uva-dare (digital academic repository) imaging of critical ... · critical limb ischemia peripheral...

UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Imaging of critical limb ischemia

Jens, S.

Publication date2015Document VersionFinal published version

Link to publication

Citation for published version (APA):Jens, S. (2015). Imaging of critical limb ischemia. Boxpress.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:30 Mar 2021

https://dare.uva.nl/personal/pure/en/publications/imaging-of-critical-limb-ischemia(0de3ac68-c360-489d-955d-c9b0872be3a5).html

ImagIng of

CrItICal l Imb

IsChemIa

Isbn 978-94-6295-116-7

Ima

gIn

g o

f C

rIt

ICa

l l

Imb

IsC

he

mIa

sjo

er

d je

ns sjoerd jens

SJRD kaft.indd 1 05-03-15(w10) 09:14

Imaging of Critical Limb Ischemia

Sjoerd Jens

This thesis was prepared at the Department of Radiology, Academic Medical Center, University of Amsterdam, the Netherlands.

Copyright 2015 © Sjoerd Jens, Amsterdam, the NetherlandsNo part of this thesis may be reproduced, stored, or transmitted in any form or by any means, without prior permission of the author.

Financial support by the Dutch Heart Foundation for the publication of this thesis is gratefully acknowledged.Printing of this thesis was financially supported by the Department of Radiology and Surgery (Academic Medical Center, Amsterdam, the Netherlands), Intersocks, JR Foundation, Cheng Shin Tire, Amsterdamsche Football Club, ChipSoft and ABN Amro.

Cover: Aaf Meijer en Thomas DiebenPrinted & Lay Out by: Proefschriftmaken.nl || Uitgeverij BOXPressPublished by: Uitgeverij BOXPress, ’s-HertogenboschISBN: 978-94-6295-116-7

Imaging of Critical Limb Ischemia

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctoraan de Universiteit van Amsterdamop gezag van de Rector Magnificus

prof. dr. D.C. van den Boomten overstaan van een door het college voor promoties

ingestelde commissie, in het openbaar te verdedigen in de Agnietenkapel

op dinsdag 21 april 2015, te 14:00 uur

door

Sjoerd Jensgeboren te Naarden

Promotiecommissie

Promotores: Prof. dr. J.A. Reekers Prof. dr. D.A. LegemateCo-promotores: Dr. M.J.W. Koelemay Dr. S. BipatOverige leden: Prof. dr. N.C. Schaper Prof. dr. R.J. de Haan Prof. dr. M.M. Levi Prof. dr. J.A.W. Teijink Prof. dr. O.M. van Delden

Faculteit der Geneeskunde

5

Table of contents

Chapter 1

Introduction and outline of thesis 7

Chapter 2

Diagnostic Performance of Computed Tomography Angiography and Contrast-Enhanced Magnetic Resonance Angiography in Patients with Critical Limb Ischaemia and Intermittent Claudication: Systematic Review and Meta-analysis. Eur Radiol. 2013 Nov;23(11):3104-14. 17

Chapter 3

Outcomes of Infrainguinal Revascularizations with Endovascular First Strategy in Critical Limb Ischemia. Cardiovasc Intervent Radiol. Accepted 2014 Aug 12. 39

Chapter 4

Title: Lowering Iodinated Contrast Concentration in Infrainguinal Endovascular Interventions: a Three-armed Randomized Controlled Non-inferiority Trial. Submitted 53

Chapter 5

Three-dimensional rotational angiography of the foot in critical limb ischemia: a new dimension in revascularization strategy. Cardiovasc Intervent Radiol. 2013 Jun;36(3):797-802. 69

Chapter 6

Randomized trials for endovascular treatment of infrainguinal arterial disease: systematic review and meta-analysis (Part 1: Above the knee). Eur J Vasc Endovasc Surg. 2014 May;47(5):524-35. 81

Chapter 7

Randomized trials for endovascular treatment of infrainguinal arterial disease: systematic review and meta-analysis (Part 2: Below the knee). Eur J Vasc Endovasc Surg. 2014 May;47(5):536-44. 103

6

Chapter 8

Perfusion Angiography of the Foot in Patients with Critical Limb Ischemia: Description of the Technique. Cardiovasc Intervent Radiol. 2015 Feb;38(1):201-5. 121

Chapter 9

Assessing the Quality of Available Patient Reported Outcome Measures for Intermittent Claudication: A Systematic Review Using the COSMIN Checklist. Eur J Vasc Endovasc Surg. Accepted 2015 Jan 22. 131

Chapter 10

Summary, conclusions and implications 163

Chapter 11

Samenvatting, conclusies en implicaties 171

Chapter 12

Dankwoord, list of publications, portfolio, curriculum vitae 179

7

Introduction

1

CHAPTER 1

Introduction

8

Chapter 1

Introduction

Critical limb ischemia

Peripheral arterial disease (PAD) is a comprehensive term for arterial diseases of the extremities. For the legs, PAD can clinically result in intermittent claudication (IC) and critical limb ischemia (CLI). In IC, the patient experiences lower extremity muscle pain induced by activity, e.g. by walking. When the patient discontinues the activity, the muscle pain is relieved. In CLI, the patient has a more severe form of PAD, which presents as lower limb pain at rest, or as the inability of ulcers or gangrene to heal spontaneously.

Etiology of ischemic foot ulcers

The major cause of PAD is systemic atherosclerosis.1 The most common risk factors for developing PAD are age, gender, smoking, diabetes mellitus, hypertension and dyslipidemia.1 In the literature on treatment of PAD in patients with CLI, patients with diabetic foot disease are seen as a separate entity. In patients with atherosclerosis, arterial occlusive disease is the main cause of sustained ulcers. However, in diabetic foot disease development and healing of an ulcer is multifactorial.2 The main contributors to diabetic foot disease are a combination of neuropathy (autonomic and sensory), atherosclerosis and microangiopathy.2,3 Autonomic neuropathy causes loss of sweating of the foot, resulting in dry skin and callus formation, which increases the pressure on local pressure points. Sensory neuropathy suppresses the pain sensations, thereby increasing the risk of developing an ulcer. Moreover, when an ulcer develops, it may remain unnoticed by the patient due to the sensory neuropathy. After developing an ulcer, arterial blood supply should be sufficient to facilitate ulcer healing. However, due to the atherosclerosis and microangiopathy, which are associated with diabetes foot disease,3 a sufficient blood supply to the foot cannot be achieved in many patients, resulting in a non-healing ulcer. When untreated or inadequately treated, the ulcer can become infected or gangrenous, eventually leading to a minor or even major amputation.

Arterial status

The arterial status of patients presenting with rest pain or an ulcer is evaluated in the vascular laboratory by measuring the ankle blood pressure, toe blood pressure or transcutaneous oxygen pressure (tcPO

2). The second Trans-Atlantic Inter-Society

Consensus Document on Management of Peripheral Arterial Disease (TASC II) defined critical ischemia as an ankle pressure of

9

Introduction

1pressures can be falsely high due to medial arterial calcifications.5,6 In these patients a toe pressure of < 30 mmHg was regarded as indicative of critical ischemia. This cutoff should not be applied that strict, since it was shown recently that a toe-brachial index is not sensitive for earlier detection of ischemia in diabetes.7

Anatomic evaluation

For composing an optimal treatment strategy for the patient, the full arterial tree of the lower limb should be evaluated. At the vascular laboratory this can be performed using duplex ultrasound (DUS), which also gives functional information. For the above-the-knee arteries, DUS has a median sensitivity for detecting a stenosis of more than 50% of 88%, and a median specificity of 95%.8 In the below-the-knee arteries this is respectively 84% and 93%.8 Although DUS is considered to be operator dependent, interobserver agreement seems to be comparable to digital subtraction angiography (DSA), which is considered to be the gold standard for popliteal and tibial artery assessment.9,10

Other modalities for anatomic evaluation in PAD are computed tomography angiography (CTA) and magnetic resonance angiography (MRA), of which MRA can be subdivided into contrast-enhanced MRA (CE-MRA) and non-CE-MRA. CTA requires the use of iodinated contrast agents and for CE-MRA, gadolinium-based contrast agents are necessary. A drawback of CTA is that intravascular use of iodinated contrast has an increased risk for contrast-induced nephropathy and allergic reactions,11 and gadolinium-based contrast agents used for CE-MRA are associated with the development of nephrogenic systemic fibrosis in patients with chronic renal failure.12 Therefore, non-CE-MRA such as time-of-flight MRA would be preferred, but for this imaging modality the sensitivity and specificity is inferior to CTA and CE-MRA in detecting arterial stenosis more than 50% or occlusion in PAD. For detecting significant stenoses or occlusions in the entire arterial tree, non-CE-MRA has a sensitivity and specificity ranging between 79-94% and 74-92% respectively.13 For CTA this is respectively 95% (95%-confidence interval; 95%CI, 92%-97%) and 96% (95%CI, 93-97%).14 For CE-MRA the sensitivity and specificity is respectively 95% (95%CI, 92-96%) and 96% (95%CI, 94-97%).15

The reported diagnostic accuracy of CTA and CE-MRA is based on studies of patients with predominantly IC, a less advanced stage of PAD. Therefore it remains unclear whether the accuracy of these modalities would be the same in patients with CLI, since their arterial lesions may differ in number, extent and location from those of patients with IC.16,17 Also the unpredictable flow pattern and collateral formation in CLI may influence the accuracy. DSA is traditionally considered the reference standard in many studies,14,15 but due to its invasive character and risk for complications this modality is nowadays not indicated anymore for initial anatomic evaluation.1,18 Depending on local expertise,

10

Chapter 1

DUS, CTA and CE-MRA are currently the preferred modalities for assessment of lower extremity arteries in patients with PAD in whom an intervention is considered.

Treatment

At our institution, the first line treatment of CLI patients is endovascular revascularization, especially in those with a poor condition due to comorbidities, unfavorable anatomy for surgery, no venous material for bypass, or old age. Other patients are discussed by a multidisciplinary team to reach consensus on whether an intervention of any kind, i.e. endovascular or surgical revascularization, or primary amputation, is indicated, or whether a conservative approach should be applied. At the Academic Medical Center, endovascular revascularization is performed by an interventional radiologist. The current strategy to open an arterial stenosis or occlusion is to perform a balloon angioplasty. When several attempts of balloon angioplasty provide inadequate dilatation, or when a flow-limiting dissection occurs, a bail-out stent may be placed. Whether one should use a bare stent or a drug-eluting stent is still under debate. If transluminal recanalization of an occluded artery cannot be established, a subintimal angioplasty (SA) may be performed.19 In CLI patients, the 1 year primary patency rate of SA in patients with CLI is relatively low, i.e. about 64-74% in femoropopliteal and 46-56% in tibial artery lesions,20 which is inferior compared to an infrainguinal bypass.21 However, long-term patency in these patients seems to be of less importance as ulcer healing mainly takes place within 6-9 months, and any patency after healing is therefore not essential.22

For tibiopedal endovascular intervention, three strategies can be applied. The first strategy is to revascularize as many arteries as possible, since increasing the number of patent arteries after angioplasty increases the one year limb salvage rate.23 A second strategy is to revascularize the artery supplying the ulcer area, according to the angiosome model.24 According to this model, each tibial artery supplies a specific region in the foot. A limitation of this model is that the three tibial arteries have communicating arteries. Since in patients with CLI additional collaterals are present, it is difficult to determine which artery should be opened to promote ulcer healing.25 A third strategy would be to assess by angiography which tibial artery is most suitable to improve ulcer perfusion, taking into account the arterial pathology and inter-arterial connections. Currently this assessment may be done using DSA, but since this technique is two-dimensional, it is hard to interpret three-dimensional arterial connections.

Outcome assessment

During endovascular intervention the main criterion for success is whether the stenosis or occlusion has become fully patent with adequate outflow.22 Postinterventional parameters such as ankle-brachial index, toe systolic blood pressure and transcutaneous oxygen pressure26,27 can be used to predict wound healing.28 However, there is a need for

11

Introduction

1techniques that not only assess the state of the macrocirculation, but also the perfusion status of an ischemic foot.29 Ideally, an instrument is needed that is able to evaluate the improvement in foot perfusion during the endovascular intervention. This would allow the interventional radiologist to immediately know whether the intervention has led to sufficient increase of perfusion, or whether additional revascularization is necessary, if possible. The aim of treating patients with CLI is to achieve wound healing and prevent amputation, and these outcomes are frequently used in research. However, since patients with CLI are generally old and frail, they will not always retain ambulation and independence despite limb salvage. To be able to attend to the feelings and opinions of the patient, patient-reported outcome measures (PROMs) are increasingly considered to be important. Two patient-reported outcomes (PROs) are quality of life and functional status. Before a PROM can be used in clinical trials and practice, several domains, i.e. reliability (the extent to which scores for patients whose disease status has not changed are the same for repeated measurements), validity (the degree to which an instrument measures the outcome it intends to measure) and responsiveness (the ability to detect change over time in the measured outcome) need to be evaluated. PROMs are a new area of research for PAD and thus need further evaluation before they can be used to guide decisions.

12

Chapter 1

THESIS OUTLINE

This thesis aims to improve diagnostic evaluation, treatment selection, endovascular intervention, and outcome assessment in patients with PAD with a focus on CLI. In Chapter 2 we present a systematic review, evaluating whether CTA or CE-MRA should be performed in patients with CLI or IC for optimal diagnostic work-up. In this review DSA is used as a reference standard, and we indicate which imaging technique is preferred for the different stages of disease (i.e. CLI or IC). After optimal diagnostic work-up, a treatment strategy is selected for patients with CLI. The study reported in Chapter 3 describes the effectiveness of an endovascular treatment first strategy in patients with poor condition due to comorbidities, unfavorable anatomy for surgery, no venous material for bypass or old age. For the remaining patients, a multidisciplinary team discussed and proposed the optimal treatment strategy. We evaluate the clinical outcomes one year after initial treatment in a cohort of patients with CLI, in relation to the selected treatment strategy. If the patient is referred for endovascular intervention, the patient will be exposed to intravascular iodinated contrast. Since this contrast can affect renal function, we report in Chapter 4 whether the endovascular procedure can confidently be performed with the use of two lower iodinated contrast concentrations, comparing 240 and 140 mg iodine/ml with the standard concentration of 300 mg iodine/ml. During endovascular treatment, the interventional radiologist completely relies on the two-dimensional images provided. In patients with PAD and an ulcer, this condition complicates the assessment of which tibial artery is predominantly responsible for the blood supply to the ulcer. Therefore, we report in Chapter 5 whether availability of real-time three-dimensional angiographic images of the ankle and foot contribute to anatomical interpretation and endovascular treatment strategy. For the treatment of arterial stenosis and occlusion, drug-eluting balloons and drug-eluting stents have become available over the last decade. To evaluate whether these ‘new’ devices have additional value over non-drug-eluting balloons and stents, and whether arterial lesions should primarily be stented, we systematically review the literature for femoropopliteal lesions, and for tibial lesions. These reviews are described in Chapter 6 and Chapter 7, respectively. To be able to directly assess the change in foot perfusion before and after percutaneous transluminal angioplasty, we describe a new two-dimensional imaging technique in Chapter 8. In this pilot study, we report on our first experience with this technique, and evaluate the change in perfusion before and after treatment. In the end, the purpose of treating PAD patients is to establish an optimal clinical outcome, i.e. prevention of amputation and relief of ischemic rest pain in CLI, and restoring walking capacity and improving quality of life in IC. To accurately assess and evaluate the quality of life and functional status in patients with IC, and

13

Introduction

1to monitor their change, a valid, reliable and responsive measurement instrument is needed. In Chapter 9 we present the results of a systematic review of literature, and give an overview of the patient-reported outcome measures that are the most suitable for evaluating quality of life and functional status. Finally, in Chapter 10 we summarize the main findings of this thesis and discuss its implications for clinical practice and scientific research.

14

Chapter 1

References

1. Norgren L, Hiatt WR, Dormandy JA et al (2007) Inter-Society Consensus for the Management of Peripheral Arterial Disease (TASC II). J Vasc Surg 45 Suppl S: S5-67

2. Boulton AJ (2013) The pathway to foot ulceration in diabetes. Med Clin North Am 97: 775-790

3. Wiernsperger N, Rapin JR (2012) Microvascular diseases: is a new era coming? Cardiovasc Hematol Agents Med Chem 10: 167-183

4. Carter SA (1992) Ankle and toe systolic pressures comparison of value and limitations in arterial occlusive disease. Int Angiol 11: 289-297

5. Resnick HE, Foster GL (2005) Prevalence of elevated ankle-brachial index in the United States 1999 to 2002. Am J Med 118: 676-679

6. Chistiakov DA, Sobenin IA, Orekhov AN, Bobryshev YV (2014) Mechanisms of medial arterial calcification in diabetes. Curr Pharm Des 20: 5870-5883

7. Stoekenbroek RM, Ubbink DT, Reekers JA, Koelemay MJ (2014) Hide and seek: does the toe-brachial index allow for earlier recognition of peripheral arterial disease in diabetic patients? Eur J Vasc Endovasc Surg doi: 10.1016/j.ejvs.2014.10.020

8. Collins R, Burch J, Cranny G et al (2007) Duplex ultrasonography, magnetic resonance angiography, and computed tomography angiography for diagnosis and assessment of symptomatic, lower limb peripheral arterial disease: systematic review. BMJ 16: 1257

9. Koelemay MJ, Legemate DA, van Gurp JA, de Vos H, Balm R, Jacobs MJ (2001) Interobserver variation of colour duplex scanning of the popliteal, tibial and pedal arteries. Eur J Vasc Endovasc Surg 21: 160-164

10. Koelemay MJ, Legemate DA, Reekers JA, Koedam NA, Balm R, Jacobs MJ (2001) Interobserver variation in interpretation of arteriography and management of severe lower leg arterial disease. Eur J Vasc Endovasc Surg 21: 417-422

11. Bottinor W, Polkampally P, Jovin I (2013) Adverse reactions to iodinated contrast media. Int J Angiol 22: 149-154

12. Daftari Besheli L, Aran S, Shaqdan K, Kay J, Abujudeh H (2014) Current status of nephrogenic systemic fibrosis. Clin Radiol 69: 661-668

13. Collins R, Cranny G, Burch J et al (2007) A systematic review of duplex ultrasound, magnetic resonance angiography and computed tomography angiography for the diagnosis and assessment of symptomatic, lower limb peripheral arterial disease. Health Technol Assess 11: 1-184

14. Met R, Bipat S, Legemate DA, Reekers JA, Koelemay MJ (2009) Diagnostic performance of computed tomography angiography in peripheral arterial disease: a systematic review and meta-analysis. JAMA 301: 415-424

15. Menke J, Larsen J (2010) Meta-analysis: Accuracy of contrast-enhanced magnetic resonance angiography for assessing steno-occlusions in peripheral arterial disease. Ann Intern Med 153: 325-334

16. Ozkan U, Oguzkurt L, Tercan F (2009) Atherosclerotic risk factors and segmental distribution in symptomatic peripheral artery disease. J Vasc Interv Radiol 20: 437-441

17. Graziani L, Silvestro A, Bertone V et al (2007) Vascular involvement in diabetic subjects with ischemic foot ulcer: a new morphologic categorization of disease severity. Eur J Vasc Endovasc Surg 33: 453-60

18. Singh H, Cardella JF, Cole PE et al (2003) Quality improvement guidelines for diagnostic arteriography. J Vasc Interv Radiol 14: S283-288

19. Reekers JA, Kromhout JG, Jacobs MJ (1994) Percutaneous intentional extraluminal recanalisation of the femoropopliteal artery. Eur J Vasc Surg 8: 723-728

15

Introduction

120. Met R, Van Lienden KP, Koelemay MJ, Bipat S, Legemate DA, Reekers JA (2008) Subintimal angioplasty

for peripheral arterial occlusive disease: a systematic review. Cardiovasc Intervent Radiol 31: 687-697

21. Klinkert P, Post PN, Breslau PJ, van Bockel JH (2004) Saphenous vein versus PTFE for above-knee femoropopliteal bypass. A review of the literature. Eur J Vasc Endovasc Surg 27: 357-362

22. Reekers JA, Lammer J (2012) Diabetic foot and PAD: the endovascular approach. Diabetes Metab Res Rev 28 Suppl 1: 36-39

23. Peregrin JH, Koznar B, Kovác J et al (2010) PTA of infrapopliteal arteries: long-term clinical follow-up and analysis of factors influencing clinical outcome. Cardiovasc Intervent Radiol 33: 720-725

24. Alexandrescu VA, Hubermont G, Philips Y et al (2008) Selective primary angioplasty following an angiosome model of reperfusion in the treatment of Wagner 1-4 diabetic foot lesions: practice in a multidisciplinary diabetic limb service. J Endovasc Ther 15: 580-593

25. Brownrigg JR, Apelqvist J, Bakker K, Schaper NC, Hinchliffe RJ (2013) Evidence-based management of PAD & the diabetic foot. Eur J Vasc Endovasc Surg 45: 673-681

26. Ubbink DT, Spincemaille GH, Reneman RS, Jacobs MJ (1999) Prediction of imminent amputation in patients with non-reconstructible leg ischemia by means of microcirculatory investigations. J Vasc Surg 30: 114-121

27. Ubbink DT, Tulevski II, de Graaff JC, Legemate DA, Jacobs MJ (2000) Optimisation of the non-invasive assessment of critical limb ischaemia requiring invasive treatment. Eur J Vasc Endovasc Surg 19: 131-137

28. Schaper NC, Andros G, Apelqvist J et al (2012) Specific guidelines for the diagnosis and treatment of peripheral arterial disease in a patient with diabetes and ulceration of the foot 2011. Diabetes Metab Res Rev 28 Suppl 1: 236-237

29. Apelqvist JA, Lepäntalo MJ (2012) The ulcerated leg: when to revascularize. Diabetes Metab Res Rev 28 Suppl 1: 30-35

17

Diagnostic Performance of CTA and CE-MRA in CLI and IC: Systematic Review

2CHAPTER 2

Diagnostic Performance of Computed Tomography Angiography and Contrast-Enhanced Magnetic Resonance Angiography

in Patients with Critical Limb Ischaemia and Intermittent Claudication: Systematic Review and Meta-analysis.

Sjoerd Jens, Mark J.W. Koelemay, Jim A. Reekers, Shandra Bipat

Eur Radiol. 2013 Nov;23(11):3104-14.

Online Supplementary Figures and Tables can be found at URL: http://www.boxpress.nl/proefschriften/ebooks/sjoerd_jens/

18

Chapter 2

Abstract

OBJECTIVE: To evaluate the diagnostic performance of computed tomography angiography (CTA) and contrast-enhanced magnetic resonance angiography (CE-MRA) in detecting haemodynamically significant arterial stenosis or occlusion in patients with critical limb ischaemia (CLI) or intermittent claudication (IC).

METHODS: Medline and Embase were searched for studies comparing CTA or CE-MRA with digital subtraction angiography as a reference standard, including patients with CLI or IC. Outcome measures were aortotibial arterial stenosis of more than 50 % or occlusion. Methodological quality of studies was assessed using QUADAS.

RESULTS: Out of 5,693 articles, 12 CTA and 30 CE-MRA studies were included, respectively evaluating 673 and 1,404 participants. Summary estimates of sensitivity and specificity were respectively 96 % (95 % CI, 93-98 %) and 95 % (95 % CI, 92-97 %) for CTA, and 93 % (95 % CI, 91-95 %) and 94 % (95 % CI, 93-96 %) for CE-MRA. Regression analysis showed that the prevalence of CLI in individual studies was not an independent predictor of sensitivity and specificity for either technique. Methodological quality of studies was moderate to good.

CONCLUSION: CTA and CE-MRA are accurate techniques for evaluating disease severity of aortotibial arteries in patients with CLI or IC. No significant differences in the diagnostic performance of the two techniques between patients with CLI and IC were found.

19


2

Introduction

Digital subtraction angiography (DSA) is considered the reference standard for evaluating arterial stenosis or occlusion in patients with peripheral arterial disease (PAD).1 In current practice, however, DSA is not used as a diagnostic tool, because of the invasive character and risk of complications,1,2 and because meta-analyses have shown that both computed tomography angiography (CTA)3 and contrast-enhanced magnetic resonance angiography (CE-MRA)4 are highly accurate non-invasive imaging techniques. Some studies even suggest that CE-MRA is superior to DSA in visualising arteries of the lower leg and foot.5,6

Although CTA and CE-MRA are widely used for diagnostic imaging of PAD patients,7 it is unclear whether these techniques are sufficiently accurate in patients with critical limb ischaemia (CLI). Studies included in the meta-analyses of CTA and MRA mainly comprised patients with intermittent claudication (IC) or even asymptomatic patients, and only a small number of CLI patients.3,4 It is important, however, to study both groups separately, as CLI and IC are two different entities. First of all, patients with IC have a less severe form of PAD with mostly single-level lesions as opposed to multilevel disease in patients with CLI.8 Moreover, arterial lesions in IC are located more proximally, i.e. aortoiliac and femoropopliteal segments,9 whereas lesions in CLI patients are mainly located in the femoral and tibial arteries, especially in patients with diabetes or end-stage renal disease.10 As distal artery diameters are smaller it is likely that they are more difficult to assess, especially with concomitant proximal disease. The objective of our systematic review and meta-analysis was to compare the diagnostic performance of CTA and CE-MRA to detect haemodynamically significant arterial stenosis or occlusion, with DSA as the reference standard, in patients with CLI and IC.

Materials and methods

Inclusion criteria and methods of the analysis were specified in advance in a formal protocol and were based on the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines.11 The review protocol was not published or registered.

Literature search

Studies were identified by searching the electronic Medline and Embase databases on 3 August 2012. The search was limited to publication dates from January 1995 to present. The search strategy was developed in collaboration with a clinical librarian. The strategy was broad and consisted of three components, with search terms defined for three components: (1) PAD, (2) CTA or CE-MRA and (3) DSA. Databases

20

Chapter 2

were searched by combining the search terms of individual components using ‘OR’ and subsequently combining the three components using ‘AND’. The searches were checked for completeness by verifying whether all potentially relevant articles from three systematic reviews3,4,12 were identified. See Online Supplementary Tables 1 and 2 for the detailed search strategies.

Study selection

One author (S.J.) first screened titles and abstracts for eligibility. Articles were excluded if the article was a review, case report, comment or letter, if no participants had CLI or IC, or if no CTA or CE-MRA had been performed. The remaining titles and abstracts were assessed independently by two authors (S.J. and M.K.) to identify potentially eligible papers.

Eligibility criteria

A score form was developed, which was first pilot-tested on multiple included studies. The full text of all potentially eligible articles was retrieved and assessed for eligibility by three authors. One observer (S.J.), a PhD student with 4 years of experience in performing systematic reviews, checked all articles. Two observers (M.K. and S.B.), a vascular surgeon and epidemiologist, respectively, both with more than 10 years of experience with performing systematic reviews, checked a subset of articles. Discrepancies were resolved by discussion. All studies comparing CTA or CE-MRA with DSA and including 10 or more patients older than 18 years with CLI or IC were included. Studies including patients with conditions other than CLI or IC, including asymptomatic PAD (Fontaine stage I or Rutherford grade 0), were excluded when outcome measures could not be extracted for patients with either CLI or IC. The language used in full-text articles was restricted to English, Spanish, French, German, Italian and Dutch if translations of articles in other languages could not be obtained. All reviewers possessed enough language experience to understand the studies in these languages. Outcome measures had to be haemodynamically significant stenosis or occlusion of segments between the abdominal aorta and foot arteries. For studies to be included in this review it had to be possible to construct two-by-two contingency tables, i.e. (1) normal or less than 50 % stenosis and (2) more than 50 % stenosis or occlusion, or three-by-three contingency tables, i.e. (1) normal or less than 50 % stenosis, (2) more than 50 % stenosis, and (3) occlusion, to compare CTA or CE-MRA with DSA as a reference standard.

21


2

Data extraction

Three observers extracted data from each article, using the aforementioned score form. One observer (S.J.) checked all articles. Two observers (M.K. and S.B.) checked a subset of articles. Discrepancies were resolved by discussion. Data extracted were characteristics of study design, study participants, type of imaging and outcome measures, i.e. data for two-by-two or three-by-three contingency tables for the main and subgroups.

Methodological quality and risk of bias in individual studies

Methodological quality and potential bias of included studies were assessed independently by three authors (S.J. and M.K. or S.B.). We used the Quality Assessment of Diagnostic Accuracy Studies (QUADAS) tool.13

The QUADAS items scoring the spectrum of patients, assessment of target condition by DSA, avoidance of differential verification and incorporation bias were scored as ‘adequate’ in advance in accordance with our strict selection criteria. Avoidance of misclassification bias was defined as ‘adequate’ if the time period between CTA or CE-MRA and DSA was less than 30 days. All items of methodological quality were scored as adequate (‘yes’), inadequate (‘no’) or not reported (‘unclear’). Disagreements were resolved by discussion.

Group analyses

The main groups analysed were the aortotibial segments evaluated by CTA and CE-MRA. These groups were analysed for the following: summary estimates of sensitivity and specificity to identify arterial stenosis more than 50 % or occlusion; whether prevalence of CLI in the study population was an independent predictor of logit-transformed sensitivity and specificity; proportions of correct diagnosis; understaging and overstaging of arterial segments; interobserver agreement; and publication bias. The subgroups analysed were the aortopopliteal and tibial segments evaluated by CTA and CE-MRA. Additional subgroups for CE-MRA were tibial segments evaluated with a bolus chase or a separate tibial imaging technique. Subgroups were analysed for summary estimates of sensitivity and specificity.

Planned methods of analysis

Two-by-two tables were used to calculate summary estimates of sensitivity and specificity. Data were analysed using a fixed effects, mixed effects or random effects bivariate statistical model depending on the statistical inconsistency assessed with the I2 statistic.14 A random effects model was used if both I2 values were higher than 25 %, fixed effects model if both I2 values were lower than or equal to 25 %, and mixed effects model if one of the two I2 values was lower than or equal to 25 % and the other higher than 25 %.

22

Chapter 2

Differences between summary estimates of the main or subgroups were analysed using the z test15 and a p-value less than 0.05 was considered statistically significant. Prevalence of CLI in the study population was evaluated as an independent predictor of logit-transformed sensitivity and specificity by linear regression analysis incorporated in the same bivariate models.15 A p-value less than 0.05 was considered statistically significant. Three-by-three tables were used to calculate summary estimates of correct diagnosis, understaging and overstaging. These data were obtained using a multivariate approach previously described by Bipat et al,16 using either a random effects or fixed effects multivariate approach, depending on the Akaike Information Criterion (AIC) values. A lower AIC value indicates a better fit.17

For the contingency tables, calculations of summary estimates were based on the averaged outcome if a study compared different imaging techniques of CTA or CE-MRA in the same study population, or if data were available for multiple observers. To assess the possibility of publication bias we constructed funnel plots of the included studies and performed a modified Egger’s linear regression test.18 The funnel plot was constructed and a linear regression test was performed by respectively plotting and analysing the natural logarithm of the diagnostic odds ratio (DOR) per study against the sample size of the individual studies. To prevent the DOR becoming infinite, i.e. no false negative or true negative results, 0.5 would be used instead of zero to calculate the DOR.19

Bivariate analyses, z test and linear regression analyses were performed in SAS (version 9.2; SAS Institute, Cary, NC, USA). I2 statistics were calculated using Excel (Microsoft Office 2003; Microsoft, Redmond, WA, USA). Multivariate analyses of three-by-three tables were performed in WinBUGS (version 1.4; MRC Biostatistics Unit, Cambridge, UK). The modified Egger’s linear regression test was performed using IBM SPSS statistics (version 19.0; IBM, Chicago, IL, USA).

Results

Study selection

Our initial search yielded 3,135 potentially eligible articles in Medline and 2,558 articles in Embase. In total 5,693 articles were identified. Removal of 953 duplicates and screening of title and abstract led to the exclusion of another 4,516 articles. Of the remaining 224 articles, 182 did not fulfil the eligibility criteria and were therefore excluded. In total, 42 articles were included in this review.5,20–60 See Figure 1 for the flow diagram of the article selection process.

23


2

Study characteristics

Of the 42 studies, 12 studies evaluated CTA20–31 and 30 evaluated CE-MRA 5,32–60 compared with DSA. None of the studies compared both CTA and CE-MRA with DSA.

CTA

Twelve CTA studies were included; 6 were prospective26–31 and 6 were retrospective20-25 studies. In total 673 participants with CLI or IC were studied; the median sample size was 35 (range, 18–279). Of 299 participants (44 %) for whom disease status was available, 51 % had CLI and 49 % had IC. The number of slices of the CT systems studied ranged between 1 and 64 slices. See Table 1 for detailed CTA study characteristics.

CE-MRA

Thirty CE-MRA studies were included, of which 28 were performed prospectively.5,32–40,42–45,47–60 In total 1,404 participants with CLI or IC were studied;

Figure 1. Flow diagram of the article selection process

Abbreviations: CE-MRA, contrast-enhanced magnetic resonance angiography; CLI, critical limb ischaemia; CTA, computed tomography angiography; DSA, digital subtraction angiography; IC, intermittent claudication

24

Chapter 2

Tabl

e 1.

Com

pute

d to

mog

raph

y an

giog

raph

y (C

TA) a

nd c

ontr

ast-

enha

nced

mag

neti

c re

sona

nce

angi

ogra

phy

(CE-

MRA

) stu

dy c

hara

cter

isti

cs

CT

A s

tudi

es

Aut

hor

Des

ign

No.

of

pati

ents

(%

men

)A

ge, m

ean

(SD

or

ran

ge),

YFo

ntai

ne s

tage

II

/III

/IV

(%

)C

T s

lices

Segm

ents

(no

. per

pt

)B

ilate

ral

asse

ssm

ent

Foti

adis

201

1 [2

0]R

etro

spec

tive

41(5

6)66

(12

)27

/22/

5164

Aor

to-p

oplit

eal (

35)

Yes

Kau

201

1 [2

1]R

etro

spec

tive

58(6

0)73

(38

-89)

31/2

2/47

2x32

Aor

to-t

ibia

l (35

)Ye

sLi

GC

200

8 [2

2]R

etro

spec

tive

31(N

A)

NA

(51

-90)

NA

/NA

/23

64Fe

mor

o-ti

bial

(6)

No

Cia

200

7 [2

3]R

etro

spec

tive

279(

NA

)59

(38

-82)

NA

16A

orto

-tib

ial (

17)

Yes

Li X

M 2

007

[24]

Ret

rosp

ecti

ve30

(NA

)N

AN

A64

Aor

to-t

ibia

l (24

)Ye

sB

ui 2

005

[25]

Ret

rosp

ecti

ve25

(96)

63 (

10)

56/2

0/24

4A

orto

-tib

ial (

31)

Yes

Will

man

n 20

05 [

26]

Pros

pect

ive

39(6

9)65

(44

-81)

82/1

8/0

16A

orto

-tib

ial (

35)

Yes

Cat

alan

o 20

04 [

27]

Pros

pect

ive

50(7

8)67

(43

-89)

6/48

/46

4A

orto

-tib

ial (

23)

Yes

Mar

tin

2003

[28

]Pr

ospe

ctiv

e41

(NA

)67

(45

-84)

NA

4A

orto

-tib

ial (

35)

Yes

Ofe

r 20

03 [

29]

Pros

pect

ive

18(8

3)64

(50

-79)

78/1

7/6

NA

Aor

to-t

ibia

l (25

)Ye

sPu

ls 2

001

[30]

Pros

pect

ive

31(5

4)53

(38

-75)

97/3

/04

Aor

to-p

oplit

eal (

7)Ye

sR

ieke

r 19

97 [

31]

Pros

pect

ive

30(N

A)

62 (

42-8

5)87

/10/

31

Fem

oro-

popl

itea

l (7)

Yes

CE

-MR

A s

tudi

es

Aut

hor

Des

ign

No.

of

pati

ents

(%

men

)A

ge, m

ean

(SD

or

ran

ge),

YFo

ntai

ne s

tage

II

/III

/IV

(%

)M

R p

roto

col (

Stre

ngth

, Te

sla)

Segm

ents

(no

. per

pt

)B

ilate

ral

asse

ssm

ent

Anz

idei

201

1 [3

2]Pr

ospe

ctiv

e35

(60

)66

(18

)71

/14/

14St

atio

n by

sta

tion

(1.

5)Fe

mor

o-ti

bial

(16

)Ye

sB

ui 2

010

[33]

Pros

pect

ive

333

(69)

64 (

10)

NA

Bol

us-c

hase

(1.

5)A

orto

-ilia

c (7

)Ye

sG

erre

tsen

201

0 [3

4]Pr

ospe

ctiv

e31

(N

A)

NA

84/1

0/6

Stat

ion

by s

tati

on (

1.5)

Aor

to-t

ibia

l (23

)Ye

sW

ang

2010

[35

]Pr

ospe

ctiv

e31

(67

)72

(46

-87)

0/29

/71

Hyb

rid

(1.5

)A

orto

-tib

ial (

32)

Yes

Posc

henr

iede

r 20

09 [

36]

Pros

pect

ive

20 (

70)

67 (

11)

NA

Bol

us-c

hase

(1.

5)Fe

mor

o-ti

bial

(6)

No

Ow

en 2

009

[37]

Pros

pect

ive

30 (

73)

69 (

37-9

0)0/

40/6

0H

ybri

d (1

.5)

Aor

to-t

ibia

l (13

)N

oB

erg

2008

[38

]Pr

ospe

ctiv

e30

(63

)N

A (

43-8

1)60

/30/

10H

ybri

d (3

.0)

Aor

to-t

ibia

l (30

)Ye

sA

ndre

isek

200

7 [3

9]Pr

ospe

ctiv

e31

(74

)67

(43

-81)

84/3

/13

Bol

us-c

hase

+ h

ybri

d (1

.5)

Fem

oro-

tibi

al (

20)

No

Die

hm 2

007

[40]

Pros

pect

ive

10 (

60)

82 (

8)10

0/0/

0B

olus

-cha

se (

1.5

and

3.0)

Fem

oro-

tibi

al (

8)N

oD

euts

chm

ann

2006

[41

]R

etro

spec

tive

38 (

55)

68 (

49-8

4)N

AB

olus

-cha

se (

1.5)

Aor

to-t

ibia

l (N

A)

NA

Gjo

nnae

ss 2

006

[42]

Pros

pect

ive

58 (

62)

NA

(47

-80)

100/

0/0

Bol

us-c

hase

(1.

5)A

orto

-pop

litea

l (15

)Ye

s

25


2

Jank

a 20

05 [

43]

Pros

pect

ive

25 (

NA

)N

AN

AH

ybri

d (1

.0)

Aor

to-t

ibia

l (26

)Ye

sLa

peyr

e 20

05 [

44]

Pros

pect

ive

31 (

70)

65 (

35-8

3)0/

0/10

0H

ybri

d (1

.5)

Fem

oro-

tibi

al (

10)

No

Lein

er 2

005

[45]

Pros

pect

ive

152

(65)

62 (

10)

96/4

/0B

olus

-cha

se (

1.5)

Aor

to-p

oplit

eal (

7)Ye

sSc

hmit

t 200

5 [4

6]R

etro

spec

tive

69 (

73)

67 (

10)

54/1

3/33

Stat

ion

by s

tati

on (

1.5)

Tib

ial (

8)N

ode

Vri

es 2

005

[47]

Pros

pect

ive

38 (

71)

63 (

10)

89/8

/3B

olus

-cha

se (

1.5)

Aor

to-t

ibia

l (29

)Ye

sB

ezoo

ijen

2004

[48

]Pr

ospe

ctiv

e15

(86

)66

(52

-77)

80/1

3/7

Bol

us-c

hase

(1.

5)A

orto

-tib

ial (

29)

Yes

Lein

er 2

004

[5]

Pros

pect

ive

23 (

52)

72 (

56-8

6)0/

100/

0B

olus

-cha

se (

1.5)

Aor

to-i

liac

(21)

No

Cro

nber

g 20

03 [

49]

Pros

pect

ive

35 (

45)

78 (

50-9

8)9/

3/89

Hyb

rid

(1.5

)T

ibia

l (13

)N

oH

uber

200

3 [5

0]Pr

ospe

ctiv

e40

(N

A)

61 (

43-7

8)78

/23/

0St

atio

n by

sta

tion

(1.

5)A

orto

-tib

ial (

19)

Yes

Stef

fens

200

3 [5

1]Pr

ospe

ctiv

e50

(58

)65

(35

-86)

NA

Bol

us-c

hase

(1.

5)A

orto

-tib

ial (

19)

Yes

Wyt

tenb

ach

2003

[52

]Pr

ospe

ctiv

e56

(69

)67

(35

-89)

82/1

8/0

Bol

us-c

hase

(1.

5)A

orto

-tib

ial (

NA

)Ye

sLe

nhar

t 200

0 [5

3]Pr

ospe

ctiv

e20

(N

A)

NA

NA

Stat

ion

by s

tati

on (

1.5)

Aor

to-t

ibia

l (21

)Ye

sLu

ndin

200

0 [5

4]Pr

ospe

ctiv

e39

(53

)67

(51

-87)

87/1

0/3

Bol

us-c

hase

(1.

0)A

orto

-ilia

c (7

)Ye

sSu

eyos

hi 2

000

[55]

Pros

pect

ive

13 (

100)

72 (

61-8

7)10

0/0/

0B

olus

-cha

se (

1.5)

Aor

to-i

liac

(8)

Yes

Lenh

art 1

999

[56]

Pros

pect

ive

17 (

NA

)N

AN

AB

olus

-cha

se (

1.5)

Aor

to-t

ibia

l (20

)Ye

sSu

eyos

hi 1

999

[57]

Pros

pect

ive

23 (

86)

68 (

52-8

5)83

/17/

0St

atio

n by

sta

tion

(1.

5)A

orto

-tib

ial (

19)

Yes

Win

tere

r 19

99 [

58]

Pros

pect

ive

76 (

56)

66 (

36-9

6)87

/13/

0B

olus

-cha

se (

1.5)

Aor

to-t

ibia

l (31

)Ye

sLa

issy

199

8 [5

9]Pr

ospe

ctiv

e20

(85

)53

(42

-62)

100/

0/0

Stat

ion

by s

tati

on (

1.0)

Fem

oro-

tibi

al (

26)

Yes

Rof

sky

1997

[60

]Pr

ospe

ctiv

e15

(60

)66

(49

-89)

0/N

A/N

ASt

atio

n by

sta

tion

(1.

5)A

orto

-tib

ial (

21)

NA

Abb

revi

atio

ns: C

T, c

ompu

ted

tom

ogra

phy;

MR

, mag

neti

c re

sona

nce;

NA

, not

ava

ilabl

e; p

t, pa

tien

t; Y,

yea

r

26

Chapter 2

the median sample size was 31 (range, 10–333). Of 901 (64 %) participants for whom disease status was available, 31 % had CLI and 69 % had IC. The MR system had a magnetic strength of 1.5 Tesla (T) in 25 studies, 1.0 T in 3, and 3.0 T in 1. One study compared 1.5 T with 3.0 T.40 The MR protocols consisted of a single-injection bolus chase in 15,5,33, 36,40–42,45,47,48,51,52,54–56,58 multi-injection station-by-station imaging in 8,32,34,46,50,53,57,59,60 and of a hybrid protocol with separate imaging of the tibial arteries in 6 studies.35,37,38,43,44,49 One study compared a bolus chase with a hybrid protocol.39 See Table 1 for detailed CE-MRA study characteristics.

Methodological quality and risk of bias in individual studies

Selection criteria were clearly described in 42 % (5/12) for CTA and 50 % (15/30) for CE-MRA studies. Risk of misclassification bias was avoided in 42 % of CTA (5/12) and 80 % of CE-MRA studies (24/30). Risk of partial verification bias was avoided in 50 % (6/12) of CTA and 77 % (23/30) of CE-MRA studies. Execution of DSA was clearly described in 58 % (7/12) of CTA and in 43 % (13/30) of CE-MRA studies. None of the CTA studies and 3 % of the CE-MRA studies (1/30) clearly had clinical data available during image interpretation. See Figure 2 for overall methodological quality of CTA and CE-MRA studies and Online Supplementary Figures 1 and 2 for the individual methodological quality of the studies.

Figure 2. Methodological quality of CTA and CE-MRA studies used for meta-analysis.

Abbreviations: CE-MRA, contrast-enhanced magnetic resonance angiography; CTA, computed tomography angiography; DSA, digital subtraction angiography

27


2

Syntheses of CTA results

Two-by-two contingency tables

Aortotibial segments

Two of the 12 included studies had an additional segment analysed outside the aortotibial segments, i.e. renal arteries and dorsal pedal artery.24,29 The summary estimate of sensitivity was 96 % (95 % CI, 93–98 %; I2 = 92 %) and that of specificity was 95 % (95 % CI, 92–97 %; I2 = 97 %). Of the arterial segments studied by CTA 1.0 % were non-diagnostic. Regression analysis was possible in eight studies reporting the disease status of 292 patients (43 % of the total study population). The prevalence of CLI in these patients was not a statistically significant independent predictor of logit-transformed sensitivity (p = 0.77) and specificity (p = 0.65) in the aortotibial segments.

Aortopopliteal segments

Five studies reported the outcome of the image analysis of aortopopliteal segments.20,21,26,30,31 The summary estimate of sensitivity was 97 % (95 % CI, 83–100 %; I2 = 86 %) and specificity was 95 % (95 % CI, 90–98 %; I2 = 86 %). Of the segments studied by CTA 1.1 % were non-diagnostic.

Tibial segments

Three studies reported the outcome of image analysis of the tibial arterial segments.20,21,26 The summary estimate of sensitivity was 95 % (95 % CI, 91–97 %; I2 = 70 %) and specificity was 91 % (95 % CI, 60–98 %; I2 = 99 %). Of the segments studied by CTA 1.6 % were non-diagnostic. See Table 2 for the summary estimates of the sensitivity and specificity of CTA and Online Supplementary Tables 3–5 for two-by-two contingency tables of CTA main and subgroups.

Aortopopliteal versus tibial segments

No statistically significant differences were found between the summary estimates of the sensitivity (P = 0.58) and specificity (P = 0.52) of aortopopliteal and tibial segments.

Three-by-three contingency tables

Ten studies allowed reconstruction of three-by-three contingency tables for outcome measures.20,22–25,27–31 CTA correctly diagnosed segments as normal or stenosis less than 50 %, stenosis more than 50 % and occlusion in respectively 96 % (95 % CI, 94–97 %), 89 % (95 % CI, 86–92 %) and 94 % (95 % CI, 92–96 %). Stenoses more than 50 % on DSA were understaged by CTA in 7 % (95 % CI, 4–10 %), whereas 4 % (95 % CI,

28

Chapter 2

Tabl

e 2.

Sum

mar

y es

tim

ates

of C

TA a

nd C

E-M

RA

Sum

mar

y es

tim

ates

fro

m t

wo-

by-t

wo

tabl

esTe

chni

que

Segm

ents

No.

of

stud

ies

Non

-dia

gnos

tic

(%)

Sens

itiv

ity

in %

(95

%C

I)Sp

ecifi

city

in %

(95

%C

I)C

TA

aort

o-ti

bial

121.

096

(93

-98)

95 (

92-9

7)ao

rto-

popl

itea

l5

1.1

97 (

83-1

00)

95 (

90-9

8)ti

bial

31.

695

(91

-97)

91 (

60-9

8)C

E-M

RA

aort

o-ti

bial

301.

293

(91

-95)

94 (

93-9

6)ao

rto-

popl

itea

l22

0.2

93 (

88-9

5)95

(94

-96)

tibi

al19

2.1

94 (

92-9

6)93

(90

-96)

tibi

al b

olus

-cha

se*

107.

393

(88

-96)

92 (

89-9

5)ti

bial

sep

arat

e**

101.

395

(91

-97)

94 (

88-9

7)Su

mm

ary

esti

mat

es f

rom

thr

ee-b

y-th

ree

tabl

esD

SA N

orm

alD

SA s

teno

sis

DSA

occ

lusi

onM

odal

ity

Nor

mal

Sten

osis

Occ

lusi

onN

orm

alSt

enos

isO

cclu

sion

Nor

mal

Sten

osis

Occ

lusi

onC

TA

(%

)96

(95%

CI:

94-

97)

4(9

5%C

I: 3

-5)

0.1

(95%

CI:

0-0

.3)

7(9

5%C

I: 4

-10)

89(9

5%: 8

6-92

)4

(95%

CI:

3-6

)1

(95%

CI:

0.4

-2)

5(9

5%C

I: 3

-6)

94(9

5%C

I: 9

2-96

)C

E-M

RA

(%

)96

(95%

CI:

94-

97)

4(9

5%C

I: 3

-6)

0.4

(95%

CI:

0.2

-0.6

)6

(95%

CI:

4-9

)89

(95%

CI:

86-

92)

4(9

5%C

I: 3

-6)

3(9

5%C

I: 2

-5)

5 (9

5%C

I: 3

-7)

92(9

5%C

I: 8

9-94

)

Abb

revi

atio

ns: C

E-M

RA

, con

tras

t-en

hanc

ed m

agne

tic

reso

nanc

e an

giog

raph

y; C

TA

, com

pute

d to

mog

raph

y an

giog

raph

y.*

bolu

s-ch

ase

tech

niqu

e st

arti

ng fr

om th

e ao

rtic

or

fem

oral

leve

l **

Im

agin

g of

the

tibi

al a

rter

ies

usin

g a

sepa

rate

tibi

al im

agin

g te

chni

que

29


2

3–6 %) were overstaged. See Table 2 for the three-by-three contingency table summary estimates of CTA studies and Online Supplementary Table 6 for individual study data.

Interobserver agreement

Interobserver agreement of CTA was assessed using ĸ statistics in four studies, evaluating stenosis grades ranging from two to five categories.21,25–27 The median ĸ value was 0.80 (range, 0.66–0.93).

Syntheses of CE-MRA results

Two-by-two contingency tables

Aortotibial segments

In 5 out of 30 studies pedal arterial segments were analysed in combination with aortotibial segments. The summary estimate of sensitivity was 93 % (95 % CI, 91–95 %; I2 = 87 %) and that of specificity was 94 % (95 % CI, 93–96 %; I2 = 91 %). Of the segments studied by CE-MRA 1.2 % were non-diagnostic. Regression analysis was possible in 23 studies reporting the disease status of 901 patients (64 % of the total study population). Prevalence of CLI in these patients was not a statistically significant independent predictor of logit-transformed sensitivity (p = 0.60) and specificity (p = 0.16).

Aortopopliteal segments

A total of 22 studies reported the outcome of CE-MRA image analysis compared with DSA.32,33,35–45,47,48,52–56,58,59 The summary estimate of sensitivity was 93 % (95 % CI, 88–95 %; I2 = 74 %) and that of specificity was 95 % (95 % CI, 94–96 %; I2 = 67 %). Of the segments studied by CE-MRA 0.2 % were non-diagnostic.

Tibial segments

Outcome of image analyses was available in 19 studies. 32,35–41,43,44,46–49,52,53,56,58,59 Nine of these studies did not focus their imaging on tibial arteries,36,41,47,48,52,53,56,58,59 i.e. they used a bolus chase technique that started at the aortic or femoral level, nine studies imaged the calf separately,32,35,37,38,40,43,44,46,49 and one study analysed the tibial arteries using both techniques,39 i.e. bolus chase and separate imaging of tibial arteries. Summary estimates of sensitivity and specificity of all tibial segments analysed by CE-MRA were 94 % (95 % CI, 92–96 %; I2 = 66 %) and 93 % (95 % CI, 90–96 %; I2 = 89 %) respectively. CE-MRA was non-diagnostic in 2.1 % of arterial segments.

30

Chapter 2

Bolus chase or separate tibial imaging technique for tibial segments

Tibial segments imaged using the CE-MRA bolus chase technique starting from the aortic or femoral level had summary estimates of sensitivity of 93 % (95 % CI, 88–96 %; I2 = 57 %) and specificity of 92 % (95 % CI, 89–95 %; I2 = 83 %). About 7.3 % of the tibial segments were non-diagnostic. Imaging of tibial arteries using a separate tibial imaging technique had summary estimates of sensitivity of 95 % (95 % CI, 91–97 %; I2 = 72 %) and specificity of 94 % (88–97 %; I2 = 92 %). Only 1.3 % of these segments were non-diagnostic. No statistically significant differences were found between the summary estimates of sensitivity (p = 0.39) and specificity (p = 0.56) of the bolus chase and separate imaging of the calf technique. See Table 2 for the summary estimates of sensitivity and specificity of CE-MRA and Online Supplementary Tables 7–9 for the two-by-two contingency tables of the CE-MRA main and subgroups.

Aortopopliteal versus tibial segments

Summary estimates of the sensitivity (p = 0.43) and specificity (p = 0.25) between the aortopopliteal and tibial segments showed no statistically significant differences.

Three-by-three contingency tables

Eleven studies reported outcome measures to construct three-by-three contingency tables.32,43,44,48,51,53,54,56–59 Of the segments scored by DSA as normal or stenosis less than 50 %, stenosis more than 50 %, or occluded, 96 % (95 % CI, 94–97 %), 89 % (95 % CI, 86–92 %) and 92 % (95 % CI, 89–94 %) respectively were correctly classified by CE-MRA. Of the segments scored as stenosis more than 50 %, about 6 % (95 % CI, 4–9 %) were understaged, and 4 % (95 % CI, 3–6 %) were overstaged. See Table 2 for three-by-three contingency table summary estimates of CE-MRA studies and Online Supplementary Table 6 for individual study data.

Interobserver agreement

Interobserver agreement of CE-MRA was assessed using ĸ statistics in 16 studies, reporting on 22 MRA techniques, evaluating stenosis grades ranging from two to six categories.5,32,34–40,43–45,50–52,59 The median ĸ value was 0.82 (range, 0.60–1.00).

Comparison of CTA versus CE-MRA

Summary estimates of the sensitivity (P = 0.15) and specificity (P = 0.61) of CTA and CE-MRA of aortotibial segments did not differ significantly. No statistically significant differences were found between CTA and CE-MRA for aortopopliteal (sensitivity, p = 0.34; specificity, p = 0.22) and tibial (sensitivity, p = 0.74; specificity, p = 0.73) segments.

31


2

Publication bias for CTA and CE-MRA

Linear regression analyses showed non-significant regression coefficients for CTA of 6.5 (95 % CI, −16.7 to 29.7; p = 0.55) and for CE-MRA of −5.4 (95 % CI, −18.1 to 7.3; p = 0.39). The funnel plots of natural logarithm DOR plotted against the sample size per study are presented in Online Supplementary Figures 3 and 4.

Discussion

Summary of evidence

Our meta-analyses show that both CTA and CE-MRA are accurate tools for detecting haemodynamically significant stenoses and occlusions in aortotibial arterial segments in patients with either CLI or IC. CTA had a sensitivity and specificity of 96 % (95 % CI, 93–98 %) and 95 % (95 % CI, 92–97 %) respectively, whereas CE-MRA had a sensitivity of 93 % (95 % CI, 91–95 %) and specificity of 94 % (95 % CI, 93–96 %). Moreover, CTA and CE-MRA can distinguish confidently between segments with a stenosis more than 50 % and occlusion, and both have good to excellent interobserver agreement. The number of non-diagnostic segments was small, i.e. about 1–2 %, for both CTA and CE-MRA. Diagnostic accuracy did not change significantly for both CTA and CE-MRA when prevalence of CLI increased in the study population, or when only tibial arteries were evaluated. For tibial arteries the summary estimates of CTA are inaccurate because of the limited data. Bolus chase or separate imaging of the calf arteries, used for CE-MRA in lower leg assessment, had similarly high diagnostic accuracy. However, separate imaging resulted in fewer non-diagnostic tibial arterial segments. The methodological quality of both CTA and CE-MRA studies was moderate to good and there were no indications for the presence of publication bias.

Limitations

An important aim of our study was to estimate the summary sensitivity and specificity of CTA or CE-MRA for patients with CLI or IC separately, but this was not possible. It is striking that disease status was reported for only 58 % of patients in the included studies, which significantly limits the external validity of such studies. For CTA, all 12 study cohorts consisted of a mix of CLI and IC patients. For CE-MRA, four studies reported the outcomes of patients with solely IC,40,42,55,59 and five studies patients with CLI only.5,35,37,44,60 We therefore decided to perform linear regression analyses to study the association between prevalence of CLI in the inception cohort, and sensitivity and specificity. Four CTA and seven CE-MRA studies could not be used for these analyses, as they did not report the prevalence of CLI. Linear regression analyses showed no significant associations, meaning that the prevalence of CLI patients in the study may

32

Chapter 2

not affect either sensitivity or specificity. However, the heterogeneity of studies was great, making these analyses less reliable. Preferably we would also have performed subgroup analyses of the aortoiliac, femoropopliteal and tibial segments. Unfortunately, because of poor reporting, such analyses were not possible and we combined, therefore, the aortoiliac and femoropopliteal segments into one subgroup, i.e. ‘aortopopliteal’. The summary estimates of sensitivity and specificity were based on the diagnostic arterial segments; non-diagnostic arterial segments were not included in these analyses. Yet, only 1–2 % of the arterial segments were non-diagnostic for both CTA and CE-MRA. We believe, therefore, that exclusion of non-diagnostic segments will not have a major impact on the diagnostic accuracy of CTA and CE-MRA. Imaging of tibial arteries using a bolus chase technique had a 7.3 % rate of non-diagnostic segments, as opposed to 1.3 % with separate imaging of the tibial arteries. Separate imaging of the tibial arteries would, therefore, be preferred. Specificity may be overestimated given that most studies analysed both legs, resulting in around two thirds (72 %) of arterial segments without haemodynamically significant lesions. Readers of CTA or CE-MRA images were, therefore, likely to be biased towards scoring a negative result, i.e. no occlusion or stenosis more than 50 %. Vice versa, this may have led to an underestimation of sensitivity. However, overall data and all subgroups showed high sensitivity values and therefore the underestimation seems to be limited or even absent. Because of our restrictive selection criteria, several studies were not included, resulting in the exclusion of potentially relevant, but also disruptive data. The language restriction resulted in the exclusion of 13 potentially eligible articles. These articles concern the Chinese, Japanese or Russian languages, making translations by computer programmes difficult.61–73 Other articles were excluded as outcomes for just CLI or IC patients were not reported separately, but also for asymptomatic patients or patients with aneurysms. Furthermore, authors were not contacted and articles were immediately excluded if two-by-two contingency tables could not be constructed, because from previous experience it seems that few authors are able to provide unpublished data. In summary, both computed tomography angiography and contrast-enhanced magnetic resonance angiography are accurate techniques in evaluating disease severity of arterial segments from the aorta to the tibial arteries in patients with critical limb ischaemia or intermittent claudication. No significant difference was demonstrated in the diagnostic performance of computed tomography angiography and contrast-enhanced magnetic resonance angiography between patients with these conditions. For tibial arteries a separate imaging technique by contrast-enhanced magnetic resonance angiography is preferred. To perform subgroup meta-analyses in the future, studies should report outcomes separately for patients with critical limb ischaemia or intermittent claudication, and for the aortoiliac, femoropopliteal and tibial segments.

33


2

Acknowledgements

Joost Daams, MA (clinical librarian at Academic Medical Center Amsterdam, the Netherlands), provided assistance with the study search. Mr Daams did not receive compensation for his contribution. The authors have no conflict of interest.

34

Chapter 2

References

1. Norgren L, Hiatt WR, Dormandy JA, Nehler MR, Harris KA, Fowkes FG (2007) Inter-society consensus for the management of peripheral arterial disease (TASC II). J Vasc Surg 45:S5–67

2. Singh H, Cardella JF, Cole PE et al (2003) Quality improvement guidelines for diagnostic arteriography. J Vasc Interv Radiol 14:S283–S288

3. Met R, Bipat S, Legemate DA, Reekers JA, Koelemay MJ (2009) Diagnostic performance of computed tomography angiography in peripheral arterial disease: a systematic review and meta-analysis. JAMA 301:415–424

4. Menke J, Larsen J (2010) Meta-analysis: accuracy of contrast-enhanced magnetic resonance angiography for assessing steno-occlusions in peripheral arterial disease. Ann Intern Med 153:325–334

5. Leiner T, Kessels AG, Schurink GW et al (2004) Comparison of contrast-enhanced magnetic resonance angiography and digital subtraction angiography in patients with chronic critical ischemia and tissue loss. Invest Radiol 39:435–444

6. Dorweiler B, Neufang A, Kreitner KF, Schmiedt W, Oelert H (2002) Magnetic resonance angiography unmasks reliable target vessels for pedal bypass grafting in patients with diabetes mellitus. J Vasc Surg 35:766–772

7. Cao P, Eckstein HH, De RP et al (2011) Chapter II: diagnostic methods. Eur J Vasc Endovasc Surg 42(Suppl 2):S13–S32

8. Matzke S, Lepantalo M (2001) Claudication does not always precede critical leg ischemia. Vasc Med 6:77–80

9. Ozkan U, Oguzkurt L, Tercan F (2009) Atherosclerotic risk factors and segmental distribution in symptomatic peripheral artery disease. J Vasc Interv Radiol 20:437–441

10. Graziani L, Silvestro A, Bertone V et al (2007) Vascular involvement in diabetic subjects with ischemic foot ulcer: a new morphologic categorization of disease severity. Eur J Vasc Endovasc Surg 33:453–460

11. Moher D, Liberati A, Tetzlaff J, Altman DG (2009) Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 339:332–336

12. Collins R, Cranny G, Burch J et al (2007) A systematic review of duplex ultrasound, magnetic resonance angiography and computed tomography angiography for the diagnosis and assessment of symptomatic, lower limb peripheral arterial disease. Health Technol Assess 11:iii–xiii

13. Whiting P, Rutjes AW, Reitsma JB, Bossuyt PM, Kleijnen J (2003) The development of QUADAS: a tool for the quality assessment of studies of diagnostic accuracy included in systematic reviews. BMC Med Res Methodol 3:25

14. Higgins JP, Thompson SG, Deeks JJ, Altman DG (2003) Measuring inconsistency in meta-analyses. BMJ 327:557–560

15. Altman DG (1991) Practical statistics for medical research, 1st edn. Chapman and Hall, London

16. Bipat S, Zwinderman AH, Bossuyt PM, Stoker J (2007) Multivariate random-effects approach: for meta-analysis of cancer staging studies. Acad Radiol 14:974–984

17. Akaike H (1974) A new look at the statistical model identification. IEEE Trans Automat Control 19:716–723

18. Egger M, Davey SG, Schneider M, Minder C (1997) Bias in meta-analysis detected by a simple, graphical test. BMJ 315:629–634

19. Glas AS, Lijmer JG, Prins MH, Bonsel GJ, Bossuyt PM (2003) The diagnostic odds ratio: a single indicator of test performance. J Clin Epidemiol 56:1129–1135

35


2

20. Fotiadis N, Kyriakides C, Bent C, Vorvolakos T, Matson M (2011) 64-section CT angiography in patients with critical limb ischaemia and severe claudication: comparison with digital subtractive angiography. Clin Radiol 66:945–952

21. Kau T, Eicher W, Reiterer C et al (2011) Dual-energy CT angiography in peripheral arterial occlusive disease-accuracy of maximum intensity projections in clinical routine and subgroup analysis. Eur Radiol 21:1677–1686

22. Li G-C, Deng G, Qin Y-L et al (2008) The comparative study of 64-slices spiral CT angiography with DSA hi lower extremity arterial occlusive diseases. J Interv Radiol 17:336–339

23. Cia S, Jiang H, Cui X-Y, Wang D-W, Sun Y-T (2007) Application of multi-slice spiral CT angiography in diagnosis of arteriosclerotic occlusive disease of lower extremity. Chin J Med Imaging Technol 23:1022–1025

24. Li X-M, Xiao Y, Tian J-M, Guang J-Z, Tian J-L, Gong J (2007) The diagnostic value of 64-multislice CT in patients with peripheral arterial occlusive diseases: comparison with digital subtraction angiography. J Interv Radiol 16:371–374

25. Bui TD, Gelfand D, Whipple S et al (2005) Comparison of CT and catheter arteriography for evaluation of peripheral arterial disease. Vasc Endovascular Surg 39:481–490

26. Willmann JK, Baumert B, Schertler T et al (2005) Aortoiliac and lower extremity arteries assessed with 16-detector row CT angiography: prospective comparison with digital subtraction angiography. Radiology 236:1083–1093

27. Catalano C, Fraioli F, Laghi A et al (2004) Infrarenal aortic and lower-extremity arterial disease: diagnostic performance of multi-detector row CT angiography. Radiology 231:555–563

28. Martin ML, Tay KH, Flak B et al (2003) Multidetector CT angiography of the aortoiliac system and lower extremities: a prospective comparison with digital subtraction angiography. Am J Roentgenol 180:1085–1091

29. Ofer A, Nitecki SS, Linn S et al (2003) Multidetector CT angiography of peripheral vascular disease: a prospective comparison with intraarterial digital subtraction angiography. Am J Roentgenol 180:719–724

30. Puls R, Knollmann F, Werk M et al (2001) Multi-slice spiral CT: 3D CT angiography for evaluating therapeutically relevant stenosis in peripheral arterial occlusive disease. Rontgenpraxis 54:141–147

31. Rieker O, Duber C, Neufang A, Pitton M, Schweden F, Thelen M (1997) CT angiography versus intraarterial digital subtraction angiography for assessment of aortoiliac occlusive disease. Am J Roentgenol 169:1133–1138

32. Anzidei M, Napoli A, Zaccagna F et al (2011) First-pass and high-resolution steady-state magnetic resonance angiography of the peripheral arteries with gadobenate dimeglumine: an assessment of feasibility and diagnostic performance. Invest Radiol 46:307–316

33. Bui BT, Miller S, Mildenberger P, Sam A, Sheng R (2010) Comparison of contrast-enhanced MR angiography to intraarterial digital subtraction angiography for evaluation of peripheral arterial occlusive disease: results of a phase III multicenter trial. J Magn Reson Imaging 31:1402–1410

34. Gerretsen SC, le Maire TF, Miller S et al (2010) Multicenter, double-blind, randomized, intraindividual crossover comparison of gadobenate dimeglumine and gadopentetate dimeglumine for MR angiography of peripheral arteries. Radiology 255:988–1000

35. Wang CC, Liang HL, Hsiao CC et al (2010) Single-dose time-resolved contrast enhanced hybrid MR angiography in diagnosis of peripheral arterial disease: compared with digital subtraction angiography. J Magn Reson Imaging 32:935–942

36. Poschenrieder F, Hamer OW, Herold T et al (2009) Diagnostic accuracy of intraarterial and i.v. MR angiography for the detection of stenoses of the infrainguinal arteries. Am J Roentgenol 192:117–121

36

Chapter 2

37. Owen AR, Robertson IR, Annamalai G et al (2009) Critical lower-limb ischemia: the diagnostic performance of dual-phase injection MR angiography (including high-resolution distal imaging) compared with digital subtraction angiography. J Vasc Interv Radiol 20:165–172

38. Berg F, Bangard C, Bovenschulte H et al (2008) Feasibility of peripheral contrast-enhanced magnetic resonance angiography at 3.0 Tesla with a hybrid technique: comparison with digital subtraction angiography. Invest Radiol 43:642–649

39. Andreisek G, Pfammatter T, Goepfert K et al (2007) Peripheral arteries in diabetic patients: standard bolus-chase and time-resolved MR angiography. Radiology 242:610–620

40. Diehm N, Kickuth R, Baumgartner I et al (2007) Magnetic resonance angiography in infrapopliteal arterial disease: prospective comparison of 1.5 and 3 Tesla magnetic resonance imaging. Invest Radiol 42:467–476

41. Deutschmann HA, Schoellnast H, Portugaller HR et al (2006) Routine use of three-dimensional contrast-enhanced moving-table MR angiography in patients with peripheral arterial occlusive disease: comparison with selective digital subtraction angiography. Cardiovasc Intervent Radiol 29:762–770

42. Gjonnaess E, Morken B, Sandbaek G et al (2006) Gadolinium-enhanced magnetic resonance angiography, colour duplex and digital subtraction angiography of the lower limb arteries from the aorta to the tibio-peroneal trunk in patients with intermittent claudication. Eur J Vasc Endovasc Surg 31:53–58

43. Janka R, Fellner C, Wenkel E, Lang W, Bautz W, Fellner FA (2005) Contrast-enhanced MR angiography of peripheral arteries including pedal vessels at 1.0 T: feasibility study with dedicated peripheral angiography coil. Radiology 235:319–326

44. Lapeyre M, Kobeiter H, Desgranges P, Rahmouni A, Becquemin JP, Luciani A (2005) Assessment of critical limb ischemia in patients with diabetes: comparison of MR angiography and digital subtraction angiography. Am J Roentgenol 185:1641–1650

45. Leiner T, Kessels AG, Nelemans PJ et al (2005) Peripheral arterial disease: comparison of color duplex US and contrast-enhanced MR angiography for diagnosis. Radiology 235:699–708

46. Schmitt R, Coblenz G, Cherevatyy O et al (2005) Comprehensive MR angiography of the lower limbs: a hybrid dual-bolus approach including the pedal arteries. Eur Radiol 15:2513–2524

47. de Vries M, Nijenhuis RJ, Hoogeveen RM, de Haan MW, van Engelshoven JM, Leiner T (2005) Contrast-enhanced peripheral MR angiography using SENSE in multiple stations: feasibility study. J Magn Reson Imaging 21:37–45

48. Bezooijen R, van den Bosch HC, Tielbeek AV et al (2004) Peripheral arterial disease: sensitivity-encoded multiposition MR angiography compared with intraarterial angiography and conventional multiposition MR angiography. Radiology 231:263–271

49. Cronberg CN, Sjoberg S, Albrechtsson U et al (2003) Peripheral arterial disease. Contrast-enhanced 3D MR angiography of the lower leg and foot compared with conventional angiography. Acta Radiol 44:59–66

50. Huber A, Scheidler J, Wintersperger B et al (2003) Moving-table MR angiography of the peripheral runoff vessels: comparison of body coil and dedicated phased array coil systems. Am J Roentgenol 180:1365–1373

51. Steffens JC, Schafer FK, Oberscheid B et al (2003) Bolus-chasing contrast-enhanced 3D MRA of the lower extremity. Comparison with intraarterial DSA. Acta Radiol 44:185–192

52. Wyttenbach R, Gianella S, Alerci M, Braghetti A, Cozzi L, Gallino A (2003) Prospective blinded evaluation of Gd-DOTA-versus Gd-BOPTA-enhanced peripheral MR angiography, as compared with digital subtraction angiography. Radiology 227:261–269

37


2

53. Lenhart M, Herold T, Volk M et al (2000) Contrast media-enhanced MR angiography of the lower extremity arteries using a dedicated peripheral vascular coil system. First clinical results Rofo 172:992–999

54. Lundin P, Svensson A, Henriksen E et al (2000) Imaging of aortoiliac arterial disease. Duplex ultrasound and MR angiography versus digital subtraction angiography. Acta Radiol 41:125–132

55. Sueyoshi E, Sakamoto I, Matsuoka Y, Hayashi H, Hayashi K (2000) Symptomatic peripheral vascular tree stenosis. Comparison of subtracted and nonsubtracted 3D contrast-enhanced MR angiography with fat suppression. Acta Radiol 41:133–138

56. Lenhart M, Djavidani B, Volk M et al (1999) Contrast medium-enhanced MR angiography of the pelvic and leg vessels with an automated table-feed technique. Rofo 171:442–449

57. Sueyoshi E, Sakamoto I, Matsuoka Y et al (1999) Aortoiliac and lower extremity arteries: comparison of three-dimensional dynamic contrast-enhanced subtraction MR angiography and conventional angiography. Radiology 210:683–688

58. Winterer JT, Laubenberger J, Scheffler K et al (1999) Contrast-enhanced subtraction MR angiography in occlusive disease of the pelvic and lower limb arteries: results of a prospective intraindividual comparative study with digital subtraction angiography in 76 patients. J Comput Assist Tomogr 23:583–589

59. Laissy JP, Debray MP, Menegazzo D et al (1998) Prospective evaluation of peripheral arterial occlusive disease by 2D MR subtraction angiography. J Magn Reson Imaging 8:1060–1065

60. Rofsky NM, Johnson G, Adelman MA, Rosen RJ, Krinsky GA, Weinreb JC (1997) Peripheral vascular disease evaluated with reduced-dose gadolinium-enhanced MR angiography. Radiology 205:163–169

61. Li J, Zhao J-G, Zhu Y-Q, Li M-H, Wang J, Qiao R-H (2011) Evaluation of diabetic peripheral arterial disease in lower limb by using 3.0 T contrast-enhanced MR angiography with simultaneous calf compression. J Interv Radiol 20:231–236

62. Wu Q-Y, Lin J, Li D, Zeng M-S (2011) Imaging evaluation of calf arteries in patients with peripheral arterial occlusive disease by using time-resolved angiography with interleaved stochastic trajectories on MR scanner. Chin J Radiol 45:560–565

63. Zhang L, Chang J, Shi D-C et al (2010) An in vitro experimental study and clinical applications of MR angiography with low-dose contrast agent of lower limb arteries at 3.0 T. Chin J Radiol 44:1078–1083

64. Li D-W, Guo W-L, Lu Z-M, Qi W-X, Guo Q-Y (2009) Application of energy subtract angiography of dual source CT in diagnosis of arterial diseases of the lower extremities. Chin J Med Imaging Technol 25:1806–1809

65. Yang M, Teng G-J, Liu B, Wu M, Jin J-Y, Deng G (2008) Imaging technique and postprocessing of lower extremity arteries using 64-slice CT. J Interv Radiol 17:353–356

66. Yuan F, Liu Y-S, Dong S-Y, Zhao J, Feng K-L (2008) Diagnostic value of 64-slice computed tomography angiography for peripheral arterial occlusive diseases. Chin J Med Imaging Technol 24:1767–1770

67. Wang G-Y, Zhao B, Wang G-B, Zhang Z-F, Qiu X-L (2007) Three-dimensional contrast-enhanced MR angiography in the classification of peripheral arterial occlusive disease. Chin J Radiol 41:598–601

68. Yuan F, Liu Y-S, Long M-M, Yuan B, Gu X, Gao G-F (2006) Multiposition dynamic contrast enhanced MR angiography in peripheral vessel diseases at 3.0T MR. Chin J Med Imaging Technol 22:726–729

69. Volodiukhin MI, Ibatullin MM, Mikhailov IM, Malinovskii MN, Ignat’ev IM, Bredikhin RA (2005) Combined bolus magnetic resonance angiography and two-dimensional time-of-flight magnetic

38

Chapter 2

resonance angiography in patients with occlusive diseases of lower limb arteries. Angiol Sosud Khir 11:29–36

70. Saito Y, Noda H, Itabashi Y et al (2004) Table-moving MRA of the lower extremities in patients with arterial occlusive disease. Japan J Clin Radiol 49:547–554

71. Zhang L, Jin Z, Xu Z-R (2004) Diagnostic value of magnetic resonance angiography for diabetic foot and arterial disease of lower leg. Chin J Clin Rehabil 8:3626–3627

72. Tomihira A, Hino Y, Sugihara M (2002) Stepping table gadolinium-enhanced three-dimensional MR angiography in arterial occlusive disease of the pelvic and lower extremity arteries. Japan J Clin Radiol 47:525–533

73. Matsumura K, Sato K, Hashida K, Utsumi N, Ishizawa T (2001) Contrast-enhanced subtraction MR angiography of the systemic aorta and lower extremity arteries for vascular intervention: usefulness of stepping-table method on a 0.5 T system. Japan J Interv Cardiol 16:130–136

39

Outcomes of Infrainguinal Revascularizations with Endovascular First Strategy in CLI

3CHAPTER 3

Outcomes of Infrainguinal Revascularizations with Endovascular First Strategy in Critical Limb Ischemia

Sjoerd Jens, Anne P. Conijn, Franceline A. Frans, Marieke B.B. Nieuwenhuis, Rosemarie Met, Mark J.W. Koelemay, Dink A. Legemate, Shandra Bipat,

Jim A. Reekers

Cardiovasc Intervent Radiol. 2014 Aug 12. [Epub ahead of print]

Online Supplementary Figures and Tables can be found at URL: http://www.boxpress.nl/proefschriften/ebooks/sjoerd_jens/

40

Chapter 3

Abstract

PURPOSE: This study was designed to study the outcome of infrainguinal revascularization in patients with critical limb ischemia (CLI) in an institution with a preference towards endovascular intervention first in patients with poor condition, unfavourable anatomy for surgery, no venous material for bypass, and old age.METHODS: A prospective, observational cohort study was conducted between May 2007 and May 2010 in patients presenting with CLI. At baseline, the optimal treatment was selected, i.e., endovascular or surgical treatment. In case of uncertainty about the preferred treatment, a multidisciplinary team (MDT) was consulted. Primary endpoints were quality of life and functional status 6 and 12 months after initial intervention, assessed by the VascuQol and AMC Linear Disability Score questionnaires, respectively.RESULTS: In total, 113 patients were included; 86 had an endovascular intervention and 27 had surgery. During follow-up, 41 % underwent an additional ipsilateral revascularisation procedure. For the total population, and endovascular and surgery subgroups, the VascuQol sum scores improved after 6 and 12 months (p < 0.01 for all outcomes) compared with baseline. The functional status improved (p = 0.043) after 12 months compared with baseline for the total population. Functional status of the surgery subgroup improved significantly after 6 (p = 0.031) and 12 (p = 0.044) months, but not that of the endovascular subgroup.CONCLUSIONS: Overall, the strategy of performing endovascular treatment first in patients with poor condition, unfavourable anatomy for surgery, no venous material for bypass, and old age has comparable or even slightly better results compared with the BASIL trial and other cohort studies. All vascular groups should discuss whether their treatment strategy should be directed at treating CLI patients preferably endovascular first and consider implementing an MDT to optimize patient outcomes.

41

Outcomes of Infrainguinal Revascularizations with Endovascular First Strategy in CLI

3

Introduction

Patients with critical limb ischemia (CLI) suffer from ischemic rest pain or nonhealing ulcers,1 severely affecting their quality of life.2 Treatment in these patients is initially directed at preventing amputation by achieving ulcer healing and at relieving ischemic rest pain. When treatment is successful, quality of life is expected to improve significantly.3

The choice of treatment, i.e., endovascular or surgical revascularization, primary amputation, or conservative treatment, differs per patient. The BASIL trial randomized patients with CLI to primary bypass surgery or endovascular balloon angioplasty and showed less morbidity and costs for balloon angioplasty within the first 2 years. However, beyond 2 ye

uva-dare (digital academic repository) imaging of critical ... · critical limb ischemia peripheral...

Documents