Survival extrapolation validation-based case study (Bullement et al.)
A method to check how accurate long-term survival predictions are in cancer immunotherapy by comparing early predictions with later, more complete data from the same trial.
At a glance
Use when
Conducting HTA for cancer immunotherapies with immature survival data; evaluating the reliability of survival extrapolations in economic models.
Avoid when
Long-term survival data are already available; for non-oncology indications with different survival patterns.
Inputs
Early and late data cuts from randomized controlled trials, including Kaplan-Meier survival curves, time-to-event data, and fitted parametric models.
Outputs
Comparison of predicted vs. observed survival outcomes, assessment of extrapolation error, implications for cost-effectiveness results.
How it works
A retrospective validation approach that assesses the credibility of survival extrapolations in oncology HTA by comparing model projections based on early data cuts with actual outcomes from later, mature data cuts, particularly in the context of immunotherapy where survival curves may exhibit non-proportional hazards and long-term plateaus.
- HTA domains
- Clinical Effectiveness, Costs & Economic Evaluation, Safety
- Assumptions
- Later data cuts represent a more accurate reflection of true long-term survival; the trial population remains consistent across data cuts; the treatment effect is stable over time.
- Strengths
- Provides empirical validation of commonly used extrapolation techniques; highlights risks of bias in long-term projections; supports more robust decision-making in HTA.
- Limitations
- Limited to therapies with available mature follow-up data; findings may not generalize to novel agents without long-term data; dependent on quality and timing of data cuts.
- Also known as
- validation-based survival extrapolation case study, Bullement et al. survival validation method
Questions this answers
- › How accurate are long-term survival extrapolations from early clinical trial data in immunotherapy?
- › Do extrapolated survival curves converge with real-world observed survival when more mature data become available?
- › What are the implications of survival extrapolation errors for cost-effectiveness estimates in HTA?
- › Which parametric models perform best in predicting long-term survival in immune-oncology?
- › How sensitive are economic models to the choice of extrapolation time horizon?
- › Can early HTA decisions be trusted given uncertainty in survival projections?
References & sources
Similar by meaning
- Survival extrapolation incorporating general-population mortality using excess-hazard and cure models (Sweeting et al. tutorial)
- survextrap
- NICE DSU Technical Support Document 14
- Change-point survival models for relative treatment effect extrapolation
- INES (INteractive model for Extrapolation of Survival and cost)
Beta record. Based on the original catalogue summary; primary-source enrichment pending.