Rodney A. Radtke, MD: The guidelines of the American Academy of Neurology and the American Epilepsy Society are helpful in that they are based on evidence-based literature so that they’re the best-quality studies we have and they produce information about what we can prove using these drugs to treat epilepsy. Their weakness is that they’re very limited in terms of these prospective controlled studies that don’t frequently resemble how we take care of patients in the clinic. And so obviously you need clinical experience and expert experience to supplement that in order to make a decision about how to treat a patient.

For someone like me who’s an expert or an epileptologist, they really don’t have much impact on our practice. In other words, that’s the literature that we already know and that we’ve used to build our practice. For the less experienced or the clinician who’s less invested in the treatment of epilepsy patients, it can be very useful to know what’s in the literature to support the use of the various drugs in the various situations. So it’s helpful in a limited way, but the more experience you get, I think they become less and less important in terms of affecting your treatment decisions.

I’ve been practicing for a long time, and when I initially started the practice, we had only 5 drugs to choose from, so it wasn’t too complicated. And we had to use those drugs in patients in order to gain control of their seizures. The biggest change over the years with the advent of many new medicines is improved tolerability. We now have drugs that are much better tolerated, have fewer drug-drug interactions, and likely have much less in the way of long-term adverse effects. So when I’m choosing drugs now, I pretty much use only the newer drugs, and by that, I mean drugs from 1995 on, where tolerability is better, safety is good, and efficacy is at least comparable with what we had back in the good old days.

It’s important to understand that the use of blood levels in monitoring antiepileptic drugs is not evidence-based. There are few if any studies that are really class 1 studies looking at the role that blood levels should play in managing patients. And there is a wide variation in clinical practice among epileptologists in the field. I think I’m on the side that likes to use drug levels more. I find them useful. I think we all agree that drug levels are important and helpful early in the management of a patient, where we establish what we think is an appropriate therapeutic dose. We check a level, and that reinforces that this is the patient on 200 mg and this is their level; and then, as you go forward, you can make comparisons. It helps guide maybe what dose you choose for some agents, and it also helps check for compliance or adherence down the road when they show up in the ER [emergency department] for a breakthrough seizure. You do a level, their level is nondetectable, and it’s very clear that nonadherence has been a major contributor to their seizure.

There are certain drugs for which I think levels are more helpful or more important, and those are drugs that have a lot of drug-drug interactions or that have a wide interindividual variation between dose and level. For an example, lamotrigine is a drug in which I find levels very helpful because there are drug interactions in terms of enzyme inducers lowering the level, estrogen lowering the level, and valproate raising the level. In addition, if I give 1 patient 200 mg, their level could be 4 and another patient might be 12. And so that interindividual variation and the drug-drug interactions make AED [antiepileptic drug] levels very helpful in a drug like lamotrigine.

Conversely, levetiracetam, I think, is a drug in which levels are limitedly useful in that there are no drug-drug interactions. Usually, there’s a very close correlation between patients and dose in terms of what the level is, and also the adverse effects don’t seem to be parallel with the dose or the level. So it’s a drug that can be used very readily without the use of levels and not be very limited.

Many people ask how the mechanism of action should affect drug selection. For me, not much. There are some people who rely heavily on taking 1 from column A, a sodium channel blocker; 1 from column B, a GABA [gamma-aminobutyric acid] enhancer; and maybe 1 from column C, a different mechanism. But we have to understand that the mechanisms are proposed, not usually proven or documented. And even if you’ve used a sodium channel drug, it’s very clear that a different sodium channel drug has efficacy and may be the next-best agent. That said, I do like to spread around the options after I’ve used a couple of sodium channel drugs, say lacosamide and lamotrigine. I’m probably not going to go there for a third sodium channel drug, and I’m going to move to something different, such as levetiracetam with SV2A [synaptic vesicle protein 2A] binding or maybe a GABAergic drug or a multipurpose mechanism like topiramate.