HUMAN INSULIN MASKING OF HYPOGLYCEMIA SYMPTOMS

Executive Summary

HUMAN INSULIN MASKING OF HYPOGLYCEMIA SYMPTOMS occurs no more frequently than with animal insulin, FDA's Endocrinologic and Metabolic Drugs Advisory Committee concluded at its March 27 meeting. The committee unanimously identified as "weak" or "equivocal" the "evidence suggesting that human insulin therapy is associated with fewer warning symptoms of hypoglycemia than comparable animal-derived insulin." Indicative of the committee's consensus, panel member Steven Winters, MD, Montefiore Hospital, Pittsburgh, commented: "I think from today's testimonies I'm really not convinced that human insulin is associated with fewer warning symptoms of hypoglycemia than animal insulin preparations....I feel that the bulk of the evidence suggests that the phenomenon is not a real one." Winters noted that "the data suggesting [greater hypoglycemia unawareness with human insulin] in part come from retrospective studies, and I doubt really that retrospective studies are useful because it seems to me that one cannot recall being unaware." FDA arranged the advisory committee review following U.K. reports last fall of 17 unexplained deaths in diabetics. The deaths were reported by toxicologist Patrick Toseland, MD, Guy's Hospital in London, who claimed that many of the mortalities occurred in diabetics who had switched from animal to human insulin. His hypothesis was that the human insulin masked the signs of hypoglycemia. The committee did not review Toseland's data, but did examine a wide range of clinical and epidemiologic data from individual researchers, FDA, NIH, the Centers for Disease Control and human insulin manufacturers Lilly and Novo-Nordisk. Based on the committee's review of the studies, the panel also unanimously agreed that the available data do not "indicate a significant difference between the incidences of hypoglycemia, unexplained death, or death attributable to hypoglycemia for human and animal insulin therapies." Nonetheless, a couple of committee members cited one study as inconclusively raising the possibility of a greater number of hypoglycemic reactions with human insulin than animal insulin. Referencing a case-control study presented by Arthur Teuscher, MD, a professor at University Medical Department, Inselspital, in Bern Switzerland, committee chairman Saul Genuth, MD, Case Western Reserve University, Cleveland, said the data "might indicate a significant difference between the incidence of hypoglycemia for human insulin therapies." Genuth, however, pointed to potential patient selection biases in the study and the need for a new statistical analysis "that brings out the significant odds ratio for prior human insulin use increasing the risk" of hypoglycemia. * Most panel members emphasized the need for further studies to clarify the issue of hypoglycemia symptom masking, suggesting a variety of studies ranging from epidemiologic review to prospective clinical studies. Novo-Nordisk has started to look into the issue and plans to conduct several additional epidemiologic and clinical studies. The labeling for Lilly's Humulin human insulin product has contained a warning statement regarding hypoglycemia masking since 1983. Novo-Nordisk's human insulins will carry similar labeling statements by May, according to FDA. The advisory committee generally felt that further labeling statements were not warranted, at present, given the data. A few members, however, did suggest amending insulin product labeling. Their recommendations included: inserting a statement that hypoglycemia symptom masking has not been shown to be greater with human insulins; adding further patient information; and adding further information on the "generic" issue of hypoglycemia unawareness.