Initial analytic validation of automated bone scan measures for treatment response assessment in patients with metastatic castration-resistant prostate cancer (CRPC).

Abstract

e15174 Background: Evaluation of novel therapies for patients with CRPC has been limited by the lack of validated quantitative measures of bone tumor burden. Tools including visual assessment for determining progression based on lesion count and semi quantitative measures including the Bone Scan Index for predicting survival have been proposed. To date, a validated quantitative measure that can be used to assess either progression or response has not been available. We have developed a computer-aided (CAD) system for measurement of bone scan tumor burden including lesion count (LC), mean intensity (MI) and total area (TA). This study is to assess the range of normal variation and potential clinically meaningful cut points for changes in bone scan tumor burden. Methods: From an anonymized image data set, 20 CRPC patients were selected with a pair of bone scans acquired 6 to 24 weeks apart. All patients had failed at least 2nd line drug therapy; 10 were on no therapy, 10 received drug therapy between the two scans. Scans were acquired at 12 sites using a standard of care protocol including 20-25 mCi of Tc MDP, injected 2-4 hours before whole body scintigraphy. An automated CAD system was run for each patient on bone scans. Degenerative joint disease was removed manually by one of two board certified Nuclear Medicine Physicians. A two-sample Wilcoxon rank sum (Mann-Whitney) test was used to test for differences between the two groups in the ratio of LC (visit 2 / visit 1), change in MI and % change in tumor TA. Results: Table shows the median (IQR) for ratio of LC, change in MI and % change in TA. Differences between groups were significant for all measures. Conclusions: This preliminary data suggests total bone scan tumor area may be useful as an early surrogate outcome for CRPC patients with 30% change from baseline as a potential cut point to distinguish responders from non-responders. This work lays a foundation for validation of these measures against other clinically relevant outcome measures in a larger cohort. Ratio of LC Change in MI % change in TA No therapy 1.24 (0.56) -1.40 (14.80) 7.13 (27.61) Drug therapy 0.54 (0.68) -17.60 (16.30) -73.76 (45.38) P value 0.0012 0.0019 0.0003