MILES' GLUCOBAY FOR DIABETES NOT RECOMMENDED FOR APPROVAL: CMTE. CITES LACK OF CLINICAL EFFICACY AND LONG-TERM USE DATA AS WELL AS LIVER TOXICITY POTENTIAL

Miles' Glucobay (acarbose) alpha-glucosidase inhibitor for diabetes failed to receive a recommendation for approval by FDA's Endocrinologic & Metabolic Drugs Advisory Committee on July 25. The seven voting members of the advisory committee unanimously voted "no" to four safety and efficacy questions relating to Glucobay and cited questions about the lack of clinical efficacy data, long-term use data and the inadequately-explored hepatotoxic potential of the drug. According to Miles, Acarbose modulates the digestion and absorption of dietary carbohydrates, inhibits postprandial glycemic response to ingested food, causes no increase in insulin secretion and decreases the daily insulin intake requirement of diabetics. Miles said the drug works in the small intestine to competitively inhibit the binding of complex carbohydrates by occupying enzyme binding spots. FDA agreed with Miles that the drug's mechanism of action is "unique." Robert Josse, MD, St. Michael's Hospital, Toronto, a consultant to Miles, said "the whole action of this drug has been succinctly suggested as converting a gorger into a nibbler." Josse said acarbose can mirror the clinical effect in diabetics "who are prepared to take up to 15, 16, 17 meals a day" to equalize their absorption of glucose. Miles came to the advisory committee seeking a broad label for Glucobay both in insulin and non-insulin dependent diabetics and presented data on more than 1,400 patients from six short-term, double-blind, placebo-controlled studies that measured acarbose's effect on glycated hemoglobin (HbA1-c) as their primary efficacy endpoint. Secondary efficacy endpoints measured by Miles were the number of hypoglycemic incidents, weight loss (in the case of Type I), and the plasma glucose concentration measured after fasting, postprandial and Cmax and RISE values. Although acarbose patients showed a consistent improvement in postprandial hypoglycemia, none of the other endpoints consistently showed a statistically significant response to the drug. Miles had intended the effect of acarbose on lipid levels to be another primary efficacy endpoint, but later changed to using only HbA1-c. The drug showed no significant effect in reducing lipid levels. Miles offered the following tentative indication to the committee: "Acarbose is indicated for lowering the blood glucose and glycated hemoglobin levels in patients with Non-Insulin Dependent Diabetes Mellitus [Type II(BRACKET). In addition to diet therapy, acarbose may be used alone or in combination with sulfonylureas therapy. Acarbose is also indicated as an adjunct to diet and insulin for lowering the blood glucose and glycated hemoglobin levels in patients with Insulin Dependent Diabetes Mellitus [Type I(BRACKET)." Although five of the six studies demonstrated a statistically significant reduction of HbA1-c levels, the committee was unconvinced that a statistically significant reduction also constituted a clinically significant reduction. When asked if acarbose made diabetics feel better, Miles Director Medical Research Robert Coniff, MD, said: "We did not ask that global question." Glucobay's side effects, as reported in the studies, were "mild to moderate" gastrointestinal events, Miles said, particularly flatulence, diarrhea and abdominal pain. Toxicology studies also indicated the possibility of hepatotoxic effects of the drug. Radiolabeling of acarbose showed that unexcreted drug metabolites settled most commonly in the liver, spleen, adrenal gland and kidney. Miles presented data that showed acarbose to significantly increase SGOT (serum glutamic-oxaloacetic transaminase) and SGPT (serum glutamic-pyruvic transaminase). Miles claimed that patients with the elevated liver enzymes were asymptomatic and the condition was reversible with discontinuation of the drug. The committee said, however, that the hepatotoxic potential of the drug had not been studied enough, and when later voting on two questions regarding the safety of the drug -- "Is the demonstrated or potential toxicity of Glucobay therapy acceptable, given its efficacy?" and "Has the safety of systemically absorbed drug breakdown products been adequately investigated?" -- unanimously answered "no" to both questions. Many committee members voiced concern that since diabetes is a chronic illness, patients would take acarbose for many years. The committee did not like the fact that none of the studies presented by Miles looked at patients for more than 24 weeks. The committee and FDA had other problems with the Glucobay data as well. FDA statistician Daniel Marticello, PhD, said that Miles' claim that acarbose demonstrated a significant reduction in patients' need for daily insulin intake in two studies was "suspect" because "it is not clear whether or not this positive result is due to a beneficial effect of acarbose or a detrimental effect, since patients were instructed to reduce their insulin dose if clinical signs and/or symptoms of hypoglycemia occurred during double-blind treatment." Acarbose is currently marketed in six European countries. Miles filed its NDA (number 20-086) for acarbose on Aug. 6, 1990. Commenting on the outcome of the committee meeting, a spokesperson for Miles said, "obviously, we are very disappointed that the FDA advisory committee acted as it did on acarbose." Noting that "millions of diabetes patients in Europe use this product now daily," he said the company would "possibly" consider submitting European clinical use data. He added: "I think there's a lot that we have to look at before . . . [we can] determine our future course of action."