When the Food and Drug Administration pressed for a recall last month of two diet drugs taken by millions, it cited data from five medical centers showing an alarming rate of heart-valve defects: 32% of the 291 patients tested.
But doctors at 21 other medical sites across the country say they're finding a far lower incidence of valve problems. The doctors performed tests on nearly 750 patients after the recall of Redux and Pondimin was announced, and they found possible valve leaks in just 57 of them -- a rate of only 8%.
The conflicting data raise the question of whether the FDA and American Home Products Corp., which markets Redux and Pondimin, might have acted prematurely in withdrawing the drugs. And a lower rate, if further research bears it out, would provide some reassurance to former diet-drug users and could reduce American Home's potential legal liability as it fights dozens of lawsuits filed in the days after the recall. The company declined to comment.
Like the FDA's study, the new batch of data isn't stringently scientific. Like the FDA, The Wall Street Journal simply surveyed doctors whose data were readily available, and came up with a sample that isn't necessarily representative of all patients who have taken the diet drug.
In addition, echocardiograms by their nature are subjective, and some doctors, fearing liability, may have been tempted to play down side effects when they reported their test results. Other investigators are pursuing more rigorous research that promises a more conclusive understanding of the diet-drug mystery.
Perhaps the best data to date will be presented next month at the North American Association for the Study of Obesity meeting in Cancun, Mexico, by Mehmood Khan, a researcher at Hennepin County Medical Center in Minneapolis. Dr. Khan, who has previously said he found abnormality rates "consistent" with FDA numbers in hundreds of patients, will present his most recent figures, as well as a much-awaited comparison with a "control" group of overweight patients who did not take diet drugs.
For the moment, the new data from doctors who have found lower abnormality rates offer some support to those who criticize the FDA's handling of the diet drugs. "It was premature for the agency to even try to derive anything statistically meaningful from its data," says Morton Maxwell, a physician and longtime proponent of diet-drug therapy who heads the obesity center at the University of California, Los Angeles. He says he has tested 25 of his patients without finding a heart problem. He also argues that an 8% rate may be an acceptable risk for severely fat people, who have a very high incidence of diabetes, hypertension and heart disease and whose obesity can be a greater risk for them than the drugs.
FDA officials defend the recall and express confidence that the 32% rate of heart-valve problems is correct. The original batch of 291 cases was provided by five clinicians, and they have since forwarded more test results that are consistent with the 32% risk, the FDA says. "We've done our job. We've done what is necessary to determine there is a problem," says the FDA's Robert O'Neill, director of the FDA's office of epidemiology. "Other information is up to [American Home] to find out."
Some doctors support the recall. A common refrain: Redux and Pondimin were prescribed for millions of people who weren't dangerously obese and therefore may have been running an unacceptable risk just to lose a few pounds. Altogether, six million people in the U.S. took one of the two drugs.
Jack L. Crary, an interventional cardiologist at MeritCare Medical Center in Fargo, N.D., was among the first to identify heart-valve abnormalities in diet-drug patients. He says problems are showing up in almost 30% of the echocardiograms he has performed, though he declines to disclose exact numbers.
Louis Aronne, an obesity specialist and clinical associate professor of medicine at Cornell University, adds: "The FDA was absolutely right in taking them off the market. Even if [the rate of side effects] is 2%, they did the right thing."
The new reports on 750 cases come from interviews with doctors who have been performing echocardiograms on diet-drug patients in the weeks since the recall. Many of these cardiologists are puzzled by why their tests are spotting a lower valve-trouble rate than the FDA first found.
"I think this [valve problem] is a real issue, but I strongly, strongly doubt the incidence is anywhere near as high as a third," says cardiologist Daniel Kulick of Mission Internal Medical Group in Mission Viejo, Calif. He and his colleagues have performed "echoes" on some 50 diet-drug patients, but only one "possibly" satisfies the FDA criteria for heart-valve leakage, he says.
"We're just not seeing it. I have talked to a number of my colleagues, and they aren't seeing it, either," says Washington, D.C., cardiologist Joel Rosenberg, who has performed heart tests on about 20 patients at George Washington University without finding any leaks. At Duke University, cardiologist Thomas Ryan reports testing about 28 patients and finding no problems.
In Tyler, Texas, obesity specialist Gary Huber has tested 92 patients and found leaking valves in only eight, or about 9%. Even in those suspect cases, he says, it isn't clear whether the trouble was caused by the diet drugs or by other risk factors associated with obesity, such as high blood pressure.
Redux, the first new prescription diet pill in more than two decades, was known to pose some risks when the FDA approved it in April 1996. It can increase the incidence of a rare but often deadly lung disease known as primary pulmonary hypertension. But the risk was considered low enough to be overshadowed by the higher risks of being overweight.
Few experts had expected a risk of leaky heart valves. Then last July, researchers reported problems in 24 previously health women who had taken "fen-phen," a two-pill combination of fenfluramine (sold as Pondimin and similar to Redux) and phentermine (sold under various names). At the time, American Home added a warning to its packaging while dismissing the findings as "a clinical observation, not a study." But alarm grew in August when the FDA said it had learned of 28 more cases. American Home mounted the recall on Sept. 15.
One reason the newer batch of data is finding a far lower rate of problem: The medical centers providing results may be testing patients who used the drugs for only short periods of time or in low doses. Some of the doctors conducting the tests aren't the same ones who prescribed the drugs, so some patients who were tested may have been on drugs other than the two that were recalled. One theory gaining currency: The heart-valve problems may fade away quickly after the drugs are stopped.
The biggest reason for the discrepancy may be the high degree of subjectivity in interpreting echocardiograms. The tests produce beating-heart images by bouncing high-frequency sound waves off the heart to get a fuzzy picture of the organ. Blood flow in and out of the heart is displayed on a video screen in different colors, allowing an examiner to determine whether leakage is occurring.
Leakage is roughly classified as trivial, mild, moderate or severe, but the definitions can vary among doctors. Some critics argue the FDA used an overly broad definition of what constitutes a valve abnormality, including people with only "mild" aortic-valve leaks that might normally be classified as nothing to worry about.
"No echocardiographer in his right mind would read [mild aortic leakage] as a problem," contends T. Anthony Don Michael, a clinical professor at the UCLA School of Medicine, who has performed echocardiograms on 70 diet-drug patients without finding any significant abnormalities.
But Dr. Michael also says the cumulative doses of fen-phen his patients received were about half of those commonly prescribed, and that higher doses may very well result in valve leaks.
The FDA still says that the consistency of results from the original five centers suggests the findings are on target. "If [the doctors] were using different definitions, then how could we get such consistent findings" from each of the sources providing data for the study, asks the FDA's Mr. O'Neill.
Another doubt: It isn't clear exactly what portion of the general population -- particularly among the overweight -- typically has leaky heart valves. That makes it harder to know what rate would be out of whack among diet patients. Some doctors estimate that only 1% to 5% of the general population has leaky valves. But others believe the frequency may be higher than that -- especially in obese people, regardless of whether they have taken diet drugs.
Moreover, the FDA never independently evaluated the echocardiogram reports from the five clinicians. "The science is a muddle," says Nelson Schiller, a professor and echocardiography expert at the University of California at San Francisco. The FDA responds that it lacks the expertise to independently evaluate echocardiograms.
Ultimately, only formal studies will determine for sure whether a link between the diet drugs and heart-valve disease exists. American Home is sponsoring a wide-scale study comparing the hearts of hundreds of Redux and Pondimin users to control subjects of similar ages and weights. In this study, heart-valve images will be read by doctors who don't know which subjects received the drugs and which didn't, eliminating the possibility of bias. The results aren't expected for several months.
Copyright © 1997 Steven J. Milloy. All rights reserved. Site developed and hosted by WestLake Solutions, Inc.