Rezulin’s Effect on Heart Was Also Seen as Concern
Yet another unsettling chapter in the rise and fall of the diabetes pill Rezulin is buried in government files: The unresolved question of whether the drug could contribute to heart failure.
Three years ago, the Food and Drug Administration approved Rezulin while ignoring warnings of danger to the liver. It was not until last week--after thousands of liver injuries and deaths--that the FDA announced withdrawal of the drug.
Now, previously undisclosed documents and interviews show that concerns also were raised about Rezulin’s effect on the heart.
The FDA “should have delayed approval [of Rezulin] until all the questions were addressed,” said Guston Turner, a pharmacist from the FDA’s scientific-investigations division who had found irregularities in research measuring Rezulin’s effect on the heart.
The concessions made by FDA officials as they granted “fast-track” approval to Rezulin on Jan. 29, 1997, demonstrate what can happen when the agency’s commitment to speed vies with scientific prudence.
The FDA’s handling of the liver and heart dangers of Rezulin promises to focus a renewed national debate over Congress’ demand for ever-faster reviews of new prescription drugs.
“I would hope that this unfortunate experience with Rezulin could be the basis for an oversight investigation, if not a [congressional] hearing, so that we can try to learn from this,” said Rep. Henry A. Waxman (D-Los Angeles).
The concern over heart effects also underscores questions remaining about Avandia and Actos, two diabetes drugs that, while similar chemically to Rezulin, were last week termed “safer alternatives” by the FDA.
An FDA spokeswoman, Laura Bradbard, said the agency would not comment. The FDA announced Friday that it will present in more detail its rationale for withdrawing Rezulin, manufactured by Warner-Lambert Co., at a public meeting in late May.
Still unknown is how many Rezulin users died or were injured since the drug for adult-onset diabetes was introduced in March 1997.
The FDA has linked 63 liver-failure deaths to the diabetes pill, but this total is based on voluntary reports that reflect 1% to 10% of actual fatalities, experts say.
*
As far back as the mid-1990s, FDA pharmacologists were struck by what they saw in animals given doses of Rezulin: Discolored, overweight hearts.
“These changes were drug-related, and were responsible for the early mortality in both sexes,” wrote two FDA pharmacologists, Ronald W. Steigerwalt and Herman M. Rhee.
The findings were consistent with those from other early animal research of Rezulin and Actos, which are of the same fledgling chemical class, nicknamed the “glitazones.”
“In rats, in dogs, in monkeys--I’ve never seen a class of drug that had such a consistent pattern of cardiopulmonary toxicity,” recalled Dr. John L. Gueriguian, a retired FDA medical officer who in the 1990s studied the glitazones and Rezulin, in particular. He recommended its rejection.
However, Gueriguian noted that a drug’s effect on an animal does not necessarily predict how it will impact a human being.
And for this reason, the team of FDA specialists studying Rezulin as it approached approval in late 1996 wanted to see whether the drug adversely affected the heart functions of humans.
“My primary concern about [Rezulin] is related to its potential for cardiac toxicity,” wrote an FDA diabetes specialist, Dr. Robert I. Misbin, in an Oct. 18, 1996, e-mail to Gueriguian.
In a Dec. 5, 1996, e-mail to another colleague, Misbin said the specialists reviewing Rezulin were seeking “some measure of assurance that the cardiac events in animals were not also observed in patients.”
Answering this question before Rezulin got onto the market was particularly important: The diabetic patients who would be treated are often overweight and at far higher risk of developing congestive heart failure.
An estimated 15 million Americans have adult-onset diabetes, a condition characterized by high blood-sugar levels. There are several classes of drugs that treat the disease. Some diabetes pills lower blood-sugar levels by stimulating the secretion of insulin; the glitazones can improve the body’s overall use of insulin.
But, if Rezulin or another glitazone caused abnormal fluid retention and weight gain, this would only worsen a delicate preexisting vulnerability to heart failure.
Yet outside the controls of a clinical trial, it would be far more difficult to distinguish whether problems observed in diabetic patients were caused by Rezulin or a preexisting malady.
All of which elevated the significance of one particular clinical trial, as FDA specialists faced a short deadline to finish their fast-track review. Known as the “Echo Study,” short for echocardiogram, this 1994 Warner-Lambert trial assessed Rezulin’s impact on heart functions from five sites: Buffalo, N.Y.; Omaha, Chicago, St. Louis and Pittsburgh.
The primary objective of the Echo Study was to determine whether use of Rezulin for 48 weeks would result in a change in the left ventricle, the thick-walled chamber of the heart that pumps blood through the aorta. Changes to the left ventricle were of interest because heart failure can result if it cannot contract forcefully enough.
The Echo Study was structured so that only patients with relatively normal heart function were enrolled--a profile that experts say is not reflective of the larger diabetic population. Of the 114 who completed this study, none suffered heart failure. But when the FDA scrutinized the data, they found more questions than answers. A high number, 26%, of the original 154 patients dropped out.
Suspecting possible irregularities in the data, the FDA in early January 1997 dispatched Turner, the veteran agency pharmacist, to Buffalo and Omaha. Turner had little time to spare: The agency had only about three weeks remaining to complete the review of Rezulin.
After digesting Turner’s promptly reported findings, Dr. Solomon Sobel, director of the FDA’s endocrine-drug division, voiced concern. Sobel wrote in a Jan. 13, 1997, e-mail that he found it “disturbing” that variations existed in how different Warner-Lambert consultants had interpreted the same echocardiogram data recorded at Omaha.
Turner also found cause for concern at Buffalo: The echocardiograms indicated an increase in left-ventricular-wall thickness.
But in both cases, senior FDA officials made concessions in Warner-Lambert’s favor. The officials decided that the company’s “central reader” of the echocardiograms was better qualified than specialists at Omaha to interpret the data from that site.
As for Buffalo?
“This site’s data was excluded due to poor local technique,” according to the FDA’s minutes of a Jan. 16, 1997, meeting with Warner-Lambert executives. “FDA attendees acknowledged that [Warner-Lambert] may have had ‘bad luck’ with this site selection. It was also noted that [Rezulin] patients in the study had an increase in LVM,” shorthand for left-ventricular mass.
Reached in Buffalo, the endocrinologist who served as Warner-Lambert’s principal investigator for the Echo Study, Dr. Paresh Dandona, said he had delegated much of the work to heart specialists. Dandona said these specialists deviated from the Echo Study procedures while studying up to a third of the patients.
“The Echo Study was not properly conducted, according to the standard protocol,” Dandona said in an interview. He added: “I was not made aware of the fact that the patients from this site, in particular, for whatever reason, whether it was due to technique or whatever, that they had increased left ventricular mass.”
Senior FDA officials concluded that the conduct of the study was flawed. A senior Warner-Lambert executive who oversaw the clinical trials conducted in support of Rezulin’s approval, Dr. Randall W. Whitcomb, declined to be interviewed last week.
When the company presented its case for Rezulin to an FDA advisory committee on Dec. 11, 1996, it stated: Rezulin “does not increase cardiac mass or impair cardiac function.”
When the FDA granted approval to Rezulin on Jan. 29, 1997, officials negotiated a nonbinding pledge from Warner-Lambert to start a new study, designed to assess the drug’s effect on patients with preexisting evidence of heart disease.
But few patients were enrolled, and the study was never completed.
Meanwhile, scores of patients who took Rezulin have died of heart failure, according to reports filed voluntarily with the FDA. But doctors say that because so many of these patients had preexisting heart problems or were simultaneously treated with a battery of other drugs, it would be difficult to prove that Rezulin was the culprit.
Looking back at the handling of the Echo Study, Turner, the now-retired FDA pharmacist, said, “The real problem there was whether there was any heart enlargement or not. They ignored it.”
*
Times researcher Janet Lundblad in Los Angeles contributed to this report.
