Cause-specific Long-Term Mortality After Physician-Modified Branched/Fenestrated Endovascular Aortic Repair

Abstract

Physician-modified endografts (PMEG) are being increasingly used for treatment of complex aortic diseases. Several series describe short and mid-term outcomes, but limited long-term data is available. Here we sought to evaluate the long-term causes of death after PMEG, and the association with type and complexity of repair. A single, high-volume, academic aortic referral center retrospectively reviewed all consecutive branched/fenestrated PMEG procedures from 2010 to 2016, which allowed all subjects the opportunity for 5 or more years of survival time. All subjects were deemed either high anatomic and/or physiologic risk for open repair. Commercial custom and chimney EVAR procedures were excluded. Death certificates, obituary websites, and electronic medical records were reviewed to assign cause-specific mortality. The primary end point was all-cause mortality. Secondary end points included cause-specific-death categorized as: major adverse cardiovascular event (myocardial infarction, malignant arrythmia, congestive heart failure, sudden death), malignancy, pulmonary (chronic obstructive pulmonary disease/respiratory failure), infection, aorta-related, and other (including metabolic, stroke, and trauma). Kaplan-Meier methodology was used to estimate survival. A total of 232 consecutive PMEG procedures were analyzed (male, 72%; mean age, 73 ± 8 years). Indications included thoracoabdominal aneurysm (extent I-V; 58%), pararenal AAA (33%), postsurgical pseudoaneurysm or failed EVAR (6%), chronic dissection (2%), and penetrating paravisceral ulceration (1%). A majority were performed electively (75%, n = 174) while the remaining were completed for symptomatic or intact (20%, n = 47) and emergent or ruptured (5%, n = 11) presentations. We included 762 target vessels (celiac, n = 179; superior mesenteric artery, n = 204; renal, n = 379) and most procedures had four-vessel revascularization (61%, n = 142). The 30-day mortality rates were: elective, 3% (n = 5/174) and nonelective, 14% (n = 8/58). Secondary interventions occurred in 18% (n = 42). The mean survival was 4.9 ± 3.2 years (median, 5.4; interquartile range, 1.9-7.5; 95% confidence interval [CI], 4.9-6.1) and 57% (n = 132/232) died during follow-up. Among 132 documented deaths, 10% (n = 13/132) occurred before hospital discharge and/or at 30 days or less, whereas 90% (n = 119/132) occurred after 30 days and/or after hospital discharge. Cause-specific mortality was attributed to major adverse cardiovascular events (21%), other causes (18%), aorta related (13%), unknown (13%), pulmonary (12.5%), infection (12.5%), and malignancy (10%). Causes of late aorta-related death (n = 16/232, 6.9% overall; 12.1% of all deaths recorded) were rupture (n = 5) and mesenteric ischemia (n = 2), respectively. Freedom from long-term all-cause and aorta-related mortality was: 5-years, 56 ± 3% and 90 ± 2%; 8-years, 43 ± 4% and 87 ± 3%, respectively (Fig 1). Nonelective presentation had higher long-term risk of all-cause and aorta-related mortality (all-cause: hazard ratio [HR], 1.6; 95% confidence interval, 1.1-2.4; aorta-related: HR; 3.2; 95% confidence interval, 1.2-8.5) (Fig 2). Four-vessel repairs were associated with improved survival (vs fewer than 4 vessels: HR, 0.6; 95% confidence interval, 0.4-0.9). PMEG in high-risk patients was durable and provided excellent long-term outcomes in patients deemed high risk for open repair. Cardiovascular morbidity is the most significant factor leading to late mortality after PMEG. Aorta-related mortality is low; however, elective repairs with more proximal coverage appear to be protective from poor long-term outcomes.Fig 2Freedom from aorta-related death after physician-modified endograft (PMEG) by presentation (includes perioperative deaths).View Large Image Figure ViewerDownload Hi-res image Download (PPT)