More accurate prediction of treatment response than mean apparent diffusion coefficient value in multiple myeloma using whole body MRI histogram analysis

Abstract

Aim: To assess baseline histogram parameters from apparent diffusion coefficient (ADC) images in predicting early treatment response in newly diagnosed multiple myeloma (NDMM) patients. Methods: The histogram parameters of lesions in 68 NDMM patients were obtained with the Firevoxel software. The presence of deep response after two cycles of induction was recorded. Results: Some parameters were significantly different between the two groups, for example, ADC 75% in lumbar spine (p = 0.026). No significant difference in mean ADC for any anatomic site was found (all p > 0.05). The combination of ADC 75, ADC 90 and ADC 95% in lumbar spine; ADC skewness and ADC kurtosis in rib achieved a sensitivity of 100% in predicting deep response. Conclusion: Histogram analysis of ADC images can describe NDMM heterogeneity and accurately predict treatment response.