Panels discourage black box warning on AEDs; Antiepileptic drug patients should get medication guide explaining suicidality risk, members suggest.

BELTSVILLE, MD.--A boxed warning about an increased suicidality risk with the use of antiepileptic drugs should not be added to the labels of drugs in this class, but patients prescribed these drugs should receive information about this risk with every prescription, according to the majority of two federal advisory panels.

At a joint meeting of the Food and Drug Administration's Peripheral and Central Nervous System Drugs and Psychopharmacologic Drugs advisory committees, the panel members agreed with the evidence indicating an increased suicidality risk and that this information needed to be communicated to physicians and patients. But they voted 14-4 with 3 abstentions against the FDA's proposal to include it in a black box warning in the labels of all antiepileptic drugs (AEDs). However, most (17-4) voted that patients should receive a medication guide describing the finding with each AED prescription filled.

Among their concerns were that a black box could reduce appropriate prescribing of AEDs and could affect compliance among patients. For a physician, a black box warning is "a big deal" and "comes with a very negative connotation," that could influence a physician to question whether a patient needs a drug, said the acting panel chair, Dr. Larry Goldstein, professor of medicine, Duke University, Durham, N.C.

Dr. Daniel Pine, chief of child and adolescent research at the National Institute of Mental Health's mood and anxiety disorders program, Bethesda, Md., said he was concerned that a black box would discourage patients from taking their medications. But he advised the FDA to come up with "creative ways short of a black box" to communicate this information to physicians and patients. He said he hoped that a medication guide would encourage physicians to discuss the issue of suicidality with patients, without frightening patients away from taking the medication.

The panels reviewed the results of an FDA analysis of data on 11 AEDs that compared the rates of suicidality (episodes of suicidal ideation, suicidal behavior, or completed suicide) between patients in treatment and placebo groups. In a metaanalysis of 199 prospective, randomized, parallel-arm, placebo-controlled trials of 27,863 patients on treatment and 16,029 on placebo, the overall odds ratio for suicidality with treatment was a statistically significant 1.80. Odds ratios for individual AEDs ranged from 0.57 to 2.75. There were four completed suicides among people on treatment (0.1%), and none among those on placebo, and the rate of suicidal behavior or ideation was 0.37% among patients on an AED, and 0.24% among those on placebo.

Based on these results, which the FDA made available to the public in January 2008, the agency concluded that there was a signal for increased suicidality for the class of AEDs and proposed that the finding be included in a boxed warning added to the labels of all currently marketed AEDs that are used long term, including those not in the meta-analysis, in the warnings and precautions section of the label, and in a patient medication guide.

The AEDs in the meta-analysis were carbarmazepine (Carbatrol, Equetro), divalproex sodium (Depakote, Depakote ER), felbamate (Felbatol), gabapentin (Neurontin), lamotrigine (Lamictal), levetiracetam (Keppra), oxcarbazepine (Trileptal), pregabalin (Lyrica), tiagabine (Gabitril), topiramate (Topamax), and zonisamide (Zonegran). Of the 199 trials, 31% evaluated AED use in epilepsy, 28% were for AED use for 8 psychiatric indications, and 41% were for 11 other indications. Those enrolled were at least 5 years of age, at least 20 patients were in each arm, and treatment lasted for at least 7 days. In two-thirds of the studies, there were no suicidality events.

When analyzed separately, the odds ratios exceeded 1 (indicating an increased risk) for all but 3 of the 11 AEDs. (The odds ratio was under 1 for carbamazepine, which had the second fewest patients in studies; and for divalproex; an odds ratio for felbamate could not be calculated because there were not many patients in the trials and there were no suicidality events among patients on placebo or the drug.)

Based on this analysis, "we're quite comfortable saying there's causality between suicidality and AEDs," Dr. Russell Katz, director of the FDA's division of neurology products, said at the meeting. The signal was detected across different mechanisms of action, and appeared to be independent of the AED's mechanism, he said.

The statisticians on the panel agreed that the FDA's analyses were solid. Because of the size and randomized nature of the dataset, this was "a signal that is an important one for the field to be aware of," said Andrew Leon, Ph.D., professor of biostatistics in psychiatry, Weill Cornell Medical Center, New York.

The panels agreed in a 20-0 vote with 1 abstention with the FDA's conclusion that an overall increase in suicidality was shown for the AEDs analyzed and most agreed (15-5 with one abstention) that this finding should apply to all currently approved, chronically administered AEDs. Dr. Katz said the FDA had proposed applying the warning to all marketed AEDs because limiting the warning to the 11 AEDs in the meta-analysis could shift prescribing to other AEDs that may have the same signal, but weren't included in the analysis, which would also include older, less well studied drugs that have other safety and efficacy issues.

This was also a concern when the FDA added the suicidality black box warning to antidepressants, and for this reason the warning was added to the older tricyclic antidepressants because of concerns that the warning would drive physicians back to prescribing these older drugs.

During a press briefing after the meeting, Dr. Tom Laughren, director of the FDA's division of psychiatry products, explained that when the FDA panel voted in favor of a boxed warning for suicidality for antidepressant labels, safety and efficacy were considered, and the pediatric efficacy data "were quite weak," but in the case of the AEDS, the data for efficacy are robust, and the size of the effect was small compared with antidepressants.

He acknowledged that after the black box warning was added to the labels of antidepressants, there was evidence of a drop in prescriptions. However, he said in some cases, that dropoff might have been a positive development, because there was evidence of overprescribing.

The FDA usually follows its panels' recommendations, which are not binding.

BY ELIZABETH MECHCATIE

Senior Writer