ASCO 2017: Phase 3 prognostic analysis of the automated bone scan index in men with bone-metastatic castration-resistant prostate cancer

Chicago, IL ( The vast majority of men with metastatic castrate resistant prostate cancer (mCRPC) have bone metastasis. Bone scans indirectly measure tumor activity and efficacy by measuring osteoblastic activity in the tumor microenvironment. Changes in bone scan may reflect flare/healing in the setting of effective systemic treatment rather than true progression leading to variability in interpretations of response vs. progression. The Prostate Cancer Working Group 3 defines bone scan eligibility and progression metric but remains semi-quantitative.

In 1998 the bone scan index (BSI) was developed in Memorial Sloan Kettering (MSK) as a method to quantify total skeletal area on imaging invaded by metastasis. Recently, a technique involving a computational approach of BSI that employs artificial intelligence called the automated BSI (aBSI) was presented. Both the the manual and aBSI have been independently associated with reduced survival in men with mCRPC in phase 2 studies.

Dr. Armstrong presented a prospective phase 3 clinical study attempting to extend the prognostic validation of aBSI in a multinational men with bone-mCRPC. In this study, whole-body bone scans were acquired at screening in a placebo-controlled phase 3 trial of men with mCRPC and bone metastases who were treated with tasquinimod/placebo (n = 1,245). All scans generated at 241 trial sites in 37 countries were assessed for image quality and analyzed using the EXINI aBSI software version 2. All scans were blindly associated with outcomes. Primary objective was to clinically validate aBSI at baseline as a prognostic biomarker for survival in mCRPC patients. The authors analyzed overall survival (OS), cancer specific survival (CSS), radiographic progression-free survival (rPFS), and symptomatic skeletal related events (SSEs).

The final aBSI-population (721 pts) was representative of the entire trial population based on patient characteristics at screening and OS outcomes. Median aBSI was 1.07 (range 0-32.6). The aBSI-population was divided into quartiles (n = 180-181) with aBSI median levels of 0.05 (Q1), 0.58 (Q2), 2.06 (Q3), and 6.72 (Q4). Baseline aBSI was strongly correlated with manual BSI, number of bone metastasis, alkaline phosphatase (ALP) and moderately associated with PSA. aBSI was significantly associated with OS and CSS (HR 1.2 per doubling of BSI; p < 0.0001) and remained independently associated with OS after adjustment for treatment, PSA, CRP, LDH and albumin. Baseline aBSI was also strongly associated with rPFS (p = 0.0061), time to symptomatic progression (p < 0.0001), and time to SSE (p = 0.0068). However, it was not correlated with progression of soft tissue metastasis (p=0.25).

In summary, aBSI is associated with overall and prostate cancer specific survival as well symptomatic and bone radiographic progression in mCRPC men. Dr. Armstrong concluded his presentation suggesting the usage of aBSI in clinical trials as a promising endpoint.

Presented By: Andrew J. Armstrong, Division of Medical Oncology and Urology, Duke Cancer Institute, Duke University, Durham, NC

Written by: Hanan Goldberg, MD, Urologic Oncology Fellow (SUO), University of Toronto, Princess Margaret Cancer Centre
Twitter: @GoldbergHanan

at the 2017 ASCO Annual Meeting - June 2 - 6, 2017 – Chicago, Illinois, USA