This JAMA Guide to Statistics and Methods reviews how and under what conditions surrogate outcomes can replace patient-centered outcomes in randomized trials and stresses the importance of properly validating outcomes as surrogates for direct measures of patient experiences.
Randomized clinical trials have a long history of success in many medical arenas. Many trials that change clinical practice use clinical outcomes that are direct measures of how a patient feels, functions, or survives. The substantial resources required by trials using such end points are powerful incentive to pursue designs that reduce the numbers of patients required, the length of follow-up, and the trial costs.1
WHAT ARE INTERMEDIATE OUTCOMES AND WHY ARE THEY USED?
Patient-centered outcomes—direct measures of how a patient feels, functions, or survives—often reflect the effects of multiple factors, reducing the expected treatment effect and thus increasing the required trial size. To reduce the trial resources, a frequent approach has been to use a biomarker or another replacement end point that is an intermediate outcome thought to capture the causal pathway through which the disease process affects the patient-centered outcomes. Intermediate outcomes, which may be physiological measures, laboratory test results, imaging results, or other such measures, are appealing because trials that use these outcomes are shorter, smaller, and statistically more powerful than those that evaluate patient-centered outcomes.
WHAT ARE THE LIMITATIONS OF INTERMEDIATE END POINTS?
Despite the appeal of using replacement end points, two fundamental requirements must be met2 to ensure that the replacement end point is a “valid surrogate,” ie, the effect of the intervention on the replacement end point reliably predicts its effect on the patient-centered outcome. The first requirement is that the replacement end point and the patient-centered outcome are strongly correlated. The second is that effects of an intervention on the replacement end point should fully capture its net effect on the patient-centered outcome. There are several reasons this second requirement is often not met; thus, “a correlate does not a surrogate make.”3
HOW HAVE INTERMEDIATE OUTCOMES BEEN USED?
An article published in JAMA Oncology4 by Ritchie and colleagues illustrates these issues in the immuno-oncology setting, where checkpoint inhibitors have frequently been evaluated for the treatment of advanced solid cancers. The objective response rate (ORR), determined from decreases in tumor size, is often thought to be a valid surrogate for overall survival simply because, on a patient-specific level, treatment responders live longer than nonresponders. The article evaluated the primary end points in 24 randomized controlled phase 2 trials of checkpoint inhibitors. Across these trials of checkpoint inhibitors, the correlation between the observed ORR odds ratio and the overall survival hazard ratio was modest (0.57 [95% CI, 0.23-0.89]). Ritchie et al4 suggest avoiding ORR as a primary end point in phase 2 trials of checkpoint inhibitors.