Supine bottom-up extralevator abdominoperineal excision for anorectal adenocarcinoma is not inferior to standard approach and may be thus safely performed.

BACKGROUND: Extralevator abdominoperineal excision (APE) for rectal carcinoma has been described in order to improve pathological and oncological results compared to standard APE. To obtain the same oncological advantages as extralevator APE, we have previously described a new procedure starting by a perineal approach: the supine bottom-up APE. Our objective is to compare oncological and surgical outcomes between the supine bottom-up APE and the standard APE.

METHODS: All patients with low rectal adenocarcinoma requiring APE were retrospectively included and divided into 2 groups: supine bottom-up APE (Group A) and standard APE (Group B).

RESULTS: From 2008 to 2016, 61 patients were divided into Groups A (n = 30) and B (n = 31). Postoperative outcomes and median length of stay were similar between groups. Patients from Group A had a significantly longer distal margin (30 [8-120] vs. 20 [1.5-60] mm, p = 0.04) and higher number of harvested lymph nodes (14.5 [0-33] vs. 11 [5-25], p = 0.03) than those from Group B. Circumferential resection margin involvement was similar between groups (28 vs. 22%, p = 0.6), whereas tumors from Group A were significantly larger and more frequently classified as T4 than those from Group B. Operative time was significantly shorter in Group A (437.5 [285-655] minutes) than in Group B (537.5 [361-721] minutes, p = 0.0009). At the end of follow-up, local recurrence occurred in 7 and 16% of patients from Groups A and B (p = 0.68). Three-year overall and disease-free survival rates were similar between groups (87 vs. 90%, p = 0.62 and 61 vs. 63%, p = 0.88, respectively).

CONCLUSION: Our findings suggest that supine bottom-up APE doesn't impair surgical outcomes, pathological results, overall and disease-free survivals in comparison with standard APE. This new procedure may be thus safely performed and decrease the operative time. Further randomized multicentric studies are required to confirm these results.