Vaginal Dose Is Associated With Toxicity in Image Guided Tandem Ring or Ovoid-Based Brachytherapy

Purpose To calculate vaginal doses during image guided brachytherapy with volume-based metrics and correlate with long-term vaginal toxicity. Methods and Materials In this institutional review board–approved study, institutional databases were searched to identify women undergoing computed tomography and/or magnetic resonance–guided brachytherapy at the Duke Cancer Center from 2009 to 2015. All insertions were contoured to include the vagina as a 3-dimensional structure. All contouring was performed on computed tomography or magnetic resonance imaging and used a 0.4-cm fixed brush to outline the applicator and/or packing, expanded to include any grossly visible vagina. The surface of the cervix was specifically excluded from the contour. High-dose-rate (HDR) and low-dose-rate (LDR) doses were converted to the equivalent dose in 2-Gy fractions using an α/β of 3 for late effects. The parameters D0.1cc, D1cc, and D2cc were calculated for all insertions and summed with prior external beam therapy. Late and subacute toxicity to the vagina were determined by the Common Terminology Criteria for Adverse Events version 4.0 and compared by the median and 4th quartile doses, via the log-rank test. Univariate and multivariate hazard ratios were calculated via Cox regression. Results A total of 258 insertions in 62 women who underwent definitive radiation therapy including brachytherapy for cervical (n=48) and uterine cancer (n=14) were identified. Twenty HDR tandem and ovoid, 32 HDR tandem and ring, and 10 LDR tandem and ovoid insertions were contoured. The median values (interquartile ranges) for vaginal D0.1cc, D1cc, and D2cc were 157.9 (134.4-196.53) Gy, 112.6 (96.7-124.6) Gy, and 100.5 (86.8-108.4) Gy, respectively. At the 4th quartile cutoff of 108 Gy for D2cc, the rate of late grade 1 toxicity at 2 years was 61.2% (95% confidence interval [CI] 43.0%-79.4%) below 108 Gy and 83.9% (63.9%-100%) above ( P =.018); grade 2 or greater toxicity was 36.2% (95% CI 15.8%-56.6%) below 108 Gy and 70.7% (95% CI 45.2%-96.2%) above ( P =.004); and grade 3 or worse toxicity was 9.9% (95% CI 0.0%-23.6%) below 108 Gy and 30.0% (95% CI 4.7%-55.3%) above ( P =.025). This association was maintained on multivariate analysis, independent of covariates such as applicator type, age, and dose rate. Conclusions Vaginal dose was associated with all grades of vaginal toxicity. Confirmation at other sites using this methodology will be necessary to establish reproducibility; however, the inte