Survival Following Ruptured Abdominal Aortic Aneurysm ... - Core

Survival Following Ruptured Abdominal Aortic Aneurysm Before and During the IMPROVE Trial: A Single-centre Series G.K. Ambler *, C.P. Twine, J. Shak, K.E. Rollins, K. Varty, P.A. Coughlin, P.D. Hayes, J.R. Boyle Department of Vascular Surgery, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

WHAT THIS PAPER ADDS This study addresses the changes in results following repair of ruptured abdominal aortic aneurysms at a single centre following introduction of the IMPROVE trial protocol. While survival following open surgical repair remained unchanged, survival following endovascular repair deteriorated significantly when compared with previous results. This suggests that whereas there may be a survival benefit for endovascular repair in selected patients, this may not be the case in an unselected cohort. Objectives: The first large-scale randomised trial (Immediate Management of the Patient with Rupture: Open Versus Endovascular repair [IMPROVE]) for endovascular repair of ruptured abdominal aortic aneurysm (rEVAR) has recently finished recruiting patients. The aim of this study was to examine the impact on survival after rEVAR when the IMPROVE protocol was initiated in a high volume abdominal aortic aneurysm (AAA) centre previously performing rEVAR. Methods: One hundred and sixty-nine patients requiring emergency infrarenal AAA repair from January 2006 to April 2013 were included. Eighty-four patients were treated before (38 rEVAR, 46 open) and 85 (31 rEVAR, 54 open) were treated during the trial period. A retrospective analysis was performed. Results: Before the trial, there was a significant survival benefit for rEVAR over open repair (90-day mortality 13% vs. 30%, p ¼ .04, difference remained significant up to 2 years postoperatively). This survival benefit was lost after starting randomisation (90-day mortality 35% vs. 33%, p ¼ .93). There was an increase in overall 30-day mortality from 15% to 31% (p ¼ .02), while there was no change for open repair (p ¼ .438). There was a significant decrease in general anaesthetic use (p ¼ .002) for patients treated during the trial. Randomised patients had shorter hospital and intensive treatment unit stays (p ¼ .006 and p ¼ .03 respectively). Conclusions: The change in survival seen during the IMROVE trial highlights the need for randomised rather than cohort data to eliminate selection bias. These results from a single centre reinforce those recently reported in IMPROVE. Ó 2014 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. Article history: Received 5 September 2013, Accepted 15 January 2014, Available online 16 February 2014 Keywords: Abdominal aortic aneurysm, Survival analysis, Selection bias INTRODUCTION There has been intense debate over the role of randomised trials for endovascular repair of ruptured abdominal aortic aneurysms (rAAAs). Since the technique was first described in 1994, initial uptake in specialist centres led to the publication of combined world data showing favourable early results.1 However, meta-analysed cohort data from more comprehensive literature have produced mixed results.2e4 Some enthusiastically found short-term survival benefits compared with open repair,2 while others were more pragmatic, citing heavy selection and publication bias in

* Corresponding author. G.K. Ambler, Department of Vascular Surgery, Cambridge University Hospitals NHS Foundation Trust, Hills Road, Cambridge CB2 0QQ, UK. E-mail address: [email protected] (G.K. Ambler). 1078-5884/$ e see front matter Ó 2014 European Society for Vascular Surgery. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejvs.2014.01.011

meta-analysed studies.3,4 The implication was that true “gold standard” evidence was needed before definitive recommendations on widespread adoption could be made. The Immediate Management of the Patient with Rupture: Open Versus Endovascular repair (IMPROVE) randomised trial is an attempt to provide this evidence, and has recently completed recruitment in the UK.5,6 While the 30-day results from IMPROVE have been recently reported,7 outcomes from centres which were already performing endovascular repair of rAAA (rEVAR) successfully before switching to randomisation under the trial protocol are unknown.8 Detractors of randomised studies for rAAA have suggested that the trial is ethically questionable, as rEVAR results were likely to stay the same while patients randomised to open surgical repair (OSR) would have worse outcomes, resulting in net harm to the enrolled patient population.9 Proponents suggest that outcomes are impossible to predict, citing recognised problems with the current retrospective literature and the fact that

European Journal of Vascular and Endovascular Surgery

Volume 47 Issue 4 p. 388e393 April/2014

previous small randomised trials for rEVAR had equivocal results.10,11 The aim of this study was therefore to examine rAAA outcomes in a high volume AAA centre previously performing rEVAR before and after the implementation of the IMPROVE trial protocol to assess impact on patient survival. METHODS All patients were treated at Addenbrooke’s Hospital as part of the local vascular network. The study period included January 2006 to April 2013, with follow-up until July 2013. This marks the time period since out of hours EVAR was available at Addenbrooke’s. Recruitment for IMPROVE started in September 2009.5,6 All patients were assessed and treated by specialist vascular surgeons. If EVAR was performed, specialist interventional radiologists were always involved. Permissive hypotension at 100 mmHg systolic pressure was allowed for all patients. A senior anaesthetist was present for all procedures, including those under local anaesthetic. EVAR was performed with a C arm in theatre by a team consisting of at least one consultant vascular surgeon and one consultant vascular interventional radiologist. Postoperatively patients went to high dependency or intensive care depending on clinical need. Initial experiences of rEVAR at Addenbrooke’s, including results from the early part of the study period, have been published previously.8,12 A retrospective analysis of a prospectively maintained database was performed. Since randomising into IMPROVE, these data have been held by the local trial co-ordinator. The database was cross-referenced with the electronic medical record (eMR) browser for error checking and mortality statistics. eMR is linked to the UK Office for National Statistics for mortality reporting. Any data not available on eMR were obtained directly from patient notes. No patients were lost to follow up. Analysis was principally performed as treated rather than on an intention to treat IMPROVE protocol. The Manne

389

Whitney U test was used to assess differences between lengths of stay. The Fisher exact test was used to test for significance between individual group characteristics. KaplaneMeier estimates were used to examine results for 5 years of follow up, but 2-year data are reported due to the high attrition rate after this time.13 The log rank test was used to compare survival estimates. Confounder adjustment was performed using Cox proportional hazards modelling to account for differences in age; gender; history of diabetes, ischaemic heart disease, or cerebrovascular disease; and preoperative medications. Data analysis was carried out with the R statistical package14 version 3.0.1 with the “survival” add-on package15 version 2.37-4 loaded for survival analysis. The IMPROVE trial is registered as ISRCTN 48334791, and local ethical approval has been granted for recruitment within this trial. The present study was carried out as part of routine local service evaluation within the trial and so was not deemed to require separate ethical approval. RESULTS One hundred and seventy-nine patients presented with ruptured AAA from January 2006 to April 2013 (Fig. 1). Basic demographics for patients treated before and during IMPROVE are presented in Table 1. Sixty-nine (39%) of these underwent rEVAR and one hundred (56%) underwent OSR. Ten patients (6%) were not offered or were refused treatment and were palliated so were not analysed. Thirtyseven patients were randomised into the IMPROVE trial, of which thirty-five underwent surgery. Fifty patients (28%) were treated during the IMPROVE time period but were not randomised. The most common reason for this was that they had already undergone computed tomography (CT) scanning and suitability for EVAR was known at the time of referral. Two (3%) EVAR and four (4%) OSR died in theatre. One hundred and sixty-nine patients (94% of those presenting with ruptured AAA) were therefore available for analysis.

Figure 1. Flow chart showing division of patients into groups. Note. rAAA ¼ ruptured abdominal aortic aneurysm; EVAR ¼ endovascular aneurysm repair; OSR ¼ open surgical repair.

390

G.K. Ambler et al.

Improve Before During

10 17

.04

21:10

.62

14:17

.002

88 46

.57

86 27

.65

0.6 0.4

.06 1.0 .08

Survival

0.8

p

0.2

Age (median IQR) Gender (M:F) ITU days (median IQR) Total admission days 14 23.5 (median IQR) AUI:bifurcated graft 23:15 (EVAR only) General:local 31:7 anaesthetic (EVAR only) Preoperative markers Lowest systolic BP 90 46.5 (median IQR) Highest heart rate 80 35 (median IQR) 42% Ischaemic heart disease Cerebrovascular 9% disease Diabetes 7% Creatinine 129 70.2 (median IQR) Preoperative medication Antiplatelet/ 57 anticoagulation (%) Beta blocker (%) 23 Statin (%) 52 Note. IQR ¼ interquartile range; ITU ¼ AUI ¼ aorto-uni iliac.

During IMPROVE (n ¼ 85) 75.4 11.6 72:13 25

0.0

Before IMPROVE (n ¼ 84) 78.4 7.3 71:13 3 15

1.0

Table 1. Demographics for patients treated before and during IMPROVE.

0

5

10

15

20

25

30

Time (Months)

Figure 2. Comparative survival of ruptured abdominal aortic aneurysm repair before and during the IMPROVE trial. The solid line shows survival pre-IMPROVE, the dashed line shows survival during the trial period.

40%

.88

7%

.59

6% 133 63

.77 .30

1 hour 13 minutes, IQR 39 minutes to 1 hour 36 minutes for OSR; p ¼ .87).

48

.28

Survival before and during IMPROVE

23 1.0 46 .54 intensive treatment unit;

Eighty-four patients (50%) were treated for rAAA preIMPROVE and 85 (50%) were treated during the trial. A total of 69 patients underwent rEVAR: 38 (55%) preIMPROVE and 31 (45%) during the trial. A total of 100 patients underwent OSR: 46 pre-IMPROVE and 54 during the trial. During the trial, 16 patients were randomised to CT scan followed by EVAR if appropriate, of whom 10 (63%) went on to have EVAR and six had OSR. Nineteen patients were randomised to OSR, of whom 16 (84%) had OSR and three were treated with EVAR. The difference in crossover rate was not significant (p ¼ .25). Patients who were randomised within the trial had significantly reduced total admission days (p ¼ .002) and intensive treatment unit (ITU) stay (p ¼ .01) when compared with patients who were not randomised. The delay from presentation to EVAR was a median of 3 hours 31 minutes (interquartile range [IQR] 1 hour 24 minutes to 7 hours 55 minutes) before IMPROVE, which was significantly longer than for patients undergoing OSR (median 1 hour 15 minutes, IQR 45 minutes to 2 hours 18 minutes; p < .0001). This improved during the trial, with no significant difference in time to theatre being observed between groups during the trial period (median 1 hour 1 minute, IQR 43 minutes to 2 hours 26 minutes for EVAR vs.

Thirty-day mortality for the 84 patients treated preIMPROVE was 15%. Thirty-day mortality for the 85 patients treated during the trial was 31%. There was a significant survival advantage for patients treated preIMPROVE at 30 days (log rank chi square 5.5, p ¼ .019), but the difference became non-significant at later time points. This survival advantage was present exclusively in the rEVAR group (log rank chi square 8.4, p ¼ .004), and persisted in this group beyond 2 years (log rank chi square 4.9, p ¼ .03). This survival advantage persisted throughout follow-up after confounder adjustment (hazard ratio of death following OSR over rEVAR 2.8, p ¼ .006). The survival benefit of the rEVAR group before the trial was also reflected in improved survival comparison with OSR during this period (log rank chi square 4.2, p ¼ .04 at 2 years). This benefit of rEVAR over OSR has been lost since the trial started (log rank chi square 0.1, p ¼ .79) after adjusting for confounders. The analysis was repeated for patients randomised within the trial, but no survival difference between procedures was seen in this subgroup on an as-treated basis, and there was no survival difference between patients who were randomised and those who were not. If the survival advantage of rEVAR observed during the pre-IMPROVE period had been maintained during the trial, the study would have had 98% power to detect a survival benefit at the 5% level, so the study was adequately powered. Fig. 2 shows survival curves pre-IMPROVE; Fig. 3 shows separate survival curves for the different procedures.



391

undergoing OSR before and during IMPROVE. There was no significant difference in mortality before IMPROVE compared with the group treated during the trial (log rank chi square 0.6, p ¼ .438): pre-IMPROVE cumulative survival at 3 and 12 months was 70% and 63%, during it was 67% and 63% respectively. Patients undergoing OSR within the trial also had a significant reduction in ITU stay (p ¼ .047) and a reduction in total admission days, though this was not significant (p ¼ .055). EVAR follow-up and reintervention

Figure 3. Comparative survival before and during IMPROVE by procedure. The solid lines show survival prior to commencement of the IMPROVE trial, the dashed lines show survival during the trial. Black lines show ruptured EVAR survival, grey lines show ruptured OSR survival. Note. EVAR ¼ endovascular aneurysm repair; OSR ¼ open surgical repair.

There was no significant difference between reinterventions before and during IMPROVE groups (p ¼ .59). There were two inpatient reinterventions during the primary admission in each EVAR group: one type 2 endoleak requiring intervention due to sac expansion in both groups; and one limb occlusion requiring femoral crossover grafting in both groups. There were 11 (29%) long-term reinterventions in the group pre-IMPROVE: four type 1 endoleaks, two type 2 endoleaks, two limb occlusions requiring crossover grafting, two limb kinks requiring stenting, and one limb dislocation. There have been six reinterventions for the IMPROVE group (19%): two type 1 endoleaks, two infected crossover grafts requiring removal and axillo-femoral bypass, one occluded crossover graft requiring thrombectomy, and one limb kink requiring stenting.

EVAR for rAAA

DISCUSSION

Patients treated with EVAR during the IMPROVE trial were in general younger (median age 74 vs. 79 years, p ¼ .03), and were more likely to be treated under local anaesthetic (p ¼ .002). Although there was no significant difference in length of stay before and during the trial, there was a significant reduction in ITU stay (p ¼ .03) and total admission days (p ¼ .006) for the patients treated with rEVAR within the trial, when compared with those who were not treated within the trial. There was a trend towards shorter neck length and increased neck angulation in the group treated during the IMPROVE trial, though these trends did not reach significance (mean SD neck length in IMPROVE group 2.9 1.4 cm vs. 3.6 2.8 cm, p ¼ .20; mean SD neck angulation in IMPROVE group 47 22 vs. 40 21, p ¼ .18). Overall 30-day mortality for all EVAR patients was 13%. Three-, 12-, and 24-month survival was 77%, 74%, and 66% respectively. Before IMPROVE these figures were 87%, 84%, and 76%, which fell to 65%, 61%, and 53% respectively during the trial. Fig. 3 shows comparative survival for patients undergoing EVAR or OSR before and during IMPROVE.

The major finding was that after starting randomisation into the IMPROVE protocol, the survival benefit for rEVAR over OSR repair was lost. Survival for rEVAR showed a significant decrease while survival for OSR was unchanged compared with pre-IMPROVE. There was a significant decrease in ITU stay (p ¼ .01), total admission days (p ¼ .002), and general anaesthetic use (p ¼ .002) for patients in the IMPROVE trial. The strengths of these data are that the database was maintained prospectively from electronic medical records used to code Hospital Episode Statistics data which appears to be accurate for rAAA.16 The data since IMPROVE have been maintained by the trial investigators, again reducing concern over error in retrospective data. Having said this, the pre-IMPROVE data were collected retrospectively for some variables and is therefore subject to the usual limitations. Not all patients presenting since 2009 have been randomised, and this may produce bias in the dataset. However, this was almost entirely due to either individual consultant selection, with some randomising all patients, or prior knowledge of suitability for EVAR, making them ineligible, reducing this concern to some extent. Therefore, implementing the IMPROVE protocol reduced survival in rEVAR patients without changing it for patients undergoing OSR. There was no reason for this found when patient demographics were compared (Table 1). The most probable explanation is that this effect was due to bias from

Open survival Overall 30-day mortality in the OSR group was 30%. Three-, 12- and 24-month survival was 68%, 63%, and 58% respectively. Fig. 3 shows comparative survival for patients

392

selecting patients with more suitable anatomy for EVAR pre-IMPROVE. This may have caused operators to push the limits of EVAR devices by aggressive utilisation outside the instructions for use (IFU) during the trial.17,18 This only adds to arguments for the value of the IMPROVE trial, as previously published survival results for rEVAR may be better than true outcomes when patients are unselected.1 The larger worldwide cohort studies for rEVAR show a 30day mortality ranging from 21% to 32%.1,2,19 Thirty-day mortality pre-IMPROVE in this centre was 14%, which has now increased to 22%. This may mean the higher mortality centres were already treating more unwell or unstable patients. Importantly, however, results of all patients presenting here have worsened since randomisation. There is an argument that this would always be the outcome of a randomised trial for rAAA, as patients with borderline suitability for rEVAR are treated in a way in which they were not before the trial. Pushing the limits of the technology and the EVAR teams’ knowledge of rEVAR has worsened EVAR outcomes by selecting patients who will not be treated in this way in the future, irrespective of the results of the trial. In addition, it is possible that outcomes for OSR may have been subtly improved by the inclusion of patients with anatomy suitable for EVAR, as it has been shown that these patients also do better when treated with OSR.20 Indeed, both neck length and angulation were more adverse in the EVAR group treated during the trial, although these results did not reach significance; so it is possible that anatomic unsuitability has resulted in deteriorating outcomes. One possible source of bias is that, while a specialist vascular surgeon was always involved in both assessment and treatment of all patients, those patients treated with rEVAR also benefitted from the involvement of a specialist interventional radiologist. This may explain better early results of rEVAR before IMPROVE, although it cannot explain the subsequent findings. A reduction in intensive care stay and hospital stay are two benefits of rEVAR over open repair. Although not assessed here, postoperative quality of life for rEVAR patients may be faster to return to preoperative levels than for open patients, which is important in a group with such poor long-term survival.21,22 As a consequence, even if there is no survival benefit of rEVAR over OSR, quality of life improvements may make it a superior therapy overall, at least for a subset of patients.23 The 30-day outcomes from the IMPROVE trial have been recently reported and have demonstrated no overall difference in mortality between an endovascular first strategy compared with open repair.7 The reported mortalities in both groups were very similar to those in our patients after we started randomisation, despite the majority of our cohort being treated outside the trial protocol. The Cambridge Vascular Unit was relatively high on the learning curve before starting randomisation, illustrated by the increasing utilisation of local anaesthetic and reducing time to theatre for those patients treated with rEVAR during the trial period, though small delays in getting patients

G.K. Ambler et al.

to theatre have previously been shown not to impact survival significantly.24 The centre also had better survival for open ruptures than national UK data even before randomisation, and results may be different in the full trial dataset.25 Additionally, crossover from EVAR to open after randomisation was low, which may not be the case nationwide leading to a poorer open survival in line with previous cohort data. In conclusion, the introduction of randomising patients with rAAA led to a loss in the survival benefit of rEVAR over OSR. This was explained by a decreased survival in patients treated by EVAR and unchanged survival in those undergoing OSR. The effect has been to worsen outcomes of rEVAR away from the results commonly quoted from cohort data. These findings reflect those recently reported in IMPROVE, and selection bias may explain the low mortality rates for rEVAR in the historical cohort. ACKNOWLEDGEMENTS All of the authors would like to thank M.S. Gohel, C. Cousins, A. Winterbottom, T.C. See, B. Koo, and D. Appleton, who contributed to the management of these patients. CONFLICT OF INTEREST None. FUNDING G.K.A. received funding as an Academic Clinical Fellow from the UK National Institute for Health Research. REFERENCES 1 Veith FJ, Lachat M, Mayer D, Malina M, Holst J, Mehta M, et al. Collected world and single center experience with endovascular treatment of ruptured abdominal aortic aneurysms. Ann Surg 2009;250(5):818e24. 2 McPhee J, Eslami MH, Arous EJ, Messina LM, Schanzer A. Endovascular treatment of ruptured abdominal aortic aneurysms in the United States (2001e2006): a significant survival benefit over open repair is independently associated with increased institutional volume. J Vasc Surg 2009;49(4):817e26. 3 Dillon M, Cardwell C, Blair PH, Ellis P, Kee F, Harkin DW. Endovascular treatment for ruptured abdominal aortic aneurysm. Cochrane Database Syst Rev 2007;(1):CD005261. 4 Hinchliffe RJ, Powell JT, Cheshire NJ, Thompson MM. Endovascular repair of ruptured abdominal aortic aneurysm: a strategy in need of definitive evidence. J Vasc Surg 2009;49(4): 1077e80. 5 Powell JT, Thompson SG, Thompson MM, Grieve R, Nicholson AA, Ashleigh R, et al. The immediate management of the patient with rupture: open versus endovascular repair (IMPROVE) aneurysm trial e ISRCTN 48334791 IMPROVE trialists. Acta Chir Belg 2009;109(6):678e80. 6 Powell JT. Time to IMPROVE the management of ruptured abdominal aortic aneurysm: IMPROVE trialists. Eur J Vasc Endovasc Surg 2009;38(2):237e8. 7 IMPROVE Trial Investigators. Endovascular or open repair strategy for ruptured abdominal aortic aneurysm: 30 day outcomes from IMPROVE randomised trial. BMJ 2014;348:f7661.



8 Sadat U, Hayes PD, Kullar PJ, Cousins C, Varty K, Boyle JR. An emergency EVAR service reduces mortality in ruptured abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2009;37(2):189e93. 9 Veith FJ. Part two: against the motion: it is not necessary to perform a randomised trial to compare open and endovascular repair of ruptured abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2010;40(4):424e7. 10 Powell JT, Hinchliffe RJ. Part one: for the motion: a randomised controlled trial is the best way to determine whether endovascular repair is the preferred management strategy in patients with a ruptured aortic aneurysm. Eur J Vasc Endovasc Surg 2010;40(4):421e4. 11 Reimerink JJ, Hoornweg LL, Vahl AC, Wisselink W, van den Broek RA, Legemate DA, et al. Endovascular repair versus open repair of ruptured abdominal aortic aneurysms: a multicentre randomised controlled trial. Ann Surg 2013;258(2):248e56. 12 Noorani A, Page A, Walsh SR, Varty K, Hayes PD, Boyle JR. Midterm outcomes following emergency endovascular aortic aneurysm repair for ruptured abdominal aortic aneurysms. Eur J Vasc Endovasc Surg 2012;43(4):382e5. 13 Kaplan EL, Meier P. Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958;58:457e81. 14 R Core Team. R: a language and environment for statistical computing. Vienna: Austria; 2013. 15 Therneau TM, Grambsch PM. Modeling survival data: extending the Cox model. New York: Springer; 2000. 16 Johal A, Mitchell D, Lees T, Cromwell D, van der MJ. Use of Hospital Episode Statistics to investigate abdominal aortic aneurysm surgery. Br J Surg 2012;99(1):66e72. 17 Holt PJ, Karthikesalingam A, Patterson BO, Ghatwary T, Hinchliffe RJ, Loftus IM, et al. Aortic rupture and sac expansion after endovascular repair of abdominal aortic aneurysm. Br J Surg 2012;99(12):1657e64.

393

18 Reichart M, Geelkerken RH, Huisman AB, van Det RJ, de Smit P, Volker EP. Ruptured abdominal aortic aneurysm: endovascular repair is feasible in 40% of patients. Eur J Vasc Endovasc Surg 2003;26(5):479e86. 19 Giles KA, Pomposelli FB, Hamdan AD, Wyers MC, Schermerhorn ML. Comparison of open and endovascular repair of ruptured abdominal aortic aneurysms from the ACSNSQIP 2005e07. J Endovasc Ther 2009;16(3):365e72. 20 Dick F, Diehm N, Opfermann P, von Allmen R, Tevaearai H, Schmidli J. Endovascular suitability and outcome after open surgery for ruptured abdominal aortic aneurysm. Br J Surg 2012;99(7):940e7. 21 Schermerhorn ML, Bensley RP, Giles KA, Hurks R, O’malley AJ, Cotterill P, et al. Changes in abdominal aortic aneurysm rupture and short-term mortality, 1995e2008: a retrospective observational study. Ann Surg 2012;256(4):651e8. 22 Coughlin PA, Jackson D, White AD, Bailey MA, Farrow C, Scott DJ, et al. Meta-analysis of prospective trials determining the short- and mid-term effect of elective open and endovascular repair of abdominal aortic aneurysms on quality of life. Br J Surg 2013;100(4):448e55. 23 Rollins KE, Shak J, Ambler GK, Tang TY, Hayes PD, Boyle JR. Midterm cost effectiveness of open and endovascular repair for ruptured abdominal aortic aneurysms. Br J Surg 2014;101: 225e31. 24 Boyle JR, Gibbs PJ, Kruger A, Shearman CP, Raptis S, Phillips MJ. Existing delays following the presentation of ruptured abdominal aortic aneurysm allow sufficient time to assess patients for endovascular repair. Eur J Vasc Endovasc Surg 2005;29(5):505e9. 25 Lees T, Stansby G. The National Vascular Database report 2009. The Vascular Society of Great Britain and Ireland; 2010. Retrieved from: http://www.vascularsociety.org.uk/vascular/ wp-content/uploads/2012/11/NVDREPORTFINAL-10Nov.pdf.