COGNEX EXPANDED ACCESS AVAILABILITY SHOULD AWAIT START OF NEW CLINICAL TRIALS, FDA ADVISORY CMTE. RECOMMENDS; PANEL CONTINUES TO QUESTION DRUG's EFFICACY

Warner-Lambert must begin another efficacy trial of its Alzheimer's drug Cognex (tacrine) before the product is released under an expanded access program, FDA's Peripheral and Central Nervous System Drugs Advisory Committee recommended at its July 15 meeting. If Warner-Lambert were willing to conduct a new well- controlled efficacy trial, "some version of a Treatment IND would make some sense," commented committee panel member Herbert Weingartner, PhD, National Institute on Aging. The committee was not asked to formally vote on the expanded access comment, but an informal poll of the committee indicated majority support for the concept with the caveat that new efficacy studies would have to be done. The advisory committee meeting was called by FDA to discuss a Treatment IND proposal for Cognex from Warner-Lambert. The idea of a tacrine Treatment IND was initiated by FDA in a March 22 letter to the company following the advisory committee's vote on March 15 not to recommend approval of the drug ("The Pink Sheet" March 25, p.4). Although a majority of committee members favored some sort of expanded access program for the Alzheimer's therapy, most were reserved in their recommendations. "I would favor expanded access, but I am not especially enthusiastic about it," remarked committee member Sid Gilman, MD, University of Michigan. Warner-Lambert had hoped to generate enough safety data from the proposed Treatment IND to allow the company to gain approval for Cognex without having to conduct further efficacy studies. FDA Office of Drug Evaluation I Director Robert Temple, MD, told the committee in his opening statement that the company wanted to know "that if [the proposed] expanded access program provides great reassurance on safety, whether on that basis alone [FDA] would be prepared, with [the committee's] advice, to approve the drug." The committee unanimously rejected Warner-Lambert's proposal to bypass any further efficacy studies. The members, in an eight- to-zero vote, agreed that the effect of tacrine in Alzheimer's patients "constitutes an insufficient demonstration of efficacy." The committee members' hesitation to support an expanded access program for Cognex was due largely to the drug's marginal efficacy in Alzheimer's patients in controlled trials to date. Committee Acting Chairman Steven Ferris, PhD, New York University Medical Center, who was against a Treatment IND for tacrine, said "I would be in favor of open access only when we have more evidence of efficacy." Some committee members noted that allowing the company to distribute tacrine under a Treatment IND without any real proof of efficacy would be setting the wrong precedent. "Our decisions here will be precedent setting in that if we agree that the company should go to a Treatment IND on the basis of what I would venture to call a minor change on the ADAS-COG (Alzheimer's Disease Assessment Scale), then additional medications that are requested for approval would only have to pass that same measure," Gilman said. A representative of an Alzheimer's patient advocacy group even was reluctant to endorse a Treatment IND for tacrine. "From a family member's standpoint, the results are very disappointing," Alzheimer's Association Chairman Richard Gehring said. However, he noted that the members of his association would probably prefer to have access to the drug as soon as possible. Some committee members also expressed concern that a Treatment IND would impede enrollment of a concurrent controlled trial. A Treatment IND for tacrine might have a similar impact on controlled studies that parallel track had on ddI (Bristol-Myers Squibb's Videx) by slowing enrollment of AIDS patients in the clinical studies, FDA consultant Paul Meier, PhD, University of Chicago, noted. Calling the impact of parallel track "nothing less than disastrous," Meier said that "we learned much less than what we could have learned." FDA maintains that parallel track may have delayed enrollment somewhat but that the setback was not substantial. A new efficacy trial for Cognex could take up to two years, one committee member estimated. Furthermore, the Treatment IND could not begin under FDA regulations until the clinical trial was underway. The agency indicated that the ball is now in Warner- Lambert's court as to whether it would want to proceed with the trial and Treatment IND. "Obtaining additional effectiveness data is plainly a long term task," Temple commented. "We have no certain knowledge as to whether [the company] would be interested in that." In a July 17 statement of sales and earnings, Warner-Lambert said that it is "evaluating these recommendations and will be working closely with the FDA to determine what steps will be taken next regarding the drug." Initially, Warner-Lambert proposed to the committee a Treatment IND protocol in which 50,000 patients would be enrolled and receive doses of up to 120 mg/day for 12 weeks. The purpose of the Treatment IND study was to collect safety data with regard to tacrine's liver toxicity and other side effects. The treatment protocol required that physicians test patients' liver enzyme levels once every six weeks. Depending on how high liver enzyme levels rose, therapy would be stopped until the levels returned to normal and then patients would be redosed with tacrine. If patients demonstrated liver enzyme levels over 10 times the upper limit of normal, tacrine would be discontinued. The company had planned to charge for the drug under the Treatment IND cost-recovery provisions at a price to wholesalers of $ 3 per daily patient dose, or approximately $ 4 per day to the consumer. "It would be approximately the introduction price of the drug when it is approved," Warner-Lambert Research Division VP and Chairman Ronald Cresswell said. However, if Warner-Lambert were to be required to conduct new clinicals as recommended by the committee, the company would have to wait to begin charging for the drug "until enrollment in those trials is substantially complete," Temple said. The advisory committee was still not persuaded that Cognex is significantly efficacious in treating Alzheimer's despite hearing new analyses of the data in the tacrine NDA from Warner-Lambert. In addition, the company also provided information from a French study that showed that Cognex had positive, albeit weak, effect in one cognitive parameter. The panel concluded that not much had changed since the advisory committee's March meeting where the panel rejected recommending approval of the drug. The committee found at the March meeting that the two tacrine pivotal trials submitted by Warner-Lambert were plagued by design problems, confounding factors and generally weak efficacy data. The panel members agreed that tacrine treatment caused some patients to improve several points on the ADAS-COG cognition test. However, no significant improvement was seen in the Clinical Global Impression of Change (CGIC). In addition, the panel found that there was only a trend toward improvement as assessed by patients' caregivers. Committee acting chairman Ferris reiterated that he "(BRACKET)does] not feel that the quite small change on a scale on the ADAS-COG alone is sufficient to warrant a judgment that there is a sufficient demonstration of clinically meaningful efficacy." The committee is "not even sure about the actual effect size, that is two points or three points on the particular scale," he noted. Ferris said that they are also unsure about "how much of that effect is a treatment effect" because of problems in the study design. Responding to the anecdotal claims of dramatic improvement of some patients receiving tacrine, committee member Michael Brooke, MD, MacKenzie Center for Health Sciences, Alberta, remarked that "the days of snake oil, and the days when a drug could be espoused because the enthusiasm of one or another individual . . . are over. We would welcome the data which would allow us to say yes this drug does work, no this drug doesn't work; we don't have that data yet," he said. One of the problems identified by the committee was that Warner-Lambert's efficacy trials were probably conducted at too low a dose; the tacrine dose was lowered from 160 mg/day in the studies to either 40 mg/day or 80 mg/day when it was discovered that a large percentage of patients were suffering from liver toxicity. Even at the lower doses, about 25% of patients experienced elevated liver enzyme levels three times the upper limit of normal. About 40% of patients showed any clinical response to the treatment. Warner-Lambert told the committee it plans to do a study at higher doses with a longer patient treatment period. Proposed at the July 15 meeting as a Phase IV study, the trial protocol calls for doses of up to 160 mg/day and would take more than two years to complete. The company said it plans to begin the trial, which will enroll 400 patients, in the fourth quarter. The high-profile advisory committee meeting featured opening remarks by FDA Commissioner Kessler, who emphasized that any decision about a Treatment IND must be based on scientific data only. "Our despair over a patient's disease must not cloud our vision," he said. "Our anguish over a patient's plight must not allow us to forget the scientific basis on which our decisions are made." The commissioner added that "no doubt, [FDA] will be criticized, whatever decision we make." Kessler added, however, that he "would like to see every promising Alzheimer' therapy designated "1AA," a rating currently reserved for therapies for such life-threatening illnesses as AIDS and cancer. FDA wants to "turn over protocols in weeks [and] measure review times in months," Kessler noted. FDA Neuropharmacologic Drugs Division Director Paul Leber, MD, announced at the meeting that he has created a special unit within the division dedicated to the expeditious review of Alzheimer's drugs.