More and more biomarkers are being used in lieu of clinical outcome measures in order to accelerate the approval of desperately-needed drugs.
Legend has it that the first clinical trial was conducted around 600 B.C. by King Nebuchadnezzar II. The aim of his trial was to compare the effects of diet on the health of his royal family. So, he ordered family members to consume a restricted diet of meat and wine for a period of three years. However, after only a few days on this diet, four of the royal children requested that they be allowed to exchange the restricted diet for one of bread and water. The king allowed them to make the switch, but the rest of the family was asked to remain on the meat and wine diet. Ten days later, those individuals who switched to the bread and water diet reported that they felt well-nourished compared to those who remained on the restricted diet.
Clinical trials have come a long way since this simple experiment. Randomization, placebo groups, and multi-center trials are all part of modern-day trials. Another very important and crucial addition to clinical trial design over the years is the clinical trial endpoint, which can be a clinical outcome (e.g., relief of symptoms or prolonged survival) or a surrogate endpoint.
The FDA defines a surrogate endpoint as a laboratory measurement or physical sign that is used in therapeutic trials as a substitute for a clinically-meaningful endpoint. “The effect on the surrogate endpoint alone is not itself a clinical benefit to the patient, but if it is a valid surrogate, an effect on it predicts such a benefit (e.g., the way lowering blood pressure reduces the likelihood of a stroke, heart attack, kidney failure, or death). Biomarkers include a wide range of physiological, pathological, and even anatomical measurements. When a biomarker serves as a basis for approval, it is called a surrogate endpoint (blood pressure, cholesterol). Biomarkers can be many other things, e.g., identifying responders (tumor markers can predict who will be likely to respond to the breast cancer treatment Herceptin), or identifying people at high risk of an adverse outcome,” says Robert Temple, MD, director of the Office of Medical Policy and acting director of the Office for Drug Evaluation I, FDA Center for Drug Evaluation and Research, White Oak, Md. “A biomarker is not a surrogate endpoint unless [the FDA] is willing to base approval and marketing on it.”
The ultimate surrogate
According to Giora Feuerstein, MD, assistant vice president and head, Discovery Translation Medicine, Wyeth Research, Collegeville, Pa., there are 10 biomarkers currently approved by the FDA for use as surrogate endpoints. Some of these markers include LDL cholesterol, blood pressure, hemoglobin A (1C), glucose, insulin, human immunodeficiency virus type 1 messenger RNA, and CD4-positive T-cells. For example, drugs for hypertension are approved entirely on the basis of the blood pressure endpoint and statins are approved on their ability to the lower LDL cholesterol.
Low-density lipoprotein (LDL) cholesterol is an example of a laboratory measure whereas blood pressure would be considered a physical sign. And the FDA defines a clinically-meaningful endpoint as a direct measurement of a how a person feels, functions, or survives. “A surrogate is valid to the extent that an effect on it predicts such a clinical benefit,” says Temple.
Measuring LDL cholesterol as a surrogate endpoint for cholesterol-lowering drugs is a classic example. This is because lowering LDL cholesterol has been shown to reduce the risk of heart attack and stroke. “The biomarker of measuring cholesterol levels can be a relatively quick and inexpensive way to show that a [cholesterol-lowering] drug has benefit, without actually having to show that it reduces heart attack and stroke,” says Andrew Feigin, MD, an associate professor of neurology at New York University School of Medicine, New York, and a physician at the Feinstein Institute for Medical Research, North Shore LIJ Health System, Manhasset, N.Y.
Drug companies would love to base drug approval on a surrogate endpoint because the effect of a drug on a surrogate can be detected very rapidly in significantly smaller populations. “The FDA recognizes this, and their Accelerated Approval Rule allows use of less than fully validated surrogates to approve drugs for serious and life-threatening diseases with no good treatment. The rule allows drugs designed to treat diseases to reach approval sooner than if it had gone through the traditional route. In the case of drugs approved under this rule, the FDA will ask the company to perform studies post-approval to show the clinical benefit of the drug, i.e., improving cancer survival,” says Temple.
Just a biomarker
As discussed above, “biomarker” is a much broader term than “clinical surrogate.” And despite the fact that there are few biomarkers currently approved for use as surrogate endpoints, it does not stop some researchers from trying to validate more of them. And the FDA is completely supportive of such endeavors.
However, according to Feigin, “The big disadvantage of the biomarker is that unless you have proven beyond the shadow of a doubt that the biomarker is reflecting a clinical benefit, then people can always argue about whether the biomarker is useful or not.” Validation includes conducting lengthy, rigorous clinical trials. And this is surely the case for the biomarkers Feigin uses in his own research.
Feigin is interested in Parkinson’s disease (PD) and Huntington’s disease—degenerative neurological diseases in which there is a selective loss of neurons over time. Feigin uses positron emission tomography (PET) to measure the rate of that loss. By using PET imaging, he is able to quantify the number of these cells in PD patients.
“There has been some controversy about the use of imaging biomarkers in PD over the last five years that have really called into question their value,” says Feigin. “A biomarker is only useful when it has been firmly linked to beneficial clinical outcome.” However, he says some imaging biomarkers have started to be used in clinical trials for PD drugs.
“The issue is that symptomatic therapies, especially for PD, can make people look like they have milder disease while they are taking the therapy,” says Feigin. “So the idea of a biomarker would be to give you a measure of how severe a person’s PD is independent of how they clinically look at any given moment. That’s one advantage of using an imaging biomarker to assess efficacy of a PD drug.”
Feigin has usedPET biomarker to measure brain metabolism in a Phase 1 trial for a form of gene therapy called glutamic acid decarboxylase in 12 patients with PD. They got proof-of-concept where PET showed the changes in the brain expected from the therapy, a fact that led them to conclude it might be a useful outcome measure in the Phase 2 trial.
“A major advantage that I can think of for a biomarker in PD is that it could potentially provide a more objective and reproducible measure so that you might be able to show a benefit in a smaller population of patients rather than standard clinical trials in PD, which require dozens of patients (if it’s symptomatic therapy) or hundreds of patients (if it is a protective kind of therapy),” says Feigin.
Although biomarkers may not usually be used as endpoints for a drug trial, they can be used to confirm diagnosis and help distinguish PD from other diseases that mimic it. “So it can be used theoretically for screening patients for a clinical trial to make sure you are dealing with classical PD,” says Feigin. “And again I am interested in imaging to achieve that.”
In summary, the greatest benefit of the biomarker is that it can be used as a clinical trial surrogate endpoint. One look at LDL cholesterol’s reign as a surrogate endpoint says it all. And with the FDA on the side of approving more and more biomarkers as surrogate endpoints in the future, and drug companies and researchers working hard to get that approval, one can predict that surrogate endpoints will be utilized with increasing frequency in the future.
This article was published in Drug Discovery & Development magazine: Vol. 11, No. 2, February, 2008, pp. 30-32.
Filed Under: Drug Discovery