Surrogate markers not up to the task, studies find

December 11th, 2018

When cancer patients make decisions about their treatment, they want to know whether a certain drug will make them live longer or improve their quality of life. Clinical trials are meant to provide information on these outcomes, but in recent years, trials for new cancer drugs have focused more on “surrogate endpoints” and less on outcomes that matter to patients.

A “surrogate endpoint” is an outcome that is not clinically meaningful by itself, but correlates with a more important outcome. In recent years, the US Food and Drug Administration has approved more and more drugs based on surrogate endpoints rather than overall survival or quality of life. Using surrogate endpoints — such as “progression-free survival,” tumor size, or response rate — shortens the time it takes to conduct a clinical trial and bring a drug to market, which some argue is essential for patients whose lives are in imminent danger from diseases like cancer.

However, other oncologists contend that surrogate endpoints do not actually correlate strongly with the outcomes patients care about. In a recent article in JAMA Internal Medicine, a group of researchers from Canada, Japan, and other countries, conducted a systematic review of randomized cancer drug trials to find out whether there is an association between progression-free survival (a commonly used surrogate endpoint for cancer drug approval) and health-related quality of life. Their analysis of 38 randomized clinical trials found no significant association between the two measures. The findings “suggest that [quality of life] should be measured directly and accurately in future studies,” the authors write. 

Another recent study in the European Journal of Cancer provided an updated “umbrella review” of surrogate endpoint correlations. Researchers analyzed studies that evaluated how strongly certain surrogate markers correlate with overall survival. More than a third of studies showed a weak correlation between the surrogate marker and overall survival. Only 12% of studies found a strong correlation for all markers examined. Despite the increasing reliance on surrogate markers in cancer drug approval, this updated review shows that surrogate markers are no more useful in predicting better survival than in earlier years.

This research confirms what many clinicians have feared — that in approving the majority of new cancer drugs based on surrogate endpoints, we are sacrificing quality for speed.