The early survival advantage of endovascular aneurysm repair (EVAR) compared with open repair reverses over time, possibly because of higher rates of reintervention related to endoleaks and aneurysm sac expansion. Therefore, we sought to examine the association between sac behavior, endoleaks, reintervention, and long-term survival.

We reviewed all patients undergoing EVAR in the Vascular Quality Initiative between 2003 and 2017 with an imaging study at 1 year postoperatively (±6 months). We defined aneurysm sac changes by Society for Vascular Surgery guidelines (change ≥5 mm) and determined independent predictors of sac behavior, new endoleak, and reintervention using hierarchical logistic regression. We employed Cox regression to examine the association between sac behavior and long-term survival. We performed propensity matching between patients with sac regression and those with failure to regress as a secondary analysis.

Of 30,074 EVAR patients, 14,817 (49%) had a 1-year imaging study and were included in this study. At 1 year, 40% of sacs regressed, 35% remained stable, and 25% expanded. Factors independently associated with sac expansion were age (by decade: odds ratio [OR], 1.07; 95% confidence interval [CI], 1.01-1.13; P = .02), appearance of new endoleak (OR, 1.23; 95% CI, 1.10-1.37; P = .001), smaller aortic diameter (diameter <5 cm: OR, 1.37; 95% CI, 1.21-1.55; P < .001), anemia (OR, 1.47; 95% CI, 1.20-1.80; P < .001), rupture (OR, 1.33; 95% CI, 1.07-1.65; P = .01), and chronic kidney disease (OR, 1.15; 95% CI, 1.05-1.25; P < .01), whereas former smoking (OR, 0.86; 95% CI, 0.76- 0.96; P < .01), cerebrovascular disease (OR, 0.82; 95% CI, 0.67-0.99; P = .04), and statin therapy at discharge (OR, 0.83; 95% CI, 0.75-0.91; P < .001) were associated with lower risk of expansion. Both sac expansion (OR, 2.3; 95% CI, 2.0-2.7; P < .001) and a stable sac (OR, 3.1; 95% CI, 2.7-3.5; P < .001) were associated with the development of new endoleaks. Any failure of the aneurysm sac to regress was associated with long-term mortality compared with sac regression (stable sac size: hazard ratio, 1.2; 95% CI, 1.03-1.4; P = .02; sac expansion: hazard ratio, 1.6; 95% CI, 1.3-2.1; P < .001). This association persisted in patients without documented endoleaks and remained robust after accounting for reinterventions and endoleaks seen in follow-up or on completion angiography. In the propensity-matched cohort, patients with failure to regress experienced lower long-term survival (77% at 10 years compared with 82% for patients with sac regression; P = .01).

After EVAR, aneurysm sac behavior is associated with the development of new endoleaks, reintervention, and long-term mortality. Not only sac expansion but any failure of the sac to regress is associated with higher long-term mortality, independent of reinterventions or endoleaks, and may be modified by statin therapy. Further study is needed to establish whether patients with stable sacs could benefit from selective reintervention.