The UK endovascular aneurysm repair (EVAR) trials: Randomised trials of EVAR versus standard therapy

L. C. Brown, J. T. Powell, S. G. Thompson, D. M. Epstein, M. J. Sculpher, R. M. Greenhalgh

Research output: Contribution to journalArticlepeer-review


Objective: To assess the efficacy of endovascular aneurysm repair (EVAR) against standard alternative management in patients with large abdominal aortic aneurysm (AAA). Design: Two national, multicentre randomised trials - EVAR trials 1 and 2. Setting: Patients were recruited from 38 out of 41 eligible UK hospitals. Participants: Men and women aged at least 60 years, with an AAA measuring at least 5.5 cm on a computerised tomography scan that was regarded as anatomically suitable for EVAR, were assessed for fitness for open repair. Patients considered fit were randomised to EVAR or open repair in EVAR trial 1 and patients considered unfit were randomised to EVAR or no intervention in EVAR trial 2. Interventions: EVAR, open repair or no intervention. Main outcome measures: The primary outcome was mortality (operative, all-cause and AAA related). Patients were flagged at the UK Office for National Statistics with centrally coded death certificates assessed by an Endpoints Committee. Power calculations based upon mortality indicated that 900 and 280 patients were required for EVAR trials 1 and 2, respectively. Secondary outcomes were graft-related complications and reinterventions, adverse events, renal function, health-related quality of life and costs. Cost-effectiveness analyses were performed for both trials. Results: Recruitment occurred between 1 September 1999 and 31 August 2004, with targets exceeded in both trials: 1252 randomised into EVAR trial 1 (626 to EVAR) and 404 randomised into EVAR trial 2 (197 to EVAR). Follow-up closed in December 2009 with very little loss to follow-up (1%). In EVAR trial 1, 30-day operative mortalities were 1.8% and 4.3% in the EVAR and open-repair groups, respectively: adjusted odds ratio 0.39 [95% confidence interval (CI) 0.18 to 0.87], p = 0.02. During a total of 6904 person-years of follow-up, 524 deaths occurred (76 AAA related). Overall, there was no significant difference between the groups in terms of all-cause mortality: adjusted hazard ratio (HR) 1.03 (95% CI 0.86 to 1.23), p = 0.72. The EVAR group did demonstrate an early advantage in terms of AAA-related mortality, which was sustained for the first few years, but lost by the end of the study, primarily due to fatal endograft ruptures: adjusted HR 0.92 (95% CI 0.57 to 1.49), p = 0.73. The EVAR procedure was more expensive than open repair (mean difference £1177) and not found to be cost-effective, but the model was sensitive to alternative assumptions. In EVAR trial 2, during a total of 1413 person-years of follow-up, a total of 305 deaths occurred (78 AAA related). The 30-day operative mortality was 7.3% in the EVAR group. However, this group later demonstrated a significant advantage in terms of AAA-related mortality, but this became apparent only after 4 years: overall adjusted HR 0.53 (95% CI 0.32 to 0.89), p = 0.02. Sadly, this advantage did not result in any benefit in terms of all-cause mortality: adjusted HR 0.99 (95% CI 0.78 to 1.27), p = 0.97. Overall, EVAR was more expensive than no intervention (mean difference £10,222) and not found to be cost-effective. Conclusions: EVAR offers a clear operative mortality benefit over open repair in patients fit for both procedures, but this early benefit is not translated into a long-term survival advantage. Among patients unfit for open repair, EVAR is associated with a significant longterm reduction in AAA-related mortality but this does not appear to influence all-cause mortality. Trial registration: Current Controlled Trials ISRCTN 55703451. Funding: This project was funded by the NIHR Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 16, No. 9. See the HTA programme website for further project information.

Original languageEnglish
JournalHealth technology assessment
Issue number9
Publication statusPublished - 1 Feb 2012

Cite this