KMDATA: a curated database of reconstructed individual patient-level data from 153 oncology clinical trials

Restricted mean survival time based on Wu-Kolassa estimator compared to Kaplan-Meier estimator

OBJECTIVE

The purpose of the paper is to introduce the Wu-Kolassa estimator (WKE) and the RMST based on it for its less biased estimation and substantial power gain, compared to the Kaplan-Meier estimator (KME), to researchers working in medical and health sciences to evaluate and compare patient survival times.

RESULTS

The seven numerical studies showed that the power gain in WKE-based RMST analysis can reach more than 80 %, depending on the size of the study and the trend of failure rate. For the phase III study of iniparib plus gemcitabine and carboplatin (GCI) versus gemcitabine and carboplatin (GC) in patients with metastatic triple-negative breast cancer, GCI is superior to GC demonstrated by WKE-based RMST analysis (point estimate of treatment difference in RMST = 0.807; 95 % CI: 0.0214 to 1.592, 1.5 times higher than the estimate based on KME = 0.542; 95 % CI: -0.385 to 1.470). For the phase III trial that studied ovarian suppression (os) with tamoxifen or exemestane, tamoxifen plus os is superior to tamoxifen alone using WKE-based RMST (point estimate of difference = 0.228; 95 % CI: 0.016 to 0.439, more than 2-fold higher than the estimated difference based on KME = 0.113; 95 % CI: -0.008 to 0.234).

CONCLUSIONS

The Wu-Kolassa estimator is superior to the Kaplan-Meier estimator as it reduces the estimation bias and increases the power in the RMST analysis when the censoring rate is high or when the reference group has more censoring data than the test group.

Cancer patient survival can be parametrized to improve trial precision and reveal time-dependent therapeutic effects

Individual participant data (IPD) from oncology clinical trials is invaluable for identifying factors that influence trial success and failure, improving trial design and interpretation, and comparing pre-clinical studies to clinical outcomes. However, the IPD used to generate published survival curves are not generally publicly available. We impute survival IPD from ~500 arms of Phase 3 oncology trials (representing ~220,000 events) and find that they are well fit by a two-parameter Weibull distribution. Use of Weibull functions with overall survival significantly increases the precision of small arms typical of early phase trials: analysis of a 50-patient trial arm using parametric forms is as precise as traditional, non-parametric analysis of a 90-patient arm. We also show that frequent deviations from the Cox proportional hazards assumption, particularly in trials of immune checkpoint inhibitors, arise from time-dependent therapeutic effects. Trial duration therefore has an underappreciated impact on the likelihood of success.

Analysis of more than 150 Phase 3 oncology clinical trials supports parametric statistical analysis, significantly increasing the precision of small early-phase trials and relating deviations from the Cox proportional hazards model to trial duration.