Abstract
BACKGROUND AND PURPOSE: This study aimed to evaluate the impact of pelvic radiation therapy (RT) on the occurrence of severe radiation-induced lymphopenia (SRIL) and identify its clinical and dosimetric predictors using voxel-wise analysis. Understanding these impacts is crucial for improving patient outcomes and optimizing treatment protocols in radiation oncology. MATERIALS AND METHODS: A retrospective analysis was conducted on 122 patients who underwent pelvic RT. Absolute lymphocyte counts (ALC) were measured before treatment and within one month of RT initiation. Patients were classified into SRIL and non-SRIL groups on the basis of their lowest recorded ALC during treatment. The associations between SRIL and clinical/dosimetric parameters were assessed via univariable (UVA) and multivariate (MVA) analysis. The influence of regionally detailed dose was assessed by voxel-based analysis (VBA) on spatially normalized 3D dose maps and CT images, focusing on the sacrum, femoral heads, and pelvic bones. RESULTS: SRIL was associated with clinical and dosimetric factors. The baseline ALC was the most significant clinical predictor, with a lower baseline ALC increasing SRIL risk (OR = 0.996, p = 0.001). VBA further revealed localized highly related regional dose patterns, with 92.17% of the left femoral head and 91.32% of the right femoral head showing significant SRIL associations, whereas the associations were significantly lower in the sacrum (10.39%) and pelvic bones (left: 30.01%, right: 31.52%). CONCLUSION: This study identified key clinical and dosimetric factors influencing SRIL in patients undergoing pelvic radiotherapy. Baseline ALC was the most significant clinical factor, and VBA showed that regional dose pattern changes within the femoral head were significantly associated with SRIL. CLINICAL TRIAL NUMBER: Not applicable.