Abstract:Medical Decision Making, Ahead of Print. BackgroundSurvival extrapolation is essential in cost-effectiveness analysis to quantify the lifetime survival benefit associated with a new intervention, due to the restricted duration of randomized controlled trials (RCTs). Current approaches of extrapolation often assume that the treatment effect observed in the trial can continue indefinitely, which is unrealistic and may have a huge impact on decisions for resource allocation.ObjectiveWe introduce a novel methodology as a possible solution to alleviate the problem of survival extrapolation with heavily censored data from clinical trials.MethodThe main idea is to mix a flexible model (e.g., Cox semiparametric) to fit as well as possible the observed data and a parametric model encoding assumptions on the expected behavior of underlying long-term survival. The two are "blended" into a single survival curve that is identical with the Cox model over the range of observed times and gradually approaching the parametric model over the extrapolation period based on a weight function. The weight function regulates the way two survival curves are blended, determining how the internal and external sources contribute to the estimated survival over time.ResultsA 4-y follow-up RCT of rituximab in combination with fludarabine and cyclophosphamide versus fludarabine and cyclophosphamide alone for the first-line treatment of chronic lymphocytic leukemia is used to illustrate the method.ConclusionLong-term extrapolation from immature trial data may lead to significantly different estimates with various modelling assumptions. The blending approach provides sufficient flexibility, allowing a wide range of plausible scenarios to be considered as well as the inclusion of external information, based, for example, on hard data or expert opinion. Both internal and external validity can be carefully examined.HighlightsInterim analyses of trials with limited follow-up are often subject to high degrees of administrative censoring, which may result in implausible long-term extrapolations using standard approaches.In this article, we present an innovative methodology based on "blending" survival curves to relax the traditional proportional hazard assumption and simultaneously incorporate external information to guide the extrapolation.The blended method provides a simple and powerful framework to allow a careful consideration of a wide range of plausible scenarios, accounting for model fit to the short-term data as well as the plausibility of long-term extrapolations.

Survival Analysis Using a 5-Step Stratified Testing and Amalgamation Routine in Randomized Clinical Trials

Is inadequate risk stratification diluting hazard ratio estimates in randomized clinical trials?

Covariate-adjusted Group Sequential Comparisons of Survival Probabilities

Optimality of testing procedures for survival data

Augmenting the logrank test in the design of clinical trials in which non-proportional hazards of the treatment effect may be anticipated

Supplementary Figure S1 from Proportional Hazards Violations in Phase III Cancer Clinical Trials: A Potential Source of Trial Misinterpretation

Robust Design and Analysis of Clinical Trials With Non-proportional Hazards: A Straw Man Guidance from a Cross-pharma Working Group

Combining Survival Trials Using Aggregate Data Based on Misspecified Models

Unbiased Prediction and Feature Selection in High-Dimensional Survival Regression

Stratified modestly-weighted log-rank tests in settings with an anticipated delayed separation of survival curves

Design and Analysis of group-sequential clinical trials based on a modestly-weighted log-rank test in anticipation of a delayed separation of survival curves: A practical guidance

Blended Survival Curves: A New Approach to Extrapolation for Time-to-Event Outcomes from Clinical Trial in Health Technology Assessment

A new method for clustered survival data: Estimation of treatment effect heterogeneity and variable selection

CaseCohortCoxSurvival: an R Package for Case-Cohort Inference for Relative Hazard and Pure Risk under the Cox Model

Blended Survival Curves: A New Approach to Extrapolation for Time-to-Event Outcomes from Clinical Trials in Health Technology Assessment

Studentized Permutation Method for Comparing Two Restricted Mean Survival Times with Small Sample from Randomized Trials.

Semiparametric Additive Modeling of the Restricted Mean Survival Time

Statistical Considerations for Evaluating Treatment Effect under Various Non-proportional Hazard Scenarios

A Novel Stratified Analysis Method for Testing and Estimating Overall Treatment Effects on Time-to-Event Outcomes Using Average Hazard with Survival Weight

Targeted Estimation of Heterogeneous Treatment Effect in Observational Survival Analysis

Leveraging machine learning: Covariate-adjusted Bayesian adaptive randomization and subgroup discovery in multi-arm survival trials