Bone scan index reliably predicts prostate cancer outcomes
CHICAGO — An automated bone scan index has been confirmed as a prognostic biomarker of OS, radiographic PFS and skeletal-related events among men with castration-resistant prostate cancer and bone metastases in a phase 3 trial presented at the ASCO Annual Meeting.
“Early bone scan progression is associated with reduced overall survival. The automated bone scan index permits a faster and more fully quantitative biomarker that can be used for clinical trials without limiting the clinical trial design to prevent both response and progression perturbations,” Andrew J. Armstrong, MD, associate professor of medicine, pharmacology and cancer biology, and surgery at Duke University School of Medicine, said in a presentation. “The primary objective of this analysis was to clinically validate the baseline-motivated bone scan index as a prognostic biomarker for overall survival.”
The automated bone scan index is a computational approach to bone scan imaging that uses artificial intelligence to determine the percentage of skeletal mass that is affected by cancer. The trial aimed to further validate the prognostic value of this tool.
Armstrong and colleagues enrolled 1,245 men with metastatic, castration-resistant prostate cancer and bone metastases who were treated with tasquinimod (Active Biotech) or placebo and conducted whole-body scans at screening on all participants. The researchers locked the prospective biomarker analysis of the automated bone scan index before unblinding treatment. They analyzed automated bone scan indexes for independent prognostic correlation with OS, radiographic PFS and symptomatic skeletal-related events.
The researchers evaluated scans from 241 trial locations in 37 countries. The patients who received automated bone scan indexes (n = 721) appeared characteristic of all participants involved in the study based on OS outcomes and patient characteristics at the time of screening.
The median automated bone scan index was 1.07 (standard error = 0.05). The researchers grouped the different indexes into quartiles (n = 180-181); levels ranged from Q1 to Q4 (Q1, 0-0.3; Q2, > 0.3-1.1; Q3, > 1.1-4; and Q4, > 4).
Median OS ranged from 35 months in Q1 to 13 months in Q4 (P < .0001). The researchers observed significant associations between baseline automated bone scan index and OS (HR, 1.2 per doubling of bone scan index; P < .0001); these correlations maintained independent associations with OS after the researchers adjusted for treatment, PSA, C-reactive protein, LDH and albumin. The baseline automated bone scan index also correlated with radiographic PFS (P = .0005), time to symptomatic progression (P < .0001) and time to symptomatic skeletal-related events (P = .001).
This trial is the largest to date to show “clinically important outcomes associated with the bone scan index at baseline,” Armstrong said.
He also discussed the ways in which this study may inform practice.
“The automated bone scan index could be considered for clinical trial eligibility or for risk stratification based on these defined thresholds; they may be promising as endpoints. Ongoing work to define bone scan index criteria for bone progression will inform future work that may create a more quantifiable bone marker,” Armstrong said. “This is not meant to replace the radiographic interpretation by nuclear medicine physicians who read a bone scan. It is intended to quantify what they’re measuring so we can incorporate this into clinical trial practice and the designing of endpoints.” – by Julia Ernst, MS
Armstrong AJ, et al. Abstract 5006. Presented at: ASCO Annual Meeting; June 2-6, 2017; Chicago.
Disclosures: Armstrong reports he is a consultant or advisor for Bayer, Dendreon, Eisai, Janssen Biotech, Medivation, Novartis and Sanofi; receives honoraria from and serves on the speakers bureau for Dendreon and Sanofi; receives travel, accommodations and expenses from Dendreon, Janssen Biotech, Medivation and Sanofi; receives research funding through his institution from Astellas Pharma, Bayer, Bristol-Myers Squibb, Dendreon, Gilead Sciences, Janssen Oncology, Medivation, Novartis, Pfizer and Sanofi; and receives patents, royalties and other intellectual property through his institution for circulating tumor cell novel capture technology. Please see the full study for a list of all other researchers’ relevant financial disclosures.