genetic contributions whichare also modified by environmental factors. It is evident that largerscale analysis of significant numbers of whole genomes/exomes willbe necessary to better understand the molecular aetiology of TOF.

136 SHOULD FAMILIAL SCREENING BE ROUTINELY OFFERED TOPATIENTS WITH BICUSPID AORTIC VALVE DISEASE?

doi:10.1136/heartjnl-2011-300198.136

R Panayotova, S Hosmane, A Macnab, P Waterworth. University Hospital of SouthManchester, Manchester, UK

Background Bicuspid aortic valve (BAV) disease is one of the mostcommon congenital cardiac abnormalities with prevalence in the


BCS Abstracts 2011

general population of up to 2%. There has been growing evidencesupporting its familial predisposition with an autosomal dominantpattern of inheritance. It is often associated with ascending aorticdilatation and dissection, occurring at a younger age than in patientswith idiopathic aortic aneurysms. BAV disease carries a 6% lifetimerisk of aortic dissection, 9 times higher than that of the generalpopulation. Thus, the presence of BAV and dilatation of theascending aorta requires regular monitoring with a view to timelypre-emptive surgery. Current ACC/AHA guidelines state thatechocardiographic screening of first degree relatives of patients withBAV is recommended. This, however, to our knowledge, is notroutinely done within the UK.Methodology and Results We set out to explore the practicalities ofrunning a routine echocardiographic screening programme for firstdegree relatives of patients with BAV disease. We identified a total of47 patients who had undergone aortic valve surgery performed bythe same Consultant Cardiothoracic Surgeon in the context of BAVdisease in the period May 2007eSeptember 2009. Screening of firstdegree relatives was offered to these patients. 24 patients (51%) gaveus information regarding family members who would like to attendfor an echocardiogram. A total of 75 first degree relatives werereferredean approximate average of 3 per patient. Out of these, 52relatives (70%) actually attended for an appointment. Theremainder did not undergo testing with us as they either lived in adifferent geographic region or expressed a personal preference not tobe scanned at this time. The incidence of newly diagnosed bicuspidaortic valve disease in our cohort of first degree relatives was 8% (4out of 52 relatives). One of these asymptomatic individuals had asignificant ascending aortic aneurysm, which required promptsurgery. Among the relatives of the 24 index patients, there were atotal of 8 cases (3: 1 ratio) of bicuspid aortic valve diseasedeitherknown or newly diagnosed via screening.Conclusions There is a relatively high prevalence and incidence ofbicuspid aortic valve disease among first degree relatives of patientswith this common congenital cardiac abnormality. Routine echo-cardiographic screening should be offered to these families. Imple-menting such a programme is limited by adequate motivation toattend for a screening test if well, and by varying clinical practice indifferent geographic regions. Patients with bicuspid aortic valvedisease should be made aware of its familial pattern of inheritanceand screening of their first degree relatives should be activelypursued in order to reduce the potential morbidity and mortalityassociated with this condition and its related aortopathy.

137 A CITED2->VEGFA PATHWAY COUPLES MYOCARDIAL ANDCORONARY VASCULAR GROWTH IN THE DEVELOPINGMOUSE HEART

doi:10.1136/heartjnl-2011-300198.137

1S D Bamforth, 1S T MacDonald, 1J Braganca, 1C-M Chen, 1C Broadbent,1J E Schneider, 2R Schwartz, 1S Bhattacharya. 1University of Oxford, Oxford, UK;2Texas A&M Health Science Centre, Houston, Texas, USA

Introduction Myocardial development is dependent on the concom-itant growth of cardiomyocytes and a supporting vascular network.The coupling of myocardial and coronary vascular development ismediated in part by VEGFA signalling. Cited2 is a transcriptional co-factor that can inhibit hypoxia-activated transcription and also actsas a co-activator for transcription factors such as TFAP2. Geneticevidence indicates that Cited2 is essential for cardiac left-rightpatterning via regulation of the Nodal-Pitx2c left-right patterningpathway. Zygotic and epiblastic deletion of Cited2 results in atrio-ventricular septation, outflow tract and aortic arch abnormalities, aswell as left-right patterning defects such as right-isomerism. Cited2is also essential for adrenal, neural crest, liver, lung, lens andplacental development. However, the early requirement of Cited2 in

left-right patterning and placental development makes it difficult toidentify a later specific role for Cited2 in myocardial development. Toovercome this problem we therefore investigated the role of Cited2in the myocardium by conditional deletion in cardiomyocyte precursors.Methods Cited2 was selectively deleted from cardiomyocytes byintercrossing mice transgenic for Cited2 and Nkx2-5Cre. Embryoswere collected and processed for analysis by histology, MRI, X-Galstaining, quantitative reverse transcriptase PCR (Q-RTPCR), chro-matin immunoprecipitation and transient transfection assays.Results The cardiomyocyte specific knockout of Cited2 results inabnormal myocardial compact zone growth and ventricular septaldefects. This is associated with a decreased ratio in the number ofsmall vessels to large vessels, and a reduction in Vegfa expression. Wealso show that CITED2 is present at the Vegfa promoter in mouseembryonic hearts, and that it stimulates human VEGFA promoteractivity in cooperation with TFAP2 transcription factors in tran-sient transfection assays. However, we observed no change in themyocardial expression of the left-right patterning gene Pitx2c, aknown target of Cited2.Conclusions The myocardial and capillary defects observed inmyocardial loss of Cited2 are not associated with Pitx2c deficiencyand suggests that Cited2 can cause myocardial and vascular defectsvia a mechanism that is distinct from its effect on the left-rightpatterning pathway. Our results delineate a novel mechanism ofVegfa regulation by CITED2 and TFAP2 transcription factors, andindicate that coupling of myocardial and coronary vascular growthin the developing mouse heart occurs, at least in part, through aCited2->Vegfa pathway. This pathway may be targeted for thetreatment of heart failure resulting from ischaemic heart disease.

138 CELL-SPECIFIC ROLE OF NOX2 NADPH OXIDASE INDEVELOPMENT OF ANGIOTENSIN II-INDUCED CARDIACFIBROSIS IN VIVO

doi:10.1136/heartjnl-2011-300198.138

1S Chaubey, 1C E Murdoch, 1A Ivetic, 1B Yu, 2D Vanhoutte, 2S Heymans, 1A Brewer,1A M Shah. 1Kings College London BHF Centre of Excellence, London, UK; 2UniversityHospital Maastricht, Maastricht, The Netherlands

Introduction Mice globally deficient in Nox2 are protected againstcardiac fibrosis in response to chronic AngII infusion even thoughthe degree of hypertrophy was unaltered. The selective effect ofNox2 on fibrosis may reflect its activation in a non-cardiomyocytecell type. We hypothesised that Nox2, which is expressed in endo-thelial cells and inflammatory cells, may be important for cardiacfibrosis in these cell types.Methods To investigate the role of Nox2 in inflammatory cells, wegenerated chimeric mice by irradiation (10Gy, 15 min) to depleteresident bone marrow cells, followed by bone marrow (BM) trans-plantation, using the following permutations: wild-type (WT)recipient with either KO or WT BM, and KO recipients with WTBM. To assess the role of endothelial Nox2, we used transgenic micewith endothelial-specific overexpression of Nox2 (TG) utilising thetie2 promoter construct.Result AngII (1.1 mg/kg/day, 14-day) infusion caused similarincrease in systolic hypertension and cardiac hypertrophy in all 3chimeric groups. However, cardiac fibrosis assessed by Sirius redstaining was significantly lower in KO mice receiving WT BM(0.560.1%) compared to the WT:WT group (2.760.7%) or in WTreceiving KO BM (2.360.6%). These data suggested that residentNox2-expressing cells are responsible for the protective effectobserved in global Nox2 KO mice. TG mice developed the same levelof systolic hypertension and hypertrophy as WT littermates afterAngII infusion. However, the extent of cardiac fibrosis was signifi-cantly greater in TG than WT by w2-fold (p<0.05). This wasassociated with a greater degree of infiltration by CD45+


BCS Abstracts 2011

inflammatory cells (w1.8-fold, p<0.05) and 30% (p<0.05) moreVCAM-1 positive blood vessels in AngII treated TG hearts.Furthermore, isolated TG endothelial cells recruited w2-fold(p<0.05) more leukocytes than WTupon AngII treatment.Conclusion These results indicate there is a cell-specific role ofendothelial Nox2 in the development of fibrosis. Endothelial Nox2enhances AngII-induced cardiac fibrosis, possibly by enhancinginflammatory cell recruitment and influx via VCAM-1 expression.Although inflammatory cells may be important for the developmentof fibrosis, our results indicate that Nox2 in these cells is notessential for any pro-fibrotic effect.

139 ENDOTHELIAL SPECIFIC INSULIN RESISTANCE PROMOTESTHE DEVELOPMENT OF ATHEROSCLEROSIS

doi:10.1136/heartjnl-2011-300198.139

1M C Gage, 1N Yuldasheva, 2C Jackson, 1M Kearney, 1H Imrie, 1H Viswambharan,1M Kahn, 1J Smith, 1S Galloway, 1R Cubbon, 1P Sukumar, 1A Aziz, 1S Wheatcroft.1Leeds University, Leeds, UK; 2University of Bristol, Bristol, UK

Background Global insulin resistance and endothelial dysfunctionhave been identified as predisposing factors for atherosclerosis.However, it is unclear whether selective insulin resistance in endo-thelial cells alone, is sufficient to promote atherosclerosis. Here weaddressed this question by crossing Endothelial Specific MutantInsulin Receptor Over-expressing (ESMIRO) mice with ApoE nullmice. ESMIRO mice over-express a human insulin receptor withAla-Thr1134 mutation in the tyrosine kinase domain (whichdisrupts insulin signalling) selectively in endothelial cells under thecontrol of the tie-2 promoter/enhancer.Methods Male ApoE�/�ESMIRO mice were compared with sex-matched littermate ApoE�/� mice (both on a C57Bl6 background)after feeding a Western-style diet for 12 weeks.Results ApoE�/�ESMIRO mice were morphologically indis-tinguishable from ApoE�/� control littermates, with normal devel-opment and no difference between groups in body mass. Heart rate,systolic blood pressure, glucose tolerance, insulin sensitivity andfasting glucose levels were similar in ApoE-/-ESMIRO and ApoE�/�

mice. Aortic lipid deposition, assessed by en-face oil red O staining,was similar in ApoE�/�ESMIRO and ApoE�/� mice (6.4%60.5% vs5.8%60.5%; p¼0.39). However, atherosclerotic lesion area in crosssections of aortic sinus was significantly increased in ApoE�/�

ESMIRO mice compared to ApoE�/� controls (24.8%62.4% vs16.6%62.4%; p¼0.02). Absolute plaque size was also significantlyincreased in ApoE�/�ESMIRO mice compared to ApoE controls(226 448.9616 154 mm2 vs 149 424.41624 221 mm2; p¼0.01).Conclusions Endothelial specific insulin resistance is sufficient topromote atherosclerosis and increase lesion area in ApoE null mice.This suggests that enhancing endothelial insulin sensitivity may bean appropriate target to prevent atherosclerosis in insulin-resistantconditions.

140 IN VIVO DEPLETION OF ENDOGLIN RESULTS INSIGNIFICANT LEFT VENTRICULAR REMODELLING

doi:10.1136/heartjnl-2011-300198.140

1B J Davison, 1R Redgrave, 1B Keavney, 2A Blamire, 1H Arthur. 1Institute of HumanGenetics, Newcastle University, Newcastle Upon Tyne, UK; 2Institute of CellularMedicine, Newcastle University, Newcastle Upon Tyne, UK

Endoglin, a TGFb co-receptor, is essential for cardiovascular devel-opment. However, endoglin also has an important role in fibrosis inadult life. Endoglin heterozygous mice have been shown to havereduced fibrosis in response to renal injury. They also have signifi-

cantly reduced cardiac function following myocardial infarction. Inrat cardiac fibroblasts, endoglin expression is up regulated followingstimulation with angiotensin II and TGFb, resulting in reducedexpression of MMP1 and increased expression of collagen I. Theseeffects are inhibited by an endoglin specific antibody. Using ourconditional endoglin knockout mice we sought to investigate therole of endoglin in cardiac healing following myocardial infarction.Adult Eng fl/fl Rosa26-CreERT2 or control (Eng fl/fl) mice weretreated with intraperitoneal injection of tamoxifen for 5 days toactivate CreERT2 and deplete endoglin by Cre-lox recombination.Mice then underwent surgical coronary artery ligation or shamoperation. Cine cardiac MRI was performed 28 days after injury.Measurement of left ventricular (LV) volumes and myocardial masswere made using ImageJ, and parameters of cardiac function werecalculated. We found that LV volumes and mass were significantlyincreased (p<0.001) and ejection fraction significantly reduced(p¼0.005) in endoglin deficient mice compared to controls.However, we also noticed LV volume and mass were increased insham operated endoglin deficient mice. This led us to investigate theeffect of endoglin knockdown on normal heart structure and func-tion in adult mice. Cine cardiac MRI was therefore performed onmice without any surgical procedure after endoglin knockdown. Wefound that in the endoglin deficient mice, LV volume and mass wereagain significantly increased (p<0.03). However, ejection fractiondid not differ significantly from controls. These results demonstratethat depletion of endoglin results in significant left ventricularremodelling and suggest that endoglin plays an essential role in themaintenance of normal cardiac structure. The fact that cardiacfunction was preserved indicates that this is not a cardiomyopathicprocess and we hypothesise that the increased left ventricularvolume in the endoglin-deficient mice may be the result of alter-ations in the extracellular matrix. We are currently investigating thispotential molecular mechanism for left ventricular remodelling inthe absence of endoglin.

141 TISSUE FACTOR PATHWAY INHIBITOR REGULATESVASCULAR DEVELOPMENT IN ZEBRAFISH

doi:10.1136/heartjnl-2011-300198.141

1E W Holroyd, 2C K Pierret, 2V M Bedell, 3J Townend, 2S C Ekker, 2R D Simari. 1QueenElizabeth Hospital, Birmingham, UK; 2Mayo Clinic, Rochester, UK; 3Queen ElizabethHospital, Rochester, UK

Introduction Angiogenesis requires the coordinate regulation ofmultiple biological pathways, including haemostasis. Tissue factorpathway inhibitor (TFPI) is a potent anticoagulant molecule,inhibiting tissue factor-led coagulation. However, the role of TFPI invascular development is unknown. Zebrafish (Danio rerio) provide aunique model system to study vascular development in vertebrates.Despite the divergence of jawed fish (teleosts) over 430 million yearsago, there is notable conservation of the constituent molecules ofthe clotting cascade. Multiple features make this vertebrate modelunique, including its genetic accessibility, the ability to titrate thedegree of genetic knock-down, external embryonic development,and the transparent nature of the embryos.Methods Using in-situ hybridisation techniques, we demonstrateTFPI expression during early vertebrate development. We thenutilise transgenic fish with labelled endothelium (Fli1GFP) anderythrocytes (GATA1dsRed), to study in real time, concomitantfluorescent imaging of both structural development and dynamicblood flow observation, in living zebrafish embryos.Results TFPI expression was identified at 24 h post fertilisation(hpf) in the pronephros (Abstract 141 figure 1ddark blue stainingdenotes TFPI expression; none seen in control embryos). Subse-quently, TFPI mRNA became more abundant, localising to


BCS Abstracts 2011

developing kidney, gut, and vasculature. Morpholino (MO)-basedknockdown of TFPI resulted in coagulopathy and disorderedvascular development (Abstract 141 figure 2. left column: greenendothelial Fli1GFP, right column: red erythrocytes. AB: uninjectedcontrols. CD: MO controls. EeP: TFPI knockdown). Abnormallytargetted (ie, vessels sprouting from normal site but growing inabnormal direction; grey arrows) and extra vessels (ie, superfluousvessels not seen in controls; white arrows in Abstract 141 figure

2EFHIK&L) were seen by 48hpf. TFPI MO induced coagulopathy (aspontaneous clot or bleed; white arrows Abstract 141 figure 2GJMP)in 25.262.3% (p<0.01 cf. uninjected controls) at 3 ng and23.865.8% (p<0.01) at 9 ng. Control embryos did not demonstratesignificant signs of coagulopathy (3.363.3%). Extra arteries occurredin 26.462.6% (p<0.001 by ANOVA cf. uninjected controls) ofembryos injected with 3 ng of TFPI MO. To further define the role ofTFPI in vascular function, RNAi-mediated knockdown of TFPI wasperformed in human endothelial cells (EC). Knockdown of TFPIresulted in enhanced EC tube formation on Matrigel and ECmigration in injury model associated with increased phosphor-ylation of Vascular Endothelial Growth Factor Receptor-2.Conclusion These data represent the first demonstration of TFPIexpression in zebrafish and the first description of a unique pheno-type following TFPI knockdown. They support a model in whichTFPI acts a molecular break to angiogenesis both in vivo, duringearly vertebrate embryogenesis, and in vitro in mature humanendothelial cells, secondary to constitutive regulation of VEGFsignalling.

142 ATRIAL SOURCES OF REACTIVE OXYGEN SPECIES VARYWITH THE SUBSTRATE AND DURATION OF ATRIALFIBRILLATION: IMPLICATIONS FOR THE ANTIARRHYTHMICEFFECT OF STATINS

doi:10.1136/heartjnl-2011-300198.142

1S Reilly, 1R Jayaram, 2C Anroniades, 3S Verheule, 1K M Channon, 1N J Alp,3U Schotten, 1B Casadei. 1University of Oxford, John Radcliffe Hospital, Oxford, UK;2First Department of Cardiology, University of Athens, Athens, Greece; 3Department ofPhysiology, University of Maastricht, Maastricht, The Netherlands

Background Atrial fibrillation (AF) is the most common sustainedcardiac arrhythmia and is associated with altered nitric oxide (NO)-redox balance. The molecular mechanisms and implications of thisphenomenon in the management of patients with AF are poorlyunderstood. Statins improve NO-redox imbalance and decrease theoccurrence of postoperative AF but are less effective in the secon-dary prevention of AF, suggesting that the sources of reactive oxygenspecies might vary with the substrate and duration of AF.Methods and Results We investigated atrial tissue from 130 patientsundergoing cardiac surgery (26 with permanent AF, 32 who devel-oped AF post-operatively and 72 who were in normal sinus rhythmbefore and after surgery), and from goats in sinus rhythm (SR,n¼19) with or without atrial structural remodelling secondary tosurgical AV block (AVB, n¼10) or after 2 weeks (2W, n¼15) or6 months (6M, n¼10) of pacing-induced AF. Atrial NADPH oxidaseactivity (chemiluminescence and 2-OH ethidium, Abstract 142figures 1 and 2), NOX2 & p22phox protein level were increased after2W-AF and in patients who developed AF post-operatively (n¼32).In contrast, the increased superoxide production in atrial tissue fromgoats with AVB or 6M-AF was exclusively driven by mitochondrialoxidases and uncoupled NOS (secondary to a reduction in atrial BH4level and an increase in arginase activity). These findings wererecapitulated in the right atrial appendage of patients. Increase inbasal superoxide production in postoperative AF was associatedwith an apocynin-reversible increase in NADPH oxidase activity andprotein level of the NOX2 and p22phox subunits. NOS activityremained coupled despite the increase in superoxide production. Inline with this, atrial BH4 content was unaltered. In contrast, inpatients with permanent AF, increased superoxide production wasnot reversed by apocynin, and was maintained by mitochondrialoxidases and uncoupled NOS (secondary to BH4 deplition). Ex-vivoinhibition of HMG-CoA reductase with atorvastatin (20 mMol/l)inhibited NADPH oxidase activity (via reducing activity of Rac1 andmembrane translocation of cytosolic subunit p47phox and p67phoxof NADPH oxidase) and caused a mevalonate-reversible reduction in




BCS Abstracts 2011

superoxide release in atrial samples of patients with post-operativeAF but had no effect in patients with permanent AF. Similarly,atorvastatin did not induce a mevalonate-reversible changes in theatrial BH4 concentration and NOS uncoupling in neither group.Conclusions Together, these findings indicate that upregulation ofNOX2-NADPH oxidases is an early but transient event in thenatural history of AF, as mitochondrial oxidases and uncoupled NOSaccount for the statin-resistant increase in atrial superoxideproduction in permanent AF. Variation in atrial sources of reactiveoxygen species with the duration and substrate of AF may explainthe reported variability in the effectiveness of statins in theprevention and management of AF.

143 TISSUE FACTOR PATHWAY INHIBITOR REGULATESANGIOGENESIS INDEPENDENTLY OF TISSUE FACTOR VIAINHIBITION OF VASCULAR ENDOTHELIAL GROWTH FACTORSIGNALLING

doi:10.1136/heartjnl-2011-300198.143

1E W Holroyd, 2K Larsen, 2R G Vile, 2D Mukhopadhyay, 2R D Simari. 1Queen ElizabethHospital, Birmingham, UK; 2Mayo Clinic, Rochester, Minnesota, USA

Introduction The biological systems of coagulation and angiogenesisshow considerable interdependence. Proteases and inhibitors withinthe tissue factor (TF) pathway of coagulation have emerged aspotential regulators of angiogenesis. Tissue factor pathway inhibitor(TFPI), as the primary physiological inhibitor of tissue factor (TF)-mediated coagulation, is ideally situated to modulate the pro-angiogenic effects of TF. However, TFPI may also have effects onangiogenesis independent of its anti-TF ability.Methods We determined the effects of altered TFPI expression onthe regulation of angiogenesis in vivo using genetically-modifiedmurine models of vascular overexpression (SM22áTFPI strain) andendothelial-specific deletion of the TF-binding domain of TPFI(Tie2TFPI). We then defined the mechanism of these effects in vitrousing Human Umbilical Vein Endothelial Cells (HUVECs) over-expressing TFPI or via exogenous addition of TFPI-derived peptidesin assays of angiogenesis.Results Vascular-directed over-expression of TFPI (SM22áTFPIstrain) inhibited angiogenesis in vivo (Abstract 143 figure 1).SM22áTFPI showed significantly impaired recovery from ischaemiain the hindlimb ischaemia model after 3 days (p<0.05, n¼5 pergroup), which persisted throughout the experiment. Survival (until1-cm tumour dimension) of SM22áTFPI mice vs wild-type control(median survival 14 cf. 10 days) following s.c. B16 melanomainjection (n¼7 per group, c2¼4.325, *p<0.05). Endothelial-specificdeletion of the TF-binding domain of TFPI failed to reveal a pro-angiogenic phenotype. This led us to suspect that the anti-angio-genic action of TFPI may be independent of TF. Systemic delivery ofthe murine TFPI carboxyl-terminus (mTFPIct) replicated the effectsof endogenous overexpression. In vitro, overexpression of TFPIinhibited endothelial cell tube formation on Matrigel and migrationusing an injury migration model. Human TFPIct (hTFPIct) inhibitedtube formation and migration through inhibition of VascularEndothelial Growth Factor Receptor-2 (VEGFR2) tyrosine-951phosphorylation, a key event in migration. hTFPIct did not inhibitVEGF121-induced migration, which lacks the heparin-bindingdomain of VEGF165. Utilising the chimeric receptor, EGDR, whichcontains the extracellular domain of epidermal growth factor (EGF)and the intracellular domain of VEGFR2/KDR, a direct effect ofTFPIct on the intracellular domain of VEGFR2 was excluded(Abstract 143 figure 2) TFPIct did not block phosphorylation ofEGDR when stimulated with EGF.



Conclusion Angiogeneis is a key biological system in health anddisease; enabling cells in a hypoxic environment to stimulate newblood vessel growth. These data demonstrate, both in vivo and invitro, an inhibitory role for TFPI in angiogenesis that is TF-inde-pendent. In addition to it classical role as a TF-antagonist, TFPI, viaTFPIct, interferes with the interaction of VEGF165 with theextracellular domain of VEGFR2, thereby limiting angiogenesis.

144 A DRUGGABLE INHIBITOR OF CARDIAC HYPERTROPHYIDENTIFIED THROUGH AN INNOVATIVE CHEMICAL LIBRARYSCREEN

doi:10.1136/heartjnl-2011-300198.144

R Abou Leisa, T M A Mohamed, D Oceandy, S Prehar, M Zi, F Baudoin, L Neyses,E J Cartwright. Manchester University, Manchester, UK

Cardiac hypertrophy is a prerequisite for the development of heartfailure. It currently affects almost one million people in the UK. Feweffective anti-hypertrophic agents with druggable properties havebeen identified. Recently, our group showed that plasma membranecalciumATPase isoform 4 (PMCA4) knockoutmice showed a reducedresponse to hypertrophic stress prompting us to hypothesise that anovel PMCA4 specific inhibitor would modify the development ofcardiac hypertrophy. A library of 1280 medically optimisedcompounds was screened using a novel in vitro assay which measuresthe Ca2+ dependent ATPase activity of PMCA4. The compound AP2was identified, which inhibited PMCA4 activity with high affinity(IC50¼300 nM) but not other PMCAs (PMCA1, PMCA2and PMCA3)or related ATPases which are expressed in the heart including thesarcoplasmic reticulum calcium ATPase and Na/KATPase. In isolatedneonatal rat cardiomyocytes (NRCM), AP2 showed dose dependentinhibition of phenylephrine-induced hypertrophy, indicated by an85% reduction in cell surface area as well as in BNP activity. In vivostudies showed that AP2 (5 mg/kg body weight/day IP) significantlyreduced pressure-overload induced hypertrophy following 2 weekstransverse aortic constriction (TAC) (heart weight/tibia length (mg/mm): sham, 5.560.3, vehicle treated TACmice, 8.760.2, AP2 treatedTAC mice, 7.0 60.5, n¼10 in each group, p<0.01). AP2 treated TAC


BCS Abstracts 2011

mice showed a significant reduction in the cardiomyocyte crosssectional area (sham, 26763.4 mm2, vehicle treated TAC mice,48065.8 mm2, AP2 treated TAC mice, 31963.9 mm2). A significantreduction in the expression of the hypertrophic marker ANP and BNPand in the percentage of fibrosis was also observed in these micecompared with vehicle treated TAC mice. AP2 treatment led to asignificant reduction in the expression of the bona fide calcineurintarget RCAN1.4 and a reduction in theNFATphosphorylation level invivo and the NFAT transcriptional activity in vitro. In conclusion, wehave identified AP2 as a novel PMCA4 specific inhibitor and shown itspotential to modify the development of cardiac hypertrophy likelythrough inhibition of calcineurin/NFAT signalling. This compoundhas drug-like properties and thus lays the basis for a novel approach fortreating cardiac hypertrophy and failure through PMCA4 inhibition.

145 CHARACTERISATION OF FRACTIONATED ATRIALELECTROGRAMS CRITICAL FOR MAINTENANCE OF AF: ARANDOMISED CONTROLLED TRIAL OF ABLATIONSTRATEGIES (THE CFAE AF TRIAL)

doi:10.1136/heartjnl-2011-300198.145

R J Hunter, I Diab, M Tayebjee, L Richmond, S Sporton, M J Earley, R J Schilling. Bartsand The London NHS Trust, London

Introduction Targeting complex fractionated atrial electrograms(CFAE) in the ablation of atrial fibrillation (AF) may improveoutcomes, although whether this is by eliminating focal drivers orsimply de-bulking atrial tissue is unclear. It is also uncertain whatelectrogram morphology should be ablated. This randomised studyaimed to determine the impact of ablating different CFAEmorphologies compared to normal electrograms (ie, de-bulkingnormal tissue) on the cycle length of persistent AF (AFCL).Methods After pulmonary vein isolation CFAE were targetedsystematically throughout the left then right atrium, until termi-nation of AF or abolition of CFAE prior to DC cardioversion. 10 selectrograms were classified by visual inspection according to a vali-dated scale, with Grade 1 being most fractionated and grade 5 normal.Patients were randomised to have CFAE grades eliminated sequentially,from grade 1 to 5 (group 1) or grade 5 to 1 (group 2). Because grade 5electrograms were considered normal, only 5 were ablated. MeanAFCL was determined manually over 30 cycles from bipolar electro-grams recorded at the left and right atrial appendages before and aftereach CFAE was targeted. Lesions were regarded as individual obser-vations, and a resultant increase in mean AFCL$5 ms was regarded assignificant. The randomised strategy first controlled for any cumu-lative effect of ablation on AFCL, and second allowed assessment ofthe order of ablation on the number of CFAE lesions required.Results 20 patients were randomised. The CFAE grade determined byrapid visual inspection for the 968 electrograms targeted agreed withthat at off-line manual measurement in 92.7% (l¼0.91). AFCLincreased after targeting 49.5% of grade 1 CFAE, 33.6% of grade 2,12.8% of grade 3, 33.0% of grade 4, and 8.2% of grade 5 CFAE(p<0.0001 for grades 1, 2, and 4 vs 5, 3 vs 5 not significant). Binarylogistic regression confirmed the effect of CFAE grade, but showed noeffect of electrogram amplitude, location in the left or right atrium, orthe order in which CFAE were targeted. There was no differencebetween groups in the number of grade 1 or 2 CFAE encountered, butthere were fewer grade 3 and 4 CFAE in group 2 than group 1 (bothp<0.01), translating to fewer CFAE targeted per patient in group 1compared to group 2 (37614 and 58618 respectively; p¼0.015).Conclusion Targeting CFAE is not simply atrial de-bulking. Ablatingcertain grades of CFAE caused AFCL prolongation, suggesting theyare more important in maintaining AF. Targeting these CFAE mayreduce unnecessary left atrial destruction. (ClinicalTrials.govnumber, NCT00894400).

Abstract 145 Figure 1 Impact of CFAE grade on the proportion oflesions causing AF cycle length prolongation.

146 IS THERE AN ASSOCIATION BETWEEN THROMBOGENESISMARKERS AND ATRIAL FIBRILLATION BURDEN INPACEMAKER POPULATION?

doi:10.1136/heartjnl-2011-300198.146

C W Khoo, S Krishnamoorthy, G Dwivedi, B Balakrishnan, HS Lim, G Y H Lip. UniversityDepartment of Medicine Centre for Cardiovascular Sciences, City Hospital,Birmingham, UK

Background and Objectives Contemporary pacemaker devices areable to quantify atrial high-rate episodes (AHREs) and atrial fibril-lation burden (AFB) accurately. In this study, we aim to assess therelationship of thrombogenesis markers in association with AHREsand AFB.Methods We studied 87 patients with dual-chamber pacemaker.Patients on warfarin were excluded. AHREs were defined as atrial-rate $220 beats/min and $5 minutes. AFB and percentage ofcumulative pacing were derived from pacemaker diagnostics. Plasmalevels of von Willebrand factor (vWf), tissue factor (TF), solubleP-selectin (P-sel) and D-dimer (DDM) were analysed using ELISA.Results Baseline characteristics and co-morbidities were comparablebetween groups (Abstract 146 table 1). Patients with AHREs hadsignificantly higher cumulative percentage ventricular pacing(p¼0.012). There were no significant differences in levels of vWf, TF,P-sel and DDM between patients with and without AHREs. TheAFB ranged from 0 to 99% in AHRE group. TF (r¼0.516, p¼0.086),P-sel (r¼0.795, p<0.001) and DDM (r¼0.643, p¼0.045) correlatedwith AFB. On linear regression analysis, both P-sel and DDM wereindependently associated with AFB (p<0.05).


No AHRE (n[70) AHRE (n[17) p value

Age, years 71.0611.6 75.468.8 0.096

Hypertension, (%) 38, (54) 12, (71) 0.116

Antiplatelet, (%) 53, (76) 14, (82) 0.739

Percentage atrial pacing 34.6 (6.8e81.5) 22.1 (6.9e65.0) 0.414

Percentage ventricular pacing 21.9 (1.8e99.0) 98.6 (41.0e99.9) 0.012

vWf, IU/dl 94.2616.2 93.9633.7 0.977

TF, ng/ml 0.2 (0.1e0.3) 0.1 (0.0e0.2) 0.105

P-sel, ng/ml 47.6615.8 63.4629.7 0.055

DDM, ng/ml 180.0 (82.0e390.0) 152.5 (82.5e307.5) 0.553


BCS Abstracts 2011

Conclusion AFB is independently associated with increased indicesof P-sel and D-dimer which indicate platelet activation and throm-bosis respectively.

147 THROMBOEMBOLIC RISK STRATIFICATION, ANTI-THROMBOTIC AND ANTICOAGULATION USE FOR PATIENTSWITH ATRIAL FIBRILLATION, A CLINICAL AUDIT

doi:10.1136/heartjnl-2011-300198.147

R A Veasey, R Kulanthaivelu, P Patel, D W Harrington. Kent and Sussex Hospital,Tunbridge Wells, UK

Introduction Atrial fibrillation (AF) is the most prevalent arrhythmiain primary and secondary healthcare settings. Thromboembolic (TE)risk assessment and initiation of anti-thrombotic or anticoagulation(AT/AC) therapy, according to level of risk, is recommended in bothnational and international guidelines. NICE guidance stratifiespatients with AF in to low, moderate or high risk categories andrecommends “aspirin”, “aspirin or warfarin” or “warfarin” therapyrespectively. ACC/ESC guidance endorses use of the CHADS2scoring system and for scores of 0, 1, or $2 recommends “aspirin”,“aspirin or warfarin” or “warfarin” therapy respectively. In addition,it is recommended that AF episode frequency or subtype (parox-ysmal (PAF), persistent (PersAF) or chronic (CAF)) does not influ-ence TE risk assessment. We audited UK cardiologists and generalpractitioners (GPs) to assess adherence to these guidelines.Methods We designed an audit questionnaire assessing: (1) use ofrisk stratification tools, (2) choice of AT/AC for increasing levels ofrisk, and (3) choice of therapy for a number of hypothetical patientswith variable TE risk and variable AF subtype. The questionnairewas distributed by electronic or postal mail to 1176 cardiologists and621 randomly selected GPs.Results In total, 421 responses were received (306 cardiologists, 115GPs). Overall, 91.4% of responders reported use of TE risk strat-ification tools (97.1% cardiologists, 76.5% GPs, p<0.001). NICE riskassessment is used by 26.6% of responders (24.5% cardiologists,32.2% GPs, p¼0.14), CHADS2 by 79.3% (90.2% cardiologists, 50.0%GPs, p<0.001). The frequency of reported use of AT/AC for each risklevel of the NICE assessment and CHADS2 score are shown inAbstract 147 tables 1 and 2 respectively. Type of AF (PAF/PersAF/CAF) reportedly influences the use of AT/AC for 34.3% orresponders (24.2% cardiologists, 46.3% GPs, p¼0.001). Abstract 147figure 1 demonstrates AT/AC usage for each of the followinghypothetical patients: 1. 61 year old, hypertension, PAF episodestwice a year lasting 1e2 h (NICE risk: mod, CHADS2 score 1/6). 2.43 year old, diabetes, PAF episodes weekly lasting 10e12 h (NICErisk: mod, CHADS2 score 1/6). 3. 53 year old, hypertension, CAF(NICE risk: mod, CHADS2 score 1/6). 4. 78 year old, no other riskfactors, CAF (NICE risk: mod, CHADS2 score 1/6). 5. 76 year old,hypertension, diabetes, PAF episodes 3e4 times per year lasting<1 hour (NICE risk: high, CHADS2 score 3/6). 6. 77 year old,hypertension, diabetes, PAF episodes occurring weekly and lastingseveral hours (NICE risk: high, CHADS2 score 3/6). 7. 80 year old,previous TIA, CAF (NICE risk: high, CHADS2 score 3/6).


NICE Risk None (%) Aspirin (%) Aspirin or Warfarin (%) Warfarin (%)

Low 16.7 78.3 2.9 2.1

Moderate 0.6 3.5 66.9 28.7

High 0.0 0.0 4.4 95.6


CHADS2 Score None (%) Aspirin (%) Aspirin or Warfarin (%) Warfarin (%)

Zero 27.0 70.3 1.3 1.3

One 4.7 45.1 43.2 7.0

Two 0.0 7.8 32.2 60.0

Three 0.0 0.8 11.5 87.7

Four 0.0 0.0 5.3 94.7

Five 0.0 0.0 2.9 97.1

Six 0.0 0.0 2.9 97.1


Conclusions TE risk stratification tools are reportedly widely used inUK clinical practice. AT/AC use for NICE and CHADS2 risk levelsare mostly appropriate, although warfarin is under recommendedfor patients with a CHADS2 score of 2/6. In addition, the use of AT/AC is influenced, inappropriately, by AF episode frequency andsubtype.

148 THE ASSESSMENT OF TRANSIENT LOSS OFCONSCIOUSNESS: WE’RE STILL NOT ASKING THE RIGHTQUESTIONS

doi:10.1136/heartjnl-2011-300198.148

A E Bewick, A Gasson, L Ala, R A Bleasdale. Royal Glamorgan Hospital, Cardiff, UK

Accurately diagnosing patients with TLOC can be achieved in mostcases with a detailed clinical history. We set out to assess howpatients were assessed in the setting of a district general hospital(DGH) with 570 beds, receiving an unselected intake via generalpractice and an A&E. Using the ESC guidelines of 2009 we generateda 22-question study proforma for a retrospective review of themedical records. We identified 322 cases for possible inclusion over a4 month period. 26 of the case notes were not available to analyse, 8had insufficient details to identify the relevant patient. Therefore intotal 288 notes were reviewed. Inclusion required the TLOC to becomplete, of rapid onset and short duration with spontaneouscomplete recovery. A further 123 patients were therefore excluded.This left 165 data sets (58% male). The age distribution was atypical bimodal distribution with 16% between 10 and 29 years ofage and 48% over the age of 70 years. 73% were assessed in A&E,18% were assessed in the Acute Medical Unit (AMU) and 7% wereassessed in rapid access ambulatory clinics. Only 4% of the initialassessments were undertaken by consultants, 12% by a SpecialistRegistrar (SpR), 21% by a year 1 foundation program (FP1) doctorand the majority was assessed by FP 2 or core medical trainees(CMT). Key diagnostic elements of the history are still beingneglected. For example, the symptoms at the onset of the TLOCwere documented in only 58% of cases; the recovery symptom


BCS Abstracts 2011

profile was reported in only 37%. Only 47% (n¼78) of recordsdescribed a witness account. Within the witness accounts that wererecorded, key elements remained un-reported for example skincomplexion was only reported in 35% of the 78. The duration of theTLOC was recorded in only 44%, Tongue biting in 27% and thepresence or absence of abnormal movements was recorded in only12% of this 78 patients. The presence or absence of a family historyof sudden cardiac death was only reported in 2% cases. The familyhistory of a cardiomyopathy was only recorded in 1% and a familyhistory of TLOC was recorded in 1%. A patient past history ofcardiac disease was asked about in 40% of cases while a past historyof TLOC was only asked about in 35%. In this majority elderlystudy population, a recent change in drug therapy was only askedabout in 2% of cases. This study highlights that in a DGH envi-ronment, the initial assessment of patients with TLOC is under-taken by junior medical staff who often do not document keydiagnostically differentiating elements of the history and examina-tion indicating an ongoing lack of adequate training regarding themost appropriate and accurate techniques for differentiating thecauses of TLOC.

149 AUTOMATED ANALYSIS OF ATRIAL ABLATION-SCAR USINGDELAYED-ENHANCED CARDIAC MRI

doi:10.1136/heartjnl-2011-300198.149

1L Malcolme-Lawes, 1R Karim, 2C Juli, 2P B Lim, 2T V Salukhe, 2D W Davies,1D Rueckert, 1N S Peters, 2P Kanagaratnam. 1Imperial College London, London, UK;2Imperial College Healthcare, London, UK

Introduction Visualisation of the ablation-related atrial scar usingdelayed-enhanced MRI (DE-MRI) may reveal important underlyingcauses for atrial fibrillation (AF) recurrence following ablation. Inorder to develop and objective method for delineating ablation-scarwe compared pre and post DE-MRI after Cryo-balloon lesion on thebasis that a more predictable lesion set would be created forvalidation.Methods and Results 12 patients undergoing cryoablation for PAFwere enrolled in the study, and underwent pre-ablation DE-MRIscans. Pulmonary vein isolation (PVI) was confirmed in all patientsat the end of the cryoablation procedure using a circular mappingcatheter. Additional ablation by RF or Freezer Max was required toachieve PVI in 59%. No ablation was performed in any region otherthan the PV ostia. Post-ablation DE-MRI was performed at3 months. An automatic segmentation of the LAwas produced withcustom software from the MRA sequence. The preablation andpostablation free breathing late gadolinium enhanced sequence wasregistered to the MRA and the maximum intensity within the LAwall was projected onto the post ablation LA surface. The bloodpool was identified automatically using custom software as theregion 1 cm inside the wall of the LA, and its mean (BPM) and SDused as a baseline. To identify a universal threshold for scar, regionsof brightest myocardium were initially selected in pre and postablation MRIs. The brightest regions were 1.961.2 vs 8.763.1 SDsabove the BPM in pre-and post-ablation MRIs respectively(p¼0.001). A threshold of 5 SDs above the BPM was thereforeprogrammed into our custom software to identify regions of scar forall patients. The ostial regions were defined as extending 1 cm bothproximal and distal to the PVeLA junction, and selected manuallyfor left and right sided veins prior to scar projection. (See Abstract149 figure 1). The scar proportion within these regions was calcu-lated using commercially available software ITK-SNAP. Total LAscar proportion was 0.260.02% vs 6.360.75% in pre and postablation scans respectively. The increase in scar seen in the PV ostiawas 24.661.38% compared with 2.661.28% in the rest of the LA(p¼0.01) (See Abstract 149 figure 2).

Abstract 149 Figure 1 Comparison of pre-ablation and post-ablation %scar using fixed threshold.


Conclusion We have demonstrated the feasibility an objective,automated method of DE-MRI analysis of left atrial ablation-scar.This technique will now need to be validated against clinicaloutcomes.

150 IMPLANTABLE CARDIOVERTER-DEFIBRILLATORLEAD COMPLICATIONS AND CLINICAL EFFECTIVENESSIN PATIENTS WITH INHERITED CARDIACCONDITIONS

doi:10.1136/heartjnl-2011-300198.150

1,2R Bastiaenen, 1S Ben-Nathan, 2S Jones, 2D Ward, 2M Gallagher, 1,2S Sharma,1,2E R Behr. 1St George’s University of London, London, UK; 2St George’s Hospital,London, UK

Background Implantable cardioverter-defibrillator (ICD) therapy canreduce sudden death due to ventricular arrhythmia (VT/VF) but isnot without complication, particularly in young patients who livefor many years with a device in situ. We aimed to determine theICD complication rate in our inherited cardiac condition (ICC)population compared with international reports. Particular impor-tance was given to inappropriate shock therapy due to lead failure asthere are new ICD technologies available.Methods Patients with ICCs who had ICD implantation or boxchange between January 2006 and September 2009 were included.Data on clinical characteristics, complications and ICD therapieswere obtained from pacing and hospital records. We compared ourdata with several ICD studies of patients with specific ICCs(Abstract 150 table 1).


BCS Abstracts 2011


SGH ICCpatients(n[101)

Long QTSyndromepatients(n[51)

HCMpatients(n[506)

ARVCpatients(n[106)

BrugadaSyndromepatients(n[220)

Follow-up (months; mean6SD) 74653 87 44633 58635 38627

Appropriate therapy (%) 26 24 20 24 8

Inappropriate therapy (%) 18 29 27 19 20

Lead failure (%) 21 25 7 2 9

Complication rate excludinglead failure (%)

26 31 n/a 34 20

Results 101 patients (mean age 44.1614.8 years; 59 male) wereincluded (idiopathic VF 15%; DCM 17%; ARVC 22%; HCM 21%;long QT syndrome 17%; Brugada syndrome 6%; others 2%). Duringa mean follow-up of 74.0653.2 months 2 patients died (1 inappro-priate shocks; 1 stroke). Indications were secondary prevention in71.3% of patients. ICD types were 56.4% single chamber; 39.6%dual chamber; 4.0% biventricular. Appropriate therapy successfullyterminated VT/VF in 27 (26.7%) patients 34.7% of secondary and6.9% of primary prevention patients received appropriate therapy.Inappropriate therapy occurred in 18 (17.8%) patients and leadfailure (noise/wear/fracture) in 22 (20.8%) patients (Abstract 150table 2). 12 out of 18 inappropriate shocks were due to lead failure, 5sensing errors (1 T-wave oversensing; 4 AF), 1 generator fault. 10/22leads that failed were Medtronic Sprint Fidelis and these wereresponsible for 8/12 patients receiving inappropriate shocksincluding one death due to lead fracture. Comparison with otherstudies indicates a high lead failure rate due to the long follow-upperiod, similar to the LQT Study which reports 25% lead failure over87 months (Abstract 150 table 1). With lead failure excluded thecomplication rate is comparable to shorter follow-up studies. Inap-propriate and appropriate therapy rates are similar among all studies.


Complication Number of patients % of patients

Lead failure 21 20.8

Inappropriate shock 18 17.8

Lead displacement 5 4.9

Infection 5 4.9

Pneumothorax/Haemothorax 5 4.9

Box/Wound/Other revision procedure 7 6.9

Thrombosis (venous/lead) 2 1.9

Haematoma 5 4.9

Chronic abdominal cavity post-explant

1 0.9

Conclusions There is a significant rate of ICD lead failure in patientswith ICCs, which may be expected given the high frequency ofSprint Fidelis leads implanted during this period and the long follow-up. Our results compare favourably to other similar studies. Thehigh rate of appropriate therapy highlights the clinical effectivenessof ICD intervention in secondary prevention. Lead complicationsmay be lower with the use of new ICD technology in selected patients.

151 RISK OF RECURRENCE FOLLOWING EXTRACTION OF CARDIACIMPLANTABLE ELECTRONIC DEVICES FOR INFECTION: WHENSHOULD A NEW DEVICE BE RE-IMPLANTED?

doi:10.1136/heartjnl-2011-300198.151

H E Thomas, M Das, D Twomey, C J Plummer, E J Shepherd. Freeman Hospital,Newcastle upon Tyne, UK

Background The recommended management of cardiac implant-able electronic device (CIED) infection is complete system

extraction. There are limited clinical data on the optimal time fordevice re-implantation. A small series reported good results withsimultaneous contralateral implantation. We evaluated thisapproach in our institution for patients without signs of systemicsepsis. We present clinical outcomes and completeness of extraction.Methods The clinical records of all patients undergoing leadextraction in our institution since January 2008 were reviewed.Results 68 patients underwent CIED extraction for infection duringthis time period (see Abstract 151 table 1). In 34 cases, the devicewas removed with simple traction, 9 with locking stylet, 22 withlocking stylet and laser sheath, 1 with locking stylet and mechanicalsheath and 2 with femoral snare. There was complete hardwareremoval in 64 cases (94%). One patient with lead related endo-carditis required a subsequent surgical procedure to remove a leadfragment and in 4 other patients who had erosion, pocket infectionor threatened erosion, a small fragment of lead remained. 18/68patients were re-implanted with a new device on the contralateralside on the same day as the extraction. 28/68 patients received anew device between 1 and 227 days later and 22/68 have notundergone reimplantation. An active fixation bipolar TPW(temporary pacing wire) was used in 6 patients for a mean7.862.7 days. 3 patients had a further device related procedureduring a mean follow-up of 4456304 days: 1 lead reposition, 1pocket washout and 1 extraction. Of the 2 procedures carried out forrecurrent infection, 1 was managed with a TPW for 7 days prior toreimplantation and 1 underwent reimplantation at 14 days withoutTPW. In addition, the patient requiring pocket washout had afragment of lead remaining following their initial extraction.


Indication for device extraction Number of patients, n[80 (%)

Erosion 31 (39)

Pocket infection 25 (31)

Lead infection 7 (9)

Threatened erosion 4 (5)

Pain 1 (1)

Conclusion We report low rates of recurrent infections followingCIED extraction. None of the 18 individuals simultaneously re-implanted with a new device on the contralateral side needed anyfurther procedures during the follow-up period. This approach may beappropriate, particularly in pacing dependant patients who wouldotherwise require a TPWwith its associated risks. In those individualswho required a TPW, the risk of recurrent infection in our series was17% despite our use of an active fixation pacing lead and externalisedpulse generator which has a lower reported complication rate. Onlyone of the 4 patients with a residual lead fragment required re-intervention for recurrent infection. This provides some supportiveevidence that in patients with high surgical risk and pocket abnor-malities, if fragments of lead may remain, the patient may be treatedconservatively and monitored for signs of recurrent CIED infection.

152 REAL-TIME CARDIAC MR ANATOMY AND DYSSYNCHRONYOVERLAY TO GUIDE LEFT VENTRICULAR LEAD PLACEMENTIN CRT

doi:10.1136/heartjnl-2011-300198.152

1,2A Shetty, 1,2S Duckett, 1,2M Ginks, 1,2Y Ma, 1,2M Sohal, 1,2P Mehta, 1,2S Hamid,1,2J Bostock, 1,2G Carr-White, 1,2K Rhode, 1,2R Razavi, 1,2C A Rinaldi. 1Guys and StThomas’ Hospital NHS Foundation Trust, London, UK; 2King’s College London, London, UK

Introduction Optimal left ventricular (LV) lead placement via thecoronary sinus (CS) is a critical factor in defining response to cardiacresynchronisation therapy (CRT). Using novel semi-automatedimage acquisition, segmentation, overlay and registration software


BCS Abstracts 2011

we set out to guide lead placement by avoiding scar and targetingthe region of the LV with the latest mechanical activation.Methods 17 patients underwent cardiac magnetic resonance (CMR)scans. 3Dwhole heart imageswere segmented to produce high fidelityanatomical models of the cardiac chambers and coronary veins. 2, 3, 4chamber and short axis cine images were processed using Tomtecsoftware to give a 16 segment time volume-dyssynchrony map. Inpatients with myocardial scar the late gadolinium enhancementimages were manually segmented and registered to the anatomicalmodel along with the dyssynchrony map. The 3 latest mechanicallyactivated segments with <50% scar were identified and this infor-mation was overlaid at CRT implant on to live fluoroscopic imagesusing a prototype version of the Philips EP Navigator software.Subsequently, the x-ray C-arm and table could be moved freely whileautomatically maintaining a registered roadmap. We used a highfidelity pressure wire to assess the acute haemodynamic response topacing in different regions of the overlaid 16 segmentmodel. All dP/dtmeasurements were compared to baseline AAI or VVI (for thosepatients in AF) pacing at 5e10 beats/min above intrinsic rate.Results 15 of the 17 patients underwent successful placement of a LVpacing lead via the CS with satisfactory pacing parameters and nophrenic nerve stimulation at implant. The mean time from insertionof the CS guide catheter into the venous sheath to successfulcannulation of the CS was 1.361.0 min. In 2 patients we were unableto place a LV lead successfully in any branch of the CS. We paced in atleast one of our 3 target segments in 11 patients. 67% of patients wereresponders as defined by a 10% increase in +dP/dt over baseline. Themean change in +dP/dt for the best lead position vs baseline+dP/dtwas 15.9611.3% for DDDLV pacing. This compares to a meanchange in+dP/dt of 14.9612.3% when the CMR dyssynchrony-mapdefined target region was paced DDDLV. The region of best+dP/dtresponse was postero-lateral, lateral or posterior in all cases.Conclusion We have shown it is feasible to acquire, overlay andaccurately register cardiac MR data on to fluoroscopic images at thetime of CRT implant. Our data suggest that it is also possible toidentify and place the LV lead in at least one target region in mostpatients. This appears to give close to the best acute haemodynamicresponse that can be achieved in any branch of the CS. The initialresults of this pilot study suggest that a MR dyssynchrony guidedapproach to LV lead placement may allow ideal LV lead positioning(Abstract 152 figures 1 and 2).



153 VENTRICULAR PACING ALONG INDIVIDUAL BRANCHES OFTHE CORONARY SINUS USING A QUADRIPOLAR LV PACINGLEAD

doi:10.1136/heartjnl-2011-300198.153

1,2A K Shetty, 1,2P Mehta, 1,2S Duckett, 1,2J Bostock, 1,2M Ginks, 1,2S Hamid,1,2M Sohal, 1,2R Razavi, 1,2Y Ma, 1,2K Rhode, 1,2A Arujuna, 1,2C A Rinaldi. 1Guys and StThomas’ Hospital NHS Foundation Trust; 2King’s College London, London, UK

Introduction Cardiac resynchronisation therapy (CRT) usuallyinvolves placing the left ventricular (LV) pacing lead in the postero-lateral or lateral region of the LVepicardial surface as this is thoughtlikely to re-coordinate myocardial contraction most effectively. TheLV lead is standardly placed in a position with the best pacingparameters and satisfactory stability. It is not known, however,whether there is a significant difference in haemodynamic response toLV pacing in different regions of the same coronary sinus (CS) vein. Inthis study we aimed to evaluate the difference in acute haemody-namic response to pacing along individual branches of the CS.Methods 16 patients underwent an acute haemodynamic studyduring their CRT-defibrillator implant. We used a high fidelitypressure wire to assess the acute haemodynamic response (AHR) topacing in different branches of the coronary sinus. We used a novelquadripolar lead (Quartet, St Jude Medical, Sylmar, California, USA)that has four poles on the LV lead―distal tip and 3 ring electrodes.The 3 ring electrodes are spaced 20 mm, 30 mm and 47 mm fromthe distal tip electrode and the four poles allow bipolar pacingbetween them. It was thus possible for us to test pacing parametersand AHR along a significant proportion of a CS branch withouthaving to reposition the LV lead.Results DDDLV pacing was attempted in as many different CSbranches as possible in each patient (total 56 different positionsused). The mean overall percentage difference in AHR (measured bychange in +dP/dt compared to baseline AAI pacing or VVI pacing inAF patients) between an individual CS branch bipole with thelowest +dP/dt and that with the highest was 6.665.6%. Muchlarger differences in change in +dP/dt were seen, however, betweendifferent branches of the CS in the same patient with a meandifference in change in +dP/dt in the best CS vein compared to theworst CS vein of 16.766.3%. Although the difference in AHR seenbetween different bipoles within the same vein were not large, wedid find that in some cases no pacing capture was found with one


BCS Abstracts 2011

bipole but was found with another. Furthermore, differences inwhether phrenic nerve stimulation (PNS) occurred were seen whenusing different LV lead bipoles within the same branch of the CS.Conclusion Our data suggest that only a small difference in AHR isseen when pacing along the same branch of the CS compared topacing within different branches of the CS within the same patient.This means that although the site of LV lead placement is impor-tant, a proximal or distal position within a CS branch is much lessimportant than choosing the right branch in terms of acutehaemodynamic response. A choice of bipoles on the LV lead maymean, however, that problems with capture thresholds or PNS canbe overcome without the need to reposition the LV lead.

154 PATIENTS RECEIVING STANDARD PACEMAKERGENERATOR REPLACEMENTS FREQUENTLY HAVEIMPAIRED LEFT VENTRICULAR FUNCTION AND EXERCISEINTOLERANCE, RELATED TO THE PERCENTAGE OF RIGHTVENTRICULAR PACING

doi:10.1136/heartjnl-2011-300198.154

1G A Begg, 1J Gierula, 1Z L Waldron, 2K K Witte. 1Leeds General Infirmary, Leeds, UK;2University of Leeds, Leeds, UK

Background Right ventricular (RV) pacing is an accepted treatmentfor symptomatic bradycardia. However, long-term RV pacing isincreasingly recognised to be detrimental to left ventricular (LV)systolic function. We wanted to establish the prevalence, associatedfeatures and predictors of LV systolic dysfunction (LVSD) andoutcome in a contemporary group of patients with longeterm RVpacemakers.Methods We prospectively recruited consecutive patients listed forPGR between 2008 and 2010 at Leeds General Infirmary. Weperformed echocardiography, exercise testing and recorded indica-tions for pacing, pacing variables and duration of pacing, co-morbidities, current medication and renal function.Results Of 399 PGR procedures 342 subjects (86%), 184 men,attended. Non-attendees had similar pacing variables and were ofsimilar age as attendees. Mean age (SE) was 76 (1), and meanduration of pacing was 10 (0.3) years. Comorbidites were common:diabetes mellitus in 11%, previous myocardial infarction in 15%,previous cardiac surgery in 26% and atrial fibrillation (AF) in 26%.Medical therapy included b-blockers in 60% and ACE inhibitors in70%. Dual chamber devices were implanted in 77% (45% of allpatients had rate responsive (RR) pacing programmed). Meanpercentage of ventricular pacing (%VP) was 61 (2)%. Mean leftventricular ejection fraction (LVEF) was 49 (1)%, (44% had an LVEF<50%). Mean peak oxygen uptake (pVo2) (in 107 subjects) was 17(1) ml/kg/min and mean creatinine was 108 (3) mmol/l. There wasan inverse relationship between LVEF and %VP (0.42; p<0.0001),and years since first implanted (p¼0.09) but there was no effect onLVEF of age, the presence of AF and the pacing mode. In singlechamber devices, RR pacing was associated with higher %VP(p¼0.01), and a trend to worse LVEF (p¼0.09). These differenceswere not seen in RR programmed dual chamber devices. There was anegative relationship between pVo2 and %VP (r¼0.21; p<0.03). Evenwith a short follow-up period of 16 (0.5) months, 23 (7%) patientsare dead. Patients dead at the censor date were older at the time ofthe assessment (p<0.005), had a higher %VP (p<0.03) and worserenal function (p<0.001), but did not have significantly worse LVEFor pVo2. The presence of a single chamber device was associatedwith a poorer outcome (p<0.002) despite patients with a singlechamber device being of similar age as those with a dual chamberdevice.Conclusions Patients receiving standard pacemaker generatorreplacements frequently have cardiovascular comorbidities, leftventricular dysfunction and impaired pVo2 and suffer a high

mortality rate. In an unselected population of patients with pace-makers, we have established that the amount of RV pacing is relatednot only to important surrogate measures of outcome such asexercise tolerance and LVEF but also mortality. Whether an aggres-sive policy of limiting RV pacing in patients at risk reduces mortalityis unknown.

155 INCIDENCE SCREENING OF PATIENTS FOLLOWING STELEVATION MYOCARDIAL INFARCTION FOR PRIMARYPREVENTION IMPLANTABLE CARDIOVERTERDEFIBRILLATOR (ICD) IMPLANTATION HAS A LOWTHERAPEUTIC YIELD

doi:10.1136/heartjnl-2011-300198.155

E L Berry, H C Padgett, A J Ahsan, A D Staniforth. Nottingham University HospitalsNHS Trust, Nottingham, UK

Introduction The ICD implant rate for the United Kingdom is lowcompared with the European Union and United States of America.National Institute of Clinical Excellence guidance TAO95 (NICE2006) makes recommendations for primary prevention ICDimplantation. Our study investigated the feasibility of systemati-cally screening patients following an acute ST elevation myocardialinfarction (STEMI) to improve local ICD implant rates.Method A prospective single centre study was performed over 14-months, in tertiary centre setting. All patients with a diagnosis of anacute STEMI had an echocardiogram at 6 weeks to assess leftventricular ejection fraction (LVEF). Patients with impaired LVEFthen underwent screening for primary prevention ICD as per TA095recommendations (Abstract 155 figure 1).



BCS Abstracts 2011

Results 326 STEMI patients were identified. Of these 12(3.7%)declined investigation. 25(7.8%) died during the investigation period(22 died during their initial acute event, 3 died of non cardiac causesfollowing discharge). 10(3.1%) requested follow-up in anothergeographical area. 26(8%) patients were identified as LVEF<35%;2(0.6%) patients were assessed as not clinically suitable for furtherinvestigation. 2(0.6%) had LVEF<30% and QRS>120 ms, bothproceeded to have a primary prevention ICD implanted. 24(7.4%)patients had Holter monitors; 2(0.6%) were identified as havingepisodes of NSVT. Both patients had EPS; 1(0.3%) had inducible VTand proceeded to have a primary prevention ICD implanted.1 patient (0.3%) self presented with a cardiac arrest beforecompletion of their screening tests and received a secondaryprevention ICD. In total, 3(0.9%) primary prevention ICDs wereimplanted (Abstract 155 figure 2).

Abstract 155 Figure 2Conclusion The yield from this study was low; 3 patients (0.9%)proceeded to primary prevention ICD. It should also be noted that themethodology resultant from TA095 guidelines was labour and resourceintensive. An alternative approach of opportunistic screening inpatient groups with a high prevalence of impaired LV function mightgive a higher yield than our approach looking at disease incidence.

156 A SINGLE CENTRE EXPERIENCE OF IVABRADINE ANDCLONIDINE FOR INAPPROPRIATE SINUS TACHYCARDIA

doi:10.1136/heartjnl-2011-300198.156

P P Sadarmin, T R Betts. John Radcliffe Hospital, Oxford, UK

Introduction Inappropriate sinus tachycardia (IST) is a relatively raredisease that manifests with resting tachycardia, a rapid increase inheart rate (HR) with minimal exertion, a normal ECG and absenceof structural heart disease. Treatment options include b-blockade orsinus node modification which are not 100% successful. Neweragents like sinus node inhibitor (Ivabradine) or a centrally acting a-2sympathomimetic (Clonidine) can be used but there is no successoutcome data for either and there is also no evidence that one isbetter than the other. We present our experience of managing 6patients with a diagnosis of ISTwith either Ivabradine or Clonidineor both.Methods We identified 6 patients from 2005 to 2009 with a diag-nosis of IST (according to accepted international guidelines) whohad been treated with either Ivabradine or Clonidine or both.Medical case records were reviewed for each patient.Results 5 out of 6 patients were women with a mean age of27.5 years (range 16e40 years). All patients had been symptomaticfor alteast 6 months before presentation to our tertiary centre. 2

patients had associated symptoms of hyperadrenergic surges. Holtermonitoring prior to treatment demonstrated sinus tachycardia.Resting pre-treatment mean 24 h HR was 94610 (range 75e100)and mean HR on minimal exertion was 157620 (range 130e176).All patients had a structurally normal heart on echocardiogram. Tilttable testing was considered in 3 patients due to their symptomsand it excluded postural orthostatic tachycardia syndrome. Pre-treatment with b-blockers had been unsuccessful in 5/6 patients.The remaining patient had symptomatic asthma and was thereforeunable to tolerate b-blockers. Ivabradine was exclusively used in 3patients and clonidine in 2. 1 patient was started with Ivabradinebut later switched over to clonidine as it was ineffective. All 4patients taking Ivabradine failed to gain symptom relief with nosignificant reduction in mean 24-h HR parameters. Mean resting HRafter 3 months of Ivabradine therapy was 9569 (range 88e105) andmean HR on exertion was 159623 (range 128e180). 2/4 patientssubsequently had complete sinus node ablation and AAIR pace-maker. In contrast, the 3 patients on clonidine had greater symptomresolution and fall in resting and exercise heart rates at 3 monthsfollow-up. Resting mean HR was 8163 (range 78e83) and mean HRwith exertion was 144618 (range 132e164). The HR variability preand post treatment is shown in Abstract 156 figure 1.


Discussion In our case series of 6 patients, Clonidine was moreeffective than Ivabradine both in terms of reducing heart rate andtreating symptoms for patients with inappropriate sinus tachy-cardia. Patients with coexisting hyperadrenergic symptoms maybenefit the most. A trial of Clonidine can be recommended beforeconsidering sinus node ablation. Formal randomised controlled trialsare needed to confirm our findings.

157 AN INSIGHT INTO IMPLANTERS’ PRACTICES OF ICDIMPLANTATION: A PHYSICIAN SURVEY

doi:10.1136/heartjnl-2011-300198.157

P P Sadarmin, K C K Wong, K Rajappan, Y Bashir, T R Betts. John Radcliffe Hospital,Oxford, UK

Introduction The Implantable cardioverter defibrillator (ICD) is themainstay of treatment for the prevention of sudden cardiac death(SCD) and the management of tachyarrhythmias. Informed patient


BCS Abstracts 2011

consent is an essential part of the implant process. Our aim was toget an insight into implanters’ (Imp) practices prior to an ICDimplantation.Methods A questionnaire survey was sent to UK ICD Imp to testtheir knowledge of the risk and benefits of an ICD in patients whosatisfy trial and national guideline criteria and the incidence ofimplant complications. Information of the style and language ofconsent was requested. This questionnaire was specifically aimed atImp and was part of the larger questionnaire looking at knowledge,attitudes and factors influencing ICD prescription in the UK.Results Replies were received from 23 implanters. 35% of theresponders were between the age of 30e39 years and 39% werebetween 40 and 49 years. 83% of the responders were Consultantsand 96% were working in an implantating centre. 83% of Imp werefully aware of Primary Prevention (PP) NICE guidelines while 78%were fully aware of Secondary Prevention (SP) NICE guidelines.There was widespread use of information leaflets (87%) andspecialist ICD nurses (83%) to disseminate information to patients.All responders said they would personally discuss the therapy withthe patient prior to the implantation regardless of the referralsource. A discussion regarding the prevention of SCD, inappropriateshocks and driving restrictions were performed by 96% ofresponders and device infections and lead failures discussed by 91%.Use of absolute risk reduction in percentages and number needed totreat while explaining the risks and benefits gained from ICDs wereused by 22% and 26% respectively. There was widespread use ofphrases like “small risk” or “moderate risk” (61%) and life prolon-gation (eg, lets you live longer by an average of 3 months) (30%).Replies also indicated that Imp under-estimate overall mortality inmedicallytreated and ICD-treated patients, lead dislodgementrequiring re-positioning and major haematoma requiring reoperation.Imp overestimate infections leading to device removal and the incidenceof pneumothorax when compared to published trial or study data.Conclusion The majority of implanters are aware of UK ICDguidelines. The patient consent process is not universal. Guidelinesand awareness about end-of-life care in ICD patients is needed andshould be part of the initial consent process. Evidence based use ofrisk and benefit terminologies like ARR and NNT are needed tobetter inform the patient rather than abstract phrases. Increasingawareness of ICD complication rates can help patients and physi-cians balance risk against benefit which could lead to improvedpatient satisfaction with their therapy.


Estimate of ICDcomplications Mean % Published/Trial data %

Death as a complication ofdevice implant

0.3760.48 0.77% (Circulation.1998;98:663e670);2.08% (Br Heart J.1995:73:20e24)

Lead dislodgement requiringlead repositioning

3.562.08 5% (PACE.2005; 28:926e932);10% (Circulation.1998;98:663e670)

Lead failure requiring extractionor additional lead insertion

5.467.28 4.3% (PACE.2005;28:926e932)

Major haematoma requiringreoperation

2.7263.07 5.8% (JAMA.2006:295:1901e1911)

Device infection requiringremoval/extraction

2.2762.4 0.5% (PACE.2005:28:926e932); 0.77%(Circulation.1998;98:663e670);0.7%(MADIT2 trial)

Cardiac tamponade 0.761.07 0.2% (PACE.2005;28:926e932);0.64% (Circulation.1998;98:663e670)

Pneumothorax 1.6861.17 1.1% (PACE.2005; 28:926e932);0.89% (Circulation.1998;98:663e670)

Inappropriate shocks 14.8610.92 12% (PACE.2005; 28:926e932); 14.91%(Circulation.1998;98:663e670);18% (Z Kardiol.1996;85:809e819)

Psychological problemsassociated with the device

22.6626.68 13e38% (Clin Cardiol 1999;22:481e9)

158 IS IT COST EFFECTIVE TO USE A PLUGGED LV PORT?

doi:10.1136/heartjnl-2011-300198.158

M A Jones, T R Betts, K Rajappan, Y Bashir, K C K Wong, N Qureshi. John RadcliffeHospital, Oxford, UK

Background Many patients receiving ICD implants do not meetcriteria for CRT therapy, yet are often felt likely to benefit from CRTin the future. The reasons for this include less severe NYHA class ofHF symptoms at the time of implant, narrow QRS, and (progres-sive) atrio-ventricular conduction delay. Management optionsinclude only implanting DDD / VVI devices, and then upgrading toCRT if required; implanting CRT-D devices but without an LV lead,with the LV port “plugged”, such that if an upgrade were to becomenecessary, only a new LV lead (and implant kit) would be required;and finally, implanting CRT-D devices with LV leads in all patientsin the first instance, as has been suggested by the recent Madit-CRTand RAFTstudies. It is not clear which of these strategies is superiorin terms of the cost-benefit ratio.Purpose This study analyses a retrospective cohort of patients whoreceived CRT-D devices but without LV leads, to examine the costimplications of this approach, and to compare this cost to that of merelyimplanting a DDD device, or implanting a full CRT-D system initially.Method A retrospective analysis of all patients receiving CRT-ICDswith plugged LV ports between September 2004 and June 2009 atour institution. Patient characteristics, indication for a plugged LVport, subsequent addition of a LV lead and reasons for doing so weretaken from patient records. The total cost (surgery and hardware)was compared with the estimated cost of initially implanting singleor dual chamber ICDs and upgrading the entire system, and to thecost of implanting full CRT-D systems up front.Results 35 patients (27 male) were identified. Mean (SD) age was6768 years. 26 had ischaemic heart disease and 9 non-ischaemicdilated cardiomyopathy. All had LV EF<30%. Indications for a pluggedLV port were LBBB and NYHA class I or II symptoms in 29 andNYHA class I or II with a narrow QRS but a high chance of becomingpacemaker dependent in 6. During a mean (SD) FU of 40 616months, 6 (17%) patients had an LV lead added, all for the develop-ment of NYHA III symptoms, at 10, 11, 15, 17, 17 and 21 monthsrespectively. Total cost at end of FU period was £ 654000. If allpatients had initially been implanted with VVI or DDD ICDs and 6new CRT systems implanted, the estimated cost would have been £

598000. If all patients had received full CRT-D the cost would havebeen £ 665000. Taking into account the time to develop symptoms, itis predicted that an upgrade rate of 26%e31% would be requiredbefore using a plugged LV port becomes cost-effective. Furthermore,full CRT-D system implantation is even less cost effective.Conclusion In this series of ICD patients with potential CRT indi-cations but minimal heart failure symptoms, only a small proportionsubsequently required biventricular pacing. Using a CRT-ICD with aplugged LV port is not a cost effective strategy (Abstract 158 figure 1).

Abstract 158 Figure 1 Per cent freedom from upgrade to LV lead.


BCS Abstracts 2011

159 PILOT STUDY EXPLORING THE REGIONAL REPOLARISATIONINSTABILITY INDEX IN RELATION TO MYOCARDIALHETEROGENEITY AND PREDICTION OF VENTRICULARARRHYTHMIA AND DEATH

doi:10.1136/heartjnl-2011-300198.159

1W B Nicolson, 1C D Steadman, 1P Brown, 2M Jeilan, 2S Yusuf, 2S Kundu,2A J Sandilands, 2P J Stafford, 1F S Schlindwein, 2G P McCann, 1G A Ng. 1Universityof Leicester, Leicester, UK; 2University Hospitals of Leicester NHS Trust, Leicester, UK

Introduction There is a need for better sudden cardiac death (SCD)risk markers. Mounting evidence suggests that the mechanismunderlying risk of ventricular arrhythmia (VA) is increased hetero-geneity of electrical restitution. We investigated a novel measure ofaction potential duration (APD) restitution heterogeneity: theRegional Repolarisation Instability Index (R2I2) and correlated itwith peri-infarct zone (PIZ) a cardiac magnetic resonance (CMR)anatomic marker of VA risk.Methods Blinded retrospective study of 30 patients with ischaemiccardiomyopathy assessed for an implantable cardioverter defib-rillator. The R2I2 was derived from high resolution 12 lead ECGrecorded during programmed electrical stimulation (PES). ECGsurrogates were used to plot APD as a function of diastolic interval;the R2I2 was the maximal value of the mean squared residuals ofthe mean points for anterior, inferior and lateral leads normalised tothe mean value for the total population. PIZ was measured from lategadolinium enhanced CMR images using the full width halfmaximum technique.Results Seven patients reached the endpoint of VA/death (medianfollow-up 24 months). R2I2 > median was found to be predictive ofVA/death independent of PES result, left ventricular ejection frac-tion and QRS duration (6/14 vs 1/15 p¼0.031). Modest correlationwas seen between the R2I2 and PIZ (r¼0.41 p¼0.057) (Abstract 159figure 1).


Conclusions In this pilot study of ischaemic cardiomyopathypatients, the R2I2 was shown to be an electrical measure of VA/death risk with a moderately strong correlation with an anatomicmeasure of arrhythmic substrate, the extent of PIZ. The R2I2 mayadd value to existing markers of VA/death and merits furtherinvestigation.


VariableWhole Group(n[30)

No VA/death(n[23)

VA/death(n[7) p

Age (years) 6769 6569 7268 0.055

Sex (% male) 97 96 100

QRSD(ms) 107620 107621 106615 0.95

EF(%) 31614 32.4615 2767.5 0.34

PES result (positive/total) 12/30 7/23 5/7 0.068

R2I2>median 14/29 8/22 6/7 0.031

EDV index (ml/cm) 1.4860.41 1.4960.41 1.4560.45 0.84

SV index (ml/cm) 0.4260.14 0.4360.14 0.3960.15 0.47

Follow-up (months) 24 (18) 24 (16) 16 (16) 0.088

PIZ % 7.8 (10.7) 7.5 (8.4) 13.6 (8.5) 0.093

Scar % 10.9 (16.5) 9.67 (13.5) 21.9 (17.8) 0.16

160 HIGH DOSE OCTREOTIDE; A NOVEL THERAPY FOR THETREATMENT OF DRUG REFRACTORY POSTURALORTHOSTATIC TACHYCARDIA SYNDROME IN PATIENTSWITH JOINT HYPERMOBILITY SYNDROME

doi:10.1136/heartjnl-2011-300198.160

A E French, C Shepherd, A Horne, C Parker, J Tagney, J Pitts-Crick, T Johnson,G Thomas. Bristol Heart Institute, Bristol, UK

Introduction Postural orthostatic tachycardia syndrome (POTS) isdefined as symptomatic orthostatic intolerance with an increase inheart rate of 30 beats per minute within 10 min of head up tilt(HUT). This dysautonomia causes wide-ranging symptomsincluding palpitations, presyncope, chronic fatigue, headache andcognitive difficulties. When POTS occurs in patients with pre-existing Joint Hypermobility Syndrome (JHS), symptoms beginapproximately a decade earlier than non-JHS patients with apreponderance of neurological features, secondary to cerebral hypo-perfusion. Vascular laxity with splanchnic venous pooling has beenimplicated as a causative factor thus measures to expand plasmavolume (thereby increasing mean arterial pressure and restoringcerebral perfusion) form the mainstay of therapy. Symptomaticimprovements have been previously reported in POTS patients withthe somatostatin analogue Octreotide, a powerful splanchnic vaso-constrictor. We report the first UK series of JHS patients with drugrefractory POTS treated with high-dose octreotide.Methods Six patients (female, aged 21e52) were referred to ourinstitution. All had known JHS (4 requiring a wheelchair), neuro-logical symptoms (headache and cognitive impairment) and diag-nostic tilt-table testing with a mean increase in heart rate of64 beats/min (range 47e73) with head-up tilt (HUT). All patientshad remained symptomatic despite pre-treatment with a mean of 5POTS medications (range 5e7) including fludrocortisone, mido-drine, propranolol, ivabradine, selective serotonin reuptake inhib-itors, gabapentin and erythropoietin. Octreotide was commencedusing a short-acting preparation given 3 times daily (dosage50e250 mg according to body mass) in conjunction with a long-acting (monthly), intramuscular injection (dosage 10e30 mg). Theshort-acting preparation was weaned following the second monthlyinjection.Results During follow-up of 3 months (range 1e8), 3 (50%) patientsreported a complete resolution of all postural and neurologicalsymptoms which corresponded with a normalised response to HUT.The remaining patients reported a dramatic improvement butongoing postural symptoms. No patients developed supine hyper-tension. Side effects including mild abdominal discomfort andtransient diarrhoea were reported in 3 (50%) patients.Conclusion Octreotide is increasingly recognised as an effectivetherapy in POTS patients. Both short-acting, subcutaneous (0.9 mg/Kg) and long-acting, intramuscular (10e20 mg) preparations have


BCS Abstracts 2011

previously been reported. We conclude that higher dosages of bothpreparations when administered together are effective and welltolerated in JHS patients with drug refractory POTS.

161 CATHETER ABLATION OF ATRIAL FIBRILLATION ONUNINTERRUPTED WARFARIN USING STANDARD AND DUTYCYCLED RADIOFREQUENCY ENERGY: SAFE AND EFFECTIVE

doi:10.1136/heartjnl-2011-300198.161

J R J Foley, N C Davidson, B D Brown, D J Fox. University Hospital of SouthManchester, Manchester, UK

Introduction Catheter ablation (CA) for atrial fibrillation (AF) isgrowing exponentially. Although ablation for paroxysmal AF (PAF)is associated with shorter procedure times and less extensive leftatrial ablation vs persistent AF thromboembolic complications canoccur in both sub-groups. Inadequate anticoagulation leads tothrombotic complications and excessive anticoagulation can lead tobleeding risks. Many centres adopt a policy of discontinuingwarfarin in the immediate run-up to the procedure, covering theprocedure with unfractionated heparin and “bridging” postoperativepatients with low molecular weight heparins (LMWH) back ontowarfarin. We wished to determine the safety of CA for AF with atherapeutic INR using both the single transseptal approach andduty cycled radiofrequency energy (RF) with non irrigated PVACcatheters and the double transseptal puncture technique using irri-gated RF catheters and either CARTO or NAVX electroanatomicalmapping.Methods A retrospective analysis of 173 patients who underwentCA for AF while taking uninterrupted warfarin. Procedural targetInternational Normalised Ratio (INR) was 2e3 with a peri-proce-dural target ACT of 300e350 s. In sub therapeutic INR patientsweight adjusted LMWH was used post procedure with warfarinuntil INR was >2. Standard technique employed was large areacircumferential ablation using conventional RF energy or pulmonaryvein isolation using duty cycled RF energy. Data was gathered fordemographics, procedural INR, total dose of unfractionated heparin,fluoroscopy time, and type of radiofrequency energy used.Endpoints were minor bleeding, major bleeding (requiring trans-fusion), vascular complications, pericardial tamponade and stroke/TIA within 28 days of the procedure.Results There were 128/173 male patients, age range between 21and 73 years (mean 57 years). 122 underwent ablation for PAF and51 for persistent AF. Mean procedural INR was 2.4 (range 1.7e3.9).Mean unfractionated heparin dose was 6000 units (range1000e14 500). Mean fluoroscopy time for the PVAC group was23.4 mins (range 8.3e50.1 mins). Mean fluoroscopy time forCARTO/NAVX group was 31mins (range 14.10e58.44 mins). Therewere no major bleeding complications. There was 1 minor bleedingcomplication with a groin pseudoaneurysm. There were 2 cases ofpericardial tamponade (2/173%e1.2%) both managed with percu-taneous pericardial drainage. There were no stroke/TIAs.Conclusion These data demonstrate that CA for AF by both singleand double transseptal technique using both standard RF and dutycycled RF while maintaining a therapeutic INR is a safe procedure.Maintaining a therapeutic INR reduces the risk of embolic eventsassociated with “bridging” heparin without an increase in bleedingcomplications. This technique is convenient for patients and avoidsswitching between LMWH and warfarin and ensures patient safetyby maintaining therapeutic anticoagulation before, during and afterthe procedure.

162 THE CLINICAL MANAGEMENT OF RELATIVES OF YOUNGSUDDEN ARRHYTHMIC DEATH VICTIMS; ICDS ARE RARELYINDICATED

doi:10.1136/heartjnl-2011-300198.162

1J C Caldwell, 1Z Borbas, 2N Moreton, 2N Khan, 2L Kerzin-Storrar, 2K Metcalfe,2W Newman, 1C J Garratt. 1Manchester Heart Centre, Central Manchester UniversityNHS Foundation Trust, Manchester, UK; 2Department of Clinical Genetics, St Mary9sHospital, Central Manchester University NHS Foundation Trust, Manchester, UK

Introduction Following National Service Framework guidance on themanagement of sudden cardiac death (SCD), regional inheritedcardiac conditions clinics were established to identify and treat thoseat increased risk of dying from an arrhythmic condition. Studies haveexamined the yield of different diagnostic strategies but the outcomeof management in these patients has not been reported.Methods We present data on 193 consecutive patients (108 families)referred to a regional inherited cardiac conditions clinic because ofSCD/aborted cardiac arrest of a relative. The mean age on referralwas 38617 yrs and mean duration of follow-up was 15 months(range 1 day to 56 months). All individuals underwent clinicalassessment by history, examination, ECG and echo. If treadmillexercise test had not previously been performed this was under-taken. Further imaging by CMR or contrast echo was performed inthose with structurally abnormal hearts or with Twave inversion inV2/V3. Ajmaline provocative testing was performed in those with ahistory and/or ECG suggestive of Brugada syndrome.Results Of the 193 patients, 43 individuals (22%) from 36 families(33%) were diagnosed with an inheritable cause of SCD and 145patients were clinically normal (see Abstract 162 table 1). Fivepatients were found to have other conditions (LV non-compaction,AVNRT, skeletal myopathy, mild AS and congenital sub-aorticmembrane). Of the 43 patients diagnosed with an inheritablecondition, 21 had medication commenced by the clinic (b-blockers(21), ACEi/ARB(2), Spironolactone[1]). ICDs were implanted as perHRUK guidelines, resulting in 4 patients having an ICD inserted onclinic recommendation (2 HCM, 1 DCM, 1 ARVC). To date noappropriate therapies have been administered (follow-up 8e29months) but there was 1 inappropriate shock from a fractured lead.Three individuals had b-blockade withdrawn after negative genetictesting for an established familial mutations (2 CPVT, 1 LQT), oneICD was removed and one deactivated (both negative for CPVT). Ofthe 145 patients thought to be clinically normal, 85 were reassuredand discharged, 13 failed to return to clinic and 47 are regularreviewed as the risk of developing an inheritable condition cannot beexcluded; this includes those with family histories of HCM (7),ARVC (12), DCM (9), CPVT (5), Brugada (4) and LQT(1). To dateno deaths have occurred in those diagnosed with inherited causes ofSCD (follow-up mean 20, 1e52 range) or those clinically normalwith ongoing review (follow-up mean 22 months, 1e56 range).


Diagnosis of patient Numbers

Clinically normal 145

LQTS 12

Brugada 2

CPVT 5

ARVC 7

DCM 7

HCM 10

Conclusion A diagnosis of an inheritable cause of SCD was obtainedin 22% of individuals and 33% families with a history of SCD/aborted cardiac arrest in a relative. The number of ICDs inserted wasvery small (2%) and there have been no appropriate therapies in this


BCS Abstracts 2011

group. Two ICDs have been removed/deactivated after exclusion ofa known familial mutation.

163 THE UNITED KINGDOM TRANSCATHETER AORTIC VALVEREGISTRY - OUTCOMES TO DECEMBER 2009 AND UPDATE

doi:10.1136/heartjnl-2011-300198.163

P F Ludman. On Behalf of the UK TAVI Steering Group, Birmingham, UK

Introduction The United Kingdom Transcatheter Aortic Valve (TAVI)Registry was established to define the clinical and procedural detailsof all patients being treated by TAVI, regardless of access route ortechnology used, and to assess their outcomes. The registry hascaptured all cases in England and Wales.Methods For every TAVI performed, all centres complete an agreeddataset. The data are encrypted and sent electronically to servers atthe Central Cardiac Audit Database (CCAD) for analysis. A uniquepatient identifier (the NHS number) is used to link with the NHSCentral Register to track mortality. This analysis is based on allprocedures performed up to 31st December 2009.Results 25 centres in England and Wales performed a total of 877procedures in 870 patients; 66 in 2007, 273 in 2008 and 538 in 2009.Median number of cases per centre was 24. Median (IQR) age was82 (78e87) yrs. 52% were male, and mean logistic Euroscore (LES)was 22.2%. 68% were transfemoral, the remainder being mainlytransapical. 90% of CoreValve implants and 46% of Edwards used atransfemoral approach. Patients needing a transapical route hadmore comorbid conditions than those treated by a transfemoralroute (LES 25.2% vs 20.9%). Mortality tracking was successful in100% of patients. Survival at 30 days was 93.1%, 78.9% at 1 year(443 at risk) and 72.3% at 2 years (114 at risk). Survival wassignificantly poorer in those needing non-transfemoral approaches(1 year survival 73.5% vs 81.4%). Survival was also poorer in thosewith poorer LV ejection fraction, with moderate or severe postprocedural aortic regurgitation and with a LES>40. Survival was notassociated with age, NYHA class, or the presence of concomitantcoronary artery disease, and not different between those with aLES<21 compared with LES 21e40. There was also no difference insurvival between patients treated with CoreValve or Edwardstechnologies, or between proctored and non proctored cases. Therewas a significant improvement in survival over the 3 years of theregistry, and a change in demographics with the proportion ofpatients with prior CABG rising from 16.4% in 2007 to 29.9% in2009. The total number of cases in the UK TAVI registry has risen to1490 as of 29 Nov 2010. More details of the 2010 cohort will beavailable at the time of presentation (Abstract 163 figure 1).


Conclusions The UK TAVI Registry has successfully captured theentire TAVI activity of England and Wales incorporating the learningcurves of every centre. Outcomes following TAVI are excellent in ahigh risk patient population. Outcomes are better in the populationwho are suitable for a transfemoral approach (with less comor-bidity) than those treated with the transapical approach. These datasuggest that careful proctoring allows the introduction of a newtreatment method without an adverse effect on patient outcome,and we have demonstrated no systematic difference in outcomebetween the two commercially available technologies.

164 EARLY HAEMODYNAMIC CHANGES AND MYOCARDIALINJURY AFTER TRANSFEMORAL TRANSCATHETER AORTICVALVE IMPLANTATION (TAVI)

doi:10.1136/heartjnl-2011-300198.164

R Dworakowski, A Bhan, B Brickham, O Wendler, M Monaghan, A M Shah,P MacCarthy. Kings College Hospital, Kings Health Partners, London, UK

Purpose Transfemoral (TF) TAVI is a novel procedure for thetreatment of severe aortic stenosis, without the need for thor-acotomy or cardiopulmonary bypass. The procedure results inalmost instantaneous normalisation of transvalvular gradients, butlittle is yet known about the periprocedural haemodynamic effects.We aimed to describe these effects using 3D and tissue Doppler (tD)transthoracic echocardiography (TTE) and Cardiac Output monitoring.Methods In 16 patients undergoing TF TAVI haemodynamics werecharacterised with a number of tD and 3D TTE measurements.



BCS Abstracts 2011

These were taken at multiple time points (baseline, 6 and 24 hourspost procedure). Calculated volumetric parameters included 3D end-diastolic volume (EDV) and end-systolic volume (ESV), strokevolume (SV) and 3D LA volume (LAV). Diastolic function wasmonitored using the indices mean E:E9 and systolic function/contractility was measured with dP/dt max and early peak systolicvelocity (S9). The FloTrac system (consisting of the Vigileo monitorand sensor), uses a clinically validated algorithm to providecontinuous cardiac output (CO), stroke volume (SV) and systemicvascular resistance in real-time.Results TAVI resulted in an immediate increase in cardiac output(3.7 (baseline), 4.6 (6 h) 4.5 l/min (24 h), p<0.5 baseline vs 6 h and24 h) with no significant change in systemic vascular resistance(1162, 1292 and 1367 dyn*s/cm5). However, 6 h post-TAVI therewas a significant decrease in systolic function as measured by dP/dtmax/EDV (see Abstract 164 figure 1A) and co-existent impairmentof diastolic function as indicated by medial E:E9 values (see Abstract164 figure 1B), which was associated with an appropriate increase inLA volume (70.3, 82.6 and 72.8 ml, p<0.05 baseline vs 6 h).Following this, there was a recovery of both systolic and diastolicindices. In addition, another marker of systolic function, S9 increasedafter 24 h (6.4, 6.6, 8.2 cm/s, p<0.05 baseline vs 24 h and 6 h vs24 h). Concurrent with this recovery, we observed a significantdecrease in EDV and ESV at 24 h post-TAVI (EDV: 94.9 to 83.4 ml(p<0.05); ESV 41.9 to 33.5 ml (p<0.05)). These changes in haemo-dynamics were associated with significant increase of troponin Ilevels at 24 h and increase in CK-MB at 6 h after the procedure(troponin: 0.06 vs 1.19 mg/l, p<0.05; CK-MB 1.6 vs 6.6 mg/l, p<0.05).Conclusion Successful TF TAVI results in an immediate improve-ment in cardiac output. However, overlying this, within the first24 h both systolic and diastolic dysfunction occurs. The rise in themarkers of myocardial injury suggest this may be due to myocardialstunning and/or some periprocedural myocardial damage. Recoveryof contractility is observed after 24 hours.

165 THE OXVALVE STUDY: ECHOCARDIOGRAPHIC SCREENINGFOR VALVULAR HEART DISEASE IN THE COMMUNITYSETTING: METHODOLOGY, FEASIBILITY AND PRELIMINARYRESULTS

doi:10.1136/heartjnl-2011-300198.165

1J L d’Arcy, 2D Ebbs, 3P Grimwade, 4A J Farmer, 4D Mant, 1B D Prendergast. 1JohnRadcliffe Hospital, Oxford, UK; 2Didcot Health Centre, Oxford, UK; 3Bampton MedicalPractice, Oxford, UK; 4Department of Public Health and Primary Care, University ofOxford, Oxford, UK

Introduction Valvular heart disease (VHD) is poorly researched incomparison with other areas of cardiovascular disease. Principlelimitations are the diverse nature of patients with VHD, inability toidentify individuals at the earliest stages of disease and lack of anappropriate investigational infrastructure. Studies addressing thecontemporary epidemiology and natural history of VHD are scarcebut demonstrate an increasing prevalence in the elderly, associatedwith significant morbidity and mortality. Cohort studies in the USAare ongoing but there are no European or UK studies to date. Wehave developed a large scale, prospective community echocardio-graphic screening study within the adult Oxfordshire population, todetermine the epidemiological characteristics of VHD in the UK forthe first time, to assess the acceptability of echocardiographic screeningfor VHD, and establish cohorts with well-characterised genetic andechocardiographic phenotypes for future study. Herein, we presentpreliminary data for the first 1050 patients, with enrolment ongoing.Methods Patients >65 years, registered with participating generalpractices (GP) and with no known VHD, were invited to attendtheir GP surgery where routine demographic and cardiac data werecollected and a focused examination undertaken. Participants under-went a standard transthoracic echocardiogram (TTE) according to

British Society of Echocardiography guidelines. The threshold forinclusion in the screen positive group was deliberately low to captureall manifestations of VHD. Participants were given preliminaryresults, before completing a shortened Spielberger STAI questionnaire.Results Uptake was 46% (age range 65e96 years; male to femaleratio 1:1.1). VHD was detected in 33% of participants and preva-lence increased with increasing age (see Abstract 165 figure 1).Mitral regurgitation and aortic regurgitation were the mostcommon lesions detected (present in 17% and 14% respectively).The majority of VHD was graded as mild (84%); only 1% of VHDdetected was severe. The majority of participants (99%) describedthemselves as calm or relaxed at the time of screening; noneexpressed significant levels of worry or tension. 98% would beprepared to undergo repeat echocardiography as screening for VHD.

Abstract 165 Figure 1 Prevalence of VHD in $65s by age group.

Conclusions The prevalence of VHD in adults aged over 65 in theOxfordshire population, using a low threshold for detection, isapproximately 33% and increases with age. Mitral regurgitation isthe most common lesion, and the majority of detected VHD is mild.Echocardiographic screening for VHD is feasible in the primary caresetting and acceptable in this group of patients.

166 CARDIOVASCULAR MAGNETIC RESONANCE (CMR)TAGGING IDENTIFIES DIFFERENTIAL VENTRICULARREMODELLING IN PATIENTS WITH BICUSPID VS TRICUSPIDAORTIC VALVE DISEASE

doi:10.1136/heartjnl-2011-300198.166

S Bull, J Suttie, N Blundell, J M Francis, T D Karamitsos, A Pitcher, J D’Arcy,B Prendergast, S Neubauer, S G Myerson. John Radcliffe Hospital, Oxford, UK

Background Bicuspid aortic valves (BAV) are a common inheritedabnormality with a very high rate of adverse cardiac events at anearlier age than tricuspid aortic valves (TAV). Risk stratification formoderate to severe aortic stenosis, in both bicuspid and tricuspiddisease, remains a significant clinical challenge. It is unknownwhether pathological left ventricular (LV) remodelling, a strongpredictor of adverse cardiac events, differs between patients withbicuspid and tricuspid valvular disease with comparable trans-valvular gradients. Cardiovascular magnetic resonance (CMR)tagging provides detailed characterisation of global and regionalcontractility, and is a powerful investigative tool in the assessmentof myocardial disease. We therefore assessed left ventricular strain(using CMR tagging), valve morphology and LV hypertrophy inpatients with bicuspid and tricuspid aortic valve disease matched fortransvalvular gradient.Methods 42 subjects were recruited in total: 24 patients withmoderate to severe BAV (age 55 615 yrs, female 21%, peak trans-


BCS Abstracts 2011

aortic velocity 3.160.6/ms, LV mass 172648 g; SBP 127 614 mmHg DBP 76 610 mm Hg) and 18 patients with velocity-matchedTAV (age 74 66 years, female 28%, velocity 3.160.6/ ms, LV mass147627 g; SBP 136617 mm Hg; DBP 7967 mm Hg; Abstract 166table 1). Patients were scanned using a 1.5 TAvanto scanner (SiemensHealthcare, Erlangen, Germany) and basal, mid-ventricular and apicalshort axis tagging images were acquired. Peak systolic globalcircumferential strain was calculated at each ventricular level usingCimTag2D software v.7 (AucklandMRI Research Group, New Zealand).

Abstract 166 Table 1 Bicuspid vs tricuspid aortic valve disease

Morphology

Bicuspidaortic valve(BAV)

Tricuspid aorticvalve (TAV) p Value

Age (yrs) 55615 7466 <0.01

Female (%) 21 28 0.31

Transvalvular velocity (ms�1) 3.160.6 3.160.6 0.50

Left ventricular mass (g) 172648 147627 0.04

Systolic blood pressure (SBP) 127614 136617 0.04

Diastolic blood pressure (DBP) 76610 7967 0.15

Basal circumferential systolic strain (%) 2063 2263 0.04

Mid ventricular circumferential strain (%) 1962 2162 0.07

Apical ventricular circumferential strain (%) 1764 1963 0.04

Results Patients with BAV had significantly greater left ventricularhypertrophy (BAV 172 648 g vs TAV 147627 g; p¼0.04) despitesimilar degrees of valve stenosis. Peak systolic circumferential strainwas lower (ie, reduced contractility) in patients with BAV than TAV(Basal BAV 2063% vs TAV 2263% P¼0.04; Mid BAV 19 6 2% vsTAV 21 6 2% p¼0.07; apical BAV 1764% vs TAV 1963%; p<0.04),despite the younger age and lower blood pressure.Conclusion Ventricular remodelling differs between BAV and TAVpatients with equivalent transvalvular gradients. BAV patients,despite being younger and having lower systolic blood pressure, havemore severe hypertrophy and lower myocardial contractility. Thisfinding may have implications for monitoring disease progressionor more timely medical or surgical intervention in patients with BAV.

Abstract 166 Figure 1 Velocity matched tricuspid and bicuspid aorticvalves showing valve morphology, tagging and cine imaging.

167 AORTIC REGURGITATION QUANTIFICATION WITHCARDIOVASCULAR MAGNETIC RESONANCE PREDICTSCLINICAL OUTCOME

doi:10.1136/heartjnl-2011-300198.167

1S G Myerson, 1J D’Arcy, 2R Mohiaddin, 3J P Greenwood, 1T D Karamitsos,1J M Francis, 1A P Banning, 4J P Christiansen, 1S Neubauer. 1John Radcliffe Hospital &University of Oxford, Oxford, UK; 2CMR Unit, Royal Brompton Hospital and the NationalHeart and Lung Institute, London, UK; 3University of Leeds, Leeds, UK; 4North ShoreHospital, Auckland, New Zealand

Background The timing of valve surgery in asymptomatic patientswith significant aortic regurgitation can be challenging. Currentindications focus on symptoms and left ventricular (LV) dilation/dysfunction, but prognosis is already reduced by this time. Quan-tification of the regurgitation has not previously been used to guidemanagement, likely due to the difficulty of achieving this withechocardiography. Cardiovascular magnetic resonance (CMR) canaccurately quantify aortic regurgitation and LV volumes, and weexamined whether either could predict symptom development andthe need for aortic valve surgery.Methods 94 asymptomatic patients with moderate or severe aorticregurgitation on echocardiography were identified from four highvolume CMR centres. CMR scans were performed to quantify aorticregurgitation and LV volume indices, and subsequent clinical follow-up occurred for up to 7 years (mean 2.662.1 years). The bestpredictors of progression to symptoms and other conventionalindications for surgery were determined.Results Aortic regurgitant fraction was the best predictor of clinicaloutcome; area under the curve (AUC) on receiver operating char-acteristics analysis 0.93 (p<0.0001), with a specificity of 93% andsensitivity of 78% for predicting the progression to symptoms andsurgery. Survival without surgery was 88% for patients with aregurgitant fraction <37%, compared to 6% for those with aregurgitant fraction $37% (see Abstract 167 figure 1). Regurgitantvolume >38 mls and regurgitant volume index >25 ml/m2 were alsogood predictors (AUC 0.91 and 0.90 respectively), though regur-gitant fraction had significantly greater predictive power (OR 1.26compared to 1.09 for regurgitant volume). LV volumetric indices alsopredicted outcome, but less strongly than measures of regurgitation:LV end-diastolic volume >267 mls (AUC 0.85), LV end-systolicvolume >88 mls (AUC 0.78). Regurgitant fraction and volume werethe only independent outcome predictors on multiple logisticregression analysis. The predictive ability of CMR applied topatients with both moderate and severe aortic regurgitation onechocardiography. Supporting data also comes from a comparison

Abstract 167 Figure 1 Kaplan-Meier survival curve showing survivalwithout surgery for conventional indications.


BCS Abstracts 2011

with patients already planning to undergo surgery at the time ofCMR scanning, which showed similar regurgitant fractions in thesurgical group (mean 6SD: 45.4%612.1%) compared to the initiallyasymptomatic patients who developed symptoms or other indica-tions for surgery (42.8%610.4%); p¼0.32. Subjects who remainedasymptomatic had a significantly lower regurgitant fraction:25.368.6% (p<0.0001 vs both the planned surgical group and thesymptom progression group).Conclusions CMR quantification of aortic regurgitation and LVvolumes accurately predicts the progression to symptoms/surgery.Its use in patients with aortic regurgitation should be encouraged.

168 b-BLOCKER THERAPY IMPROVES CLINICAL OUTCOMES INPATIENTS WITH MODERATE TO SEVERE MITRALREGURGITATION

doi:10.1136/heartjnl-2011-300198.168

A Nadir, D H Elder, M G MacArtney, S D Pringle, A M Choy, A D Struthers, C C Lang.University of Dundee, Dundee, UK

Background Volume overload seen in mitral regurgitation (MR)leads to neuro-endocrine activation including heightened sympa-thetic activity. Experimental data have reported protective effects ofbeta-blockers (BBs) on myocardial function in MR suggesting thatBB therapy may be beneficial in MR. However, the effect of BBtherapy on clinical outcomes in MR has not been defined. Hence, inthis large observational study we investigated the impact of BBtherapy on clinical outcome in patients with moderate to severeMR.Methods The Health Informatics dispensed prescribing, morbidityand mortality database for the population of Tayside, Scotland waslinked through a unique patient identifier to the Tayside echo-cardiography database (>110 000 scans). Patients with a diagnosis ofmoderate or severe MR from 1993 to 2008 were identified. Coxregression model (adjusted for age, gender, left ventricular dimen-sions and function, left atrial diameter, cardiovascular (CV) historyand concurrent CV medications) was used to assess the impact ofBB therapy on all-cause mortality and cardiovascular events (CVdeath or hospitalisations).Results A total of 4437 patients with moderate to severe MR (meanage 74 Â611 years, 46% males) were identified. MR was categorisedas functional in 2523 (57%) and organic in 1894 (43%) while 1324(30%) were on BBs. Over a mean follow-up of 3.9 years there were2287 (51%) all-cause deaths and 2333 (52%) CV events. Thosetreated with BBs had a significantly lower all-cause mortality withan adjusted HR of 0.65 (95% CI 0.56 to 0.75, p<0.0001) and fewerCV events with an adjusted HR of 0.79 (95% CI 0.69 to 0.90,p<0.0001).Conclusions This large observational study suggests that BB therapyis associated with an improved survival and a lower risk of CVevents in patients with moderate to severe MR. These observationsneeds to be confirmed in prospective studies and support therationale for undertaking a future randomised clinical trial.

169 MID-WALL FIBROSIS IS AN INDEPENDENT PREDICTOR OFMORTALITY IN PATIENTS WITH AORTIC STENOSIS

doi:10.1136/heartjnl-2011-300198.169

1,2M R Dweck, 1S Joshi, 1T Murigu, 1A Gulati, 1F Alpendurado, 1R Mohiaddin,1J Pepper, 1D Pennell, 2D Newby, 1S Prasad. 1Royal Brompton Hospital, London, UK;2Centre for Cardiovascular Sciences, University of Edinburgh, Edinburgh, UK

Introduction Predicting adverse clinical outcomes in aortic stenosis ischallenging. Late gadolinium enhancement (LGE) has been asso-

ciated with an adverse prognosis in a range of other cardiac condi-tions. Using late gadolinium enhancement, we sought to assess theprognostic significance of mid-wall and infarct patterns of myocar-dial fibrosis in aortic stenosis.Methods Between January 2003 and October 2008, consecutivepatients with moderate or severe aortic stenosis (aortic valve area<1.5 cm2) underwent cardiovascular magnetic resonance withassessment of myocardial fibrosis by late gadolinium enhancement.Patients were categorised into absent, mid-wall or infarct patterns oflate gadolinium enhancement by blinded independent observers.Patient follow-up was completed using the National StrategicTracing Scheme.Results 143 patients (aged 68614 years; 97 male) were followed upfor 2.061.4 years. 81 patients had coronary artery disease, 72underwent aortic valve replacement and 27 died. Compared to thosewith no late gadolinium enhancement (n¼49), univariate analysisrevealed that patients with mid-wall fibrosis (n¼54) had an eight-fold increase in all-cause mortality despite similar aortic stenosisseverity and coronary artery disease burden. Patients with an infarctpattern (n¼40) had a six-fold increase. Mid-wall fibrosis (HR, 5.35(95% CI, 1.16 to 24.56); p¼0.03) and ejection fraction (HR 0.96(95% CI, 0.94 to 0.99); p¼0.01) were independent predictors of allcause mortality by multivariate analysis. Conclusion: Mid-wallfibrosis is an independent predictor of mortality in patients withmoderate and severe aortic stenosis. It has incremental prognosticvalue to ejection fraction and may provide a useful method ofrisk stratification in patients with advanced disease (Abstract 169figure 1).

Abstract 169 Figure 1 Kaplan-Meier curves of cardiac mortality (left)and all cause mortality (right) according to pattern of LGE (A¼ No LGE,B¼ Infarct LGE, C¼ Mid-wall LGE).


No LGE Mid-wall LGE Infarct LGE p Value

Number of patients 49 54 40 e

Mean age yrs 64616 70611 70613 0.031

Documented CAD % 37 42 98 <0.001

Ejection fraction % 69613 58621 44618 <0.001

Aortic valve area 1.0560.37 1.0060.31 0.9160.26 0.111

Indexed LV mass g/m2 92.6*(86.0, 99.6)

113.7*(104.5, 123.8)

97.8*(90.9, 105.2)

0.005

Mortality rate (deaths /1000 pt years)

15.7 142.7 173.7

* Geometricmean (95%)


BCS Abstracts 2011

170 ETHNIC DIFFERENCES IN PHENOTYPIC EXPRESSION OFHYPERTROPHIC CARDIOMYOPATHY

doi:10.1136/heartjnl-2011-300198.170

N Sheikh, M Papadakis, N Chandra, H Raju, A Zaidi, S Ghani, M Muggenthaler, S Gati,S Sharma. St. George’s University of London, London, UK

Purpose Hypertrophic Cardiomyopathy is a heterogeneous condi-tion with variable phenotypic expression. Current studies are basedon predominantly Caucasian cohorts (white patients; WP), there-fore the phenotypic manifestations of HCM in individuals ofAfrican/Afro-Caribbean origin (black patients; BP) are not fullyrealised. Data in athletes and hypertensive patients indicate thatblack ethnicity is associated with a greater prevalence of repolar-isation abnormalities on the ECG as well as a greater magnitude ofleft ventricular hypertrophy (LVH), highlighting the importance ofdefining the HCM phenotype in this ethnic group.Methods Between 2001 and 2010, 155 consecutive patients withHCM (52 BP, 103 WP) were assessed in 3 specialist cardiomyopathyclinics in South London. All individuals underwent comprehensive

cardiac evaluation including 12-lead ECG and echocardiography.Patients subject to therapeutic interventions potentially affectingrepolarisation patterns were excluded.Results Black patients revealed significantly different echocardio-graphic patterns of LVH, with more concentric (44.2% vs 30.1%)and apical (28.8% vs 11.7%) hypertrophy compared to WP whoexhibited more asymmetric septal hypertrophy (57.3% vs 25.0%)(p¼0.004). Black patients exhibited a similar magnitude of LVHcompared to WP (17.364.9 vs 18.864.1 mm, p¼0.069). Relating toECG repolarisation abnormalities, BP exhibited more Twave inver-sions in the lateral leads (76.9% vs 60.2%, p¼0.038) and deep($�0.2 mV) T-wave inversions (69.2% vs 51.5%, p¼0.035). Blackpatients also tended to display more ST segment depression (50.0%vs 35.0%, p¼0.071), although this was not statistically significant.In contrast, WP had significantly more pathological Q waves (23.3%vs 9.6%, p¼0.039).Conclusions Ethnicity appears to exert a significant effect on ECGand echocardiographic patterns in patients with HCM. A significantproportion of black patients exhibit concentric LVH, highlightingthe diagnostic challenges in distinguishing HCM from hypertensiveheart disease and physiological adaptation to exercise in blackindividuals. The greater prevalence of deep Twave inversions and Twave inversions in the lateral leads underscores the importance offurther evaluation of black individuals with such ECG repolarisationabnormalities, which may represent the initial expression of HCM.

171 THE RIGHT VENTRICLE OF THE ENDURANCE ATHLETE: THERELATIONSHIP BETWEEN MORPHOLOGY ANDDEFORMATION

doi:10.1136/heartjnl-2011-300198.171

1D Oxborough, 2R Shave, 3G Whyte, 1K Birch, 4N Artis, 3K George, 5S Sharma.1University of Leeds, Leeds, UK; 2UWIC, Cardiff, UK; 3Liverpool John MooresUniversity, Liverpool, UK; 4Leeds Teaching Hospitals NHS Trust, Leeds, UK; 5StGeorges University Hospital, London, UK

Introduction It is well established that endurance exercise results incardiac adaptation including eccentric hypertrophy of the leftventricle which can complicate the differential diagnosis of theathletic heart from some cardiac pathologies that may pre-dispose tosudden cardiac death. The impact of physiological conditioning onRV structure and function, and a similar diagnostic challenge witharrhythmogenic right ventricular cardiomyopathy (ARVC), hasreceived less attention. A recent guideline paper from the AmericanSociety of Echocardiography (ASE) has provided a range for normalRV dimensions and functional deformation. These guidelinessuggest the RV inflow (RVI) should be <42 mm while the proximaloutflow tract (RVOT) <35 mm. A recent paper also suggested thatan RVOT dimension >36 mm or 21 mm/m2 is a major criterion forthe diagnosis of ARVC and furthermore longitudinal RV deforma-tion has been shown to be impaired in these patients. In view ofthis, the aims of this study are twofold:1. To provide a range of absolute values for RV dimensions in

102 endurance athletes as well as providing a range of dataindexed for body surface area (BSA).

2. To provide normal athlete data for indices of RV strain (3) andstrain rate (SR).

Methods and Results One hundred and two (102) endurance athletes(86 males and 16 females) with a broad age range (mean 6 SD age(range)¼36 6 11 (21e71) years) volunteered and were consecutivelyenrolled in the study. All subjects were either endurance runners orcyclists and were scanned at peak condition. Echocardiographyprovided measurements of RVI, RV length, RVOT and RV diastolicarea (RVDarea). A 2D strain technique was utilised to provideindices of RV3 and systolic and diastolic SR. The values for RVI


Black HCM(n[52)

white HCM(n[103) p Value

Demographics

Age of diagnosis (years) 48.1617.1 50.5616.5 0.552

Gender (males) 61.5% 62.1% 0.942

FH of HCM/SCD 34.6% 32.0% 0.747

NYHA functional class III or IV 7.7% 7.8% 0.987

Patients on treatment 55.8% 46.6% 0.281

B-blockers 28.8% 39.1% 0.445

Calcium antagonists 26.9% 12.6% 0.026

Amiodarone 7.7% 1.9% 0.080

Diuretics 13.5% 9.7% 0.480

Disopyramide 3.8% 9.7% 0.197

Intracardiac defibrillator in situ 5.8% 5.8% 0.989

Echocanliographic characteristics

Ao (mm) 31.363.7 33.265.8 0.123

LA (mm) 40.967.3 39.967.3 0.593

LVED (mm) 44.066.1 44.466.1 0.787

mLVWTd (mm) 17.364.9 18.864.1 0.069

LV mass (g) 279.66106.5 287.66112.7 0.767

FS (%) 40.469.1 39.868.3 0.641

E wave (m/s) 0.7060.18 0.7460.20 0.443

A wave (m/s) 0.6760.18 0.6660.27 0.851

E/A 1.1160.44 1.2260.58 0.422

SAM 23.1% 37.9% 0.064

LVOT gradient ¼ 30 mm Hg 23.1% 34.0% 0.163

LVH pattern

ASH 25% 57.3% 0.004

Concentric 44.2% 30.1%

Apical 28.8% 11.7%

No hypertrophy 1.9% 1.0%

Echocanliographic characteristics

LVH (Sokolow & Lyon) 53.8% 35.9% 0.033

Left atrial enlargement 44.2% 49.5% 0.534

Pathological Q waves 9.6% 23.3% 0.039

Left axis deviation 11.5% 17.2% 0.270

Inverted T-waves 82.7% 69.9% 0.086

T-wave inversions in V1eV4 3.8% 3.9% 0.991

T-wave inversions in inferior leads 1.9% 5.8% 0.269

T-wave inversions in lateral leads 76.9% 60.2% 0.038

Deep T-wave inversions 69.2% 51.5% 0.035

ST-segment elevation 9.6% 9.7% 0.985

ST-segment depression 50% 35.0% 0.071


BCS Abstracts 2011

ranged from 30 to 55 mm with 57% of the population having valuesgreater than the normal range. Proximal RVOT ranged from 26 to49 mm with 40% of the population above the normal range. 28% ofthe population had RVOT values greater than the proposed “majorcriteria” for ARVC. RV length ranged from 70 to 110 mm andRVDarea from 13 to 38 cm2 with values falling above ASE cut-offsin 69% and 59% of the population, respectively. When indexed (ratioscaling) for BSA proximal RVOT ranged from 13 to 25 mm/m2 with6% of the population meeting the major criteria for ARVC. Peak RV3ranged from�18 to�41% and peak RV SRS9 from�0.75 to�2.65 l/s,consistent with normal ranges proposed by the ASE. RV diastolicdeformation indices displayed marked individual variability with adominant SRE9 (mean 6 SD¼2.060.61 l/s) and smaller SRA9

(1.2560.56 l/s).Conclusion RV dimensions in endurance athletes are higher thanthose proposed as “normal” and likewise may be consistent with thecriteria for ARVC. Despite this enlargement, RV function inendurance athletes is preserved and therefore the role of RV strainimaging may provide additional diagnostic value in differentiatingphysiological from pathological adaptation.



Parameter Mean ± SD (range)ASE NormalValues

Indexed (ratioscaling) for BSA

RVOT (mm) 3465 (26 to 49) <35 1763 (13 to 25) mm/m2

RVDI (mm) 4465 (30 to 55) <42 2263 (15 to 30) mm/m2

RV Length (mm) 9269 (70 to 110) <86 4565 (32 to 61) mm/m2

RVDarea (cm2) 2665 (13 to 38) <25 12.862 (8.7 to 17.6)

RV 3 (%) �2766 (�18 to �41) �18 to �39 N/A

RVSRS9 (l/s) �1.5360.43(�0.75 to �2.63)

0.7 to 2.54 N/A

RVSRE9 (l/s) 2.060.61 (0.87 to 3.76) N/A N/A

RVSRA9 (l/s) 1.2560.56 (0.28 to 2.88) N/A N/A

RVSR E9/A9 1.8961.03 (0.49 to 7.25) N/A N/A

172 INCREASING ANTI-OXIDANT CAPACITY REVERSES IRONOVERLOAD MEDIATED DYSFUNCTION INCARDIOMYOCYTES

doi:10.1136/heartjnl-2011-300198.172

1F R Millar, 1D T Baptista-Hon, 2S C O9Neill, 1M E Dıaz. 1University of Edinburgh,Edinburgh, UK; 2University of Manchester, Manchester, UK

Introduction Iron overload-cardiomyopathy (IOCM) is an increasingclinical problem worldwide. 70% of patients who receive compul-sory blood transfusion die of IOCM, with increased susceptibility toarrhythmias and sudden death. We have previously found that ironexposure impairs cardiomyocyte Ca homeostasis. However, thecellular mechanisms responsible are unknown. Iron is known toparticipate in the Fenton reaction to produce reactive oxygen species(ROS), which mediate oxidative damage. We therefore tested thehypothesis that increasing the anti-oxidant capacity of cardiomyo-cytes, with the ROS scavenger Tempol, could be cardioprotective inthe presence of iron.Methods Single rat ventricular cardiomyocytes were loaded withfluo-3 to monitor intracellular Ca changes upon stimulation whilebathed in control Tyrode solution and after adding ferrous iron (ironII). Sarcoplamic reticulum (SR) Ca load and sarcolemmal Caextrusion rates were estimated during exposure to caffeine, whichempties SR Ca stores. The ROS scavenger tempol was used todissect ROS-mediated pathways from the direct effects of iron II onCa handling. Data are provided as mean6SEM. Significance wastested using paired student t test and defined as p<0.05.Results Iron II exposure significantly increases systolic Ca transientamplitude (mean increase 82.8621.8%, n¼9) and causes sponta-neous arrhythmogenic Ca release events (SACRE). These changescorresponded with increased SR Ca content (mean increase21.065.7%, n¼8), which is known to impact on systolic Ca releaseand spontaneous activity in cardiomyocytes. Sarcolemmal Caextrusion rate was also significantly reduced upon iron II exposure(measured as the rate of fall of the caffeine response; mean decrease48.765.5%, n¼8), consistent with an overall gain of Ca by thecardiomyocyte. The onset of these Ca disruptions was significantlydelayed in the presence of the ROS scavenger tempol (p<0.001).Without tempol, SACRE onset occurred after 6.960.6 min (n¼22)following iron II exposure. The same manoeuvre in in tempoldelayed the onset of SACRE to 17.861.8 min (n¼7). Furthermore,increasing ROS scavenging reversed the increase in systolic Catransient amplitude, as well as SACRE upon washout of iron II. Incontrast, in cardiomyocytes not exposed to tempol, the effects ofiron II were irreversible.Conclusions Our data show that iron II disrupts cardiomyocyte Cahandling. This is mediated via inhibition of sarcolemmal Ca extru-sion, leading to SR Ca overload and SACRE. These are the initiatorsof most fatal non-reentrant arrhythmias and cardiac sudden death in


BCS Abstracts 2011

experimental models. The observed effects are partly due to iron II-mediated oxidative damage. This was confirmed by the presence of aROS scavenger delaying the onset of the effects of iron II, andcrucially rendering the effects reversible upon iron-washout. Theseeffects of tempol suggest a novel therapeutic target for the treat-ment of IOCM patients.

173 A GENERIC METHOD TO ASSESS THE ADEQUACY OFINDIVIDUAL MATERNAL CARDIAC RESERVE TO TOLERATETHE DEMANDS OF PREGNANCY AND LABOUR

doi:10.1136/heartjnl-2011-300198.173

1D Barker, 2N Lewis, 2G Mason, 2L B Tan. 1Liverpool Heart and Chest Hospital,Liverpool, UK; 2Leeds General Infirmary, Leeds, UK

Introduction Clinicians often feel apprehensive when managingpregnant patients with heart disease. To complement current eval-uation, we have developed a new method of directly assessing theindividual patient’s cardiac functional reserve through stress testing.Pregnant mothers with and without heart disease were studied totest the hypothesis that pregnant cardiac patients who possesscardiac reserve equivalent to that of controls can tolerate the usualdemands of pregnancy, labour and puerperium.Methods Fifty-one pregnant women with heart disease (mean age30.766.5 (range 21e42), mean gestation 25.668.6 weeks) and 102healthy pregnant women (mean age 31.465.0, (range 19e41), meangestation 25.169.2 weeks) underwent maximal symptom-limitedtreadmill cardiopulmonary exercise testing. Fifty-nine non-pregnantwomen (mean age 32.765.1 (range 20e41) years) were similarlytested and used as a control group. Cardiac output (CO) wasmeasured at peak exercise using the CO2 re-breathing method.

Cardiac power output (CPO) was calculated as the product of COand mean arterial pressure. A composite endpoint includingmaternal death, fetal death, emergency caesarean section formaternal distress and significant morbidities was determined.Results All tests were performed without significant complications.Employing data from a previous study of haemodynamics duringlabour in healthy women, the mean CPO required during peaklabour is 2.6 W. This value was adopted for investigation as theminimum required for an average woman to cope with the circu-latory demands of normal labour. The healthy controls had a meanpeak CPO (PkCPO) of 3.7960.6 Wand all non-pregnant women hadPkCPO exceeding 2.6 W. The majority of heart disease patients wereable to achieve PkCPO values overlapping their healthy counter-parts. Only a small proportion of the cardiac patients had PkCPOvalues lower than the 2.6 W cutoff. Women were significantly morelikely to have uncomplicated pregnancy, labour and puerperium ifable to achieve PkCPO>2.6 W (OR 8.1, 95% CI 1.8 to 37.0,p¼0.023). Pregnant women in NYHA class I had PkCPO valuesindistinguishable from controls (mean 3.9860.77 W, NS); whereassymptomatic pregnant women had significantly lower values (mean3.1560.71W, p<0.005).Conclusions Direct measurement of cardiac functional reservecapacity can be performed by maximal cardiopulmonary exercisetesting with non-invasive assessment of PkCPO, which can be safelyundertaken during pregnancy. A cutoff value of PkCPO 2.6 W wasidentified as the lower limit for healthy women, corresponding tothat required for normal labour. Most cardiac patients studied hadPkCPO values comfortably above this cutoff, and all asymptomatic(NYHA I) and low risk cardiac patients had PkCPO values similar tocontrols. Measurement of PkCPO allows pregnant patients to befurther classified into those with adequate vs limited cardiac reserve,supplementary to existing risk stratification methods.


BCS Abstracts 2011

The number next to the author indicates the page number, not the abstract number.

Abbas A, A1, A3Abidin N, A70Abozguia K, A55Abu-Own H, A25Adam Z, A5Adams PC, A39Aggarwal R, A22, A30Aggarwal S, A60, A61Ahmed R, A43Ahsan AJ, A86Aitma TJ, A4Akhtar M, A9Al Fakih K, A36Ala L, A82Alahmar A, A27Alamgir MF, A23Alp NJ, A79Alpendurada F, A53Alpendurado F, A94Amersey R, A15Anand A, A7Angelini G, A24Anroniades C, A79Antoniou S, A5Appleby C, A27Archbold A, A46Archbold RA, A5Arthur H, A78Arthur HM, A75Artis N, A95Artis NJ, A72, A73Arujuna A, A85Augustine D, A63Aung N, A60, A61Austin D, A5Avery P, A75Awan M, A5Ayers L, A50Aziz A, A78

Babar J, A63Baker CSR, A18Baker S, A68Balakrishnan B, A66, A81Baliga V, A56Ball SG, A35, A41, A72Balmforth AJ, A35, A41Bamforth SD, A77Banerjee A, A40Banerjee G, A7Banerjee R, A63Banerjee S, A7Banner NR, A48, A49, A64Banning AP, A12, A93Banypersad SM, A59Bapat V, A19Baptista-Hon DT, A96Barker D, A57, A97Barker J, A7Barmby D, A21Barnes S, A60Barrington S, A68Barsan A, A49Barth J, A60Barth JH, A8, A17Baruah R, A51, A58Basavarajaiah S, A33Bashir Y, A87, A88

Bastiaenen R, A33, A38, A83Baudoin F, A80Baumbach A, A24Bawamia BR, A10Beadle RM, A55Beatt K, A24Bechar I, A70Bedell VM, A78Begg G, A56Begg GA, A60, A86Begley D, A50Behan M, A16Behar J, A15Behar JM, A25Behr E, A38Behr ER, A33, A83Bell D, A72Bell J, A45Bellamy MF, A18Ben-Nathan S, A83Bennett MR, A2, A64Berry C, A16, A73Berry EL, A86Betts TR, A87, A88Bewick AE, A82Bhabra M, A19, A40Bhan A, A91Bhattacharya S, A77Biasiolli L, A62Bijnens BH, A52Birch K, A95Blackman DJ, A21Blair E, A62Blake J, A11Blamire A, A78Bland MB, A37Blaxill JM, A21Bleasdale RA, A34, A82Bloomer LDS, A41Blundell N, A92Bonser RS, A36, A48Borbas Z, A90Borg A, A70Borg AN, A74Bostock J, A52, A84, A85Boston-Griffiths E, A30Bowater S, A55Bowen TS, A56Braganca J, A77Braganza D, A2Braund PS, A41, A42Brewer A, A77Brewster S, A46Brickham B, A91Broadbent C, A77Brown AJ, A29Brown BD, A90Brown P, A89Bryan L, A72Buchanan L, A23Bull S, A92Byrom R, A60

Caldwell JC, A90Calver A, A16Calvert PA, A2, A64Camara O, A52Campbell M, A11

Cannon DT, A56Caplin JL, A23Carr-White G, A52, A84Carre F, A38Carrick D, A16Carter J, A5Cartwright EJ, A41, A80Casadei B, A79Casey FA, A75Casey M, A73Cassar TE, A62Chalmers RTA, A3Chambers J, A43Chandra N, A33, A38, A95Channon K, A12Channon KM, A79Chaubey S, A77Chaudhry U, A27Chen C-M, A77Cheng A, A60, A61Chico TJA, A43Chinchapatnam P, A52Chinnappa S, A57Chiribiri A, A68, A71Chong E, A24Choudhury RP, A12, A62, A70Chowienczyk P, A28, A32Choy AM, A94Choy AMJ, A58Christiansen JP, A93Christie J, A16Christofidou P, A41Churchhouse AMD, A7Clack L, A19Clapp B, A24, A32Claridge S, A36Clark C, A73Clark D, A70Clarke B, A26Clarke SC, A2Cleary N, A46Clesham G, A30Codd V, A42Collerton J, A57Connelly K, A22Connolly S, A7Cook DG, A41Cook S, A43Cook SA, A4Corbett S, A16Cotton J, A19Cowie MR, A53Cox RD, A41Craig BG, A75Crean A, A72Crossman D, A13Cruddas E, A24Cubbon R, A1, A3, A60, A78Cunliffe E, A52Curzen N, A16

D’Arcy J, A92, A93d’Arcy JL, A92Dıaz ME, A96Dall’Armellina E, A12Damm E, A30Danesh J, A43Das D, A43


Author index

Das M, A84Davidson C, A14Davidson J, A35Davidson NC, A90Davies DW, A51, A83Davies J, A22, A30, A45, A54Davies JE, A10Davies MJ, A6Davis E, A37, A63Davison BJ, A78Dawson A, A34de Belder A, A45de Belder MA, A5De Silva K, A24, A28, A71Deanfield JE, A41Deaton C, A26Debiec R, A41Densem CG, A2Dent H, A16Denvir MA, A7, A53Dhillon OS, A10Diab I, A81Dick AJ, A22Dick KJ, A42Dickens C, A11Diesch J, A37, A63Digby J, A70Dinh DT, A9Dixit A, A26Dixon G, A45Dobrzynski H, A4Doherty PD, A37Donald A, A32Donald AE, A41Donin A, A41Doolin O, A73Douglas-Hill C, A19Drury-Smith M, A19Duckett S, A84, A85Duckett SG, A52Dungu J, A18, A59Durham N, A60Dutka DP, A50Dweck M, A3Dweck MR, A94Dwivedi G, A66, A81Dworakowski R, A91

Earley MJ, A81Ebbs D, A92Eccleston D, A38Edmunds L, A45Edwards R, A10, A29Eftychiou C, A21Ekker SC, A78El-Omar M, A26Elder DH, A94Elder DHJ, A34, A58Ellins E, A41Emin A, A48Engelen K, A76Englyst N, A16Eve S, A47

Fairbairn TA, A73Farmer AJ, A92Fath-Ordoubadi F, A11Ferguson E, A5Ferreira V, A12Finch S, A46Flather M, A13, A24, A38

Flint J, A60, A61Foley C, A13Foley JRJ, A90Foo F, A16Ford I, A73Forfar C, A12Forfar JC, A62Fox DJ, A90Fox K, A13Fox KAA, A7Fox KF, A7Frampton C, A11Francis DP, A49, A50, A51, A58, A60, A61, A66Francis J, A37, A63Francis JM, A12, A62, A68, A92, A93Frangi AF, A52Fraser AG, A34Fraser D, A26Freemantle N, A49French AE, A89Frenneaux MP, A55

Gage M, A1, A3Gage MC, A78Gale CP, A8, A17, A60Gale CPG, A37Gallagher M, A83Gallagher S, A15, A31Gallagher SM, A5Galloway S, A78Gamble D, A7Garden OJ, A3Garratt CJ, A90Gasson A, A82Gati S, A31, A95George DA, A40George K, A95Ghani S, A31, A95Gholap NN, A6Gibbins A, A45Gibbons SM, A41Gibbs SDJ, A59Gierula J, A60, A86Gilchrist J, A73Gill JS, A52Gill PS, A59Gillivray TJMac, A3Gillmore JD, A59Ginks M, A52, A84, A85Glen E, A76Glen EA, A75Goodall AH, A42Goodship J, A75, A76Goodship JA, A75, A76Gopalan D, A63, A64Gordon B, A49Gordon J, A75Gosling O, A20Gosling OE, A65Graham A, A13, A15Graham C, A7Graham TR, A36Gray C, A3, A43Gray H, A16Greaves M, A70Greenwood JP, A13, A21, A72, A73, A93Griffin HR, A76Grimwade P, A92Guha K, A53Guilcher A, A28Gulati A, A94

Gunn J, A13Guttmann O, A9Guttmann OP, A31

Haga KK, A53Hall A, A13Hall AS, A8, A17, A35, A41Hall ASH, A37Hall D, A75Hall DH, A75, A76Hall JA, A5Hall R, A24Hamid S, A84, A85Hamid SG, A52Hancock J, A19Handa A, A62Haq IU, A39Haran H, A19Hards R, A49Harrington DW, A82Harwood SM, A17Hawkins NM, A30Hawkins PN, A59Hayward PA, A9Hedger M, A49Hegab Z, A41, A44Heneghan C, A40Heymans S, A77Hobson A, A16Hodson J, A36Holbrook I, A60Holroyd EW, A78, A80Horne A, A89Hosmane S, A76Hoye A, A23Hughes A, A49Hughes AD, A50, A66Hunt J, A32Hunter A, A60Hunter RJ, A81Hussain S, A68

Iles-Smith H, A11Imrie H, A1, A3, A78Indermeuhle A, A24Indermuhle A, A71Ingram TE, A34Irwin RB, A70Ishida M, A68, A71Ivetic A, A77Iyengar S, A65

Jackson C, A78Jagger C, A57Jain A, A13, A15, A25, A31Jain AK, A5, A9, A15James PE, A34Jamieson S, A10Jamshidi Y, A4Jayaram R, A79Jeilan M, A89Johnson T, A89Jones D, A5Jones DA, A9, A13, A15, A25, A31Jones JD, A30Jones MA, A88Jones S, A83Joshi NV, A10Joshi S, A94Joysurry D, A41Juli C, A83

Kahn M, A1, A78


Author index

Kahn MB, A3Kaier T, A39Kanagaratnam P, A51, A83Kapetanakis S, A52Kapur A, A17, A24, A31Kapur AK, A25Karamitsos TD, A12, A68, A92, A93Karia N, A12Karim R, A83Kearney L, A60Kearney M, A1, A78Kearney MT, A3, A60Keavney B, A57, A75, A76, A78Keavney BD, A75, A76Kellman P, A12Kelly B, A37Kelly P, A22, A30Kemp I, A27Kemp S, A31Kennedy J, A62Kenny A, A57Kervio G, A38Kerzin-Storrar L, A90Kesavan S, A24Kettle AJ, A11Khan FZ, A50Khan MF, A23Khan N, A90Khan S, A39Kharbanda RJ, A12Khattar R, A26Khawaja MZ, A19Khimdas K, A46Khogali S, A19Khoo CW, A66, A81Khunti K, A6Kilcullen N, A8Kingston A, A57Kirkwood T, A57Klaassen S, A76Knight C, A5, A9, A13, A15, A25, A31Knight CJ, A25Kooner J, A43Kotecha D, A38Krishnamoorthy S, A66, A81Krum H, A38Kuehl M, A55Kuijer JPA, A69Kuker W, A62Kulanthaivelu R, A82Kumar D, A46Kundu S, A89Kylintireas I, A62, A70Kyriacou A, A50, A51, A66

Lachmann HJ, A59Lang CC, A35, A58, A94Langrish JP, A3Larsen K, A80Lazdam M, A37, A63Leadbeater P, A16Lee JM, A62, A70Lee KK, A7Lees B, A13Leeson P, A37, A62, A63Leisa RA, A80Leon FL, A55Lewandowski AJ, A37, A63Lewin BL, A37Lewinter CL, A37

Lewis N, A57, A97Lewis RJ, A75Leyva F, A48Lim HS, A66, A81Lim PB, A83Lindsay A, A12Lindsay AC, A62Ling HZ, A60, A61Lip GYH, A59, A66, A81Little A, A46Liu A, A21Llewellyn-Griffiths H, A54Loader R, A65Lockie TPE, A28Loh PH, A23Lovell MJ, A5Lowdell M, A20Lucas A, A37Ludman A, A9Ludman PF, A91Lundmark P, A42

Ma Y, A84, A85MacArtney MG, A94MacCarthy P, A91MacDonald ST, A77MacDonald TM, A35Macgillivray K, A19Macgilivray TJ, A3MacGowan G, A48, A49MacLeod M, A7Macnab A, A76Mahadevan V, A26Maher A, A19Mahmod M, A68Majumder B, A20Makri L, A21Malcolme-Lawes L, A83Malik N, A22Mamas M, A26, A44Mamas MA, A41Mamasoula V, A75Manisty C, A50Manisty CH, A58Mant D, A92Marber M, A28Margulescu A, A34Markl M, A62Marshall CJ, A11Mascaro J, A36Masi S, A41Mason G, A97Mather AN, A72, A73Mathur A, A5, A9, A13, A15, A25, A31Mavroudis C, A27Mavroudis CA, A20May S, A13Mayet J, A50, A51, A58, A60, A61, A66McCann GP, A69, A74, A89McCarthy CA, A59McCarthy MI, A43Mcclean DR, A11McClure J, A73McCormick LM, A29McCusker CG, A75McDiarmid A, A49McDonagh T, A53McEntegart MB, A73McGeoch R, A73McGill LA, A15, A25

McGowan L, A11Mckenzie JL, A11McKeown PP, A75McKillop G, A3McLenachan JM, A21McMahon M, A75McNab D, A2Mehta P, A84, A85Mehta RL, A6Menon A, A36Metcalfe K, A90Millar FR, A96Miller C, A70Miller CA, A74Mills NL, A7Mitchell AG, A64Mittal TK, A64Moccata T, A11Mohamed TMA, A80Mohammed TMA, A41, A44Mohiaddin R, A93, A94Mohiddin S, A5, A9, A13Mole G, A14Monaghan M, A91Moraldo M, A50, A66Moreton N, A90Morgan-Hughes G, A65Morrell C, A8Morrice D, A40Morrison ML, A75Morton AC, A13Morton G, A24Morton GDJ, A68, A71Muckett P, A43Muckett PJ, A4Muggenthaler M, A33, A95Muir DF, A5Mukhopadhyay D, A80Mulder B, A76Mullen LJ, A29Murdoch CE, A77Murgatroyd SR, A56Murigu T, A94Murphy A, A73Murray SA, A53Musameh MD, A35Myerson SG, A68, A92, A93

Nadir A, A35, A94Nadra I, A19Nagel E, A68, A71Nair S, A26Nallaratnam M, A11Narayan HK, A10Nayar V, A50Nelson CP, A41, A42Ness A, A53Neubauer S, A12, A37, A62,A63, A68, A92, A93

Neubauer SN, A62Neubuer S, A70Nevill AM, A40New G, A38Newby D, A94Newby DE, A3, A7Newman D, A40Newman W, A90Newton T, A70Neyses L, A26, A41, A44, A80Ng GA, A89


Author index

Ng LL, A10Nicol E, A72Nicolson WB, A89Nightingale CM, A41Nijjar M, A30Nijjer SS, A7, A49Norrie J, A16Nyawo B, A29

O’Callaghan P, A55O’Doherty M, A68O’Donnell M, A53O’Sullivan JJ, A75O’Sullivan M, A2O9Neill SC, A96Obaid DR, A2, A63, A64Oceandy D, A80Okonko DO, A60, A61Oldroyd K, A16Oldroyd KG, A73Oliver M, A54Oliver RM, A23Ordoubadi F, A26Oriolo V, A47Ormerod OO, A70Owen CG, A41Owen J, A60Oxborough D, A95

Pabari P, A49, A51, A58Pabari PA, A50, A66Padgett HC, A86Padley S, A72Pagano D, A36Palmieri V, A49Palomino-Doza J, A75Panayotova R, A76Panicker MG, A64Panoulas V, A33, A38Papadakis M, A33, A38, A95Parameshwar J, A48Parker C, A89Parry G, A49Pashaei A, A52Patel P, A82Paterson E, A7Patterson C, A72Paul GA, A22Payne AR, A73Pearson IR, A8, A17, A60Peebles C, A70Pennell D, A94Penswick A, A46Pepper J, A94Perera D, A24, A28, A68, A71Perry R, A30Perry RA, A27Peters NS, A4, A51, A83Petersen SE, A62Petrie MC, A73Pettit S, A30Pierret CK, A78Pierscionek T, A75Pinney JH, A59Pitcher A, A62, A92Pitts-Crick J, A89Plein S, A72, A73Plummer CJ, A84Poole R, A37, A63Popov AF, A49Postma A, A76

Prasad S, A53, A94Pravanec M, A4Prehar S, A41, A80Prendergast B, A12, A92Prendergast BD, A92Prescott M, A41Pretsell G, A45Pringle SD, A58, A94Pugh PJ, A50Pye M, A60

Quinn PA, A10Qureshi AC, A17Qureshi N, A88

Rahman T, A75, A76Rajappan K, A87, A88Raju H, A31, A33, A38, A95Rajwani A, A1, A3Rakhit RD, A20, A27Rampat R, A17Rana B, A67Rapala A, A41Rashid S, A3Rathod B, A15, A31Rathod K, A13, A25, A31Rathod KR, A15Rathod KS, A9, A15Rathod VS, A15Rawling A, A50Rawlins J, A33, A38Ray KK, A2Raybould A, A54Razavi R, A84, A85Razavi RS, A52Razvi NA, A69Redgrave R, A78Redwood S, A19, A24, A28, A71Rees I, A55Reid CM, A9Reid J, A53Reilly S, A79Rekhraj S, A35Rhode K, A84, A85Richards JMJ, A3Richards M, A11Richards T, A61Richmond L, A81Rider O, A70Rinaldi CA, A52, A84, A85Ring L, A67Robb SD, A73Robinson N, A49Robson MD, A12, A62, A70Rogers C, A11Rogers CA, A48Rogers T, A36Rolandi C, A28Roobottom C, A65Rooney SJ, A36Rose H, A45, A54Rossiter HB, A56Rothman A, A13Rothman M, A46Roughton M, A17Rudd JHF, A63, A64Rudnicka AR, A41Rueckert D, A83Rusk RA, A67Russell S, A46, A54

Russell SJ, A45, A55

Sabharwal NS, A70Sadarmin PP, A87Salaheen D, A43Salukhe TV, A83Samani NJ, A35, A41, A42Sambu N, A16Sammut E, A15, A25Sammut EC, A13Sanderson JE, A48Sandilands AJ, A89Sands AJ, A75Sattianayagam P, A59Saul A, A73Sayan S, A51Sayer J, A22, A30Schilling RJ, A81Schlindwein FS, A89Schmitt M, A70, A74Schneider JE, A77Schofield PM, A2Schotten U, A79Schueler S, A48, A49Schuster A, A68, A71Schwartz R, A77Sehmi J, A43Semple SI, A3Sermesant M, A52Shah A, A7Shah AJ, A63Shah AM, A77, A91Shah NH, A23Shantsila E, A59Shapiro LM, A2Shardey GC, A9Sharma R, A53Sharma S, A31, A33, A38, A83, A95Shave R, A95Sheikh N, A31, A38, A95Shelton RJ, A21Shepherd C, A89Shepherd EJ, A84Shetty A, A52, A84Shetty AK, A85Shi WY, A9Shirodaria C, A70Shome J, A5Showkathali R, A22, A30Sicard P, A24Siebes M, A28Sim V, A54Simari RD, A78, A80Simon A, A48Simon AR, A49Simpson I, A16Singh R, A19Singhal A, A37Sinha M, A47Sivananthan UM, A8, A17Skinner JS, A39Smith EJ, A25Smith J, A78Smith JA, A9Sneddon L, A75Snell KIE, A69Sohal M, A84, A85Somauroo J, A31Somers K, A21Sorbron S, A72Spath N, A33


Author index

Sporton S, A81Squire IB, A6, A10Stables RH, A27Stafford PJ, A89Staniforth AD, A86Steadman CD, A69, A74, A89Stegemann B, A49Stoll VMS, A70Strain WD, A65Strauss BH, A22Struck J, A10Struthers AD, A34, A35, A58, A94Sukumar P, A78Suresh V, A20Surr J, A1, A3Sutaria N, A66Suttie J, A62, A63, A92Suttie JJ, A68Sutton AGC, A5Swanson NM, A5Syvanen AC, A42Szwejkowski BR, A34, A58

Topf A, A76Taggu W, A22Tagney J, A47, A89Tai ES, A43Tan HL, A75Tan LB, A57, A97Tan YT, A48Tayebjee M, A81Taylor R, A29Templeton C, A34ten Dijke P, A75Thackray S, A23Thomas G, A89Thomas HE, A84Thomas HL, A48Thomas M, A19, A24, A60, A61Thompson CA, A25Tilling LM, A32Tomaszewski M, A35, A41Tooze P, A21Topf A, A75Topf AL, A75

Townend J, A78Tsui S, A48Twomey D, A84Tzemos N, A73

Ungvari T, A23Unsworth B, A58, A66

Van den Bruel A, A40van Eeden FJ, A43van Rossum AC, A69Vanhoutte D, A77Veasey RA, A82Venables P, A20Verheule S, A79Vile RG, A80Virdee MS, A50Viswambharan H, A1, A3, A78Viswanathan K, A8

Waldron ZL, A86Walker S, A7Wallace W, A3Wang WYS, A35Ward D, A83Ware JS, A4Warner T, A16Wassef N, A59Waterworth P, A76Watson D, A14Watson J, A60Watson OJ, A43Watt H, A62Wechalekar A, A59Weerackody R, A15, A25, A46Weissert S, A60, A61Wells TA, A47Wendler O, A91Wenzelburger FWG, A48West NEJ, A2, A29Westwood M, A5Wheatcroft AC, A60Wheatcroft S, A1, A3, A78Wheatcroft SB, A21Whelan CJ, A59Whincup PH, A41

Whinnett Z, A51Whinnett ZI, A58White J, A60Whyte G, A95Wicks E, A31Wilkinson S, A4Williams C, A34Williams LK, A55Williams M, A3Williams P, A26Williams R, A28, A39Willson K, A51Wilson IC, A36Wilson K, A19Wilson SJ, A20Wiper A, A26Witte KK, A56, A60, A86Woldman S, A60Wong KCK, A87, A88Woodcock T, A72Woodward R, A73Wordsworth BP, A62Wragg A, A9, A13, A15, A25, A31Wren C, A75Wright GA, A22Wright RA, A5Wrightson N, A49

Yap CH, A9Yaqoob MM, A17Yellowlees D, A53Yeo Y, A43Young C, A19Young S, A45Yousaf F, A57Yousef ZR, A45, A46, A54, A55Yu B, A77Yuldasheva N, A3, A78Yusuf S, A89

Zaidi A, A31, A33, A95Zaman A, A10, A72Zhang W, A43Zi M, A41, A80Zych B, A49


Author index

2ac heartjnl bcs abstracts 2011

Documents

higfreo mice

insulin

1 receptor

igf

huvecs

vasculature

bioavailability

assessed