Virtual Library
Start Your Search
L.R. Crack
Author of
-
+
MINI 29 - Meta Analyses and Trial Conduct (ID 156)
- Event: WCLC 2015
- Type: Mini Oral
- Track: Treatment of Advanced Diseases - NSCLC
- Presentations: 1
- Moderators:D. Morgensztern, M. Redman
- Coordinates: 9/09/2015, 18:30 - 20:00, Mile High Ballroom 2a-3b
-
+
MINI29.14 - Using a Bayesian Adaptive Phase II Trial Design to Test Multiple Genetic-Marker-Directed Drugs in the National Lung Matrix Trial (ID 1241)
19:45 - 19:50 | Author(s): L.R. Crack
- Abstract
- Presentation
Background:
The National Lung Matrix Trial is a flagship trial in the United Kingdom being the first to combine the development of a technology platform that screens for multiple genetic aberrations in tumours with testing of multiple novel genetic-marker-directed drugs. The trial is focused on patients with advanced non-small cell lung cancer and currently includes 8 different drugs and 23 different drug-biomarker combinations. The aim of statistical analysis is to determine whether there is sufficient signal of activity in any drug-biomarker combination to warrant further investigation and this paper evaluates the novel statistical design that has been implemented.
Methods:
The primary outcome measure representing signal of activity is best objective response rate (ORR) in most cases and progression-free survival (PFS) in others. Each drug-biomarker combination could be considered as a single arm phase II trial and could have been designed using standard statistical approaches such as Simon’s two stage design. However, a Bayesian adaptive design was chosen because it gives a more realistic approach to decision-making in this complex setting. It has flexibility to make conclusions without fixing the exact sample size which will be important with the uncertainty around the prevalence of each biomarker. The design allows early stopping of recruitment to any drug-biomarker combinations that do not show sufficient promise at an interim analysis to warrant continuation. It also allows, when appropriate, for information about the primary outcome measure to be shared across different biomarker cohorts within any single drug to aid decision-making. Decision-making is based on the posterior probability distribution for the primary outcome measure, given the observed data and any prior knowledge, calculated using bayesian conjugate analysis. For ORR, the critical threshold which defines a signal of activity is 30% for single drugs and 40% for combinations and for PFS the critical threshold is median 3 months. The interim and final sample sizes were selected to ensure the design achieved pre-defined desirable operating characteristics.
Results:
For both ORR and PFS as primary outcomes, sample sizes of 15 and 30 patients per drug-biomarker cohort for interim and final analyses respectively were selected as giving the required operating characteristics. Simulations of trials for single drugs with ORR as the primary outcome showed that this would ensure a high probability (82%) of correctly stopping early when the true ORR was only 10% and a high probability (90%) of correctly recommending further investigation when the true ORR was as high as 40%. Operating characteristics for the combination drug arms were similar. Operating characteristics for trial arms with PFS as the primary outcome are dependent on the recruitment rate and give high probabilities (>95%) of stopping early when the true median PFS is only 1 month and high probabilities (>80%) of recommending further investigation when the true median PFS is as high as 4 months.
Conclusion:
The Bayesian adaptive design with 15 and 30 patients per drug-biomarker cohort for interim and final analyses respectively gives a flexible, efficient design with good operating characteristics to investigate multiple genetic-marker-directed drugs within a single phase II trial.
Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.
