Pelvic exenterations for gynecological malignancies: a study of 36 cases.

OBJECTIVE: Evaluation of surgical outcomes, survival, and morbidity associated with pelvic exenteration (PE) performed for gynecologic malignancies.

METHODS: Review of 36 consecutive patients who underwent PE between June 1999 and April 2010.

RESULTS: Pelvic exenteration was performed for cancer of the cervix (n = 18), endometrium (n = 9), vagina/vulva (n = 8), and ovary (n = 1). Four patients underwent PE as primary treatment and 32 patients for recurrent disease after pelvic radiotherapy. Median age was 57 years (range, 35-81 years). Bricker (n = 17), Mainz pouch (n = 10), and augmentation after bladder resection (n = 6) were used as urinary derivations. J-pouch coloanal anastomosis was performed in 14, colostomy in 13, and side-to-end anastomosis in 4 patients. There was no operative mortality. The most important postoperative complications were rectovaginal fistula (5), urinary leakage (2), vesicovaginal fistula (1), and sepsis (3). One of the 6 patients with a partial cystectomy developed a vesicovaginal fistula, which was successfully treated with a Martius flap. With a median follow-up of 78 months (range, 2-131) months, the 5-year overall and disease-specific survival (DSS) rates were 44% and 52%, respectively. Five-year DSS for cervical, endometrial, and vaginal/vulvar cancer was 44%, 80%, and 57%, respectively. Combined operative and radiotherapeutic treatment (CORT) was performed in 3 patients with pelvic side wall relapse. Of the 15 patients 65 years or older, a 5-year DSS of 71% was observed in comparison with 42% in the younger subgroup, and their complication rates were similar to the younger patient group. Thirteen patients (36%) reported to have psychological disturbances associated with stoma-related problems. Only 3 patients requested a vaginal reconstruction during follow-up.

CONCLUSIONS: Pelvic exenteration offers a sustained survival with an acceptable morbidity in patients with advanced or recurrent gynecologic cancer. Older age was not associated with higher morbidity/mortality in this series.