Effects of Radiation On Functional Bone Marrow in Patients with Pelvic Malignancies
J Pritz, Ph.D.1*, S Gulaya, M.S.1, M Bydder, Ph.D.2, Y Liang, Ph.D.3, H Zhu, M.D.1, C Hoh, M.D.2, C M Yashar, M.D.1, M McHale, M.D.1, J Einck, M.D.1, P Fanta, M.D.3, F Vaida, Ph.D.4, G Bydder, M.D.2, L Mell, M.D.1, (1) Center for Advanced Radiotherapy Technologies, Department of Radiation Medicine and Applied Sciences, University of California San Diego, La Jolla, CA. (2) Department of Radiology, University of California San Diego, La Jolla, CA (3) Department of Medicine, Division of Hematology/Oncology, University of California San Diego, La Jolla, CA (4) Department of Family Medicine, Division of Biostatistics and Bioinformatics, University of California San Diego, La Jolla, CAWE-C-WAB-6 Wednesday 10:30AM - 12:30PM Room: Wabash Ballroom
Purpose: To test the hypothesis that pelvic bone marrow fat fraction (FF) (i.e., quantity of fat relative to water) varies with glucose metabolism and radiation dose in patients undergoing pelvic radiotherapy.
Methods: 34 patients with pelvic malignancies were enrolled from 2009-2012 on prospective trials; 31 received concurrent chemotherapy. All patients underwent baseline ¹⁸F-FDG-PET and quantitative MRI with IDEAL-IQ. 9 patients underwent PET/CT simulation, 15 PET/CT, and 10 PET-only. 18 and 14 patients underwent IDEAL-IQ at mid-treatment and 1-3 weeks post-treatment, respectively. PET-only scans were rigidly registered to the planning CT (pCT). PET/CT and IDEAL-IQ scans were deformably registered to the pCT (Velocity AI, Atlanta, GA). Each image was resampled to the pCT, creating a database of FF, dose, and body weight-standardized uptake value (SUVbw) for each voxel. Analyses were performed on pelvic bone from L5 to ischia. Spearman rank correlation coefficients (SRCC) were estimated comparing SUVbw and dose vs. FF₁, ΔFF₂₁, and ΔFF₃₁. FF₁ represents FF at baseline and ΔFF₂₁ and ΔFF₃₁ represent change in FF from baseline to mid- and post-treatment, respectively. Mean SRCCs were tested for significance using a 2-sided t-test.
Results: Data analysis was completed for 15 patients (13 gynecologic, 7 anal cancer). We observed a significant negative correlation between SUVbw and FF1, and significant positive correlations between SUVbw and ΔFF₂₁ and ΔFF₃₁ (Table), indicating that high metabolic activity correlated with lower FF (a surrogate for red marrow) and increased conversion to fat (yellow marrow) during radiotherapy. We observed significant positive correlations between radiation dose and ΔFF₂₁ and ΔFF₃₁, indicating that regions of higher dose were more likely to convert to fat (Table). No correlation was observed between dose and FF1.
Conclusion: Preliminary results suggest that red marrow has higher metabolic activity, and is more likely to convert to yellow marrow during radiotherapy, in a dose-dependent manner.
Funding Support, Disclosures, and Conflict of Interest: Funding has been provided by NIH R21 Research Grant (#CA162718-01) and by the American Society of Clinical Oncology.