Surgical outcomes after colorectal surgery for endometriosis: Systematic Review and Meta-Analysis.

Bendifallah, Sofiane; Puchar, Anne; Vesale, Elie; Moawad, Gaby; Daraï, Emile; Roman, Horace

OBJECTIVE To assess the impact of type of surgery for colorectal endometriosis -rectal shaving, discoid or colorectal segmental resection- on complications and surgical outcomes. DATA SOURCES We performed a systematic review of all English and French language full-text articles addressing surgical management of colorectal endometriosis and compared the postoperative complications according to surgical technique by meta-analysis. The PubMed, Clinical Trials.gov Cochrane Library and Web of Science databases were searched for relevant studies published before March 27, 2020. The search strategy used the following MeSH terms: ("bowel endometriosis" or "colorectal endometriosis") AND ("surgery for endometriosis" or "conservative management" or "radical management" or "colorectal resection" or "shaving" or "full thickness resection" or "disc excision") AND ("treatment", "outcomes", "long term results" and "complications"). METHODS OF STUDY SELECTION Two authors conducted the literature search and independently screened abstracts for inclusion, with resolution of any difference by three other authors. Studies were included when data on surgical management (shaving, disc excision and/ or segmental resection) were provided and when postoperative outcomes were detailed with at least the number of complications. The risk of bias was assessed according to the Cochrane recommendations. TABULATION, INTEGRATION, AND RESULTS Of the168 full-text articles assessed for eligibility, 60 were included in the qualitative synthesis. Seventeen of these were included in the meta-analysis on rectovaginal fistula, 10 on anastomotic leakage, five on anastomotic stenosis, and nine on voiding dysfunction <30 days. The mean complication rate according to shaving, discoid excision and segmental resection were 2.2%, 9.7% and 9.9%, respectively. Rectal shaving was less associated with rectovaginal fistula than discoid excision (OR=0.19; 95% CI [0.10-0.36], p<0.00001, I2=33%) and segmental colorectal resection (OR=0.26, 95% IC [0.15-0.44], p<0.00001, I2=0%). No difference was found in the occurrence of rectovaginal fistula between discoid excision and segmental colorectal resection (OR=1.07, 95%CI [0.70-1.63], p=0.76, I2=0%). Rectal shaving was less associated with leakage than disc excision (OR=0.22, 95% IC [0.06-0.73], p=0.01, I2=86%). No difference was found in the occurrence of leakage between rectal shaving and segmental colorectal resection (OR=0.32, 95% IC [0.10-1.01], p=0.05, I2=71%) or between disc excision and segmental colorectal resection (OR=0.32, 95% IC [0.30-1.58], p=0.38, I2=0%). Disc excision was less associated with anastomotic stenosis than segmental resection (OR=0.15, 95% IC [0.05-0.48], p=0.001, I2=59%). Disc excision was associated with more voiding dysfunction <30 days than rectal shaving (OR=12.9, 95% IC [1.40-119.34], p=0.02, I2=0%). No difference was found in the occurrence of voiding dysfunction <30 days between segmental resection and rectal shaving (OR=3.05, 95% IC [0.55-16.87], p=0.20, I2=0%) or between segmental colorectal and discoid resection (OR=0.99, 95% IC [0.54-1.85], p=0.99, I2=71%). CONCLUSION Colorectal surgery for endometriosis exposes patients to a risk of severe complications such as rectovaginal fistula, anastomotic leakage, anastomotic stenosis and voiding dysfunction. Rectal shaving appears to be less associated with postoperative complications than disc excision and segmental colorectal resection. However, this technique is not suitable in all patients with large bowel infiltration. Compared to segmental colorectal resection, disc excision has several advantages including shorter operating time, shorter hospital stay and lower risk of postoperative bowel stenosis. REGISTRATION OF SYSTEMATIC REVIEWS PROSPERO ID.