Whole body ¹⁸F-fluoride PET SUV imaging to monitor response to dasatinib therapy in castration-resistant prostate cancer bone metastases: secondary results from ACRIN 6687.

Objectives: Dasatinib is an SRC kinase inhibitor that decreases bone turnover in men with metastatic castration-resistant prostate cancer (mCRPC). ACRIN 6687 was a multi-center clinical trial designed to evaluate differential response of normal and tumor bone to dasatinib using ¹⁸F-fluoride PET with dynamic single field-of-view (FOV) imaging followed by static whole body (WB) scans. Previous kinetic modeling results from 12 evaluable patients from the single FOV imaging found significant differences in changes of the PET biomarkers from tumor bone compared to normal bone in response to dasatinib treatment, and changes in the 30-60m SUV metrics correlated with progression-free survival (PFS). Here we present the results of using whole body (WB) static SUV PET imaging in these patients with ¹⁸F-fluoride.

Methods: Eligible patients (n=18; median age 69 years range 48-86) with bone mCRPC underwent WB ¹⁸F-flouride PET imaging approximately 76 minutes after ¹⁸F injection (range 53-95 min) prior to and 12 weeks after treatment with dasatinib. Unlike the single FOV investigation of 12 patients with 37 selected bone lesions around the pelvis and lower spinal column, WB imaging analysis collected the 5 most ¹⁸F avid bone metastases over the whole body for 90 bone lesions in 18 patients. Bone lesion regions along with matched areas of normal bone were used for analysis by SUVmax, SUVpeak and Patlak flux. A regression model assessed differences in both tumor and normal bone between pre- and post-treatment PET and compared differences between change in normal with tumor bone. Pre-treatment values and the change from pre- to post-treatment values were evaluated in association with PFS via Cox proportional hazards modeling. Results: 14 patients of the 18 enrolled had evaluable pre- and post- dasatinib treatment WB PET for analysis of changes due to therapy. Of the 18 patients, 17 have either met progression criteria or death by the time of this analysis. In the single FOV study, bone lesions were physician selected and occurred in the pelvis or along the lower spinal column, whereas only 19 of the 70 lesions (27%) used in the WB analysis overlapped with the single FOV work. Significant decrease in both SUVmax and SUVpeak occurred in bone metastases in response to dasatinib (SUVmax p = 0.014 and SUVpeak p = 0.001, n = 14) while significant increases by Patlak flux occurred in normal bone (p = 0.047). Significant differences in changes from tumor bone compared to normal bone in response to dasatinib were noted for SUVmax (p<0.001). PET SUV measures from the 14 patients with scans at baseline and following dasatinib therapy failed to show an association with PFS.

Conclusions: Pharmacodynamic changes in normal bone and bone mCRPC can be identified by WB SUV using 18F-fluoride PET in response to therapy with dasatinib. There is a differential effect of dasatinib on normal compared to tumor bone in men with mCRPC, which appears primarily related to fluoride bone incorporation. Where the previous report using single FOV 30-60min SUV metrics showed a larger decrease in bone mCRPC uptake in response to treatment with dasatinib that marginally correlated with shorter PFS (p = 0.056), WB SUV measures collected approximately 45 min later failed to find significance. WB PET measures were not associated with PFS potentially due to tumor selection for the single FOV study by local site investigators, the large variability of uptake time, later scanning allowing more clearance of the tracer independent of the disease or fewer counts with increased noise in the WB scans. The actual reasons for lack of association of WB ¹⁸F-fluoride PET SUV measures with PFS versus the limited field of view dynamic scans are not known, but might be pursued with the goal of applying ¹⁸F-fluoride PET to prostate cancer bone metastasis response assessment. SUPPORT: NIH U01- CA079778 (ACRIN, Schnall); NIH R50-CA211270 (Muzi); SFI PI 11/1027 (O’Sullivan)