Epilepsy: Recommendations on Discontinuing Therapy

Michael R. Sperling, MD: Eric, assuming your patient is taking his medicine and doing quite well for an extended period of time, and he raises the question to you, “Do I still need to take it?” When do you decide to just recommend that someone consider discontinuing treatment? I think we all ultimately put the decision in the hands of the patient or the parents, or both actually for kids old enough to think a little bit. But when do you make that? And how do you make that decision?

Jesus E. Pina-Garza, MD: There are so many scenarios. There are some cases in which we have a history and an electroencephalogram [EEG] that tells you that this is likely to be active until adolescence, like rolandic epilepsy. This may be active for a shorter period of time, like occipital epilepsy. But the presence of the electrical abnormality in the EEG tells you it is still present. So when it’s active and it’s being treated and controlled, as long as it’s present, it makes sense to continue treatment.

There are some situations in which we have a structural problem on the brain that is not going to go away. That scar is there unless we remove it surgically. But if the person is issue-free, potentially that can be a lifetime treatment. So what you don’t want to have is something that you tolerate short term, but it can make you osteoporotic or have other problems long term. We may have epilepsies that have adolescent onset, and there it’s connected with the EEG to tell you they are still active and rarely outgrown. So in those cases, again, it’s unlikely.

So the case comes more for the kid to have a known etiology, and 2 or 3 seizures, and we decided to treat him. And in many cases we have patients who become controlled for 2 years, and we do still treat, but they don’t require treatment any more. And for us, again, the advantage is that there are kids who don’t drive. So it is very easy to remove the treatment and see if they still need the medication, with some kind of precaution. So I think with my older population, when they are driving, that makes it a little bit harder. And I know probably Kate can attest to that, that sometimes putting your driving on hold is not possible.

Michael R. Sperling, MD: Kate, let’s say a 40-year-old comes to you who has had 7 or 8 seizures over the course of 5 years, in her 20s. Her MRI [magnetic resonance imaging] was normal, her EEG was normal, they were focal-impaired awareness seizures, what we used to call complex partial seizures. They happen between ages 20 and 25, and she’s now 40. She’s been taking a single medication, say Lamotrigine at a medium dose, say 400 mg per day. And her last seizure was 14 years ago, for example. How do you help decide to discontinue or not?

Kathryn A. Davis, MD, MS, FAES: There are a lot of features that I would take into account in that situation. This woman is 40. I assume, maybe incorrectly, that she’s probably past her childbearing years at least. That plays a role. So if it was a 30-year-old woman who’s getting ready to plan a family, there’s a bigger advantage there to coming off medication. If she’s not having any adverse effects or having adverse effects, that would play a role. I assume she has been tolerating it fairly well because she’s been on it for quite some time. And then, as Eric pointed out, the driving is so key here. And what ends up happening in most of these instances is that we’ll talk to the patient in depth about whether to come off the medications—the risks of doing so, benefits and risks—and when driving comes up, and my typical recommendation is to not drive for a minimum of 3 months after stopping a seizure medication, that is a nonstarter for most patients.

Michael R. Sperling, MD: So by telling them they can’t drive you’ve made the decision for them. I must say, my own practice is quite different. I don’t tell them that they can’t drive. Because in my view, if we’re willing to contemplate stopping their medication, they should be willing to drive. And I may say, now what I do, is I will typically tell them to drive less. If you’re with another driver who can drive a vehicle, let that person drive. Avoid high-speed driving for several months because, again, the timing is very arbitrary. It also depends on how much time they spend driving. If it’s someone who is in the car for an hour or so a week, now the odds of an accident are very small. If the person is in the car for a few hours everyday driving, the odds are very high and that’s going to make a difference.

Kathryn A. Davis, MD, MS, FAES: Or if they’re driving their children around. There are a lot of factors. I think I tend to be, at least at the University of Pennsylvania, the strictest on driving. And that perhaps just reflects on my own demographics and concern about other people’s children and in cars, and my own children in my car. So I think that with driving, the way that I represent this to patients influences whether they want to come off. But honestly, to the earlier point, our newer medications are very safe in the long term. And if the patient is tolerating the medication, I don’t really know if I would personally—if I was making the decision—come off.

Michael R. Sperling, MD: I think I’m with you there. There’s a very nice prediction tool that’s available on the Web. It was developed by Herm Lamberink in the Netherlands. And if you just do a search in your favorite engine—and most of our favorite engines begin at G if we’re like most Americans—you’ll find Epilepsy Prediction Tools. You can put it on your phone, so it’s right on your home screen. You put in a few key information details—age, number of seizures, whether the EEG is normal or abnormal, family history, a few other things, such as if MRI is normal or not, and time since last seizure—it will very quickly give you an estimate of the probability of relapse within 2 years or 5 years. And it’s available for people who have not had surgery. There’s a separate prediction tool for people after surgery. And the numbers that the tool gives are actually strikingly high. There’s, like, 15-year seizure-free history in that woman with everything being normal, so probably she still probably has a 12% to 15% chance of having a relapse. Which sounds awfully high to me—a 15% percent chance.

Trevor J. Resnick, MD: Yes.

Michael R. Sperling, MD: Now that’s a 15% chance over 2 years, and how long does a seizure last? A seizure lasts a few minutes. So again, you know, what percentage in any 1 minute, what are the odds of having that seizure? Incredibly small, which is why I’m generally willing to let people drive.

Trevor J. Resnick, MD: Right.

Michael R. Sperling, MD: With still the cautions of, “Don’t drive fast, because you’re much more likely to die at 60 mph than at 30 mph.” But it’s a very interesting thing. And when I put it up on my phone, show it to the patients who invariably look at it, and nobody wants to go off medicine when they see that number.

Trevor J. Resnick, MD: Yes, sure.

Michael R. Sperling, MD: It’s remarkable, but it’s a very nice thing.