Introduction: We investigated the predictive and prognostic effects of VeriStrat, a serum or plasma-based assay, on response and survival in a subset of patients enrolled on the NCIC Clinical Trials Group, BR.21 phase III trial of erlotinib versus placebo in previously treated advanced non-small-cell lung cancer patients.
Methods: Pretreatment plasma samples were available for 441 of 731 enrolled patients and were provided as anonymized aliquots to Biodesix. The VeriStrat test was performed in a Clinical Laboratory Improvement Act-accredited laboratory at Biodesix, Inc. (Boulder, CO). The results (Good, Poor) were returned to NCIC Clinical Trials Group, which performed all the statistical analyses.
Results: VeriStrat testing was successful in 436 samples (98.9%), with 61% classified as Good. VeriStrat was prognostic for overall survival in both erlotinib-treated patients and those on placebo, independent of clinical covariates. For VeriStrat Good patients, the median survival was 10.5 months on erlotinib versus 6.6 months for placebo (hazard ratio 0.63, 95% confidence interval 0.47-0.85, p = 0.002). For VeriStrat Poor patients, the median survival was 4 months for patients receiving erlotinib, and 3.1 months for placebo (hazard ratio: 0.77, 95% confidence interval 0.55-1.06, p = 0.11). VeriStrat was predictive for objective response (p = 0.002), but was not able to predict for differential survival benefit from erlotinib (interaction p = 0.48). Similar results were found for progression-free survival.
Conclusion: We were able to confirm that VeriStrat is predictive of objective response to erlotinib. VeriStrat is prognostic for both OS and progression-free survival, independent of clinical features, but is not predictive of differential survival benefit versus placebo.