Did You Know? - Deciphering Endpoints, Biomarkers and Surrogates
July 31, 2018
One of the primary aims of a clinical trial is to demonstrate effectiveness of the intervention being studied. This is accomplished through a process of establishing Endpoints for the trial and then recording observations and outcomes during the trial that provide valuable information about these endpoints. The result of the trial is a determination of how well the intervention fared and whether the data collected on the endpoints supports the trial’s hypothesis or not. Traditionally Clinical Outcomes have been the most reliable endpoints and successful trials are able to demonstrate a beneficial effect on the patient’s clinical status through the measurement of specific clinical outcomes that have a correlation to the patient’s condition. This includes clinical signs and symptoms of the disease e.g. occurrence of pain, breathlessness or seizures; impact on survival; and other aspects that provide a measure of how the patient feels or functions.
However, there are scenarios in which measurement of clinical outcomes may not be acceptable, feasible, or practical e.g. a primary endpoint of death, an extraordinarily long trial before clinical outcomes are detectable, or the need for an extremely large pool of subjects to allow gathering of sufficient supportive data due to low event rates, etc. In such cases, alternate measures may be utilized instead of clinical outcomes, thereby earning the term of Surrogate Endpoints. It is recognized that surrogate endpoints are measuring an effect as a substitute for what would have been an endpoint based upon a clinical outcome. An example of this would be using arterial blood pressure measurements as an endpoint for patients with hypertension as a surrogate for an occurrence of myocardial infarction or heart failure. Numerous surrogate endpoints have been utilized to-date including blood pressure, cholesterol levels, findings from modalities such as MRI, PET or EEG, and a wide range of biochemical biomarkers, among others.
So, what are biomarkers and how are they related to surrogate endpoints? In layman’s terms, the FDA defines Biomarkers as “defined characteristics that are measured as indicators of health, disease, or a response to an exposure or intervention, including therapeutic interventions.” Biomarkers are extremely useful in multiple ways -- as diagnostic agents, for identifying best treatments for an individual, for reviewing progression of a disease, and for monitoring safety of a therapy on an ongoing basis. While biomarkers provide a pool of eligible options to serve as one form of surrogate, they are not synonymous with the term and are not always an appropriate choice. To be considered as a Surrogate Marker, a biomarker should be “a laboratory measurement or physical sign that is used in therapeutic trials as a substitute for a clinically meaningful endpoint that is a direct measure of how a patient feels, functions, or survives and is expected to predict the effect of the therapy.” In other words, surrogate endpoints and markers should typically have gone through a scientific process of demonstrating their reliability in correlating to a clinical benefit. They are then considered to be a Validated Surrogate. An Unvalidated Surrogate, on the other hand, is one for which evidence has not yet established an actual benefit on the desired clinical outcome but where there is reasonable likelihood of this being true. Since they can be measured more easily and earlier than clinical endpoints, and often at lesser cost, e.g. measurement of tumor size in cancer patients versus overall survival statistics, there is great interest in using both validated and unvalidated surrogate markers as primary measures of effectiveness in investigational clinical trials. Trials based upon surrogate endpoints offer the advantages of reduced trial durations, smaller sample sizes, lower cost, and the ability to bring new innovations to market at a faster pace.
It is important to note however that there is still scope for mis-interpretation around the use of surrogate markers and surrogate endpoints. While this is especially true for unvalidated markers, even validated markers may be utilized in a specific context in one instance and then later be demonstrated to be ineffective, or indeed sometimes even harmful, due to additional effects that may not have been measured or observed in the original context. The CAST study on anti-arrhythmics highlights all too well the serious mistakes that can occur by making incorrect assumptions regarding surrogate markers.
Recognizing the need to enable the use of surrogate endpoints, when appropriate, while also ensuring that there are checks and measures in place to facilitate effective monitoring, the United States Food and Drug Administration (FDA) launched its Accelerated Approval program in 1992. This program provides a channel for faster approval using surrogate endpoints for novel therapies addressing serious conditions that fill an unmet medical need. Post-approval, these drugs are still required to go through phase 4 confirmatory trials using endpoints that demonstrate clinical benefit. If the confirmatory trial is unable to demonstrate clinical benefit to patients, approval for the drug may be withdrawn by the FDA. This process promotes innovation while still following through with effective monitoring to ensure safety.
Therefore, in yet another lesson from our research universe, we are reminded that even when it comes to trial design, there is a process of “personalizing” what will work best for a specific trial and to be able to do so in an efficient and repeatable manner, we must learn from collective wisdom gathered across multiple trials. As Mike Lauer, M.D., NIH Deputy Director for Extramural Research reminds us, “Some of us believe that the best way to deal with this problem is to conduct more trials that focus on clinical endpoints, but to conduct those trials in innovative, efficient, cost-effective ways. One way to do this is to design trials using pre-existing, large-scale databases or ‘registries’.” In the bigger picture, this would enable us to be more efficient on both fronts i.e. trials based upon clinical endpoints as well as those based upon surrogate endpoints, by creating an information pool that would enlighten both categories of trials, building upon the knowledge and strength that each of those routes produce, and enabling a more efficient process of bringing novel therapies and innovative solutions to market and into the hands of patients.