Clinical and Treatment Factors Associated With Vaginal Stenosis After Definitive Chemoradiation for Anal Cancer

Abstract

Although VS (vaginal shortening and/or decreased width) is a known long-term toxicity after pelvic radiation therapy (RT) for gynecologic cancer, few studies have evaluated its incidence after definitive chemoRT for anal cancer. We retrospectively reviewed the clinical and treatment characteristics of women treated for anal cancer to identify factors associated with VS. Ninety-six consecutive women who received definitive chemoRT for anal squamous cell carcinoma between 2005 and 2012 comprised the study cohort. NCI CTCAE v. 4 was used to score for VS based on the medical record description of dyspareunia, vaginal dryness or difficult pelvic exam. Other clinical factors examined included age, race, body mass index (BMI), menopausal status, smoking, tumor size/lymph node status (AJCC 7th edition), immunocompromised status, and treatment with abdominoperineal resection (APR). The exact Wilcoxon-Mann-Whitney and Jonckheere Terpstra tests were used to assess VS risk factors. Median age was 60.4 years (range, 19-97). Patient characteristics included: 26% non-White; 14% pre-menopausal; 9% immunocompromised; and 51% with current or prior tobacco use. Stage distribution included 17% T1, 39.5 % T2, 26.5% T3, and 17% T4 tumors; 43% were node positive. Ninety-eight percent received dose-painted IMRT to a median primary tumor dose of 50.4 Gy (range, 41.4 Gy-60 Gy). Chemotherapy included: 83% concurrent mitomycin-C and 5-fluorouracil (5FU), 7% cisplatin and 5FU, 6% 5FU, and 2% capecitabine. APR was performed in 13% for recurrence or late toxicity. With a median follow-up of 2.5 years, 68 women (71%) had some degree of VS and 29% were without sufficient clinical information to assess VS. VS grade distribution was 38% grade 1, 25% grade 2 and 37% grade 3. For women with adequate VS assessment, only younger age (p = 0.016) and higher RT dose (p = 0.05) were predictors of worsening grade of VS on univariate analysis. Seventy-two percent of those with VS reported using measures to prevent it (38% vaginal dilator and 37% intercourse). Women without sufficient information to assess VS had cancers with worse prognostic features, including larger tumors (5 cm vs 3 cm, p < 0.005) and lymph node positivity (63% vs 34%, p = 0.009), had lower median BMI (23 vs 26, p = 0.047), and were more likely to require an APR (25% vs 7.4%, p = 0.04). Vaginal stenosis is a common late complication of radiation therapy in patients treated definitively with chemoradiation for anal cancer, even with IMRT. Younger age and higher prescribed dose are associated with an increased incidence of VS. Measures to prevent stenosis appeared ineffective although compliance may play an important role. We are currently evaluating dosimetric parameters, such as vaginal wall dose, which may contribute to VS.