Q: How do you proceed with AED therapy when there is uncertainty regarding the patient’s seizure type?

It is often best to use an antiepileptic drug (AED) that safely covers both types of seizures when choosing the first AED for a patient who has had a seizure of an uncertain type. In our practice, we have a full hour to spend with new patients, and we are able to make phone calls to ask witnesses about the seizure, but this may not be possible in all practice settings. Moreover, distinguishing clinically between focal onset seizures with secondary generalization and primary generalized seizures is, at times, challenging; there may not be any abnormalities on testing. Some clinical clues can suggest that a patient has 1 type over another. For instance, the presence of a reliable aura prior to the seizure, a déjà vu sensation, epigastric rising, or unusual smells or tastes all suggest a focal onset seizure, or what is now termed a focal seizure without impairment of awareness. Myoclonic jerks or absences would suggest a primary generalized onset; however, these details are not always available. Additionally, some patients with focal onset seizures have no aura prior to loss of awareness and prior to a secondarily generalized tonic-clonic seizure, and there may not be caregivers or loved ones who witnessed the seizure to describe it. It is also worth noting that some AEDs that are approved for focal onset seizures, particularly older agents such as carbamazepine and phenytoin, can exacerbate certain types of generalized seizures, such as absence seizures and myoclonic seizures. I have seen instances of patients with those types of seizures ending up in the emergency department with status epilepticus after being inappropriately prescribed those drugs. For these reasons, without an actual electroencephalogram (EEG) recording, it may be difficult to tell what kind of seizure has occurred, making it more appropriate to aim broadly, especially with the availability of broader-spectrum, safe, and potentially generic AEDS that patients should be able to get, regardless of their insurance coverage.

This is not an uncommon scenario, whether someone has had an event in their sleep, or perhaps the onset of the seizure is not witnessed, or they wake up with bladder incontinence and are clearly postictal and anamnestic for the event. We have high confidence that this is a seizure, but we have no idea what the beginning of it looked like because of the aforementioned obstacles. The results of the magnetic resonance imaging scan may be normal, along with the first and/or second routine EEG results. For these patients, historically, we often chose a broad-spectrum AED for treatment, such as valproate. However, in the modern era, this is typically not the drug we use in almost half of our patients (ie, female patients, once they reach childbearing age) because of the risks of teratogenicity (both structural and neurobehavioral). Moreover, whereas for patients who clearly have focal onset seizures, we potentially have more than 20 AEDs to choose from, that number falls to really just a handful of agents if we are not sure whether it is a secondarily generalized tonic-clonic seizure or a genetic generalized tonic-clonic seizure (ie, generalized from the start). It is helpful for clinicians to realize that, other than valproate, the only broad-spectrum options for many of these cases of uncertain seizure type are lamotrigine, topiramate, perampanel, and levetiracetam. And then, depending on the age and sex of the patient, you may be down to only 1 or 2 option(s). Valproate and topiramate, for example, are not going to be used in a teenaged female patient because of the teratogenicity. Thus, although it seems like there should always be numerous AEDs to choose from, once you introduce uncertainty about the seizure type and consider other patient factors, you may have limited options.

Ninety percent of the time in our type of practice, we obtain a nice, elegant, clear diagnosis (eg, juvenile myoclonic epilepsy, right parietal lobe epilepsy, left temporal lobe epilepsy). But that is only in our type of practice (a level 4 referral epilepsy center), where we see patients with intractable epilepsy who have been referred for EEG video monitoring. Unfortunately, even with our attempts to provoke a seizure during a 5-day stay in an EEG video monitoring unit, it is possible for seizures to be infrequent enough to escape our detection yet frequent enough to prevent a patient from driving. Additionally, in the real world (eg, in a general neurology practice), where neurologists see unselected patients with epilepsy, the seizure types may be more challenging to define based on the available data. The results of the patient’s magnetic resonance imaging scan and routine EEG are most often normal. The patient definitely has seizures, but whether they represent focal onset or primary generalized epilepsy is often unknown. The wise thing to do in these situations is to use broad-spectrum AEDs that will treat both scenarios. We have several of them to choose from, including valproate in the old generation, and newer agents such as levetiracetam, lamotrigine, topiramate, zonisamide, and perampanel. They are all broad-spectrum AEDs and they are efficacious in both types of epilepsy. In these scenarios, there is no reason to choose a narrow-spectrum AED such as phenytoin, carbamazepine, oxcarbazepine, or pregabalin.