Scientists Plan Faster Approval for New Drugs
WASHINGTON — Scientists are drafting an ambitious blueprint for evaluating medical research that could take years off the time it takes to win approval for new drugs, according to researchers involved in the effort.
While still in its earliest planning stages, the new and sometimes controversial approach could have profound implications for cardiovascular disease, rheumatoid arthritis, osteoporosis, multiple sclerosis and certain cancers, as well as other illnesses.
Dr. Harold Varmus, director of the National Institutes of Health, has convened a working group of scientists to explore the possibilities of this new direction in research, which he thinks shows “real potential.”
Dr. Daniel Hoth, a former AIDS and cancer specialist for the NIH and now a San Francisco-based consultant working on the project, agreed that “this could transform drug development.”
In recent years, as Americans have been clamoring for faster access to new medicines, Congress and federal regulators have initiated procedural changes and bureaucratic streamlining to accelerate drug approval.
But the blueprint goes beyond regulatory reform to the heart of science, putting in place a fresh look at how drugs work.
Like blood pressure lowering medications that ultimately prevent stroke, or drugs that increase a woman’s bone density, reducing her risk of later developing fractures, the new effort identifies early signs of a drug’s eventual effect. Called “surrogate markers,” these are harbingers of a successful treatment or even a cure that scientists and regulatory agencies can use to predict a health benefit rather than waiting for lengthy studies to document it.
This surrogate marker approach to drug evaluation differs significantly from traditional, often years-long clinical trials, which typically involve studying drugs until their usefulness is proved--or disproved--definitively.
The new concept already is producing results. Increasing bone density is a proven surrogate marker for osteoporosis. Lowering cholesterol is known as a surrogate marker for coronary artery disease. And reducing the level of HIV in the blood is a surrogate marker for AIDS. Only last week, AIDS dropped off the list of the 10 leading killers in the U.S., the result of powerful drugs approved using this new approach--drugs that, under traditional trials, might only just now be coming onto the market.
Focusing attention on this method, Varmus believes, means that benefits “will soon be felt for others.”
And the Food and Drug Administration, the agency that ultimately decides whether to license a drug based on research evidence, itself is exploring the surrogate marker concept. In fact, the FDA reform law passed by Congress last year directs the agency to actively look at surrogate evidence when reviewing new drugs.
Scientists Say They Are ‘Building Momentum’
Scientists involved in the project plan to examine different diseases over the next year, “selecting the conditions that are ripe--or nearly ripe--for surrogate analysis,” Varmus said.
“We’d like to identify the candidates, confirm the [surrogate] relationship and, later, begin to apply them to drug evaluation,” Hoth said. “Right now, we’re building momentum--building steam and interest in this effort.”
The possibilities are endless.
“With the new advances in magnetic resonance imaging, you actually can see areas of disease in the brain of those with multiple sclerosis,” Hoth said. “You can see areas of plaque, which is associated with the progression of the disease. Suppose you found a drug that could reduce the plaque, and you could [see] that with an MRI. That may predict a long-term health benefit.”
In rheumatoid arthritis, a crippling joint disease that afflicts an estimated 2.1 million Americans, Hoth said, MRI imaging of the joint also could provide early signs of a drug’s effect on joint inflammation, thus predicting an improvement in symptoms.
In prostate cancer, certain substances or markers found in the blood that help diagnose cancer--elevated levels of prostate-specific antigen, or PSA--also might be worth exploring as possible surrogate markers, Hoth said.
But the PSAs would work as surrogate markers only if experimental drugs changed their levels and only if the changes predicted future health improvement, he said.
“The biomarkers have to get worse when the disease gets worse and they have to get better when the drug is given,” he said. “And these changes have to precede the actual clinical benefit.”
Most experts warn that, even if the surrogate approach is successful, it will not necessarily work for every drug or every disease and it will not replace the old system of large-scale efficacy trials, which are considered the gold standard of drug research. And it could involve unexpected risks.
“On the surface, it sounds very appealing to get these drugs out there faster based on biologic activity, but it can be far more complicated to determine whether that activity is accurately predicting the intended effect,” said Dr. Thomas R. Fleming, chairman and professor of biostatistics at the University of Washington.
Some medical experts view tumor shrinkage as a surrogate marker for new cancer drugs, for instance, but Fleming said that reducing the size of tumors does not always mean longer lives for cancer patients. “Lots of agents can reduce tumors to clinically undetectable levels, but, in many people, death will occur just as quickly or tumors will come back.”
The real risk of relying on surrogate markers--and the reason they are controversial--is simply the possibility that they may not work, critics say.
In the late 1980s, cardiovascular researchers assumed that two drugs known to control dangerous cardiac arrhythmias (irregularities in the heart rhythm) could keep those who had already suffered heart attacks from having subsequent life-threatening heart attacks.
They were wrong. The drugs actually provoked fatal heart attacks in an unexpected number of individuals, and the study was stopped.
To be sure, the drugs had never been approved for that purpose--the unexpected deaths occurred during a standard clinical trial--but the logic was the same and was regarded as sound at the time.
“Sometimes you aren’t likely to understand the full range of the effects of the drugs,” said Dr. Henry Masur, an NIH researcher who chaired an FDA antiviral advisory panel during the period when AIDS drugs were first being evaluated based on surrogate marker effects.
“You really have to take it disease by disease and class of drugs by class of drugs,” he added. “Every disease needs to be evaluated individually. It will depend on how rational and well-linked the surrogate is with the clinical end-point.”
Researchers Defend Project’s Risks
But Dr. Robert T. Schooley, an AIDS specialist at the University of Colorado Health Sciences Center, insisted that surrogates are worth their risks. “There are always dangers when you try new things, but you try to balance the potential benefits with the risks. If you take no risks, you make no advances.”
Schooley’s belief--and that of others involved in the project--is based on the recent successful experience with AIDS drugs, which have been serving as the model for the research direction.
In 1991, the surrogate markers concept was used to speed experimental AIDS drugs onto the market. The drugs were approved based on certain laboratory results--increases in immune system cells and decreases in the amount of virus in the blood--without knowing in advance whether these results would translate into health benefits.
The gamble paid off, and “now we’re trying to apply some of those lessons learned in AIDS to other diseases,” said Schooley, who is participating in the NIH discussions.
The new effort differs somewhat from the AIDS approach, however, in that researchers are seeking to establish more ironclad surrogate relationships in advance. With AIDS, researchers moved more hastily because they thought the typical outcome of the disease--almost certain death--made use of the surrogate approach worth the risk.
More Known About Causes of Disease
Advocates of the new approach stress that their aim is to establish confirmed surrogate relationships before moving ahead with drug approvals, to be sure the drugs are not otherwise toxic and to continue to monitor the drugs’ efficacy.
They emphasize that today much more is known about the underlying mechanisms of disease--joint inflammation and the symptoms of arthritis, for example--that can be used to predict health outcomes.
“We simply know more and better about what causes many of these diseases and how these relationships work,” Schooley said.
And new technology, such as the MRIs and more advanced lab tests, have enabled scientists to detect and measure earlier changes that would have been impossible or extremely difficult in the past.
In the late 1980s, for example, as the AIDS epidemic burgeoned, laboratory tests were far less sophisticated than they are today. AIDS researchers could not accurately measure the amount of AIDS virus in the blood, certainly not as well as they can now. Thus, reducing “viral load” as a predictor of prolonged survival was not possible.
Today, AIDS specialists know that a drug which can drive the virus to virtually undetectable levels in the blood means a longer and better life for those with HIV.
“We know a lot more now than we did a decade ago about how diseases are caused and what some of these relationships are,” Schooley said, stressing that it is important whenever possible to move away from traditional clinical trials in which death was often a measure of whether a drug worked. “Because we know more, we want to use that knowledge to help us make important decisions--when it is appropriate--before the bodies pile up.”
