Solid Ground
by Lisa Zamosky
January 2007
Soft endpoints in CNS drug trials frustrate sponsors and deprive patients of potential treatments. The search is on to find reliable measures.
Central nervous system (CNS) drug development currently is the second–largest therapeutic segment of the pharmaceutical market, and it continues to grow at a fast pace. However, this area is one of the most challenging and risky propositions for pharmaceutical companies, due partly to the lack of hard endpoints to conclusively demonstrate a new drug’s efficacy.
Patients with CNS disorders—such as depression and schizophrenia—tend to present with highly subjective symptoms. This makes diagnosis, particularly of psychiatric disorders, difficult, since it is based on the presentation of symptoms. For example, there are no diagnostic tests to distinguish the onset of bipolar disorder from an initial presentation of depression due to similar symptoms, according to the Critical Path Opportunities List published by the U.S. Food and Drug Administration (FDA) in March 2006.
For psychiatric and many neurological disorders, the severity of the illness almost always is determined using both paper–and–pencil rating scales and a clinical assessment. The efficacy of most CNS drugs is determined through the use of such scales in clinical trials.
Patient, Rate Thyself
Although existing measurements offer indications of a patient’s improvement—or lack thereof—in response to drugs, they possess inherent problems. “Because these rating scales represent asking patients a series of questions about their view of how they are feeling, they are very subject to bias and distortion,” says Dr. Bill Potter, vice president of clinical neuroscience at Merck Research Labs, North Wales, Pa., USA. “How you rate a person on these scales depends a lot on the circumstances, or how you ask the question. This can sometimes be like leading the witness.” Unfortunately, these scales, which have been developed over the last 40 years, currently are the most effective way to measure drug efficacy in diseases such as depression, Potter says.
The impact of soft endpoints on the development of new drugs is often significant, in terms of both the costs of conducting the clinical trials, and the resources necessary to analyze the data. “Even with drugs that are known to work on depression, if you do new studies on hundreds of patients, [often] you can’t tell the difference between drugs that you know work and placebo,” Potter says. The lack of solid efficacy data due to the insensitivity of the ratings instruments can cause clinical trials to fail even in the later stages of development—a costly end to what may have been a promising therapy.
“Poor endpoints make it harder to show that a treatment is really being effective,” says Dr. Pedro Delgado, professor and chair at the department of psychiatry, and associate dean for faculty development & professionalism at the University of Texas Health Science Center’s School of Medicine, San Antonio, Texas, USA. According to Delgado, the subjectivity of existing endpoints can make it appear as if ineffective treatments are working. “It happens all the time that a drug is perceived to have a signal, but the pharmaceutical company gets timid when they have a failed trial,” he says. “They realize that in order to mount the next trial, a huge investment is needed.”
However, trying to avoid a placebo effect by conducting a trial with more patients may not be the answer. The main problem with CNS drug trials doesn’t so much rest with the scales used to determine drug efficacy, but rather with the way in which the protocols are being written, according to Dr. John Marler, associate director for clinical trials at the U.S. National Institute of Neurological Disorders and Stroke (NINDS), Bethesda, Md., USA. “Too much is blamed on the scale that is chosen,” he says. “We should focus more on whether wrong doses of drugs are being used; if the blood–brain barrier is not being crossed; if we’re treating too late; if we’re too selective of patients in the study or excluding too many patients from the studies.” Trials often are designed with the wrong sample size, and too great an effect is sought, Marler says. “People are looking for penicillin–like effects in small populations.”
Risky Business
Because conducting clinical drug trials is a major investment for pharmaceutical and biotechnology companies, the failure of drugs in late–stage trials can make bringing CNS–related treatments to market cost prohibitive.
Many drugs that show promise in early trials fail in a subsequent trial, which can stall the progress of drug development for years. The field of psychiatry has done a great job of optimizing clinical benefits from existing drugs, Potter says, but it’s unclear if “for the major psychiatric illnesses we have made major improvements over what we had 40 years ago.”
When studying a drug that ultimately may have great potential for patients, the drug often carries a risk of severe side effects, Delgado says. In such cases, the challenge is to determine whether the potential benefit to the patient outweighs the potential safety risk. “Without solid endpoints, confidence is diminished, and then it becomes a risk decision for the company,” Delgado says.
The error often lies in oversimplified outcomes, rather than unreliable endpoints, Marler says. For industry sponsors and investigators, accepting the risk and proceeding with a drug’s development may hinge upon an alternate trial design or analysis. “More creative ways of analyzing the data could be used,” Marler says.
On the Horizon
Researchers are embracing several developing areas that may dramatically improve the ability to measure drug effectiveness in CNS disorders. In the field of epilepsy, for example, significant progress has been made, and may provide a model for other illnesses. In epilepsy studies, patients who have a poorly controlled condition often are introduced to a new drug in addition to the medication they are already taking. “You put a new drug on top of the old and see if there is a change,” Potter says. Researchers are just beginning to apply this method to schizophrenia, “but nobody has brought a drug forward yet with this approach,” he says.
New measures will need to be developed and agreed upon within the scientific community, Potter says. “If we can show effect on the existing rating scales, we might be able to introduce new drugs,” he says. “Because good scales have not been available, it’s been hard to introduce any truly new drug.”
Clinical researchers are also aiming to establish biomarkers as surrogate endpoints for CNS clinical trials, as well as for diagnostic purposes. Brain imaging is one such area, and although its usefulness in determining the efficacy of drugs on various CNS–related illnesses has yet to be determined, it has shown promise in areas such as multiple sclerosis. “[Brain imaging] is beginning to stratify patients to see if we get better predictions of outcomes,” Potter says. “We’re now hoping to use brain imaging to identify [target] populations.”
By providing visibility of brain abnormalities, imaging also shows some promise for patients with Alzheimer’s disease, Potter says. “If we can get it to give us information about severity, that would be the hard measure we need.” Some combination of brain and biochemical measures will more accurately identify the severity of illnesses, he says.
However, such an approach is years away for the major psychiatric disorders, and not everyone agrees that biomarkers should be used as surrogate endpoints, though the establishment of biomarkers is encouraged by the FDA. “Unfortunately, what it takes to establish high levels of sensitivity and specificity can be a much bigger effort than the effort required to do a clinical trial,” Marler says. In order to utilize brain imaging as a surrogate endpoint, it would be necessary to specify the measure’s sensitivity for the particular population in question, Marler says, and the method would have to be highly sensitive and specific for an accepted outcome. “It’s an easy concept to think of, but labor–intensive to establish a biomarker or create a truly credible outcome.”
All in the Family?
Areas involving genetic and molecular techniques also are being studied. There is possible long–term hope for developing CNS biomarkers in the mapping of the genome, Delgado says. With this kind of scientific study comes the possibility of comparing polymorphisms in the gene. These small differences detected within the gene can be associated with differences in behavior and responsiveness to medications. By identifying uniqueness in individual genes, “we may be able to account for side effects and understand why some people respond better to certain medications than others,” Delgado says.
A similar approach is gene expression, which enables researchers to see which genes are “turned on” and which are “turned off” once a patient has been given a certain drug. The hope with gene expression, Delgado says, is that by turning on certain genes, a correlation with a response to the drug may be made. Although there has been a huge amount of activity in genome mapping and gene expression, the science is at least a decade away from being fully realized for both fields of study, he says.
“There is a lot we can be doing right now, and I don’t want to wait 10 years for effective biomarkers or surrogates,” Marler says. “All of our basic research points to the fact that neurological diseases can be treated, but it will take a lot of hard work, and we need to learn to work effectively together in groups and to involve neuro–clinicians in the process.”
Marler retains and advocates a positive outlook for the future of CNS trials. A combination of approaches—from newer ratings or biomarkers to more focused trial design—may ultimately be the driver to more successful CNS drug development.
Send us Your Comments
Web Exclusives
- Statistical Advantage
-
Gary Shorter, director of biostatistics for Quintiles, talks to Envisage editor Adam Istas about the growing importance of incorporating biostatistical analysis into all stages of drug development.
- Measure for Measure
-
A proposed Health Care Comparative Effectiveness Research Institute in the United States may force the biopharmaceutical industry to emphasize patient outcomes and quality of life issues when developing a new medicinal product.