OFF-LABEL PROMOTION IS UNLIKELY TO PROVIDE INFORMATION BENEFIT TO PHYSICIANS, FDA's TEMPLE SAYS; SINGLE PIVOTAL STUDIES ARE "NOT CHEAPER" AND "HIGH RISK"

A relaxation of FDA's prohibition on the promotion of off-label uses of pharmaceuticals is unlikely to improve the practice of medicine, FDA Office of Drug Evaluation I Director Robert Temple, MD, contended during the Third Annual Princeton Drug Development Conference in Princeton, N.J. May 5.

"The idea that massive amounts of new, unvetted information from a commercial source will help physicians and patients is unproven and improbable," Temple said. "The information age is only as good as its data," he observed. Many problems in health care arise from "not knowing, with precision, the answers to many critical questions." However, caregivers "can be effectively proactive only when they have unbiased information to use," he commented.

A "straw in the wind is that increased ability to promote off-label uses would be a great boon," Temple noted. FDA critics, most prominently the Washington Legal Foundation, have characterized the agency's attitude toward off-label promotion as a hindrance to the free flow of information among scientists and clinicians. Temple's presentation to the conference, while not explicitly a response to FDA reform proposals, took issue with several common themes of the reform efforts.

"Nothing in present promotional practices suggests that promotion of off-label information (i.e. promotion with no established limit/guidance) would be something that could be called `information dissemination,'" Temple declared.

Off-label promotion would instead be "well-targeted, selected information, just as all promotion is," Temple declared, "and it will be worse than usual because there won't be a labeling to judge it against." It is "worth emphasizing," Temple added, that "there are numerous independent sources of information for physicians" and "they can be as imaginative, bold, biased, unfair, selective and even irresponsible as they please." There is "very little value in turning the promotional environment into what I believe would quickly become a real jungle," he concluded.

"I don't think it's mean-spirited of me, or nasty, to be skeptical about what new information will mean when I see that even things we do know are not implemented when it's not in anybody's interest to promote them," Temple said. "We have cheap, unequivocally life-saving treatment for hypertension in the form of diuretics, which can be had for $10 or $15 a year, and this treatment is scorned in favor of expensive agents not known to have ever prevented a stroke, stopped an MI, or saved a kidney," Temple declared.

Temple also commented on the question of whether one pivotal trial should generally be accepted as sufficient for approval. The Biotechnology Industry Organization has made that a feature of its FDA reform proposal ("The Pink Sheet" April 24, p. 3).

The FDAer questioned the real-world impact of the proposal. Temple observed that "relatively few applications turn on the one study/two studies issue." For any symptomatic disease, Temple continued, "usually there are a variety of studies (different durations, doses, combinations), usually more than two, because they are needed for safety dose response evaluation."

"What would be the gain of a single, massive world-wide study?" Temple wondered. It is "not cheaper," he maintained, and "throwing all you eggs in one basket strikes me as very very high risk," he said. "Studies don't always come out, and taking one shot at it wouldn't be my choice."

Temple also questioned the value of relying on post-marketing studies in lieu of pre-marketing clinical trials. "In terms of post-marketing studies," Temple said, "the frequency with which sponsors incorporate new data into labels (except warnings [or] new uses) is virtually zero."

Instead of conducting fewer trials, savings could be achieved by simply collecting less data, Temple suggested: "I would argue that approvals can come with a relative minimum of data." Much of the data that FDA does receive "isn't anything we asked for anyway," he maintained. New study designs that would be welcome at the agency are "well-controlled studies in `lower-tech' environments," Temple added.

"It's very hard to see years of savings...due to harmonization" with foreign regulatory agencies, Temple commented. Harmonization "is welcome, and is clearly a benefit, but there are at present relatively few major differences between the EU and the U.S.," Temple said.

One of the "substantial differences," Temple noted, is "the extent of documentation." However, "in terms of cost of development," the discrepancy is "relatively trivial."

A second gap exists between the two communities in the extent to which one deals with "individual cases," Temple commented in a panel discussion with European Medicines Evaluation Agency Executive Director Fernand Sauer. "We are very committed to review of individual data," Temple added, noting that FDA is, historically, "the only agency that has remotely close to the number of people necessary to do that."

Acceptance of EMEA reviews in the U.S. without additional data analysis in the near term "would be difficult," Temple told the group. "I don't think until individual data are come to grips with more we will be likely to accept European reviews," he said. Referring to the fact that EU reviewers are not employed full-time by the EMEA, Temple suggested that "you can't realistically...ask people who have other jobs...to come to the database with the same intensity that full-time employees at the FDA can."

"If I were a drug developer," Temple commented, "I'm not sure I'd want there to be only one review in the whole world." With that type of situation, Warner-Lambert's Alzheimer's agent Cognex "wouldn't be approved in the world, because the majority of countries haven't approved it, whereas we have," he pointed out.

Panel moderator and former Lilly chief scientific officer, Leigh Thompson, MD, asked Sauer about the use of outside reviewers: "You've chosen 1,200 wonderful experts to evaluate drugs -- are they any less competent than the full-time employees at the FDA?" Sauer responded: "The future will tell."

Potential efficiencies for trial design that are "worth thinking about, but will not be easy," include early pharmacokinetic data, titration designs, and a willingness to question those parts of data collection that seem illogical, Temple said.

"Earlier pharmacokinetics is cheap and easy, and is probably a good idea," Temple offered. "Sometimes it's obvious in retrospect that better pharmacokinetics could have made a major difference in choosing a dose and identifying potential problems in subsets." Pharmacodynamic endpoints are "not used sufficiently to lead to a wide range of doses in clinical trials, and all too often the wrong dose is chosen," Temple continued. "That's a staggering waste of resources."

Titration designs "are incredibly efficient as an early study but are almost never used," Temple continued. He acknowledged that FDA "helped tell the world about why they're uninformative," but suggested that "people could get a better fix on what dose to study by returning to some of those."

The need for "larger, simpler studies and the need to ask why data that are being collected are being collected is real," Temple told the group. "There are very few actual rules about what needs to be collected in any FDA regulations, law or guidelines, and if something is silly or not useful, it should be challenged. Nothing bad will happen if you do that."

Other suggested efficiencies included employing a nested study design to test high doses in non-responders, use of factorial studies and use of pharmacokinetic screening. Also, "there may be ways to modify the carcinogenicity bioassay, which is a potential in savings," Temple said. "Small efficiencies probably do matter."