Fischer, A., Hernandez-Villafuerte, K., Latimer, N., and Henshall, C.
Trials of drugs for cancer treatment may cause a reduction in tumour size for a period of time but might not extend life, or do so by a very small amount. Alternatively, there may be only a slight reduction in tumour size but a sustained response that allows for a long extension of life. It is the change in length of life or overall survival (OS) that is regarded as the most important outcome, but this in-formation is often not available for some years. However, if the length of time that the tumour is not growing (called progression-free survival, or PFS) is highly correlated with OS, information about the size of OS can be inferred using PFS. This may allow an effective drug to come onto the market sooner, which is in the interests of both patients and drug manufacturers. PFS is thus an example of what is known as a surrogate endpoint. The question of interest is thus whether the advantages of faster adoption of the drug outweigh the disadvantage that the surrogate outcome might not measure OS sufficiently accurately.