Elective surgery of abdominal aortic aneurysms in octogenarians: a systematic review

Abdominal aortic aneurysm (AAA) is an age-related disease. In an aging population, the prevalence of AAA is likely to increase. Open AAA repair in patients aged >80 years is often not considered because of their advanced age as such or because of comorbidities. In addition, little is known about the natural history in such patients or survival after successful repair. We performed a systematic review of the literature to determine peri-operative and late survival after AAA repair in octogenarians The Medline, Embase, and Cochrane databases were searched to identify all studies reporting on octogenarians undergoing AAA repair published between January 1966 and June 2006. Two independent observers assessed the methodologic quality of the included studies and the data extraction. Outcomes were rates of perioperative mortality, complications, and long-term survival after open or endovascular repair (EVAR). Summary estimates with 95% confidence interval (CI) were calculated using a random effects model. Thirty-nine articles were included. The median aneurysm size was 6.7 cm in the conventional AAA repair group of 1534 patients. The perioperative mortality was 0% to 33%, with a pooled mortality of 7.5% (95% CI, 6.2% to 9.0%). The median 5-year survival rate for this group was 60% (range, 14% to 86%). In the 1045 patients treated with EVAR, the median aneurysm size was 5.9 cm. Their pooled perioperative mortality varied from 0% to 6%, with a pooled mortality of 4.6% (95% CI, 3.4 to 6.0%). We could not derive 5-year survival rates from articles describing endovascular repair of AAA. The mortality rate after open or endovascular AAA repair in carefully selected octogenarians seems acceptable but is higher than the mortality rate in younger patients. Long-term survival rates were acceptable, but small sample size, selection, and publication bias must be taken into account. Finally, selection criteria for successful surgery with low mortality and morbidity rates cannot be derived from the literature