br confounding and the development of cancer is multifac
confounding, and the development of cancer is multifac-torial and thus may be influenced by unmeasured con-founding variables. However, patient age, sex, strategic health authority, deprivation index, and medical comor-bidities were controlled for in the analysis, and the similar incidence of lung cancer and obesity-related nonabdomi-nal cancers suggests that tobacco smoking and obesity were equally distributed between open surgery and EVAR groups and thus unlikely to be confounders in this study. This study also evaluated only inpatient episodes, so those patients diagnosed with cancer in the commu-nity and not admitted to the hospital would not be captured within the data set. However, in practice, this represents the minority of patients in the United Kingdom with cancer. The HES data set is a well-validated and well-used data set17-26 that achieved 100% coverage of all NHS hospitals in England during the study Bortezomib (PS-341) and allows patients to be tracked throughout its entirety. This study was not randomized, and follow-up protocols for EVAR require up to 12 imaging studies in 5 years after surgery compared with open repair, in which none are typically performed.19 This discrepancy means that ascertainment bias cannot be excluded as an explanatory factor for the results. Surveillance protocols were obtained at hospital level from a previously published survey, and it is known that individual attendance at surveillance varies.19 The surveillance practice of hospitals is also changing over
time; this might introduce random misclassification of the CT surveillance analysis, which could dilute potential associations.
This large and population-based cohort study indicates that patients who underwent EVAR compared with open AAA repair were at a greater risk for development of later abdominal cancer, although causation cannot be inferred from the available nonrandomized and obser-vational data. DNA damage observed from radiation exposure during EVAR provides a possible mechanistic explanation for the observed association, and exploration of this phenomenon in late follow-up of the available randomized trial data would be prudent.
Conception and design: SM, HM, GH, MT, AK
Analysis and interpretation: SM, AVD, VS, HM, GH, JL, AK
Writing the article: SM, AVD, VS, AK
Critical revision of the article: SM, VS, HM, GH, MT, PH, JL
Statistical analysis: AVD
Obtained funding: Not applicable
Overall responsibility: SM
1. Schwarze ML, Shen Y, Hemmerich J, Dale W. Age-related trends in utilization and outcome of open and endovascular repair from abdominal aortic aneurysm in the United States, 2001-2006. J Vasc Surg 2009;50:722-9.
2. Park BD, Azefor NM, Huang CC, Ricotta JJ. Elective endovascular aneurysm repair in the elderly: trends and outcomes from the Nationwide Inpatient Sample. Ann Vasc Surg 2014;28:798-807.
3. Lederle FA, Freischlag JA, Kyriakides TC, Matsumura JS, Padberg FT Jr, Kohler TR, et al. Long-term comparison of endovascular and open repair of abdominal aortic aneu-rysm. N Engl J Med 2012;367:1988-97.
5. Greenhalgh RM, Brown LC, Powell JT, Thompson SG, Epstein D, Sculpher MJ. Endovascular versus open repair of abdominal aortic aneurysm. N Engl J Med 2010;362:1863-71.
6. Patel R, Sweeting MJ, Powell JT, Greenhalgh RM; EVAR trial investigators. Endovascular versus open repair of abdominal aortic aneurysm in 15 years’ follow-up of the UK endovas-cular aneurysm repair trial 1 (EVAR trial 1): a randomised controlled trial. Lancet 2016;388:2366-74.
7. Becquemin JP, Pillet JC, Lescalie F, Sapoval M, Goueffic Y, Lermusiaux P, et al. A randomized controlled trial of endo-vascular aneurysm repair versus open surgery for abdominal aortic aneurysms in low- to moderate-risk patients. J Vasc Surg 2011;53:1167-73.