Switching to a New Epilepsy Drug: Is New Always Better?

Abstract & Commentary

By Padmaja Kandula, MD, Assistant Professor of Neurology and Neuroscience, Comprehensive Epilepsy Center, Weill Cornell Medical College

Dr. Kandula reports no financial relationships relevant to this field of study.

Synopsis: This randomized, single-center trial compared the efficacy and tolerability of monotherapy substitution with levetiractam vs carbamazepine and valproic acid in patients who failed initial treatment with first-generation antiepileptic agents.

Source: Hakami T, et al. Substitution monotherapy with levetiracetam vs older antiepileptic drugs. Arch Neurol 2012;68:1563-1571.

Based on the prior work of patrick kwan and martin brodie,1 The efectiveness of the first antiepileptic drug (AED) in new-onset epilepsy is nearly 50%. Conversion to a different monotherapy is then necessary for those who fail to become seizure-free or do not tolerate the initial agent. However, despite this common clinical scenario, very few evidence-based recommendations exist in the literature to guide clinicians. This clinical conundrum becomes further complicated by the existence of more than a dozen antiepileptic medications to choose from. In this randomized, single-center, open-label trial, the authors compare substitution monotherapy with levetiracetam, vs carbamazepine or valproic acid in those patients who have failed their first AED therapy.

Patients who failed initial monotherapy with either carbamazepine, phenytoin, or valproic acid for partial epilepsy based on efficacy or tolerability were eligible for this study. Patients were excluded on the basis of severe intellectual impairment; history of major psychiatric morbidity; history of substance abuse; recent (within 3 months) use of anxiolytic, antidepressant, or mood-altering substances; and pregnancy or intended pregnancy.

Eligible patients were then randomized for substitution monotherapy with either levetiracetam or a first-generation older antiepileptic agent, either extended-release carbamazepine or valproic acid. If the initial AED was carbamazepine or phenytoin, the patient was randomized to either levetiracetam or valproic acid. If the initial AED was valproic acid, the patient was then subsequently randomized to carbamazepine or levetiracetam. During a 4-week titration phase, the initial AED was weaned and the substituted AED was increased in twice weekly increments to a target dose of 1000 mg for both levetiracetam and valproic acid and 400 mg for carbamazepine. Subsequent adjustment was then left to the discretion of the treating neurologist. In pragmatic fashion, if seizures were uncontrolled on the allocated study drug, adjunctive therapy with another agent could be initiated. In addition, if adverse effects were intolerable with the randomized agent, patients were allowed to withdraw from the study altogether.

The primary outcome measures were improvement in quality of life (QOL) and depression as compared to baseline at 3 months post-randomization, based on the Hospital Anxiety and Depression Scale (HADS) and the 89-item QOL inventory report. Secondary outcomes included depression and QOL (89-item inventory) at 12 months, changes in anxiety symptomatology, Liverpool Adverse Events Profile (LAEP) scores, formal cognitive function testing, seizure counts (excluding those during cross taper phase), medication compliance, and treatment failure. All the above assessments were performed at baseline, 3, 6, and 12 months post-randomization.

At study onset, 51 patients were randomized to levetiracetam and 48 to the older generation AED treatment group (25 and 23 to valproic acid and carbamazepine, respectively). Six patients were excluded due to study noncompliance, one to medication adverse effects (levetiracetam), one to conversion to high-grade brain tumor (levetiracetam), and two deaths (one patient secondary to sudden unexplained death in epilepsy on carbamazepine and an unrelated stomach carcinoma on levetiracetam).

With regards to primary outcome, there was no difference in the HADS depression score from baseline to 3 months in the levetiracetam group (39.5% improvement) vs the older generation AED group of carbamazepine and valproic acid (34.1% improvement). Interestingly, there was a significant proportion of patients with improvement in the 89-item QOL epilepsy inventory score in the older generation AED group (71%) vs the levetiracetam group (48.8%) at 3 months.

In terms of secondary outcomes, there was no clinically significant difference in HADS depression scores at 12 months between the levetiracetam (42.9%) vs the older AED group (46.3%). The HADS anxiety scores improved in both treatment groups at 3 months (48.8% levetiracetam group vs 54.6% old AED) and 12 months (52.4% vs 61% with levetiracetam and older AED group, respectively) with, again, no clinically significant difference. At 12 months, there was no clinically significant difference (62.8% levetiracetam vs 65.8% old AED group) between the groups in 89-point QOL inventory. No significant difference in global vs specific drug adverse side effects or neurocognitive assessments were noted between the two groups at all time points. Finally, there were no significant differences in seizure counts between the two groups at all designated follow-up intervals.


The above results show that substitution monotherapy with levetiracetam vs traditional first-generation AEDs, such as carbamazepine or valproic acid, did not show any clinically significant difference in improvement in depression, QOL, anxiety, efficacy, retention rate, adverse side effects, or neurocognitive performance. Over time, both groups showed similar and sustained improvement in QOL measures, anxiety symptomatology, and side effect profile. Although the authors failed to prove their original hypothesis that levetiracetam monotherapy substitution over older AED treatment has better neurocognitive and QOL measures, this trial does have a few important reassuring clinical implications. First, switching a patient from an ineffective or intolerable drug can result in better neurocognitive and psychosocial measures. Second, the widespread clinical belief that levetiracetam has a greater tendency to exacerbate mood disorders was not borne out in this study. Both the HADS depression and anxiety scores and mood adverse side effects (LEAP scores) were no different between both groups even at the 12-month mark. On the negative side, the lack of a clinically significant difference between new (levetiracetam) and old (carbamazepine, valproate) AED substitution monotherapy highlights the ongoing need for further controlled studies to systematically compare new AEDS with older agents in refractory partial epilepsy to formulate rational therapy.


  1. Kwan P, Brodie MJ. Early identification of refractory epilepsy. N Engl J Med 2000;342:314-319.