Abstract
Irelli EC, Cocchi E, Morano A, Gesche J, Caraballo RH, Lattanzi S, Strigaro G, Catania C, Ferlazzo E, Pascarella A, Casciato S, Quarato P, Pizzanelli C, Pulitano P, Giuliano L, Viola V, Mostacci B, Fortunato F, Marini C, Di Gennaro G, Gambardella A, Labate A, Operto FF, Giallonardo AT, Baykan B, Beier CP, Di Bonaventura C; Women With Epilepsy Treatment Options and Research (WETOR) Study Group. JAMA Neurol. 2023;80(11):1174-1181. doi:10.1001/jamaneurol.2023.3400
After the recent limitations to prescribing valproate, many studies have highlighted the challenging management of female patients of reproductive age with idiopathic generalized epilepsy (IGE). However, no study, to the authors’ knowledge, has addressed the comparative effectiveness of alternative antiseizure medications (ASMs) in these patients. To compare the effectiveness and safety of levetiracetam and lamotrigine as initial monotherapy in female patients of childbearing age with IGE. This was a multicenter, retrospective, comparative effectiveness cohort study analyzing data from patients followed up from 1994 to 2022. Patients were recruited from 22 primary, secondary, and tertiary adult and child epilepsy centers from 4 countries. Eligible patients were female individuals of childbearing age, diagnosed with IGE according to International League Against Epilepsy (2022) criteria and who initiated levetiracetam or lamotrigine as initial monotherapy. Patients were excluded due to insufficient follow-up after ASM prescription. Levetiracetam or lamotrigine as initial monotherapy. Inverse probability of treatment weighting (IPTW)–adjusted Cox proportional hazards regression was performed to compare treatment failure (TF) among patients who received levetiracetam or lamotrigine as initial monotherapy. A total of 543 patients were included in the study, with a median (IQR) age at ASM prescription of 17 (15-21) years and a median (IQR) follow-up of 60 (24-108) months. Of the study population, 312 patients (57.5%) were prescribed levetiracetam, and 231 (42.5%) were prescribed lamotrigine. An IPTW-adjusted Cox model showed that levetiracetam was associated with a reduced risk of treatment failure after adjustment for all baseline variables (IPTW-adjusted hazard ratio [HR], 0.77; 95%CI, 0.59-0.99; P = .04). However, after stratification according to different IGE syndromes, the higher effectiveness of levetiracetam was confirmed only in patients with juvenile myoclonic epilepsy (JME; IPTW-adjusted HR, 0.47; 95%CI, 0.32-0.68; P < .001), whereas no significant differences were found in other syndromes. Patients treated with levetiracetam experienced adverse effects more frequently compared with those treated with lamotrigine (88 of 312 [28.2%] vs 42 of 231 [18.1%]), whereas the 2 ASMs had similar retention rates during follow-up (IPTW-adjusted HR, 0.91; 95%CI, 0.65-1.23; P = .60). Results of this comparative effectiveness research study suggest the use of levetiracetam as initial alternative monotherapy in female patients with JME. Further studies are needed to identify the most effective ASM alternative in other IGE syndromes.Importance:
Objective:
Design, Setting, and Participants:
Exposures:
Main Outcomes and Measures:
Results:
Conclusions and Relevance:
Commentary
The pathway to approval of new anti-seizure medications (ASMs) differs methodologically from the evaluation of best “first-line” monotherapy in epilepsy. Randomized placebo-controlled double-blind clinical trials of novel ASMs consist of rigorous placebo-controlled evaluations in patients with drug-resistant frequent seizures, while monotherapy has been evaluated through pragmatic open-label trials, which inform comparative effectiveness, often in “real world” practice. 1 -4 This difference largely lies in the practical and ethical difficulties in performing a rigorous placebo-controlled randomized controlled trial of monotherapy in newly diagnosed epilepsy, where the primary outcome (persistent seizures) is infrequent, necessitating a large sample size to detect an effect, and a placebo group would be unethical, given the long double-blind period needed to assess the outcome and the lack of equipoise on the risks of delayed treatment in a newly diagnosed epilepsy. But how do we draw real world conclusions from real world study designs, and are the questions they are aiming to answer the right ones?
In this study, a multicenter cohort of 543 Italian biological women of childbearing age with idiopathic generalized epilepsy (IGE) were followed after prescription of either lamotrigine or levetiracetam as initial monotherapy. 5 In this open-label study, in which drugs were discontinued based on physician’s “best judgment,” levetiracetam was associated with a lower risk of “treatment failure,” defined as either drug discontinuation in favor of another ASM due to ineffectiveness or lack of tolerability, or the addition of a second ASM due to ineffectiveness. After a subgroup analysis of subtypes of IGE (juvenile myoclonic epilepsy [JME], absence epilepsy, and generalized tonic–clonic convulsions upon awakening), the lower rates of treatment failure with levetiracetam only occurred in the group with JME. Notably, the levetiracetam group had a significantly higher rate of behavioral events (16% with levetiracetam compared to 1% with lamotrigine), although this difference in tolerability did not appear to drive greater discontinuation of the drug. In focal epilepsy, levetiracetam has fared worse than lamotrigine on retention rates, perhaps due to tolerability. 4,6 The exacerbation of myoclonus in 3.3% of patients treated with lamotrigine was thought to potentially contribute to the greater rate of “treatment failure” in the lamotrigine group, although it is unclear if an add on medication for myoclonus led to this treatment failure, or a complete switch to a different medication was undertaken. The authors conclude from these findings that “the potentially detrimental effect of sodium-channel blockers on this seizure type [myoclonus]…support the use of levetiracetam as initial alternative monotherapy in patients with JME.”
There are questions one may ask on the specifics of this study. For example, how often were women with behavioral side effects from levetiracetam offered a switch by their treating physician or encouraged to “stick it out”? Did women with myoclonus on lamotrigine elect to add on a small dose of another medication (therefore “failing treatment” with lamotrigine), with the plan to potentially go back to lamotrigine monotherapy in the event of pregnancy planning?
If one is to take the validity of a pragmatic open-label comparative effectiveness cohort study at face value, this conclusion that levetiracetam is the appropriate first-line therapy for JME in women of childbearing age is sound. Indeed, in this study, levetiracetam does have lower treatment failure, at least in JME, in a majority white Italian population. But should we really offer levetiracetam in all women of childbearing age with JME first? Or is epilepsy care more nuanced than this conclusion would have us believe?
Take this case: a young woman with a history of treatment-resistant depression presents after her second lifetime convulsion. She experiences occasional myoclonus, which is not bothersome. Her EEG confirms a diagnosis of JME. Even with the results of this study, would one still offer levetiracetam over lamotrigine? Probably not, given the known propensity for levetiracetam to cause exacerbation of psychiatric symptoms. A person with epilepsy is more than their seizure outcome, and in this case, comorbid depression is sure to drive medication choice. Another young woman may present with frequent disabling myoclonus and voice freedom from myoclonic seizures, not just tonic–clonic seizures, as her primary goal for treatment. This woman may elect to start levetiracetam, which is an effective myoclonic drug. Lastly, a woman of childbearing age may present to clinic but clearly state she has no intention of ever being pregnant and wishes for the most effective drug for generalized epilepsy, 1 and therefore be offered valproate. In fact, there are many more permutations of various patient-specific comorbidities and impact of individual seizure types and risk of drug side effects that will influence a physician’s choice of first drug.
The pragmatic open-label comparative effectiveness trials attempt to simplify the problem of treating a heterogeneous condition in a group of patients with many comorbidities by addressing a simple question: what is the efficacy and tolerability in a group with a type of epilepsy (focal vs generalized) as a whole, by defining treatment failure in a way that encompasses efficacy and tolerability. The effort is laudable—many disciplines have evidence-based first-line therapies for various ailments. Why not epilepsy too? Isn’t the goal of classification of epilepsies to enable us putting patients in “buckets” that have specific treatment responses and improve treatment strategy? The SANAD trials in generalized epilepsy offers useful data, which is that valproate has superior efficacy over other drugs. 1 But this finding is not helpful for people of childbearing age who may desire a pregnancy, which is why this Italian study 5 was undertaken. The issue of people with pregnancy potential is not the only one though, which is why this Italian cohort study fails to answer the question—what should a neurologist offer to the individual presenting to clinic. Some have embraced these inherent challenges to picking one best drug for all, and instead developed algorithms to find a potential best first drug for one. 7,8
The authors of this study conclude that “great uncertainties still exist regarding the best ASM to prescribe as initial monotherapy in these patients.” This statement cannot be disputed. The question is how to get to this answer. If there was a “perfect” ASM—well tolerated, easy to administer, safe in pregnancy, high efficacy, then we would likely have an undisputed first-line therapy for focal or generalized epilepsies. Unfortunately, every anti-seizure medication comes with some drawback—whether risk of specific side effects, need for titration, need for blood monitoring, teratogenicity, or less than ideal efficacy. There still is no first-line monotherapy for focal or generalized epilepsy. But the reason for this may not be that we have not found the right study design, but perhaps that the right drug is not here yet…
