UPJOHN HALCION 10,000-PATIENT SAFETY STUDY SCHEDULED FOR COMPLETION BY JANUARY 1994; INVESTIGATION OF NDA CONTINUES AFTER ADVISORY COMMITTEE REVIEW

Upjohn expects to complete a 10,000-patient trial evaluating the safety of its benzodiazepine hypnotic Halcion in comparison to Sandoz' Restoril by Jan. 1994, the company told FDA's Psychopharmacologic Drugs Advisory Committee on May 18. Upjohn's protocol M2100/0235, which has been reviewed by FDA and approved by an independent review board, calls for a randomized double-blind four-week comparison of Halcion (triazolam) .125 mg and .25 mg to Restoril (temazepam) 15 mg and 30 mg in elderly and adult insomniacs. Upjohn intends to enroll 1,000 investigators at 25 sites by October with full patient enrollment planned by July 1993. The trial will use structured telephone interviews to assess side effects at baseline, two and four weeks, Upjohn said. The study will also collect information on doctor visits, hospitalizations, change in other medications and sleep evaluation. The committee supported Upjohn's proposal but expressed some concerns about the study design, particularly the use of the telephone surveys. Member Robert Hamer, PhD, Medical College of Virginia, said: "I'd like to strongly urge the sponsor and FDA to very carefully design this thing so that it's capable of addressing the questions that we are...being asked. This is the second Halcion hearing I've been to and I'd really rather not go to another one." Upjohn expects the 10,000-patient study to satisfy safety concerns of both FDA and the European Community's Committee on Proprietary Medicinal Products. The CPMP stipulated that Upjohn perform such a study in its Oct. 16 opinion on the drug, which requested labeling changes, unit-of-use packaging and a postmarketing study ("The Pink Sheet" Oct. 21, 1991, p. 9). The proposed trial represents the latest salvo in Upjohn's ongoing defense of Halcion against questions about the impact of NDA errors on the drug's risk/benefit ratio. Since the disclosure of the "transcription errors" in August, Upjohn "has devoted over 30% of its [450-person] clinical division as well as staff from other divisions to conduct...a massive reconstitution of the clinical studies database and a review of the published studies and completed unpublished studies for triazolam," VP-Clinical Development & Medical Affairs Lawrence Olanoff, MD/PhD, told the committee. Upjohn's efforts paid off on May 18 when the committee voted that Halcion is safe and effective as currently labeled (see previous story). The company's application, however, is still under scrutiny at FDA to determine "whether Upjohn failed to report results accurately," Neuropharmacologic Drug Products Division Director Paul Leber, MD, said. The adverse publicity surrounding the drug has hurt Halcion sales. According to data presented by FDA Epidemiology Branch Chief Bruce Stadel, MD, prescriptions for Halcion have declined from a peak of nearly 11 mil. in 1988 to about 7 mil. last year. Upjohn reported in April that first quarter Halcion sales by dollar volume declined 39% ("The Pink Sheet" April 20, p. 19). Upjohn's Halcion patent expires in late 1993. Other benzodiazepines do not appear to be benefiting from Halcion's sales decline. Stadel's data show that temazepam prescriptions have remained flat at about 5 mil. since 1988. Abbott, promoting ProSom (estazolam), a new entry into the market, reportedly is finding that sales growth is not coming directly from physicians who have switched from Halcion. Doctors are instead prescribing other classes of sleeping pills. The turn away from benzodiazepines could present an opportunity for new non- benzodiazepine hypnotics, such as Searle's Ambien (zolpidem), which is "approvable" at FDA ("The Pink Sheet" April 27, In Brief). FDA has sought to treat the Halcion situation as a class issue. The relabeling of Halcion in November was the model for a December requested relabeling of other benzodiazepines ("The Pink Sheet" Dec. 16, 1991, T&G-1). At least two of the affected companies, Abbott and Roche (marketer of Dalmane), have responded with questions focusing in part on whether class labeling would obscure differences among the drugs.