Comparison of Long-term Outcomes Following Endovascular Repair of Abdominal Aortic Aneurysms Based on Size Threshold

Abstract

Size threshold for operative repair of abdominal aortic aneurysms (AAA) has been determined based on risks and outcomes of open repair vs surveillance. The influence of endovascular repair (EVAR) on this threshold is less established. The purpose of this study is to determine whether long-term outcomes of EVAR are affected by maximum diameter at time of treatment. All patients who underwent EVAR with modular stent grafts between January 2000 and December 2011 at a single academic institution were identified from a prospectively maintained database. Patients were stratified based on maximum aortic diameter at time of repair: small (4.0-4.9 cm), medium (5.0-5.9 cm), and large (6.0 cm). Comparisons of demographics, indications for repair, and perioperative complications were made among the groups. Long-term follow-up was reviewed for expansion of the native aneurysm sac 5 mm, secondary intervention involving the native aneurysm or stent graft, and all-cause mortality. Statistical analyses were made using analysis of variance, χ2, and Kaplan-Meier plots. During the study period, 740 patients underwent EVAR with modular stent grafts: 157 small (21.2%), 374 medium (50.5%), and 209 large (28.2%). Preoperative patient characteristics were similar among the groups with exception to mean age (69.3 ± 8.09, 71.7 ± 8.55, and 73.6 ± 8.77 years for small, medium, and large, respectively; P < .001), history of coronary artery disease (42% small, 57% medium, 51.2% large; P = .01), prior coronary angioplasty (14.6% small, 18.2% medium, 9.6% large; P = .02), congestive heart failure (5.7% small, 15.2% medium, 19.6% large; P = .01), prior vascular surgery (7% small, 15.8% medium, 10% large; P = .016), and chronic obstructive pulmonary disease (21% small, 27% medium, 33% large; P = .038). Clinical classification differed significantly as a larger percentage of small AAAs were symptomatic (19.7% small, 7.5% medium, 8.1% large; P < .001). No difference in perioperative complication rates (P = .399) or all-type endoleak at any surveillance visit (40.8% small, 41.7% medium, 44.5% large; P = .73) was observed among the groups. However, the small AAA group developed fewer type I endoleaks (5.1% vs 6.95% medium and 14.8% large; P = .001). Aneurysm sac expansion of 5 mm was observed in 2.6% of small, 5.6% of medium, and 7.2% of large AAAs during the follow-up period but did not achieve a significant difference (P = .148). Secondary intervention rates differed significantly as 5.1% of small, 7.5% of medium, and 12.9% large AAAs required secondary intervention (P = .018). In direct comparison with small AAAs, medium (P = .39) and large (P < .001) groups both required secondary intervention more frequently, with hazard ratios of 2.32 (95% confidence interval, 1.045-5.156) and 4.74 (95% confidence interval, 2.115-10.637), respectively. Overall 10-year survival was 72% in the small, 63.1% in the medium, and 49.8% in the large group (P < .001) with no known AAA-related deaths. Age-adjusted all-cause mortality differed significantly among the group aged 75 to 84 years (30.4% small, 51.6% medium, 55.7% large; P = .017). EVAR for small AAAs demonstrates improved long-term outcomes when compared with EVAR for medium and large AAAs. These data suggest that EVAR for AAAs ranging from 4 to 5 cm may have better outcomes than EVAR for age-matched patients with larger aneurysms.