Radiation-Induced Vaginal Stenosis in Patients With Locally Advanced Cervical Cancer

Abstract

Vaginal stenosis (VS) after definitive radiation therapy (RT) to the pelvis for cervical cancer is a common toxicity that can have detrimental effects on quality of life and can limit surveillance physical examinations. Quantitative measurements of VS are lacking. We investigated radiographic vaginal dimensional changes after RT for cervical cancer as a potential noninvasive marker for VS with the hypothesis that RT would be associated with reduced apical vaginal width on computed tomography (CT) imaging.We retrospectively identified 104 patients with locally advanced cervical cancer treated with definitive RT at our institution who had pre- and post-treatment CT scans available for analysis. Post-treatment scans were acquired within one year of therapy completion. We measured the transverse width at the vaginal apex, mid-canal, and introitus, as well as superior-inferior vaginal length. Changes in mean dimensional measurements were compared between pre- and post-treatment scans. We also compared mean dimensional measurements between CT and MRI where available. Nonparametric paired Wilcoxon rank sum tests were used to test for differences in measurements. Two observers made independent measurements on 15 consecutive patients and interrater reliability was assessed by calculating the intraclass correlation coefficient (ICC). Multivariable linear regression was used to identify covariables associated with change in apical width. Covariables included menopausal status, smoking history, gravida/para status, FIGO stage, tumor grade, and tumor histology. We used a backward model selection technique with stepwise removal of covariables not meeting a significance threshold of P ≤0.20.Mean transverse width at the vaginal apex was 5.87 cm on pre-treatment CT and 5.10 cm on post-treatment CT, with a mean difference of -0.77 cm (P < 0.001). There was also a statistically significant difference in width at the mid-vaginal canal although this was of smaller magnitude (pre-treatment 4.34 cm, post-treatment 4.14 cm, P < 0.001). There were no differences in width at the introitus or with respect to vaginal length. The ICC for independent apical measurements between two raters was 0.91 (95% CI 0.75, 0.97), indicating good-excellent agreement. Apical width was slightly smaller when measured on MRI compared to CT, although a small number of available paired MRI scans limited this analysis. On multivariable linear regression, menopause was correlated with less change in apical width (P = 0.001). Conversely, a smoking history was associated with increased dimensional reduction at the apex (P = 0.04). Number of deliveries and tumor stage were not significant predictors for dimensional change.Radiation-induced vaginal stenosis after definitive RT for cervical cancer treatment was demonstrable as apical and mid-vaginal constrictions on CT imaging that were more pronounced with a smoking history and less evident in post-menopausal women.C.W. Williamson: None. N.V. Kotha: None. K. Morris: None. R. Simoes-Torigoe: None. M. Kohanfars: None. J. Chen: None. B. Li: None. S. Vaitkus: None. M.T. Makale: None. F. Talke: None. J.S. Mayadev: Consultant; Astra Zeneca, NRG Oncology, GOG Foundation, Varian Medical Systems; NRG Oncology Cervix Co-Chair, GOG Foundation Board.