ANTI-ANGINAL GUIDELINES NEED NOT AGREE WITH EC DRAFT, ADVISORY COMMITTEE SAYS; RECOMMENDS WORKSHOP TO DETERMINE EARLY STAGE ISSUES IN DRUG DEVELOPMENT

FDA should convene a workshop to discuss issues of dose-response and selection in clinical development, FDA Cardiovascular & Renal Drugs Advisory Committee temporary member Craig Pratt, MD, Methodist Hospital (Houston), suggested during the panel's discussion of anti-anginal drug development guidelines Feb. 24.

Pratt suggested that the agency bring together "the best and the brightest" to discuss the issue. He pointed to National Institutes of Health workshops as a model for FDA to follow.

The committee agreed that general issues pertaining to dose-response and drug metabolism requirements need more conclusive standards set by the agency.

"There's a whole variety of issues that I think would really help the industry if there was some meaningful consensus in the agency, between the agency and the advisory panels...as to what constituted an adequate range" for dose-response studies, committee member Lloyd Fisher, PhD, University of Washington- Seattle, said.

FDA Cardio-Renal Drug Products Division Director Raymond Lipicky, MD, said dose-response/selection "probably is one of those things that deserves a separate document." FDA, however, wants to address the topic in the anti-anginal guidelines because the issue seems "to get talked about again and again and again and the words never get put into practice." In a similar vein, the committee discussed general drug interaction issues under the rubric of the anti-anginal guidelines in October ("The Pink Sheet" Oct. 17, 1994, p. 12).

Drug metabolism standards are "sort of beyond the scope of our set of guidelines," invited speaker Jeffrey Borer, MD, Cornell University Medical Center/New York Hospital, told the committee. "Details would change every year." Borer, one of the drafters of the guidelines, suggested that the guidelines include examples of original ways to determine dose-response.

The guideline process began in 1982, producing a draft dated Jan. 10, 1989. The committee's last general discussion of the guidelines took place in March 1988.

Lipicky reminded the committee that the purpose of the guidelines is to identify "things that are not placebo," not "to figure out where [a drug] fits in the therapeutic armamentarium or what patients are likely to benefit." The meeting was spurred by the publication of draft anti-anginal guidelines by the European Community, as "those guidelines differ in many respects from those provided in our draft," FDA said.

The primary difference between the guidelines is the requirement for effect: the U.S. draft requires both anti-ischemic and anti-anginal activity from a drug, whereas the EC guideline subsumes anti-anginal effects under anti-ischemic activity. Borer said the reasoning behind the U.S. guideline was to avoid exacerbating ischemia by masking the warning symptom of pain.

"I thought the strength of your guidelines from the EC guidelines was...the duality of effect," committee member Alastair Wood, MD, Vanderbilt University, stated. Summarizing the committee's thoughts, Chairman Jeffrey Anderson, MD, LDS Hospital (Salt Lake City) said: "We prefer the FDA version....While we don't want to copy the European guidelines, we ought to be aware of them."

Committee member Michael Weber, MD, Long Beach V-A Medical Center, offered a caveat: "You stress that it is important to have an anti-ischemic effect as well as a symptomatic effect, but that's different from demonstrating some long-term clinical benefit. So why couldn't we have a situation where you have a drug that would relieve anginal symptoms but as long as it was ischemia-neutral...Why could that not be considered as an anti-anginal drug?" Committee member Barry Massie, MD, San Francisco V-A Hospital, suggested a separate indication of "analgesic to relieve pain of angina" for such agents.

The draft guidelines classify three forms of angina: stable angina of effort (SAE), vasospastic angina (VA) and unstable angina (UA) or acute coronary syndrome. Borer explained that "we took the time to describe the syndromes" because "the underlying mechanisms are or may be multiple for each syndrome and are not all that well understood....We want to provide a phenomenological description that is sufficient to allow people to set up a reasonable operational protocol."

The committee had no problems with the definition of SAE. The draft guideline described the condition as "a history of chest pain characterized by the occurrence of substernal or precordial distress...fairly reproducibly brought on by exertion or emotional stress, and/or relieved by sublingual nitroglycerin, cessation of exertion or removal of emotional stress." The guideline also notes that "there is no particular need to exclude patients with atypical angina (e.g., only exercise-induced jaw discomfort, only arm pain)" from SAE studies, although protocols "may require some modifications to ensure the presence of the expected underlying conditions."

The draft defines VA as "nonexertional chest discomfort with no particular or predictable frequency or duration....associated with transient ST-segment elevation (or depression) on the ECG." The committee was less comfortable with that definition.

Invited speaker Stephen Glasser, MD, University of South Florida College of Medicine (Tampa), opined that "if we went around and asked the definition of vasospastic angina none of us would agree, so I wouldn't try for a definition of that." The committee concluded that the definition of vasospastic angina should be kept "narrow," Anderson said.

The committee unanimously felt that UA should not be included in this guideline. Glasser called it a "different disease" with "different endpoints" and referred the committee to HHS' Agency for Health Care Policy & Research and the National Heart, Lung & Blood Institute's March 1994 clinical practice guideline on unstable angina for a complete definition.

Borer explained his reasoning as "the implication of the syndrome clinically is very different, and when we deal with chronic stable angina or vasospastic angina the goal...is relief of the symptom....The clinical presentations that we lump under the name unstable angina or acute coronary syndrome has important natural history implications and it's the relief of those that's the primary goal of treatment."

FDA asked the committee whether patients enrolled in clinical trials must "have angina in their daily lives, or is it sufficient that angina be reproducibly induced by exercise?" Massie declared that, while there should be no daily life requirement, "you wouldn't want to approve a drug in which no patients studied had it."

"One can't, in the U.S., in this therapeutic climate, develop an approach" requiring patients with angina in daily life, invited speaker Nathan Reichek, MD, Medical College of Pennsylvania (Pittsburgh), said. "Most of the patients who are available for evaluation over time...are already being treated on a multi-drug regimen and we can only elicit the possibility of spontaneous angina" by withdrawal from medication over a long time period. "I don't think it's a desirable constraint. The other factor is, a large number of patients titrate to their exercise tolerance."

At least one placebo-controlled trial is necessary for anti-anginal agents, the committee agreed. The panel acknowledged the problem that current medication of most patients poses for clinical trials but concurred with the guidance that "in order to claim monotherapy works, you really need a placebo-controlled clinical trial," invited speaker Udho Thadani, MD, University of Oklahoma, said.

"If there isn't going to be some placebo control experience," Borer said, to avoid labeling restrictions "one would like to see the placebo control on background [medication] done with several different backgrounds so that you could confidently conclude that the effectiveness of the drug isn't merely an interaction with a background drug."

Borer noted that there is some value to positive-controlled trials, including "longer term trials than might be possible if the control group were placebo," which could provide "a lot of safety information." He added that "if you don't beat the positive control that's approved at the dose you used, you cannot draw the inference that the new agent is effective."

The four invited experts -- Borer, Thadani, Glasser and Reichek -- gave what Anderson called a "qualified yes" to FDA's question on the acceptability of risk from "exposing patients with angina to placebo and then exercising them." They felt that investigators should withdraw medication cautiously. The panel unanimously agreed that nontreatment with exercise was acceptable.

In trials of vasospastic angina agents, the committee found no serious problems with the Cardio-Renal Division's policy of approval "of a drug for this indication after a single, short, placebo-controlled withdrawal trial" because VA is "rare" and "effective therapy (calcium-channel blockade) is well established." Committee member Fisher suggested that VA trials have three arms: placebo, the new drug and a calcium antagonist for interpretability of data. Borer replied that such a design would be "lovely" but should not be mandated.

The committee split on the question of expanding the drug tolerance section of the guidelines, including "the pharmacodynamics of drugs over time," "the scale of the dosing interval" and "the scale of time-in-trial." Lipicky declared that the next draft would "try to say something operational" and the committee did not make specific recommendations.

Wood expressed concern that a placebo-controlled withdrawal study, recommended in the guideline, would not be sufficient to determine tolerance. Nonetheless, he did not want the draft to go beyond "endorsing" studies of tolerance.

FDA's final series of questions to the committee dealt with the current policy of requiring "that a new drug's effects (both beneficial and adverse) be described as functions of gender, race, age, and excretory function (renal and hepatic)." The committee agreed that they would "like to see diversity when appropriate," Anderson stated, but had varying impressions of what should be specifically required.

Thadani recommended listing populations studied or not studied in the labeling, but did not feel an application should be rejected if diversity was not sufficient to meet FDA's policy. He referred to the difficulty of including women in anti-anginal drug trials, because they are predominantly older and less active than men with angina and "hate" the treadmill test.

Committee member Cynthia Raehl, University of Wisconsin School of Pharmacy, thought the population the drug would be marketed to should be represented in trials. She suggested smaller pharmacokinetic/ pharmacodynamic studies to discover if subpopulations could be expected to react differently to a drug.

Borer posited that a sponsor merely needed to show that there was no harm to the listed subpopulations, while efficacy in subpopulations was "another question." He felt labeling restrictions would be sufficient penalty for under-representation while Phase IV requirements could remedy insufficient subpopulation data. Fisher, a biostatistician, noted that subgroup studies normally were at a lesser power, so the sponsor was really looking for "extremely big differences" in response. He felt "reasonable representation," explained "cogently" by the sponsor, would be a better standard than "proportional" representation of subpopulations.

The committee did not feel that "having final published antianginal guidelines" was a priority. FDA noted in meeting materials that the draft has been "freely available" and "guidance is implicitly provided by the public discussions of this committee."