OBJECTIVES: There is interest in understanding how novel anti-cancer therapies impact healthcare disparities, but such analyses typically rely on individual patient data (IPD). This study developed and applied a framework for quantifying the impact of novel anti-cancer therapies on health inequalities based on survival data from Kaplan-Meier curves.

METHODS: For metastatic breast (mBC), metastatic colorectal (mCRC) and metastatic non-small cell lung (mNSCLC) cancer, 9 novel therapies were identified using pre-defined criteria. IPD were replicated from Kaplan-Meier curves of overall survival (OS) and progression-free survival (PFS), and pooled for interventions and comparators separately, using sample size-based weights to ensure each therapy contributed equally to the analysis. Parametric mixture distributions were considered in modelling OS and PFS data. Health inequality was defined as the difference between groups with highest and lowest survival, with 2 group and 5 group stratification. For the 5-group stratification, a linear regression model was fitted to use the regression slope as an inequality measure. Results were obtained through probabilistic analysis.

RESULTS: For OS stratified by 5 groups, novel therapies significantly increased inequalities relative to comparators by 13.9 (95%-confidence interval: 1.4; 26.6) months or 25.0% for mBC and 6.7 (3.0; 10.5) months or 40.4% for mCRC, but not for mNSCLC where inequalities decreased by 14.9 (-84.5; 19.0) months or 12.2%. Results based on the regression slope were similar. Two-group stratified OS and both 2 group and 5 group stratified PFS showed significant increases in survival inequalities across all cancers.

CONCLUSIONS: Novel metastatic cancer therapies in mBC, mCRC, and mNSCLC generally increased survival inequalities relative to comparators. Although further research using (real-world) IPD is warranted, the proposed framework is a useful tool for understanding how treatments may impact survival inequalities in absence of such data, and to inform further research.