Does 18F-sodium fluoride PET/CT have a role in the assessment of response to 223Ra therapy in men with prostatic bone metastases?

Background The REASURE study investigated imaging biomarkers in men receiving 223 Ra treatment for prostate cancer bone metastases. We used REASURE data to compare and contrast the roles of [ 18 F]NaF PET/CT measurements of maximum standardised uptake value (SUVmax) and bone metabolic flux (Ki) as markers of response. Methods Thirty-four men with prostatic bone metastases received up to six cycles of 223 Ra (55 or 88 kBq/kg) at four-weekly intervals. Whole-body diffusion-weighted MRI and [ 18 F]NaF PET/CT images were acquired at baseline and 4, 12, and 24 weeks later. Values of the apparent diffusion coefficient (ADC, a measure of tumour response), SUVmax, and a novel surrogate measurement of Ki using a PET/CT SUV measurement in the left ventricle were monitored in up to 5 bone metastases in each participant. Multilinear regression analysis (MLR) was performed to determine if changes in ADC were predicted by changes in SUVmax or Ki. Radium dose and baseline SUVmax were additional independent variables. Results ADC values increased throughout the study, while SUVmax and Ki decreased. MLR analysis showed that baseline SUVmax and 223 Ra dose predicted ADC response ( P < 0.001 and P < 0.01, respectively). Changes in SUVmax and Ki at 4, 12, and 24 weeks failed to predict the changes in ADC, showing decoupling between MRI and PET/CT measurements as markers of response. Changes in Ki were a highly significant predictor of changes in SUVmax ( P < 0.001), reflecting the strong correlation between these two measurements. Conclusions Baseline SUVmax was the best predictor of ADC response, followed by 223 Ra dose. Changes in SUVmax and Ki failed to predict changes in ADC, suggesting that the changes in the PET/CT variables reflected the effect of 223 Ra uptake on osteoblasts rather than a tumoricidal effect. A decrease in SUVmax seen on post-therapy [ 18 F]NaF PET/CT scans should not be interpreted as evidence of tumour regression. Trial Registration ISRCTN ISRCTN17805587. Registered 21/01/15.