Evidence-based medicine is a buzzword these days for major medical societies trying to offer their members the most clinically sound and meaningful recommendations for treatments. No one disputes that the federal Agency for Healthcare Research and Quality (AHRQ) has taken leadership on the issue, as well.
But an AHRQ report released last December — comparing the effectiveness of older versus newer antiepileptic drugs — has some members of the epilepsy community wondering how the federal agency fell short of that goal. The latest report has “little or no clinical value,” say representatives of the American Epilepsy Society, the Epilepsy Foundation, and the AAN.
In its report, the AHRQ compared published findings on newer antiepileptic drugs (AEDs) such as lamotrigine and levetiracetam with studies on older ones like carbamazepine and valproic acid. It also compared generic versions of AEDs with original products. The lack of quality studies comparing such drugs made the task difficult, the report authors acknowledged, and precluded making any definitive statements regarding their efficacy.
Nonetheless, they concluded: “The older drug carbamazepine had advantages in epilepsy control over newer antiepileptic medications as a class, but had more adverse effects.”
The older drugs valproic acid and phenytoin also provided similar epilepsy control to newer drugs. Adverse events occurred more often with them, but did not significantly increase the risk of a patient discontinuing their use, according to the report. The time before a first seizure was lower with newer medications, but only when compared with valproic acid.
The AHRQ researchers also found no substantive differences in benefits or harms with generic versus originally approved products, although the findings were mostly limited to carbamazepine.
Because AEDs were grouped together based on whether they were older, newer, innovator, or generic, and because types of epilepsy were also lumped together, the findings were limited, the AHRQ admitted. Nonetheless, despite important differences between the different medications, the researchers said they were not significant enough to prevent them from being “pooled” together for group comparison purposes.
THE EPILEPSY COMMUNITY SPEAKS OUT
In a Feb. 14 letter to AHRQ Director Carolyn M. Clancy, MD, members of the three professional groups faulted the report for failing to reflect many of the issues and concerns raised by the organizations during its development; lumping epilepsy, a multiform disorder, together in the analysis; and for failing to differentiate specific drugs by their intended use for different types of epilepsy.
Michel J. Berg, MD, associate professor of neurology and medical director of the Strong Epilepsy Center at University of Rochester Medical Center School of Medicine and Dentistry in New York, helped draft the key questions the AHRQ researchers sought to answer. He said the report lacked clinical relevance because the authors failed to parse out important differences between different types of epilepsy and drugs that are useful.
“Different conditions require different medications. For example, we treat focal epilepsy very differently than idiopathic generalized epilepsy. I get the impression that the report was not written by individuals who are very familiar with the disease to make such broad conclusions,” he told Neurology Today in a telephone interview.
“If you look at the data in the report, the effects of drugs were small or moderate except for one category, and that was the potential overgrowth of gum tissue caused by phenytoin. One problem is that they concluded that carbamazepine is more effective, but they failed to specify for which type of epilepsy. My concern is that this will be misinterpreted. Very few of us would use carbamazepine for idiopathic generalized epilepsy,” he said.
Such a meta-analysis is not necessarily incorrect, but the problem comes with trying to apply it clinically in patients, Dr. Berg said. And even though the authors admitted the shortcomings in the study's test, “these did not make it into the conclusions.”
“You cannot just lump everything together, and that is something they did excessively, in my opinion,” he said. “My biggest concern is declaring carbamazepine superior without any explanation about which epilepsy types it treats. It can actually exacerbate seizures in people with idiopathic generalized epilepsy.”
With regard to generic versus original versions of these drugs, there is too little data to draw any conclusion at this time, he said.
James W. Wheless, MD, who was not involved with the report, agreed. “It is a vast oversimplification. With epilepsy, one size does not fit all,” Dr. Wheless, professor of neurology at the University of Tennessee Health Science Center, and director of the Le Bonheur Comprehensive Epilepsy Program at St. Jude Children's Research Hospital in Memphis, told Neurology Today in a telephone interview.
He noted that the major epilepsy organizations and the AAN have all published recommendations on treating epilepsy in its different forms, all based on current research findings, but these were apparently not considered by the AHRQ. Moreover, the AHRQ's findings are not in alignment with those that have been made by either the Food and Drug Administration or the NIH, he said. “In fact, they are the opposite.”
The AHRQ concluded that only three out of 19 AEDs approved in the US have value from a risk-benefit standpoint.
“Those of us who treat these patients know this is not true,” said Dr. Wheless. “There are huge gaps in this report. For example, most epilepsy is first diagnosed in pre-school children and geriatric patients, yet the report failed to address treatment in either group, nor does it deal with women of childbearing age, even though some AEDs are teratogenic.”
In addition, valproic acid is known to increase the risk of hepatotoxicity in children who are under two years of age, and can cause dermatological reactions in other children.
Michael Privitera, MD, professor of neurology and director of the Epilepsy Center at the University of Cincinnati Neuroscience Institute in Ohio, said the biggest concern among epileptologists is that insurers and third-party payors might use the study to limit coverage for newer medications without understanding the report's admitted limitations and shortcomings.
“I think the AHRQ felt the need to put out some type of review, but even the authors agreed that the results were not very meaningful because of limitations inherent in combining so many different studies,” he said.
“I am pretty disappointed. Everyone is in favor of evidence-based medicine, but this is what happens when combining studies of treatments for a disease that takes a number of different forms — unlike hypertension where you are dealing with just one [form].”
He noted that several of the newer drugs included in the study are very effective in certain patients, while others are barely used.
“Lamotrigine and levetiracetam are the most widely used AEDs today, but almost no one uses gabapentin for epilepsy any more. The key to using newer drugs is knowing their adverse event profile, and that detail is lost when the drugs are lumped together as in this study. Important adverse effects like cognitive impairments and depression were not assessed in many of the older studies used in this analysis, but we now regard these effects as key factors in choosing drugs.”
THE AHRQ RESPONDS
The criticism notwithstanding, the AHRQ was not ready to write off the report or effort. Critics of the report need to understand how difficult it is to reach conclusions about the efficacy and safety of AEDs given the lack of head-to-head studies to date, said Jean Slutsky, director of AHRQ's Center for Outcomes and Evidence.
“As emphasized in the report, there are few high-quality research studies comparing various medications for treatment of epilepsy and different types of epilepsy,” she said in an e-mail to Neurology Today.
“The report points out that this limitation precluded report authors from making strong conclusions about comparative effectiveness and risks of different medications,” she continued. “However, one of the cornerstones of the Effective Health Care Program is the high priority it places on identifying future research needs. Our report cites examples of potential studies that could have significant impact on this field of research.”
On the need for more research, the neurologists interviewed here were in agreement. But as for the current findings, not so much. Asked if the report will change treatment patterns among epileptologists, Dr. Berg responded, “I hope not.”