Novel Statistical Models for NSCLC Clinical Trials

The discovery of activating gene mutations and molecular abnormalities as potential predictive markers for targeted therapy have not only revolutionized the way we approach lung cancer but has greatly advanced cancer medicine as a whole. However, as more potential treatments are being developed and biological markers discovered, a significant increase in the number of trials are necessary to test and validate the efficacy of treatments and the ability of markers to accurately predict outcomes of corresponding therapies. In order to meet the increased demands of clinical testing with limited resources, designs need to be more efficient, have smaller sample sizes, give accurate conclusions and place patients in more effective treatment arms. Novel adaptive designs, including various methods that apply the Bayesian methodology, meet these demands by proposing an attractive alternative to less efficient conventional study methods and should be more frequently considered in the design and implementation of future clinical studies.