On My Mind

The nature of evidence

Scientific abstracts for the December American Epilepsy Society annual meeting have just been published. The abstracts contain several interesting messages for the epilepsy community, which we will discuss in this and upcoming issues of the newsletter. One of the abstracts raises the important issue of what constitutes good evidence of effectiveness. How should the informed person with epilepsy decide whether something is worth trying? An illustrative example of the validation process can be found in an abstract from a hospital in Manchester, UK by Cooper and colleagues entitled "Magnetic Wrist Bracelets Are Ineffective in Epilepsy - a Randomized Controlled Trial."

A Google search using the phrase "magnetic bracelets" produces 245,000 hits, and three sponsored links. Adding the word "epilepsy" to the search results in 1420 hits. Should a consumer so wish, he or she easily can use the Internet to buy a magnetic wrist or ankle bracelet to treat epilepsy. How can we know whether this is a good idea?

Doctors, scientists, governmental agencies, and insurance companies all use a process to determine whether a possible treatment is effective. This method is called a randomized, controlled clinical trial. But that is not how the process of therapy discovery begins. Let's trace the path of identifying new treatments, and then return to explain controlled clinical trials.

The process of medical drug or device treatments begins with anecdotal evidence in the form of individual stories, testimonials, or observations suggesting that something might work. These stories may be spread by word-of-mouth or be told from patient to physician in a doctor's office. If the possible new therapy is of sufficient interest, then these anecdotal observations may find their way into the medical literature as case reports. Case reports often are selective, only reporting the instances of apparent benefit. If a physician tried a magnetic wrist bracelet on five patients for seizures and saw no effect, it is unlikely that she would write up that series for publication, or that a journal would accept it for publication if she did. On the other hand, if the five patients appeared to improve, then a paper describing the result might find its way into the medical literature. Physicians consider case reports as potentially interesting leads to new therapy, but not proof, since the number of people who failed to benefit is not known and it usually is not clear that the treatment was what produced the improvement.

Case reports and testimonials also are impacted by the placebo effect. Just the act of wanting to believe can powerfully influence the outcome of a therapy trial. Studies that compare a new drug treatment to the equivalent of a sugar pill placebo find that seizures improve in the range of 20 to 30%, just from the sugar pill. Why people improve with placebo is unclear, but it is a widely seen phenomenon and one that must be taken into account in judging the effectiveness of a new treatment.

Natural fluctuation of disease symptoms also need to be considered when judging the effectiveness of a possible treatment. Most people with epilepsy recognize that they have clusters of bad days or months, and then periods when seizures are absent for no obvious reason. If the person had put on a magnetic wrist bracelet at the start of the good interval, then the bracelet would have seemed marvelously effective; whereas, in truth, the bracelet was not responsible for the improvement. People tend to try new therapies when things are going badly. This means that natural fluctuation back to their normal seizure levels would likely happen over time, even without a new treatment.

Clinical researchers have developed a tool to more accurately determine whether a treatment is effective: the randomized, controlled clinical trial. Randomized means that a random method, such as a coin flip, determines who gets the treatment and who does not. Some trials use a cross-over design, where each patient participating gets a new treatment for some period of time and a placebo for another period of time. In a double-blind randomized trial, neither the patient nor the physician knows whether the patient is receiving the new treatment or placebo. The information is kept available in a sealed envelope in case of emergencies, since patient safety is considered more important than are the rules of a clinical trial. Statisticians who specialize in clinical trial design make sure that the trial group is big enough to be able to come up with a yes or no answer on effectiveness of the new treatment. At the end of the trial, comparison is made between the participants given the placebo and those given the new treatment. If the new treatment is truly effective, significantly more people in the treated group will have benefited compared to those in the placebo group. Most clinical trials have a long-term follow-up phase in which the placebo-treated patients can try the new treatment.

Agencies such as the United States Food and Drug Administration (FDA) or the European Regulatory Authorities usually require one or more large randomized controlled clinical trials showing safety and effectiveness of a drug or device before approving it for use. Third-party payers, such as Medicare, Medicaid and insurance companies, also look for favorable outcomes of high class clinical trials before agreeing to pay for a therapy.

Let's circle back to our report on the magnetic wrist bracelet. The abstract for the epilepsy meetings reports a double-blind placebo-controlled cross-over trial in 35 patients who wore wrist magnets or dummy bracelets matched to look and feel the same. During eight weeks of testing, the average seizure frequency was approximately 22 seizures per month for the group with the dummy magnets (control group) and 28 for the group with the active magnets. The use of magnetic bracelets had no significant beneficial effect on seizure frequency. I mention this trial not because of the importance of magnetic wrist bracelets, but as an illustration of the methods that are required to prove that a treatment is or is not effective. Consumers do not design and run clinical trials, so what can you do yourself to participate in the process of evaluating possible new treatments for epilepsy?

1. Be skeptical and look at the level of evidence. If only at the level of case reports, consider the idea potentially interesting but unproven. Existence of at least one large randomized controlled clinical trial showing effectiveness is desirable. Even then, you should recognize that a treatment successful in a group of patients might not be effective or safe for you as an individual.

2. Use your physician as a resource to help evaluate the level of evidence for possible treatments. Many physicians are trained to do this, and are willing to take the time. It is your job to be an informed consumer, but the physician's job to know how to navigate evidence discussions in the medical literature.

3. Pay attention to specifics. A drug or device may not work for some conditions but might for others. Avoid painting with a broad brush. For example, magnetic bracelets wrapped around arthritic joints may be effective against pain according to at least one randomized, controlled clinical trial. That does not mean they are effective against seizures. Magnetic fields applied to the brain, where seizures originate, might be effective when wrist bracelets might not. The specifics of the trials and of the claims matter.

4. Consider risk versus benefit for you. Not every new treatment can have a randomized, controlled medical trial because they are effortful, time-consuming and expensive. This leaves a lot of unknowns. You may be using an unproven therapy that seems to benefit you. In the absence of evidence for or against effectiveness, you will have to balance the possible benefit against safety, convenience and cost of the therapy. Sometimes, it is reasonable to continue a treatment because it seems to be helping, it seems to be safe and you can afford to pay for it yourself. This is best discussed with your medical team.

5. Volunteer for clinical trials. Your doctor may ask you to join a clinical trial, either in the early case report stage or in the later randomized controlled clinical trial stage. Consider the potential benefits and risks of participating in the trial and give it serious thought. Your participation in research should always be voluntary. However, without such participation new treatments will never be developed.

Stay informed, but avoid chasing the newest fad. Know the level of evidence behind the treatments that you are receiving. Do everything you can to support research and testing of new therapies, because that will point the way to the cure for epilepsy.

The statements in this column are solely the opinion of the author, and not necessarily positions of epilepsy.com or the Epilepsy Therapy Project.

What's New?

Generic version of levetiracetam

A generic version of levetiracetam, previously only available as brand-name Keppra from UCB Pharma, has been released by Mylan Pharmaceuticals in a generic form. Mylan announced on 11/4/2008 that the United States Food and Drug Administration (FDA) approved generic versions of levetiracetam in 250, 500 and 750 mg sizes. Generic medications typically cost less than do brand-name medications. The FDA requires the generic medication produce bloodstream levels in the range of 80-125% of the level produced by the comparable brand-name medication. According to Mylan's press release, levetiracetam sales in the United States were approximately $1 billion for the year ending 9/30/08. Individual pharmacies may require weeks or a few months to receive and stock generic levetiracetam.

A new seizure drug retigabine is effective in trials

At the 2008 meeting of the American Epilepsy Society in Seattle, investigators French and Mansbach reported that the new antiepileptic medication, retigabine, was effective against seizures in a multicenter, randomized, double-blind, placebo-controlled trial. Previously completed trials showed retigabine to be effective against seizures in a range of different dosages. Retigabine has a novel mechanism of action involving activation of potassium channels in nerve cells, resulting in reduced excitability of brain cells and less tendency of the nerve cells to be involved in seizures. In this trial, the drug was given as 400 mg tablets three times a day to people with epilepsy at least 18 years of age and with at least four partial onset seizures per month. The medication dosage was increased over six weeks. During the eight weeks of the trial, they continued their baseline antiepileptic medications or other therapies. Overall, 301 patient centered the trial and 256 entered the maintenance phase. During maintenance, 8% of the patients on retigabine were seizure free versus 2% of patients on placebo. The most common side effects causing discontinuation included dizziness, confusion and fatigue, usually occurring during the buildup phase of retigabine. From this trial and others, it appears that retigabine is a useful new antiepileptic medication, and one with a novel mechanism of action involving brain potassium channels. Retigabine is sponsored by Valeant Pharmaceuticals International. It is not currently available for use by prescription, but is under consideration by regulatory authorities.

A new seizure medication, eslicarbazepine, appears effective

Two studies of the new drug eslicarbazepine are reported in the abstracts for the Dec, 2008 meeting of the American Epilepsy Society. Eslicarbazepine, brand name Zebenix™, is a new epilepsy drug, related to carbamazepine (Tegretol™) and oxcarbazepine (Trileptal™). Discovered in 1996, it may have fewer drug interactions than do carbamazepine and oxcarbazepine. Ben-Menachem and colleagues from Sweden and Portugal report results of a randomized controlled trial in 393 patients given either 0 (placebo), 400, 800 or 1200 mg of drug in a single daily dose. Seizures were reduced by an average of 33% with the 800 and 1200 mg dose, significantly better than placebo. About 20-25% of the patients in the higher dosage groups discontinued treatment because of side effects, most commonly dizziness, sleepiness or headache. A second trial by Czapinski and colleagues, performed in Poland, Hungary, Germany and Portugal also enrolled about 400 patients, and found similar outcomes, with a 43% median seizure reduction rate at doses of 1200 mg per day. The medication is sponsored by BIAL- Portela & Co, SA. Eslicarbazepine appears to be effective for epilepsy and is under consideration for release by European and US regulatory authorities.

Remote Telemedicine for Epilepsy

A group led by Syed and associates from Canada, Pakistan and in the United States presented a report on remote medical education conferences linking physicians halfway around the world. This is but one example of the rapidly growing field of telemedicine. Use of the Internet, digital cameras and microphones and computer transmission of x-ray images and pictures also are enabling epilepsy specialist physicians to examine patients at a distance. In a 2005 article in the journal, Epilepsia, Rasmusson and Hartshorn of the University of Texas in Galveston showed that outcomes were as good for 83 patients treated by telemedicine as for 72 patients at the local clinic. The patient was presented remotely by a specially-trained nurse. In a 2008 update of data, Hartshorn and Rasmusson discussing extending telemedicine to 800 epilepsy patients in 8 remote sites. Given that most cases of epilepsy in the world are located far from epilepsy specialists, the technology of telemedicine may provide a useful new approach.

Take a Look

New Consumer Journal: Starting a New Consumer Journal: Epilepsy.com is launching a new journal about epilepsy, but written, not by doctors, but by people with epilepsy or those who care for them. To participate or to learn more, click here . . .

Videos: We continue to organize the video recordings on the site, to make them more accessible. Try the video icon showing the camera in the left column and browse at will. The new set soon will have links to transcripts, if you would like to read or have the text of the videos.

Medication Compliance: Dr. Joyce Cramer, President of the Epilepsy Therapy Project, has expanded the section on compliance: how to remember to take your medicines. Take a look under "Treatment" - "Medication Compliance." The prior section has been vastly expanded.

Seizure Safety: Seizures can be dangerous, and so can treatments to prevent seizures. A new letter/handout on seizure safety is filed under the "Living with Epilepsy" - "Safety section." Several other subsections explore safety issues in more detail.

Holiday Giving and Gifts: In the holiday season, and for the rest of the year as well, use epilepsy.com to shop on-line, by clicking on the shopping bag icon in the upper right corner of the home page. You pay no more to shop with epilepsy.com, but a percentage ranging from 2 -30% will be donated for epilepsy research. So why not buy here? Here is the extensive list of vendors on shop.epilepsy.com.

1-800-FLOWERS.com • AT&T • BabyCenter Store • Back to Basics Toys • Bare Necessities • Barnes & Noble.com • Bed & Bath • Bliss World • Blockbuster • Buy.com • Cabelas.com • Calyx & Corolla • Casual Male XL • Cheryl & Co • Choice Hotels • Dell • Designer Linens Outlet • Diapers.com • Discovery Store • Eastbay • eBags.com • figleaves.com • FootSmart • FTD.com • gap.com • GiftCertificates.com • H&R Block • Harry and David • homedepot.com • Hot Topic • hsn.com • J&R Computer/Music World • JCPenney.com • Jockey.com • Kmart.com • Kohl's • Lands' End • Lenox.com • Limited Too • Macys.com • Magazines.com • Mimi Maternity • Motherhood Maternity • Netflix • newegg.com • Nordstrom.com • Northern Tool • OfficeMax.com • oldnavy.com • Orbitz • Overstock.com • Pacific Sunwear • PETCO • Piperlime • ProFlowers • Ross-Simons • Sahalie • Saks Fifth Avenue • Sears.com • ShopNBC.com • Sierra Trading Post • Smith & Hawken • Sony Style • Staples.com • StarbucksStore.com • The Children's Place • Things Remembered • T-Mobile • Toys"R"Us • Travelocity.com • ValueMags.com • VistaPrint • VitaminShoppe.com • Walmart.com • Williams-Sonoma.

Mail-Order Pharmacies: Dr. Peter Bridgeman, a neurologist in Maine, provides some thoughts on how to deal with Mail-Order Pharmacies. Take a look at the "Treatment" - "Seizure Medicines" - "Mail Order Pharmacies" section.

Robert Fisher, MD, PhD
Editor-in-Chief