ABSTRACT
The section on surrogate endpoint gives several examples of how to evaluate a variable as a surrogacy measure of a clinical endpoint. Examples in oncology and HIV/AIDS are given to illustrate how data were pooled across trials to establish the surrogate. The analyses are useful even when a regulatory decision on surrogate endpoint is not imminent. For example, if a drug is to move forward from Phase II to Phase III, a biomarker which serves like a preliminary surrogate marker is useful in making an informed decision for the drug manufacturer about the magnitude of clinical benefit expected from clinical endpoint. Most sections are accompanied by R code for analysis. It takes time, effort, collaboration, and investment to establish a surrogate variable in a particular disease area.
