Evaluation of Applicator Geometry and Impact on Reference Point and Organ Volume Dose (D2cc) in the Era of 3-Dimensional Imaging for Cervical Cancer Brachytherapy

Abstract

Intensity modulated proton therapy (IMPT) is a promising radiation therapy (RT) modality for cervical cancer treatment, especially for its potential role in reducing hematologic toxicity. IMPT dose distribution is known to be sensitive to patient anatomical changes during the treatment course. This study quantifies the effect of the anatomical change on IMPT's target coverage for cervical cancer treatment. Materials/Methods: In this IRB-approved study, 3-beam IMPT plans were generated on the planning computed tomography (CT) image for 6 enrolled patients. Three or 4 weekly CT images were obtained for each patient using the same imaging protocol in subsequent weeks after the simulation day, for a total of 21 weekly CT images. CTs of the same patient were rigidly registered to the planning CT by matching bony anatomy. The geometric cervical and nodal target volumes were the same in subsequent CTs as in the planning CT, while major organs-at-risk including bowel, bladder, and rectum were re-contoured on each weekly CT. Patient body weights at the time of each weekly CT scan were also recorded. Dose from the IMPT plan was recomputed on these weekly CTs. Dose received by at least 95% of the cervical and nodal internal target volumes (Cervical D ITV95% and Nodal D ITV95% , respectively) were compared to their planned values. Univariate analysis was performed to discover the correlation between D ITV95% to the relative changes of various anatomical factors, as well as the correlation between the anatomical factors. Results: Body weight increase was found to be significantly correlated to the decrease of both Cervical D ITV95% (Spearman rank correlation coefficient SCCZ-0.653, PZ.001) and Nodal D ITV95% (SCCZ-0.713, P<.001). Bowel volume increase was found to be significantly correlated to the decrease of Nodal D ITV95% only (SCCZ-0.502, PZ.020), but not Cervical D ITV95% (SCC Z-0.260, PZ.256). The correlations between bladder or rectum changes and target coverage were not statistically significant. Yet, there was also a significant correlation between bowel volume and body weight (SCCZ0.576, PZ.006). By deducting the bowel and bladder volume change, body weight increase was still found to be significantly correlated to the decrease of both Cervical D ITV95% (SCCZ0.613, PZ.003) and Nodal D ITV95% (SCCZ-0.646, PZ.002). Conclusion: The decrease of target coverage in IMPT for cervical cancer was found to be significantly correlated to mainly the increase of body weight, although intercorrelations between the changes of bladder, rectum, bowel, and body weight also existed. Multivariate analysis is needed to further uncover the major anatomical factor that impacts IMPT target coverage so that patient surveillance or intervention may be applied to maintain IMPT's superior dose distribution.