Optimizing Perioperative Outcomes with Selective Bowel Resection Following an Algorithm Based on Preoperative Imaging for Bowel Endometriosis

Abstract

Study objective: To validate the algorithm for selective bowel surgery based on preoperative imaging by comparing the perioperative outcomes of patients who undergo each type of bowel surgery for deep bowel disease, and secondarily to evaluate the incidence, factors, and subsequent outcomes when the actual procedure performed deviated from the preoperative surgical plan.

Design: Retrospective study comparing 3 surgical interventions in an intention-to-treat analysis.

Setting: Tertiary care hospital.

Patients: Women with significant pain (visual analog scale [VAS] >7) who were diagnosed with bowel endometriosis from preoperative imaging and underwent laparoscopic surgery for bowel endometriosis at a large referral center between 2014 and 2017.

Intervention: Laparoscopic shaving, disc resection, or full-segment resection and reanastomosis of bowel endometriosis.

Measurements and main results: A total of 172 patients (mean age, 36.6 ± 5.2 years) underwent bowel surgery for endometriosis (n = 30 shaving, 71 disc, and 71 segmental resection). Total operative time was similar in the 3 group, but the mean length of hospital stay was longer in the segmental group (5.3 ± 1.0 days) compared with the disc group (4.6 ± 0.9 days) and the shaving group (3.8 ± 1.5 days) (p = .001). The surgical procedure was performed as planned according to the clinical algorithm in 86.5% of patients. Adherence to the proposed clinical algorithm resulted in a low incidence of overall complications (8.7% of total complications, 4.6% of minor complications, and 3.5% of major complications). The incidence of minor complications was higher in the segmental group (9.9%) compared with the discoid group (1.4%) and the shaving group (0%) (p = .0236), whereas the incidence of major complications were similar across the 3 groups (3.3%, 2.8%, and 4.2%, respectively; p = .899). There was a significantly higher frequency of pseudomembranous colitis in the segmental resection group (7 patients; 9.9%) compared with the discoid group (n = 1; 1.4%) and shaving group (0%) (p = .04). Owing to discrepancies between preoperative imaging and intraoperative findings after dissection and mobilization, deviation from the planned procedure occurred in a total of 25 of 172 cases (14.5%), with a less extensive procedure actually performed in 21 of 25 (84%) of the deviated cases. One of the 4 cases (25%) that involved a more extensive procedure resulted in a major complication of rectovaginal fistula.

Conclusion: Selective bowel resection algorithm provides a systematic approach to the surgical management of patients with bowel endometriosis. Adherence to the surgical plan according to the preoperative imaging and criteria outlined in the algorithm can be accomplished in the majority of patients; however, the surgical team should be aware that upstaging or downstaging may be required, depending on the intraoperative findings. When feasible, the team should opt for a less extensive procedure to avoid complications associated with more radical surgery.

Keywords: Bowel endometriosis; Deep endometriosis; Minimally invasive surgical procedures; Transvaginal ultrasound.