27 Background: Although new oncology treatments have the potential to improve patient health, these innovations take time to reach to real-world patients. In this study, we estimated how delays in the uptake of novel oncology treatments affect real-world patient survival. Methods: This retrospective cohort study used 1991-2013 data from a cancer-registry linked to health care claims (SEER-Medicare). Our endpoint of interest was patient overall survival. We used variability in the adoption of new cancer treatments across aggregated Metropolitan Statistical Areas (MSAs) (n = 50) as a “natural experiment” that assigned treatments randomly across patients, depending entirely on where they live. This approach recovers the effect of the new treatment on the overall survival of the “marginal patient,” who received treatment solely because she resides in an area with faster uptake. This framework was applied to six case studies: trastuzumab (breast), bevacizumab (colorectal), bevacizumab (lung), erlotinib (lung), bortezomib (myeloma) and lenalidomide (myeloma). Results: Among the 92,496 patients in the study, the difference between adoption rates among eligible patients for MSAs at the 90th percentile and those at the 10th percentile was 20.1 percentage points, with the largest differences being for bevacizumab (colorectal) (51.2% at the 90th percentile MSA vs. 20.9% at the 10th percentile), and smallest for erlotinib (10.3% vs. 2.8%). Median OS among eligible patients gaining access to these landmark therapies increased by 9.6 months. Survival improvements were largest for the lenalidomide case study (33.4 months OS improvement, p < 0.005) followed by erlotinib (21.2 mo., p < 0.001), trastuzumab (19.2 mo., p = 0.016), bevacizumab for lung (7.4 mo., p = 0.002), bevacizumab for colorectal cancer (4.0 mo., p = 0.062), and bortezomib (2.9 mo., p = 0.630). Conclusions: Improving the speed at which landmark treatments are adopted in practice—through physician education and outreach or more generous health insurance access policies—has the potential to improve real-world survival for cancer patients.
Approving cancer treatments based on endpoints other than overall survival: an analysis of historical data using the PACE Continuous Innovation Indicators™ (CII)