To explore how diabetes mellitus (DM), along with the use of metformin, shapes aneurysm sac remodeling in the aftermath of endovascular aneurysm repair (EVAR). A single-institution, retrospective cohort study was conducted among consecutive individuals who underwent elective EVAR for an infrarenal abdominal aortic aneurysm (AAA) during the period from January 2011 to December 2021. Between-group differences were assessed; Kaplan–Meier estimation was used to characterize overall survival and survival free from reintervention; and Cox regression modeling was applied to identify variables predictive of sac shrinkage. The cohort comprised 529 patients: 74 (14.0%) were living with DM and receiving metformin, 26 (4.9%) had DM but no metformin prescription, and 429 (81.1%) had no DM diagnosis. Upon evaluation at the one-year milestone, diabetic individuals exhibited a significantly lower rate of sac shrinkage relative to their non-diabetic counterparts (40.0% vs. 52.0%; P = 0.038), with a concurrent inclination toward a greater prevalence of stable sac dimensions in the diabetic subset (52% vs. 42%; P = 0.055). By the final follow-up assessment, sac shrinkage remained significantly less common among diabetics on metformin compared with non-diabetics (48.6% vs. 59.9%; P = 0.047). Comparisons between diabetic patients treated with metformin and those managed without it did not reveal any appreciable disparity in sac shrinkage. Endoleak detection rates were substantially higher within the subgroups demonstrating sac stabilization and sac enlargement. Across a surveillance period reaching nine years, overall survival was demonstrably inferior in diabetic patients compared with those free of diabetes (23.5% vs. 37.5%; P < 0.001). The evidence from this study points to a detrimental association of both diabetes mellitus and metformin therapy with aneurysm sac shrinkage following EVAR. Documentation of any endoleak category was associated with blunted sac shrinkage at both early and late observation points. Overall survival proved to be significantly compromised in the diabetic cohort compared with the non-diabetic cohort.
