Prediction of mobilized hematopoietic stem cell yield in patients with multiple myeloma: Usefulness of whole-body MRI-derived indices

INTRODUCTION: High-dose chemotherapy followed by autologous stem cell transplant is the mainstay of treatment for multiple myeloma (MM). The purpose of this study was to evaluate the ability of MRI-derived indices to predict mobilized hematopoietic stem cell yield. MATERIALS AND METHODS: In this exploratory pilot work, we retrospectively analyzed 38 mobilization procedures for MM. Successful mobilization procedure was defined as a total yield of >4.0×10(6) CD34(+) cells/kg. Univariate and multivariate analyses were performed to identify factors with a significant effect on successful mobilization from among clinical characteristics including number of prior lines of therapy, period from diagnosis to harvest, type of monoclonal protein (M protein); and radiological characteristics including total diffusion volume (tDV), median apparent diffusion coefficient (ADC) of tDV, and mean fat fraction of bone marrow calculated by MRI. RESULTS: Univariate analyses showed that relatively poor mobilization was significantly associated with M protein of Bence-Jones type and with median ADC of tDV (P = 0.02 and P = 0.004, respectively). Multivariate analyses using these two indices showed that median ADC of tDV was a significant predictive factor for adequate mobilization (P = 0.01), with an area under the curve of 0.784 (cutoff value, 1.18×10(−3) mm(2)/s; sensitivity, 72.7%; specificity, 87.5%). CONCLUSION: The present data indicate that median ADC of tDV is a predictive factor for relatively poor mobilization of hematopoietic stem cells in MM patients undergoing autologous stem cell transplant.