Abstract
Seizure Outcome after Switching Antiepileptic Drugs: A Matched, Prospective Study
Finamore JM, Sperling MR, Zhan T, Nei M, Skidmore CT, Mintzer S. Epilepsia 2016;57:1294–1300
Seizure outcome after switching antiepileptic drugs: A matched, prospective study OBJECTIVE: Outcomes after changing antiepileptic drugs (AEDs) have largely been studied in single cohort series. We recently reported the first study to examine this question in a controlled manner. Here we expand on these results by using a matched, prospective methodology applied to both uncontrolled and well-controlled patients taking any AED. METHODS: We reviewed all outpatient notes over a 9-month period and identified patients with focal epilepsy who were on monotherapy. We classified those who switched AEDs as case patients, with those remaining on the same drug serving as controls. We matched cases with controls for seizure status (seizure-free in the preceding 6 months or not), current AED, and number of failed AEDs. We subsequently assessed outcome 6 months later. RESULTS: Seizure-free patients who switched drug (n = 12) had a 16.7% rate of seizure recurrence at 6 months, compared to 2.8% among controls remaining on the same drug (n = 36, p = 0.11). There was a 37% remission rate among uncontrolled patients who switched drug compared to 55.6% among controls (n = 27 per group, p = 0.18). Uncontrolled patients who had previously tried more than one AED were somewhat less likely to enter remission (p = 0.057). Neither AED mechanism of action nor change in dosage impacted outcome. SIGNIFICANCE: Herein we provide further estimation of the modest risk (~14%) associated with switching AEDs in patients in remission compared to being maintained on the same regimen. Uncontrolled patients were no more likely to enter remission after a drug switch than they were after remaining on the same drug, suggesting that spontaneous changes in disease state, and not drug response, underlie remission in this population.
Commentary
In this article, Finnamore et al. are seeking the answer to the most fundamental question that epileptologists ask as they learn the skills of antiepileptic drug (AED) selection: does drug manipulation matter? To answer this question, the authors performed an observational prospective-matched cohort study in patients with focal epilepsy on monotherapy who were either seizure free or not seizure free at the time of a drug switch, to determine if a switch changed their status for the better or worse.
Unfortunately, their numbers were relatively small (<5% of the 401 seizure-free monotherapy patients in their practice switched drugs during the observation period, and surprisingly, only 25% of the 144 who were not seizure free had a drug switch). Also, the time frame of observation was quite short. Patients were categorized as seizure free or not based on the 6 months prior to the index date, and the 6 months following. Thus, the study does not provide definitive answers.
In a surprising turn of conventional wisdom, their conclusion was that drug switches might not make patients better, if they were having seizures on their prior regimen, but it could make them worse, if they were previously seizure free. The risk of destabilization they observed in this study is similar to what was seen in a previous study from the same center (1), which provides credence to the conclusion that patients have an approximately fifteen percent risk of seizure recurrence with a drug switch, versus approximately three percent if drugs remain the same. This is bad news for patients who have side effects, because they may have to choose between seizure security and improvement in general function. Importantly, neither the present study, nor any other study has systematically examined whether drug switches in seizure-free patients actually accomplish their goals, which is usually to reduce or eliminate treatment-related side effects, because side effects of many commonly used AEDs are overlapping.
What about the patients who were presumably switching because their seizures were not controlled in the prior 6 months? Here the story becomes more complex. The authors suggest that because many of the control patients “became” seizure free without a change in drug, and in some cases without even a change in drug dosage, that improvement of treatment-resistant patients can be explained as much from spontaneous “remission” as from drug changes. This conclusion may be premature. The International League Against Epilepsy (ILAE) committee that produced the guidelines on treatment resistance also provided a definition of treatment response (2). A patient is determined to be drug responsive only after they have remained seizure free for at least three times the previous inter-seizure interval. Because, by the authors’ own admission, “a considerable fraction” of patients in the treatment resistant group had only one seizure in the prior 6 months, and the interval from their previous seizure is unknown, 6-months or even 1-year seizure free would be insufficient to declare a “remission.” In other words, in a patient who only has a seizure frequency of one in less than 6 months, a 6-month seizure-free interval would not be that surprising. Thus, there may be too much noise in the study to determine if drug switches are beneficial or not. Another interesting aspect of the study is that there is no mention of add-on therapy, which is understandable in the seizure-free patients, but surprising in the patients with seizures. Would the patients have done better with add-on therapy than a switch?
In summary, the study provides substantive and clinically important evidence for physicians and patients contemplating a drug switch in the presence of seizure freedom. Physicians should counsel patients that there is a risk of seizure destabilization, and the risk versus the benefit of a drug change should be discussed. But what of patients who are not seizure free? Should we call it quits after two drug trials, with the understanding that “many patients with epilepsy have a disease whose mechanism is radically different from the ones addressed by all currently available AEDs, which appear to largely treat the same group of drug-amenable patients” as the authors suggest? The authors conclude, “most outcome changes seen in patients with resistant epilepsy may be spontaneous rather than due to therapeutic alterations.” This may certainly appear to be true when the rules for determining treatment response (waiting 3 times the previous seizure-free interval) are ignored, but as of yet, there is no evidence supporting or refuting the benefit of drug manipulation in a study with sufficient follow-up to determine true outcomes. There are also many other factors to consider, including the ability of AEDs to produce meaningful seizure reduction in the absence of complete seizure freedom, the option of adding rather than switching, and the desire to eliminate potentially dangerous seizures such as generalized tonic clonic seizures, to name a few. A definitive study has not been done, but until it is, it seems more than prudent to continue best efforts to improve seizures in patients who are not seizure free.