The association between radiotherapy doses to bone marrow and fatigue in prostate cancer

Abstract

Purpose or Objective Bone Marrow (BM) is the primary site of haematopoiesis and, in prostate radiotherapy, 50%–55% of active BM is found in the treatment volume. Recent studies showed a robust association between Integral Dose (ID) and increased levels of fatigue in prostate cancer (PC) patients. Although these results require further validation, from a clinical perspective, the critical structure is thought to be the BM. We have investigated the BM dose compared to the ID as a predictor of patient-reported fatigue post-radiotherapy. Materials and Methods From the REQUITE database, we retrospectively analysed 23 prostate cancer patients who received EBRT. The bone marrow volume was defined as the inner cavity of the whole pelvic bone (WPB) and the contour process used the following steps: 1) manual WPB segmentation from L5 to the inferior border of the femoral head, 2) upload the pelvic contouring to the structure template management, 3) import the template structure through ATLAS-initialization in all patients, 4) manual correction applied if necessary, 5) automatic BM contouring using contraction tool, negative 0.4mm from the whole pelvic contorting in all direction as shown in figure 1. The integral dose was defined as Body-PTV to calculate the following: ID [Gy⋅l]=D [Gy]⋅V [l]. The mean dose, V10, and V20 were calculated for both the Body-PTV and each BM PTV volumes. Fatigue was scored 1= patient-reported fatigue or 0= did not report fatigue based on self-reported questionnaires that included EORTC QLQ-C30, QLQ PR25, and MFI. Statistical analysis used descriptive statistics, t-tests and boxplots, and bee swarm graphs.Results The mean ID for patients with fatigue scores of 1 was 121.6 Gy.l, versus 104.5 Gy.l for patients with scores of 0 (p < 0.05) see figure 2. The mean dose, V10, and V20, of the BM structure for patients with fatigue scores of 1 were 10 Gy, 36.7%, and 18.7% respectively. For patients with fatigue scores of 0 the mean dose, V10, and V20 were 8.8 Gy, 33.3%, and 16.2%. None of which demonstrated significant differences ( p = 0.136, p = 0.108, and p = 0.205 respectively). Conclusion We found Atlas-based auto segmentation (ABAS) is a reliable method in WPB contouring which can be a practical tool in a large cohort. To our knowledge, this is the first study investigating BM dose and fatigue in PC patients. ID was found to be the most discriminatory dosimetric parameter correlated to fatigue. Higher doses to BM were not associated with increased fatigue reports. This could be further investigated in a larger cohort.