Clinical Trials Or Observational Studies for Asthma? FDA Advisory Panels Can't Agree

FDA wants to gather long-term safety data in long acting beta-2 adrenergic agonist asthma drugs.

Whether randomized controlled clinical trials or observational studies are best for measuring the risks of long acting beta-2 adrenergic agonists remains unresolved after a two-day FDA advisory committee meeting.

FDA's Pulmonary-Allergy Drugs and Drug Safety and Risk Management Advisory Committees met March 10-11 to discuss potential study design for trials to investigate the potential for serious outcomes - such as hospitalization, intubation or death - with the use of LABAs.

FDA announced Feb. 18 that it is taking a number of steps to control the risks of the asthma drugs, such as GlaxoSmithKline's Serevent (salmeterol) and Novartis' Foradil (formoterol), by requiring a class-wide Risk Evaluation and Mitigation Strategy and working through the agency's Safe Use partners to monitor whether actual use adjusts based on the new safety measures (Also see "FDA Calls For Class-Wide REMS For LABA Asthma Drugs" - Pink Sheet, 22 Feb, 2010.).

The agency also wants the sponsors to conduct further trials to elucidate the risk profile of the class.

But some committee members questioned whether classical clinical trials are the best way to go to gather the desired information. How FDA should gather safety information will likely continue to be an issue of concern for the agency, especially as its Sentinel data-mining project builds steam (Also see ""Mini Sentinel" With Harvard Gets FDA Halfway To Health Record Target" - Pink Sheet, 18 Jan, 2010.), and FDA expects that advisory committees will increasingly face issues about post-marketing studies (Also see "FDA Advisory Panels To See More Questions On Post-Market Safety Issues" - Pink Sheet, 21 Jan, 2010.).

Challenges In Trial Design

At the LABA meeting, advisory committee members and industry representatives pointed out that recruiting for such trials would be difficult because the informed consent document would have to tell potential subjects that they would be participating in a study to measure the risk that the drugs they are taking might kill them.

"I realize that there is a common belief that the randomized controlled trial is at the top of the hierarchy and will give a definitive answer [regarding asthma drug risks]. But that is not a universal belief, and there are serious flaws with randomized, controlled trials," Carrie Redlich, Yale University School of Medicine, said in response to a question from FDA about study endpoints. FDA asked the panel to consider a variety of questions pertaining to endpoints and study design for the proposed safety trials.

"The way these questions are structured are all assuming how to fine-tune a randomized, controlled trial. It seems to me there's a more fundamental question, which is, is that an appropriate study to do to answer the question at hand. A trial, if you're lucky, may answer a question, but it may not answer a question that has any practical or clinical relevance," Redlich added.

"We believe that observational studies, be they case control or observational cohort studies, are an important part of the armamentarium we have in post-approval drug safety surveillance," said Gerald Dal Pan, director of FDA's Office of Surveillance and Epidemiology. He said the agency is working on a guidance on best practices for observational epidemiologic studies for drug safety questions using administrative health care databases, which fits the proposal GSK presented at the meeting as an alternative to randomized, controlled trials.

However, Dal Pan took issue with the presentation by Carlos Camargo of Massachusetts General Hospital, who was speaking on behalf of GSK, which advocated a case control study using a large cohort of patients with persistent asthma.

"While we agree with Dr. Camargo that, in a textbook way, case control studies are good for rare diseases, what we're talking about here is a risk that is low, and a relative risk that would be relatively low," Dal Pan said. "Our concern here is with what is called in the trade confounding by indication or residual confounding. Will they be able to get all the confounding out, the way [that] theoretically randomization should, or will there be some confounding that will just linger? And a lot of the administrative databases, for example, don't have information on smoking status, which would be an important covariate, for example, in pulmonary outcomes."

Still Some Controversy Over The Safety Signal

FDA convened the panel to gather perspective on the best way to further study the safety issues because recent trials have shown LABAs to pose an increased risk of severe asthma symptoms, hospitalization and death, but some of these trials came under fire from industry representatives and some advisory committee members. Whether there is increased risk for patients taking a LABA with an inhaled corticosteroid was a particular bone of contention.

For example, GSK's meta-analyses of studies of its Advair (fluticasone propionate/salmeterol inhalation powder) showed "there have been no deaths and no intubations in over 22,000 patients," Katherine Knobil, vice president in charge of GSK's Respiratory Medicines Development Center, said. She added that any labeling changes FDA is considering "must not interfere with patient participation."

One recent meta-analysis found that LABAs "increase the risk for asthma-related intubations and deaths, even when used in a controlled fashion with concomitant inhaled corticosteroids," according to an article that appeared Feb. 22, 2010 in the American Journal of Medicine, whose lead author was Shelley Salpeter.

"There is no statistically significant evidence that the combination increases mortality," Thomas Platts-Mills, director of the Asthma and Allergy Disease Center at the University of Virginia Medical Center, said. "The data ... that Salpeter confuses in her paper very badly as suggesting that there's a death signal - the death signal entirely comes from salmeterol or from LABA use on its own. You would not get an answer from a controlled trial of that kind."

"We have not reached the conclusion that the risk that we've seen with salmeterol is not present at some level in combination therapy," John Jenkins, director of FDA's Office of New Drugs, responded, while noting that the agency has not yet finalized any label changes.

In calling for new prescribing restrictions in February, FDA cited a meta-analysis of 110 studies the agency conducted before a December 2008 meeting of the Pulmonary Drugs, Drug Safety and Risk Management, and Pediatric Advisory Committees (Also see "FDA Panels Favor Advair, Symbicort Over Single-Entity LABA Products For Asthma" - Pink Sheet, 11 Dec, 2008.). The meta-analysis evaluated the use of LABAs in 60,954 asthma patients, using a composite endpoint to measure severe exacerbation of asthma symptoms consisting of asthma-related death, intubation and hospitalization.

"The results suggested an increased risk for severe exacerbation of asthma symptoms in patients using LABAs compared to those not using LABAs," FDA notes. The largest risk difference was seen in children age 4-11; 61 of 1,626 patients in this age group who received LABAs suffered severe exacerbation of their asthma symptoms, compared with 39 of 1,789 patients who did not receive a LABA. The overall numbers were 381 of 30,148 LABA patients compared with 304 of 30,806 non-LABA patients.

However, FDA added, "at this time, there are insufficient data to conclude whether using LABAs with an inhaled corticosteroid reduces or eliminates the risk of asthma-related death and hospitalizations."

The need for additional data is behind FDA's request for the LABA manufacturers to conduct studies to evaluate the safety of the agents when used in conjunction with an inhaled corticosteroid.

But the new discussion about the safety of combination therapy with LABAs is "based on a reanalysis of old data from the 1980s and 90s that we don't think supports this level of concern," Bobby Quentin Lanier, executive medical director of the American College of Allergy, Asthma and Immunology, maintained in an interview after the meeting.

Lanier added that the college does not support FDA's new recommendation that patients be taken off combination therapy as soon as possible, comparing this advice to "a 17-year-old asthmatic boy who wants to get off his medications as soon as he's feeling better. I've dedicated my life to fighting that [tendency]."

The American Academy of Allergy, Asthma and Immunology has also expressed concern that FDA's new risk management plan could unduly limit physician choice and insurance coverage for LABAs (Also see "FDA Outreach To Physicians On LABAs Shows Challenges Of Safety Strategy" - Pink Sheet, 8 Mar, 2010.).

- Martin Berman-Gorvine ([email protected])