Investigating the Effects of External Beam Radiation Therapy on Female Sexual Organs at Risk

Abstract

Purpose:Sexual dysfunction resulting from external beam radiation therapy (EBRT) can significantly impact the quality of life for cancer patients. While the relationship between EBRT and sexual dysfunction has been extensively studied in men, little research has focused on how EBRT affects sexual dysfunction in women. This study aims to better define the bulboclitoris and vaginal canal as organs at risk (OARs) for EBRT and to explore how these organs are affected by radiation. Methods: This retrospective study was performed under an IRB approved protocol. The bulboclitoris and vaginal canal were contoured using T2-weighted MRI scans, and these contours were transferred to CT images. An automatic contouring atlas was then created and validated by comparing atlas-generated contours to manually drawn contours using volume measurements, Dice Similarity Coefficients, Jaccard Index, and Hausdorff Distance. The radiation dose received by these OARs during pelvic EBRT was determined by transferring the contours onto planning CT images. Treatments were replanned to evaluate the feasibility of sparing dose to these organs. Dose metrics were extracted for both the clinical and replanned treatments, and dose differences were analyzed. The relationship between dose and physiological and morphological changes in these OARs was also studied using DWI image analysis and radiomics. Changes in MRI features were correlated with dose levels to explore potential for functional avoidance. Morphological changes were assessed by comparing DSC values between T2-weighted contours at different time points and calculating volume changes. Physiological changes were evaluated by analyzing changes in median DWI signal between scans taken before EBRT and prior to brachytherapy for defined ROIs. Radiomic analysis was performed on volumes defined by specific isodose lines. Results: Thirty-four patients were included in this study with 16 with uterine cancer, 6 with vaginal cancer, 10 with cervical cancer, 1 with vulvar cancer, and 1 with urethral cancer. Of the 34 patients, 15 for the atlas validation, 30 for the treatment replanning, 16 for longitudinal study, and 18 for radiomics analysis. The bulboclitoris and vaginal canal were contoured for 34 patients using 59 T2-weighted MRI scans, with mean volumes (±SD) of 12.94 (±3.76) cc for the bulboclitoris and 16.15 (±8.23) cc for the vaginal canal. A contouring atlas was created from 31 atlas subjects and validated on 15 patients. For the bulboclitoris, the median (±SD) Dice Similarity Coefficient (DSC) was 0.7 (± 0.07), the median Jaccard Index was 0.54 (±0.08), the median Hausdorff distance was 1.3 (±0.33), and the median volume difference was 1.26 cc. For the vaginal canal, the median DSC was 0.54 (±0.11), the median Jaccard Index was 0.37 (±0.37), the median Hausdorff distance was 3.44 (±1.52), and the median volume difference was 0.13 cc. EBRT treatments were replanned for 30 patients, including those with uterine (15), vaginal (6), cervical (7), vulvar (1), and urethral (1) cancers. For the bulboclitoris, the median difference in mean dose was -2.61 Gy, and 0 Gy for the vaginal canal. Dose reduction was observed in the bulboclitoris across multiple metrics, particularly for the V30 Gy, where statistically significant reductions were noted. Morphological changes were calculated for 16 patients, showing greater volume reduction and lower DSC with increasing dose for both organs. Physiological changes, assessed for 16 patients for the bulboclitoris and 9 for the vaginal canal, revealed a greater increase in DWI signal in the high-dose group compared to the low-dose group. For the vaginal canal, the mean DWI change increased from 30.3% in the low-dose group to 55.3% in the high-dose group. For the bulboclitoris, the mean DWI change increased from 44.8% to 123.3% between the low-dose and high-dose groups. Radiomic analysis of 18 patients revealed one statistically significant feature for the bulboclitoris (long high run gray level emphasis) and 13 significant features for the vaginal canal, including cluster shade, contrast, dissimilarity, energy, gray level non-uniformity, high gray level size emphasis, inverse difference moment normalized, inverse variance, and others. Conclusion: This study demonstrates that dose can be reduced to the bulboclitoris by replanning EBRT treatments and identifies morphological and physiological changes in both the bulboclitoris and vaginal canal across different dose levels. These findings may inform prospective studies aimed at defining dose thresholds associated with sexual dysfunction.