No Increased Risk of Cancer Death after Endovascular Aortic Repair in a Nationwide Population-based Cohort Study.

Abstract

The short-term benefits of endovascular aortic repair (EVAR) compared to open repair (OR) for treatment of abdominal aortic aneurysms (AAA) are well established. However, concerns have been raised regarding a potential increased long-term cancer risk associated with EVAR, related to procedural and surveillance-related radiation exposure. The aim of this nation-wide population-based cohort study was to evaluate whether EVAR is associated with an increased long-term cancer risk compared to OR. All patients undergoing primary AAA repair for an intact AAA (ICD-10: I71.4) from January 2005-February 2024 were identified from the Swedish National Patient Register. Previous and subsequent cancer diagnoses, as well as previous comorbidities for this cohort were recorded. Cause of death was retrieved from the Cause of Death Registry. Inverse probability of treatment weighting (IPTW) was applied using propensity scores derived from baseline demographics characteristics and comorbidities. Weighted cox regression models, with EVAR as the sole regressor, were then fitted for the event of a new cancer diagnosis and survival. Some 15,509 patients were identified (mean age 73 years; 17% female; 24% previous cancer diagnosis). After weighting, standardized mean differences for comorbidities, age, gender, and recent hospital admissions were all within ± 0.1. The median survival was 8.7 years (95% CI 8.4-8.9) for EVAR patients and 9.4 years (95% CI 9.1-9.6) for OR patients. The median follow-up time was 4.9 years (IQR 6.1) for new cancer and 5.9 years (IQR 6.4) for cancer-related death. Freedom from new cancer was lower in EVAR patients (HR 0.92, 95% CI 0.86-0.98) whereas cancer related survival was similar (HR 0.93 95% CI 0.85-1.02). EVAR was not associated with an increased risk of dying of cancer, but with an increased risk of being diagnosed with a new cancer. This should be interpreted carefully as there is a clear risk of detection bias of otherwise unknown tumors due to routine imaging during EVAR surveillance.