Factors Limiting Ostomy Reversal After Cytoreductive Surgery for Ovarian Cancer: A Retrospective Study.

Abstract

Background/aim: To investigate the factors related to non-reversal of ostomy after cytoreductive surgery in ovarian cancer. In many women with ovarian cancer, transitory ostomies are performed to limit the consequences of anastomotic leak. Although intended to be temporary, a proportion of these ostomies might never be reversed.

Patients and methods: This was a retrospective study of patients with 2014 International Federation of Obstetrics and Gynecology stage IIB-IVB ovarian cancer requiring a transitory ostomy during primary or secondary cytoreductive surgery at the Bergonie Institute, France, and the University Hospital of Las Palmas, Spain, between January 2012 and December 2022. Rate of ostomy reversal, its timing (weeks) and postoperative complications were assessed. Multivariate logistic regression analysis was performed to identify limiting factors for ostomy reversal.

Results: During the study period, we reviewed data on 181 consecutive patients with ovarian cancer with transitory ostomy creation; 89 (49.2%) patients were not candidates for an ostomy reversal surgery because of disease progression (n=65), death (n=16), and patient's refusal of surgery (n=8). A total of 92 patients were candidates for reversal surgery and were therefore included in the final analysis. In total, 57 (62%) patients had their ostomy reversed. The mean time from ostomy creation to ostomy closure was 47.7 (standard deviation=33.1) weeks. Hartmann's procedure (leaving a rectal stump of 5-6 cm) was identified as an independent predictive factor for non-reversal of ostomy (odds ratio=6.42, 95% confidence interval=1.61-25.53; p=0.008). Complications after ostomy reversal occurred in 32 patients (34.8%).

Conclusion: Hartmann's procedure is a limiting factor for ostomy reversal in patients with ovarian cancer. We recommend avoiding Hartmann's procedure during cytoreductive surgery, even after colorectal anastomotic leak.