Surrogate End Points and Their Validation in Oncology Clinical Trials

A new drug must show effectiveness, usually derived from well-designed and sufficiently powered phase III clinical trials, to receive regulatory approval from the US Food and Drug Administration (FDA). 1 The effectiveness of cancer drugs is demonstrated by prolonging survival and/or improving patient symptoms and quality of life. Overall survival (OS), defined as the time from random assignment to the date of death from any cause, is objective, precise, and easy to measure and interpret, and hence hasbeenconsideredthemostreliable andclinically meaningfulend point for evaluating drug efficacy in oncology clinical trials. This end point, however, requires a large sample size and prolonged follow-up, which not only make such trials expensive, but also create the risk of the drug under investigation becoming obsolete by the time the trial is complete. In addition, OS is potentially confounded or diminished by effective postprogression therapies (including crossover). These limitations of the OS end point are the primary motivations for the use of surrogate end points in oncology clinical trials, which are particularly important for rare diseases and diseases with effective subsequent-line therapy, especially in this era of personalized medicine. What is a surrogate end point? The National Institutes of Health defines a surrogate end point as “a biomarker intended to substitute for a clinical endpoint.” 2(p91) TheFDAconsidersasurrogateendpoint of aclinicaltrialtobe“alaboratory measurementor physicalsignused as a substitute for a clinically meaningful endpoint that measures directly how a patient feels, functions, or survives and that is expected to predict the effect of the therapy.” 3(p13235) Compared with a clinical end point, a surrogate end point can usually be measured earlier and requires a smaller sample size and a shorter follow-up time. In oncology, biomarkers measuring a drug’s biologic antitumor activity, such as objective response rate (ORR) 4,5 and progression-free survival (PFS), 5-8 have been proposed and evaluated as surrogate end points in clinical trials. Before a surrogate end point can replace a clinical end point for evaluating an experimental treatment in a phase III clinical trial, it must be formally validated to show that the treatment effect on the surrogate end point reliably predicts the treatment effect on the clinical end point. In a landmark paper, Prentice 9 proposed a formal statistical definition and developed a formal statistical methodology to assess surrogate end points in clinical trials. Prentice 9 defined a surrogate end point as “a response variable for which a test of the null hypothesis of no relationship to the treatment groups under comparison is also a valid test of the corresponding null hypothesis based on the true endpoint.” 9(p432) On the basis of this definition, a biomarker must simultaneously meet two conditions (ie, operational criteria) to be a valid surrogate end point. 9 First, the biomarker must be correlated with the clinical end point (“it is necessary for the surrogate to have some prognostic implication for the true endpoint”). 9(p434) Second, the biomarker must fully capture the net effect of the intervention on the clinical end point (“require thesurrogatevariabletobefullysensitiveto any treatmentdifference in true endpoint”). 9(p434) Prentice’s 9 criteria for surrogate end points