Glucophage AEs Not Comparable To Rezulin Liver Deaths - FDA Consultant

The risk of lactic acidosis associated with Bristol-Myers Squibb's Glucophage should not be compared to liver toxicities associated with Rezulin because the Glucophage side effects can be more easily managed through education, FDA consultant Saul Genuth, MD, Case Western Reserve University, told FDA's Endocrinologic & Metabolic Drugs Advisory Committee.

"I think there's a very important difference between the two," Genuth said during the March 26 review of Rezulin (troglitazone) adverse event reports. "Most of the cases of lactic acidosis due to metformin [Glucophage] are probably preventable - they've occurred under circumstances that we know increases the risk. I think education can hopefully get that risk down to near zero."

"I don't believe we can completely prevent liver toxicity on troglitazone because of the unpredictable nature of it...and I think there is no monitoring scheme that can be dictated from above that will be carried out perfectly to prevent liver death," he continued. "I think we're going to have to accept the fact that some of us are going to write prescriptions for patients and in rare circumstances that can lead to death or hepatic transplant."

Genuth's comments preceded the committee's vote to recommend restricted availability of Rezulin for use as a second-line treatment of type 2 diabetes and only in combination with other antidiabetics. The meeting was called after a series of articles highlighting liver toxicities and deaths associated with Rezulin use were published in the Los Angeles Times (¹ (Also see "Rezulin Patient Education Info Could Reduce Liver Failure - FDA Cmte." - Pink Sheet, 29 Mar, 1999.)).

The committee heard very different portraits of the relative safety of Glucophage and Rezulin during the meeting.

Warner-Lambert maintained that postmarketing data show Glucophage had a higher risk of death in its first year on the U.S. market, with 20 of the 1 mil. patients exposed to the drug experiencing fatal lactic acidosis. Rezulin was associated with 17 deaths in its first year after approval.

Warner-Lambert consultant Philip Home, PhD, University of Newcastle (U.K.), said that two deaths per 100,000 patient years could be attributed to Rezulin overall, compared to 0.8-9.0 for Glucophage. Sulfonylureas are responsible for 1.4-9.8 deaths per 100,000 patient years, Home maintained.

"All drug treatments carry some risk, but the risk is actually low, particularly when compared to the benefits," Home said. "The risk of troglitazone is comparable to established therapies."

FDA's analysis of adverse event data presented to the committee suggested the death rate associated with Rezulin is higher than that presented by the company.

According to data presented by Office of Postmarketing Drug Risk Assessment medical officer David Graham, MD, the rate of death due to Rezulin-associated liver toxicity could be as high as 1,200-1,700 per million patient years. The death rate associated with Glucophage-associated lactic acidosis is 15 per million patient years, FDA said.

Only two population-based studies assessing the risk of lactic acidosis in type 2 diabetics exist, Graham pointed out, one in patients taking metformin, and one in patients who did not. The data show that the risk of lactic acidosis per million patient years is almost exactly the same between the two groups, Graham said. All patients who were diagnosed with the toxicity also had "recognized factors capable in and of themselves of causing lactic acidosis," he noted.

"So the question arises, when we talk about tolerable risk with metformin, maybe what we should be thinking about is what is the risk of the study base, the base population, diabetics in general," the FDAer suggested. "It may be that people who use troglitazone have the same risk of lactic acidosis as this group of patients, type 2 diabetics who don't use metformin."

Out of 166.3 mil. prescriptions for sulfonylureas dispensed between April 1994 and October 1998, and 40 mil. prescriptions for Glucophage, none resulted in a liver transplant, according to United Network for Organ Sharing registration rates, Graham said. Three liver transplant registrations in that time were associated with Rezulin use, he reported, with only 7.9 mil. Rezulin prescriptions dispensed.

The disparity between comparative adverse event rates presented by FDA and the company apparently resulted from the level of under-reporting for Rezulin figured into the calculation.

FDA's Graham contended that reported Rezulin liver toxicities likely represent only about 10% of the actual adverse events, the approximate percentage reported to FDA across all products. Using that figure, Graham estimated that the number of serious liver toxicities, transplants and deaths since Rezulin's approval could be as high as 400.

When adjusted for 10% reporting, Graham said, the cumulative risk of acute liver failure at six-months of use is approximately 1:1,872. Warner-Lambert assessed the rate at closer to 1:45,000.

The company disputed the level of underreporting cited by Graham and maintained that publicity surrounding Rezulin-associated adverse events and release of "Dear Doctor" letters have likely resulted in almost 100% reporting of serious liver toxicities.

"Rezulin has been the subject of extensive publicity in both the lay and medical media, and a substantial proportion of this publicity has focused specifically on adverse liver events," Parke-Davis VP-Clinical Research Mark Pierce, MD/PhD, told the committee. "The labeling changes, 'Dear Doctor' letters and attendant publicity in the fall and winter of 1997 and in the summer of 1998 stimulated two peaks in adverse event reporting."

Other factors contributing to higher-than-average levels of reporting are the newness of the drug, the severity of the adverse event and the level of contact between physicians and Warner-Lambert sales reps, the company maintained.

"Rezulin is part of an intensely competitive market of oral hypoglycemic agents and is actively detailed," Pierce said. "Parke-Davis has several contacts each year with approximately 80% of physicians who issue 80% of all Rezulin prescriptions. These contacts have increased awareness of Rezulin and likely have led to an increase in reporting."

"It is also worth noting," Pierce continued, "that sales representatives from other companies do not hesitate to remind physicians of Rezulin's safety issues."

According to the Parke-Davis exec, the number of reports declined significantly following the two peaks, but the decline "was not due to declining drug use, since the number of patients taking Rezulin has continued to increase since launch." Pierce said that approximately 1.58 mil. diabetics have been prescribed Rezulin since the drug's launch in March 1997.

FDA's Graham again saw a different picture. According to Graham, reporting levels for Rezulin adverse events have increased steadily over the two years the drug has been marketed, but no jump in numbers has been seen at any specific time point.

Warner-Lambert's data was derived from a survey of 600 Rezulin prescribers. The survey found that 92% of physicians said they would report a liver failure leading to transplant or death, and 61% would report jaundice. Only 49% of physicians surveyed said they were likely to report reversible liver dysfunction.

During the open public hearing portion of the March 26 meeting, 12 endocrinologists spoke in favor of keeping Rezulin on the market, and several others submitted letters to the committee. The Endocrine Society also submitted a statement in support of continued availability, noting that Rezulin "is critical in the treatment of a segment of patients with type 2 diabetes and its continued availability is essential to this population."

Public Citizen's Health Research Group Director Sidney Wolfe, MD, noted in his presentation that similar testimonials were given during the 1977 review of phenformin, "also said to have unique advantages and whose withdrawal from the market was opposed by many diabetes experts." Phenformin (Ciba's DBI and USV's Meltrol) is the only drug to be withdrawn under FDA's "imminent hazard" clause.

Warner-Lambert said it is in discussions with the agency on how to implement the committee's recommendations for relabeling Rezulin.

The advisory committee is scheduled to review two new drugs that, if approved, will expand the "glitazone" class. The committee will meet April 22-23 to discuss Takeda/Lilly's Actos (pioglitazone) and SmithKline Beecham's Avandia (rosiglitazone) (² (Also see "SB Avandia, Takeda/Lilly Actos Antidiabetics Set For Advisory Cmte. Review" - Pink Sheet, 8 Mar, 1999.)).

Parke-Davis Senior VP-Clinical R&D Robert Zerbe, MD, maintained that it is too soon to start comparing the safety profiles of Actos and Avandia to Rezulin. "It is worth noting that rare events like lactic acidosis and liver failure are usually not identified until wide exposure in clinical practice," Zerbe said. "It would be inappropriate to compare and act upon safety profiles related to serious but rare adverse events between marketed drugs which have exposure in millions of patients, and those still in investigational stages which have exposures in thousands of patients."