Title: High-Dimensional Penalized Regression Models in Time-to-Event Clinical Trials
Abstract: High-dimensional data are increasingly easy to obtain on samples obtained from patients included in clinical trials, and are used to develop models for predicting the prognosis (prognostic biomarkers) and the treatment effect (predictive biomarkers) of each patient. However, the large quantity of information has rendered false positives more and more frequent in biomedical research. For variable selection in a high-dimensional setting, the lasso penalty is commonly used in the Cox model. In the prognostic setting, an empirical extension of the lasso penalty has been proposed to be more stringent on the estimation of the tuning parameter λ in order to select less false positives. In the predictive setting, focus has been given to the biomarker-by-treatment interactions in the setting of a randomized clinical trial. Multiple approaches have been proposed for selecting these interactions and are compared. Finally, we propose a strategy to obtain an individual survival prediction with a corresponding confidence interval for a future patient from a penalized regression model, while limiting the potential overfit. The performance of the approaches is evaluated through simulation studies combining null and alternative scenarios. The methods are illustrated using gene expression data from clinical trials including patients with early breast cancer.
Download the flyer here.