Postoperative complications in Hartmann's procedure versus intersphincteric abdominoperineal excision in rectal cancer: randomized clinical trial (HAPIrect).

Abstract

Background: In patients with rectal cancer, when it is not possible to restore bowel continuity with an anastomosis, the optimal surgical method is still a matter of debate. The aim of this trial was to determine 30-day postoperative surgical complication rates after Hartmann's procedure (HP) versus intersphincteric abdominoperineal excision (iAPE) in patients with rectal cancer who were not suitable for restorative surgery.

Methods: This multicentre randomized controlled trial (HAPIrect) was performed in Sweden and Finland between 2014 and 2021. Eligible patients with adenocarcinoma of the rectum located ≥ 5 cm from the anal verge and deemed unsuitable for anterior resection with anastomosis were randomized (1:1) intraoperatively to either HP or iAPE. The primary outcome was 30-day postoperative surgical complications. Secondary outcomes were major surgical complications (Clavien-Dindo grade ≥ IIIa), perineopelvic complications, and overall complications. Logistic regression in the intention-to-treat population was the primary method used to compare the surgical approaches.

Results: Of 194 eligible patients, 163 were randomized (80 patients to HP and 83 to iAPE). The study was closed before achieving the target accrual. The main reasons for not receiving an anastomosis were advanced age, co-morbidity, or poor anal sphincter function. Mean operating time in the HP and iAPE groups was 291 and 373 minutes, respectively. In the HP and iAPE groups, the surgical complication rate was 39% and 43%, respectively (odds ratio (OR) for HP 0.83; 95% confidence interval (c.i.) 0.44 to 1.54; P = 0.549) and the rate of major surgical complications was 14% and 11%, respectively (P = 0.573). Perineopelvic complications occurred in 21% and 30% of patients in the HP and iAPE groups, respectively (OR for HP 0.63; 95% c.i. 0.31 to 1.28; P = 0.197). The overall complication rate (including both medical and surgical complications) was 45% and 49% in the HP and iAPE groups, respectively (P = 0.574). In multivariable analysis adjusted for sex, preoperative radiotherapy, and surgical procedure, there was no statistically significant difference in surgical complications between the two groups.

Conclusion: Although the trial was underpowered and did not reach accrual, in randomized patients, both HP and iAPE are practicable surgical options for patients unsuitable for anastomosis.

Registration number: NCT01995396 (http://www.clinicaltrials.gov).