ROCHE REQUESTING FDA TO REVISE DIAZEPAM LABELING

ROCHE REQUESTING FDA TO REVISE DIAZEPAM LABELING to warn against interchanging different diazepam products in patients with low acid levels. Based on the results of two clinical trials, the firm said it has "asked the FDA to revise the Valium package insert to include a statement advising that caution should be exercised when different solid oral preparations of diazepam are interchanged in a patient who is also receiving a drug known to reduce gastric acidity or is known or suspected to be achlorhydric." Roche said it is "also recommending that all other solid oral diazepam dosage forms of diazepam products be required to include a similar statement, and that the agency publication entitled Approved Drug Products With Therapeutic Equivalence Evaluation ('The Orange Book') be revised to reflect these data." Along with the request for a labeling change, Roche submitted results of two clinical studies evaluating bioequivalence characteristics of diazepam tablet products manufactured by Mylan and Roche's Valium in subjects pretreated with the H[2] blocker ranitidine. The study submission explains that Mylan diazepam products were chosen for the clinicals because in in vitro at elevated pH levels they showed the greatest differences in dissolution rates compared with Valium. In one study, 32 healthy volunteers were randomly assigned to receive either a Valium 5 mg tablet, Mylan's 5 mg diazepam tablet, or either of the tablet products following pretreatment with ranitidine. Pretreatment consistent of three 150 mg ranitidine doses during the 18 hours prior to diazepam dosing. The open-label, crossover study found that in "subjects who were not pretreated with [ranitidine], the Mylan 5 mg product and Valium showed no significant differences (p=0.05) in terms of Cmax [eg., peak concentration] or AUC [eg., area under the curve]." Discussing effects of ranitidine pretreatment on bioavailability of the products, the submission states: "Comparison of AUC and Cmax values following pretreatment . . . revealed significant differences (p=0.01) in Cmax values for both treatment groups in comparison to Valium (nonpretreated). The reduction in diazepam absorption after pretreatment was great for the Mylan product, however, than for Valium. Whereas the mean reduction in Cmax for Valium following pretreatment was only 10%, for the Mylan 5 mg tablet it was 26%." Roche noted that there were no significant differences in AUC between the two pretreated groups. The second clinical compared Mylan 10 mg diazepam to Valium 10 mg following ranitidine pretreatment in 24 healthy volunteers. Subjects were then switched over to receive either Valium 10 mg or Mylan diazepam 10 mg without ranitidine pretreatment. Discussing results of the second study, the submission states that "although no differences exist in AUC values, comparison of the peak plasma levels reveals significant differences between these two drug products. Administration of Mylan's [diazepam] to the subjects pretreated with the H[2]-blocker resulted in significantly lower mean Cmax values (21%) than Valium (p=0.05)." Both studies, Roche maintains, "have demonstrated unequivocally that diazepam products with differing dissolution rate profiles at elevated pH are not bioequivalent when administered to subjects pretreated with an H[2]-blocker known to suppress gastrict acid secretion." The submission states that "these results further demonstrate that the USP XXI Dissolution Test is not adequate to assure the bioavailability of generic diazepam drug products under conditions other than fasted normal subjects." Last year in response to a Roche petition, FDA agreed to revise its diazepam bioequivalency testing guidelines to include in vitro dissolution testing at elevated pH levels ("The Pink Sheet" July 15, 1985, T&G-2). Roche submitted the results of in vivo bioequivalence tests in October 1985 that compared different diazepam products in patients with and without ranitidine pretreatment ("The Pink Sheet" Oct. 28, T&G-6).