Lactic Acidosis After Cardiac Surgery Is AssociatedWith Polymorphisms in Tumor Necrosis Factor andInterleukin 10 GenesThomas Ryan, FFARCSI, Joanna Balding, BA, Mod (Genetics), Eilis M. McGovern, FRCSI,John Hinchion, FRCSI, Wendy Livingstone, PhD, Zeb Chughtai, FRCSI,and Owen P. Smith, FRCPIDepartments of Anaesthesia, Hereditary Coagulation Disorders, and Cardiothoracic Surgery, St. James’s Hospital, Dublin, Ireland

Background. Lactic acidosis after cardiac surgery is amanifestation of excess cytokine production. Cytokine-related genetic polymorphisms account for variability incytokine response and may predispose to the develop-ment of lactic acidosis after cardiac surgery.

Methods. Routine postoperative cardiac surgery patientswere studied. Lactic acid levels were greater than 4 mmol/Lin study patients and less than 4 mmol/L in controls.Polymerase chain reaction-based techniques were used toexamine carriage of tumor necrosis factor � (TNF-�), TNFG–308A, and interleukin 10 (IL-10) G–1082A alleles.

Results. Demographic characteristics and details ofsurgery were similar for 30 control and 21 study patients.

Lactic acid levels after intensive care admission changedover time and were related to both TNF-� and IL-10G–1082A polymorphisms. All 4 study patients homozy-gous for TNF-�1 and carrying an IL-10–1082A alleledeveloped lactic acidosis (p � 0.02). There was no rela-tion between the rate of epinephrine infusion or durationof cardiopulmonary bypass and lactic acid levels.

Conclusions. Genetic factors have a role in the devel-opment of lactic acidosis after cardiac surgery.

(Ann Thorac Surg 2002;73:1905–11)© 2002 by The Society of Thoracic Surgeons

Lactic acidosis after cardiac surgery is an ominousevent and is associated with excess mortality among

patients in shock [1]. In pediatric cardiac surgery lacticacidosis is a powerful predictor of outcome with greaterlactic acid levels associated with adverse outcome includ-ing excess mortality [2]. Lactic acidosis in cardiac surgicalpatients is a manifestation of systemic inflammation andexcess proinflammatory cytokine production [3]. Sys-temic inflammation after cardiac surgery may alsopresent as the “low systemic vascular resistance syn-drome” with hypotension and high cardiac index and isalways accompanied by lactic acidosis [3]. Lactic acidaccumulation in shocked cardiac surgical patients is a netresult of excess lactic acid production with unchangedutilization and is not related to altered carbohydratemetabolism [4]. This process may be a direct metaboliceffect of tumor necrosis factor (TNF), as lactic acid pro-duction is increased by TNF-mediated inhibition of pyru-vate dehydrogenase [5].

Interindividual variation in TNF production in patientswith sepsis has been linked to polymorphisms in theTNF-� gene [6]. Polymorphisms in the TNF-� promotergene are associated with excess mortality in septic shock[7]. Interleukin 10 (IL-10) is a potent antiinflammatorycytokine that inhibits TNF production. Polymorphisms inIL-10 promoter genes are associated with variation in

IL-10 production [8]. It is plausible that these cytokinegenomic polymorphisms modulate cytokine productionand systemic inflammation in cardiac surgical patientsand that the occurrence and severity of lactic acidosis inthese patients is influenced by the presence of TNF andIL-10 genetic polymorphism. We conducted a study totest this hypothesis.

Patients and Methods

Consenting patients scheduled for routine cardiac sur-gery with normal preoperative ventricular function anduneventful surgical procedures were recruited over a6-month period. Patients with a history of hepatic diseasewere excluded. Patients were admitted to a dedicatedcardiac surgical intensive care unit after surgery withcare determined by the referring cardiac surgical service.Patient care was not modified for the purpose of thestudy. Cardiopulmonary bypass was performed using anopen system primed with 1,500 mL of Hartmann’s solu-tion with heparin-coated circuits and roller pumps. Suc-tion systems were controlled.

Patient demographic characteristics and history ofmyocardial infarction, hypertension, congestive heartfailure, vascular surgery, and diabetes were collected.The nature of the surgical procedure, duration of cardio-pulmonary bypass and the minimum temperature oncardiopulmonary bypass (CPB), the first and last arterialblood gases on cardiopulmonary bypass, the blood flowsand hemoglobin concentrations on cardiopulmonary by-pass that corresponded with these blood gases, and the

Accepted for publication Feb 17, 2002.

Address reprint requests to Dr Ryan, Department of Anaesthesia, St.James’s Hospital, James St, Dublin 8, Ireland; e-mail: ryants@iol.ie.

© 2002 by The Society of Thoracic Surgeons 0003-4975/02/$22.00Published by Elsevier Science Inc PII S0003-4975(02)03530-0

least blood flow on cardiopulmonary bypass weredocumented.

A case-control study was performed with study pa-tients having arterial lactic acid level in excess of 4mmol/L at any time in the first 24 hours after cardiacsurgery and a control group with lactic acid level nevergreater than 4 mmol/L in the first 24 hours after surgery.Lactic acid levels, hemodynamics, inotropic requirement,arterial blood gases were recorded at the end of cardio-pulmonary bypass, on arrival in the intensive care, and 6,12, and 24 hours later. The duration of postoperativemechanical ventilation and the blood loss in the first 12hours after surgery were recorded.

Genetic analysis was performed by a person who wasunaware of group allocation and lactic acid levels. DNAwas extracted from whole blood using overnight protein-ase K (1 mg/mL) cell lysis at 37°C in the presence of 0.5%sodium dodecyl sulfate followed by extraction with phe-nol/chloroform and precipitation with ethanol. Polymer-ase chain reaction (PCR) amplification of all polymorphicsites was performed in a 50 �L total volume. The stan-dard reaction mix consisted of Taq DNA Polymerasebuffer with MgCl2 (Promega; 50 mmol/L KCl, 10 mmol/LTris-HCl [pH 9.0], 0.1% Triton X-100, and 1.5 mmol/LMgCl2), 0.4 U of DNA Taq polymerase, 2 �L of genomicDNA, 4% dimethyl sulfoxide (DMSO), 30 �mol/L each ofdeoxyribonucleoside triphosphates, and 0.2 �mol/L eachof sense primer and antisense primer (Appendix 1). Thecycling variables for each assay are listed in Appendix 2,along with any changes to the standard PCR reactionmix. Restriction enzymes used for each assay are listed inAppendix 2. The IL-6, TNF-�, IL-10–1082, and IL-10–592PCR products were digested with the appropriate en-zyme overnight at 37°C. The TNF-� PCR product wasdigested for 3 hours at 37°C, and the IL-1� PCR productwas digested for 12 hours at 65°C. Restriction digestproducts were run in the appropriate percentage ofagarose gel containing 1.6 �g/mL ethidium bromide.

Continuous variables were analyzed with Student’s ttest and analysis of variance (ANOVA). The �2 test andFisher’s exact test were used to compare categoricalvariables. The relation between lactic acid levels andindividual polymorphisms was analyzed at each timepoint using Student’s t test or ANOVA where appropri-ate. Where association between an individual polymor-phism and lactic acid levels was detected on such aunivariate test, then the interaction between polymor-phism and change in lactic acid levels with time wasanalyzed by multivariate ANOVA (MANOVA) with re-peated measures. The institutional ethics committee ap-proved this study in March 1999.

Results

There were 30 control patients and 21 patients in the studygroup. Age, gender distribution, body surface area, preop-erative chronic disease states, and the nature of surgicalprocedure were similar in both groups (Appendix 3). Theminimum temperature on cardiopulmonary bypass wassimilar in study and control groups. Cardiopulmonarybypass time was longer in study patients; however, this

difference did not reach statistical significance. Arterialpartial pressure of carbon dioxide was greater in the studygroup at the end of cardiopulmonary bypass (4.74 � 0.9 pKaversus 4.5 � 0.7 pKa, p � 0.05). Otherwise arterial bloodgases, hemoglobin, and circuit flow rates at the beginningand termination of cardiopulmonary bypass were similar inthe two groups. Postoperative blood loss was similar in thetwo groups. The duration of mechanical ventilation wasgreater in the study group (Table 1).

Lactic acid levels were significantly greater in the studygroup at all times in the first 24 hours, with the greatestdifference seen 6 hours after intensive care admission(Table 2). MANOVA for repeat measures of lactic acidover time with patient group as a factor found that lacticacid levels changed significantly over time (p � 0.0001)and found an interaction between patient group and time(p � 0.0001), indicating that patient grouping affected thetemporal change in lactic acid levels. Although the dura-tion of cardiopulmonary bypass was longer in the studygroup this duration did not correlate with lactic acidlevels. Lactic acid levels at 6 hours after intensive careadmission correlated with the duration of mechanicalventilation (lactic acid level at 6 hours � 2.6 � 0.12ventilation hours; p � 0.04, R2 � 0.08).

On univariate testing 6 hours after intensive careadmission lactic acid levels were significantly higher inpatients homozygous for the TNF-�1 allele and signifi-cantly lower in patients homozygous for the IL-10–1082Gallele (Tables 3 and 4). On MANOVA with TNF-� alleleand IL-10–1082 G allele as factors and analyzing repeatmeasurement of lactic acid at 1 and 6 hours (time) afterintensive care admission, there was a significant associ-ation between TNF-� allele and lactic acid level (p �0.03), between IL-10–1082 G allele and lactic acid level(p � 0.03), and in lactic acid level change with time (p �0.0002); the interaction between time and TNF-� allelewas also significant (p � 0.0005) as was the interactionbetween time and IL-10–1082 G allele (p � 0.01).

Patient grouping was not associated with the distributionof any individual cytokine polymorphism allele; however,all 4 patients homozygous for the TNF-�1 allele and whocarried the IL-10–1082A allele were in the study group (p �0.02). One other patient who was homozygous for theTNF-�1 allele who did not carry the IL-10–1082 A allele hadnormal lactic acid levels. There was no association betweenTNF G-308A alleles and lactic acid levels. However, only 2patients were homozygous for the TNF-308A allele. One ofthese carried the IL-10–1082A allele and developed markedlactic acidosis. The other did not carry the IL-10–1082Aallele and had normal lactic acid levels. There was noassociation between IL-1� �3953, IL-6–174, and IL-10–592polymorphisms and lactic acid levels.

There was no relation between blood pressure, centralvenous pressure, and genotype. There was no significantrelation between genotype and postoperative blood loss.There was no association between lactic acid levels 6hours after intensive care admission and infusion rate ofepinephrine (lactic acid level mmol/L � 3.5 � 11.7 epi-nephrine �g/kg per minute, p � 0.4).

1906 RYAN ET AL Ann Thorac SurgLACTIC ACIDOSIS AFTER CARDIAC SURGERY 2002;73:1905–11

Comment

In this study routine cardiac surgical patients displayedtwo distinct temporal patterns of lactic acidosis. Theduration of cardiopulmonary bypass may have ac-counted for some of this difference yet the occurrence ofpostoperative lactic acidosis and the change in lactic acidlevels over time was associated with the carriage ofspecific TNF and IL-10 alleles. A genotype was identifiedwhich was always associated with lactic acidosis, yet notall patients with lactic acidosis had this genotype.

Excess lactic acid accumulation after CPB has beenattributed to splanchnic hypoperfusion with reperfusionin the initial hours following surgery and it is not incon-ceivable that visceral regional hypoperfusion might occuron a frequent basis during cardiopulmonary bypass.However, Haisjackl and colleagues [9] measuringsplanchnic blood flow with indocyanine green and using

gastric tonometry for mucosal pH measurement foundno evidence of splanchnic hypoperfusion. Indeed post-CPB splanchnic perfusion and lactic acid levels were bothincreased compared with pre-CPB levels, suggesting thatsplanchnic lactic acid production after cardiac surgery isrelated to systemic inflammation. Cremer and associates[3] investigated the “low systemic vascular syndrome”after CPB and found that patients with low systemicvascular resistance had greatly increased levels of TNFand that this excess TNF was always associated with aconcomitant lactic acidosis. Thus lactic acid productionafter CPB can be a manifestation of TNF-mediated sys-temic inflammation rather than hypoperfusion.

The association between inflammatory cytokines andexcess lactic acid production is well recognized in sepsisrelated organ failure. In this setting excess production oflactic acid parallels excess inflammatory cytokine produc-tion in organs that are failing [10]. Vary and colleagues [5] inan animal model linked TNF with inhibition of pyruvatedehydrogenase and excess lactic acid production. Using arat model of sepsis, they observed that anti-TNF antibodyreversed both TNF-mediated pyruvate dehydrogenase in-hibition and excess lactate production.

Epinephrine and other potent �-adrenergic agonists maycause lactic acidosis [11]. The mechanism for this is unclearbut a hypothesis suggests that coupling of membranebound Na/K ATPases and anaerobic glycolytic enzymesmay be responsible. Lactic acidosis has been reported withepinephrine administration after cardiopulmonary bypass[12]. Totaro and Raper [12] reported that 6 of 18 patientswho received epinephrine after cardiopulmonary bypassdeveloped lactic acidosis whereas none of 17 patients in anorepinephrine group had lactic acidosis. The study did notdetermine why only a third of patients developed acidosisin the epinephrine group. As this study did not includepatients with lactic acidosis who did not require inotropicsupport, the occurrence of lactic acidosis in such patientswas not investigated.

TNF-� and TNF-� are similar compounds and both areactive at TNF receptors [13]. TNF-� is primarily producedby activated monocytes and TNF-� by activated lympho-cytes. The B1 allele of the TNF-� polymorphism was firstassociated with excess TNF-� production by Messer andcolleagues [14] and has been associated with greater sever-ity of colitis by Koss and associategs [15]. Many polymor-phisms in the TNF-� promoter gene have been described

Table 1. Patient Demographics and Operative Details

ControlGroup

StudyGroup

pValue

Number 30 21Age (years) 60.2 � 1.6 61.4 � 1.9 NSSex (male) 28 16 NSBody surface area (m2) 1.96 � 0.03 1.88 � 0.04 NSHemoglobin (g/dL) 14.1 � 0.3 14 � 0.4 NSUrea (mmol/L) 7 � 0.4 6.8 � 0.5 NSCreatinine (mmol/L) 106 � 3.3 105 � 4 NSAlbumin (g/dL) 38 � 2.4 36 � 2.9 NSHypertension 9 6 NSMyocardial infarction 6 4 NSDiabetes mellitus 2 2 NSCABG 27 16Valve 1 4CABG/valve 2 1CPB time (min) 95 � 4.5 109 � 5.7 0.057Minimum CPB

temperature (C°)31.4 � 0.3 31.8 � 0.4 NS

Blood loss 24 hours (mL) 657 � 90 733 � 109 NSPostoperative ventilation

time (hours)6.8 � 1.3 11.5 � 1.5 0.02

All values are quoted as a frequency for categorical values and as amean � standard error for continuous variables.

CABG � coronary artery bypass graft; CPB � cardiopulmonary by-pass; NS � not significant.

Table 2. Lactic Acid Levels (mmol/L) in Control and StudyPatients After Surgery

ControlGroup

StudyGroup

pValue

Number 30 21End cardiopulmonary

bypass1.79 � 0.17 2.96 � 0.2 � 0.0001

Intensive care hour 1 1.47 � 0.19 3 � 0.23 � 0.0001Intensive care hour 6 1.54 � 0.29 6.76 � 0.34 � 0.0001Intensive care hour 12 1.3 � 0.25 3.56 � 0.3 � 0.0001Intensive care hour 24 1.61 � 0.19 1.79 � 10.23 � 0.05

All values are mean � standard error of the mean.

NS � not significant.

Table 3. Lactic Acid Levels (mmol/L) After Cardiac Surgeryin Relation to TNF B Genotype

Genotype TNFB1/B1 TNFB1/B2 TNFB2/B2p

Value

Number 5 30 16End cardiopulmonary

bypass1.9 � 0.5 2.1 � 0.2 2.7 � 0.3 NS

Intensive care hour 1 2.2 � 0.6 2.0 � 0.2 2.4 � 0.3 NSIntensive care hour 6 7.1 � 1.3 2.9 � 0.5 4.1 � 0.7 0.01Intensive care hour 12 3.1 � 0.8 1.9 � 0.3 2.5 � 0.4 NSIntensive care hour 24 2.2 � 0.5 1.2 � 0.2 1.6 � 0.3 NS

All values are mean � standard error of the mean.

TNF � tumor necrosis factor; NS � not significant.

1907Ann Thorac Surg RYAN ET AL2002;73:1905–11 LACTIC ACIDOSIS AFTER CARDIAC SURGERY

and of these, the functional significance and disease asso-ciation of the TNF-G–308A polymorphism is best docu-mented. This TNF-� polymorphism is associated with en-hanced gene transcription [16], a 3.75-fold increase inmortality with septic shock [7], and a sevenfold excessmortality from cerebral malaria [17]. Thus TNF-� andTNF-� polymorphisms are functional and modulate in-flammation. The TNF-B1 allele occurs with greater fre-quency than the TNF–308A allele and as a consequence it iseasier to investigate in a small study such as this. Thepresent study was too small to determine the effects of theTNF-G–308A polymorphism. IL-10 is a potent antiinflam-matory cytokine that inhibits TNF production. The A alleleof a polymorphism at position –1082 in the IL-10 genepromoter region is associated with lower IL-10 production[5]. In inflammatory bowel disease carriage of the IL-10–1082A allele is associated with greater severity of disease[15]. Thus a genotype associated with excess TNF-� and adecrease in IL-10 could be characterized as proinflamma-tory. We observed that all patients with this genotype hadexcess lactic acid. If one considered an additional patienthomozygous for TNF–308A and carrying the IL-10–1082Aalleles in this proinflammatory genotype then the associa-tion would be more prominent.

This study demonstrated that a combination of cyto-kine genetic polymorphisms interact to promote lacticacidosis. It is possible that patients with this proinflam-matory genotype may develop systemic inflammationand lactic acidosis after a lower threshold stimulus.However only 20% of patients in the study group had theidentified genotype. Thus the proinflammatory genotypeidentified was sufficient but not necessary to initiatesystemic inflammation. Further study is required to de-termine whether alternate genetic factors are associatedwith lactic acidosis in cardiac surgical patients. It is likelythat the etiology underlying the generation of systemicinflammation after cardiopulmonary bypass is multifac-torial. Surgical factors such as prolonged cardiopulmo-nary bypass and temperature reduction during cardio-pulmonary bypass likely act as a trigger to precipitatesystemic inflammation. We excluded patients with pro-longed, complex, or eventful surgery and thus the effectsof these factors were minimized. Further study is re-quired to determine the relative importance of geneticand surgical factors in the occurrence of lactic acidosisafter cardiac surgery.

The small size of this study and a focus on low-riskpatients excluded any possibility of relating genotypeand outcome measures. More extensive study for anassociation between genotype and outcome will followthis study. Having genetic information on a patient’sinflammatory response before surgery may have distinctbenefits. Beside the basic science interest concerning therole and interaction of the inflammatory mediators, therecould be very practical clinical benefits as those patientswith genetic predisposition to high cytokine release maybenefit from antiinflammatory mediator strategies.

This study was funded by a grant from the Royal City of DublinHospital Fund.

References

1. Davies AR, Bellomo R, Raman JS, Gutteridge GA, Buxton BF.High lactate predicts failure of intra aortic balloon pumpingafter cardiac surgery. Ann Thorac Surg 2001;71:1415–20.

2. Duke T, Butt W, South M, Karl T. Early markers of adverseevents in children after cardiac operations. J Thorac cardio-vasc Surg 1997;114:1042–52.

3. Cremer J, Martin M, Redl H, et al. Systemic inflammatory re-sponse after cardiac operations. Ann Thorac Surg 1996;61:1714–20.

4. Chiolero RL, Revelly JP, Leverve X, et al. Effects of cardio-genic shock on lactate and glucose metabolism after heartsurgery. Crit Care Med 2000;28:3784–91.

5. Vary T, Hazen S, Maish G, Cooney R. TNF binding proteinprevents hyperlactaemia and inactivation of PDH complexin skeletal muscle during sepsis. J Surg Res 1998;80:44–51.

6. Stuber F, Petersen M, Bokelmann F, Schade U. A genomicpolymorphism within the tumor necrosis factor locus influencesplasma tumor necrosis factor concentrations and outcome ofpatients with severe sepsis. Crit Care Med 1996;24:381–4.

7. Mira JP, Cariou A, Grall F, et al. Association of TNF2,aTNF-alpha promoter polymorphism, with septic shock sus-ceptibility, and mortality: a multicenter study. JAMA 1999;282:579–81.

8. Turner D, Williams D, Sankaran D, Lazarus M, Sinnott PJ,Hutchinson IV. An investigation of polymorphism in the inter-leukin-10 promoter gene. Eur J Immunogenet 1997;24:1–8.

9. Haisjackl M, Birnbaum J, Redlin M, et al. Splanchnic oxygentransport and lactate metabolism during normothermic cardio-pulmonary bypass in humans. Anaesth Anal 1998;86:22–7.

10. Douzinas EE, Tsidemiadou PD, Pitaridis MT, et al. The regionalproduction of cytokines and lactate in sepsis related multipleorgan failure. Am J Resp Crit Care Med 1997;155:53–9.

11. James JH, Luchette FA, McCarter FD, Fischer JF. Lactate isan unreliable indicator of tissue hypoxia in injury or sepsis.Lancet 1999;354:505–8.

12. Totaro RJ, Raper RF. Epinephrine induced lactic acidosis fol-lowing cardiopulmonary bypass. Crit Care Med 1997;25:1693–9.

13. Bazzoni F, Beutler B. The tumor necrosis factor ligand andreceptor family. NEJM 1996;334:1717–25.

14. Messer G, Spengler U, Jung MC, et al. Polymorphic structureof the tumor necrosis factor (TNF) locus: an NcoI polymor-phism in the first intron of the human TNF-� gene correlateswith a variant amino acid in position 26 and a reduced levelof TNF-� production. J Exp Med 1991;173:209–19.

15. Koss K, Satsangi J, Fanning GC, Welsh KI, Jewell DP.Cytokine (TNF alpha, LT alpha and IL-10) polymorphism ininflammatory bowel diseases and normal controls: differen-tial effects on production and allele frequencies. GenesImmun 2000;1:185–90.

16. Wilson AG, Symons JA, McDowell TL, McDevitt HO, DuffGW. Effects of a polymorphism in the human tumor necrosisfactor alpha promoter on transcriptional activation. ProcNatl Acad Sci 1997;94:3195–9.

17. McGuire W, Hill AV, Allsopp CE, Greenwood BM, Kwiat-kowski D. Variation in the TNF alpha promoter region associ-ated with susceptibility to cerebral malaria. Nature 1994;371:508–10.

Table 4. Lactic Acid Levels (mmol/L) After Cardiac Surgeryin Relation to Interleukin 10 Genotype

Genotype IL-10–1082GGIL-10–1082GA or AA

pValue

Number 18 33End cardiopulmonary

bypass2.2 � 0.3 2.3 � 0.2 NS

Intensive care hour 1 1.8 � 0.3 2.3 � 0.2 NSIntensive care hour 6 2.5 � 0.7 4.3 � 0.5 0.03Intensive care hour 12 1.8 � 0.4 2.4 � 0.3 NSIntensive care hour 24 1.3 � 0.3 1.5 � 0.2 NS

All values are mean � standard error of the mean.

NS � not significant.

1908 RYAN ET AL Ann Thorac SurgLACTIC ACIDOSIS AFTER CARDIAC SURGERY 2002;73:1905–11

Appendix 1Polymerase Chain Reaction Primer Sequences With Position of 5� Base, Restriction Enzymes, and Cut Sites

Primer Sequence (5�-3�)Position of

5� BaseRestriction

Enzyme

Restriction Sites

Variant Constant

IL-6–174 sense ATGACTTCAGCTTTACTCTT �324 Hsp 92 II �174 �296IL-6–174 antisense ATAAATCTTTGTTGGAGGGT �81TNFB sense CCCTCCTGCACCTGCTGCCTGG �112 Hinf I �251 �196TNFB antisense AGAGGGG TGGATGCTTGGGTTC �833TNF-�–308 sense GGAGGCAATAGGTTTTGAGGGCCAT �334 Nco I �313 �162TNF-�–308 antisense CTGTCTCGGT TTCTTCTCCATGGCG �140IL-10–1082 sense TCTGAAGAAGTCCTGATGTC �1248 Mnl I �1085 �1196, �1191IL-10–1082 antisense CTCTTACCTATCCCTACTTCC �1059 �1081 �1174IL-10–592 sense GACTCCAGCCACAGAAGCTTA �967 Rsa I �594 �884, �842, �834IL-10–592 antisense ATATCCTCAAAGTTCCCAAGC �531IL-1� �3953 sense GTGTTGTCATCAGACTTTGACCGTA �3816 Taq I �3952 �4052IL-1� �3953 antisense GAGAGCTTTCAGTTCATATCGACCA �4073

IL � interleukin; TNF � tumor necrosis factor.

Appendix 2Polymerase Chain Reaction Cycling Parameters, Changes to Standard Reaction Mix, and Percentage Agarose Gel Used in EachAssay

Polymorphism Cycles DenaturationPrimer

Annealing Elongation

Changes toStandard Reaction

MixAgarose

Gel

IL-6–174 40 94°C, 1 min 58°C, 1 min 72°C, 90 sec 3%TNFB 37 95°C, 30 sec 68°C, 30 sec 74°C, 42 sec 1 �M of each primer 2%TNF-�–308 40 94°C, 1 min 65°C, 1 min 72°C, 1 min 2%IL-10–1082 40 94°C, 1 min 58°C, 1 min 72°C, 1 min No DMSO 4%IL-10–592 40 94°C, 1 min 64°C, 1 min 72°C, 1 min 1 �M of each primer 3%IL-1� �3953 40 94°C, 1 min 65°C, 1 min 72°C, 1 min 3%

IL � interleukin; TNF � tumor necrosis factor.

Appendix 3Arterial Blood Gases and Blood Flow Rate on Cardiopulmonary Bypass

Control Group Study Group p Value

InitialpH 7.38 � 0.01 7.4 � 0.01 NSPCO2 (kPa) 4.95 � 0.1 4.8 � 0.1 NSPO2 (kPa) 33.6 � 1.9 36.6 � 2.2 NSBicarbonate (mmol/L) 22 � 0.2 22.1 � 0.2 NSBase excess �2.8 � 0.3 �2.6�0.3 NSHemoglobin (g/dL) 9.8 � 0.3 9.2 � 0.3 NSFlow L/m2 BSA 2.4 � 0.03 2.4 � 0.04 NSMinimal flow L/m2 BSA 2.2 � 0.05 2.1 � 0.06 NS

FinalpH 7.4 � 0.01 7.39 � 0.01 NSPCO2 (kPa) 4.5 � 0.07 4.7 � 0.09 0.05PO2 (kPa) 26.4 � 1.2 25.9 � 1.42 NSBicarbonate (mmol/L) 21.2 � 0.2 21 � 0.3 NSBase excess �3.8�0.3 �4 � 0.4 NSHemoglobin (g/dL) 9.4 � 0.3 9.2 � 0.3 NSFlow L/m2 BSA 2.5 � 0.02 2.5 � 0.02 NS

BSA � Body surface area in square meters; NS � not significant.

1909Ann Thorac Surg RYAN ET AL2002;73:1905–11 LACTIC ACIDOSIS AFTER CARDIAC SURGERY