Levetiracetam May Be More Effective for Late-Onset Partial Epilepsy

Background  Many agents are available for treating epilepsy; however, population studies have failed to show overall differences in efficacy for a given seizure type. Clinical experience suggests that certain individuals will respond to a given agent while others with the same seizure type will not.

Objectives  To examine a population of patients who received one of the newer antiepileptic drugs, levetiracetam, and to identify those who had either a dramatic improvement or a significant worsening of seizures.

Methods  Retrospective medical record review of patients with refractory epilepsy.

Results  Patients who responded well to levetiracetam therapy were older at the onset of epileptic seizure than those who did not (mean [SD] age, 51 [5] vs 27 [3] years; P<.05). This was also true of the subset of patients who had localization-related epilepsy. Patients with temporal lobe onset were likely to do well whereas patients with frontal lobe onset were not.

Conclusions  These results suggest that certain subpopulations may be particularly likely to respond to levetiracetam therapy. These need to be confirmed in a larger prospective trial; however, looking for specific characteristics of patients who respond to certain drugs may lead to useful guidelines for drug choices in treating epilepsy.

MANY EFFECTIVE drugs exist for all types of epilepsy. However, numerous controlled trials have failed to show overall differences in efficacy between drugs in populations of patients who have partial epilepsy.^1-8 This makes it difficult for the clinician to choose a drug for a patient, particularly one who has experienced failure with previous agents owing to lack of efficacy. Choices within a seizure type are either based on the physician's personal impression rather than data (ie, random) or based on an adverse effect profile rather than efficacy.

Despite a lack of concrete differences in efficacy, in an individual patient it is not uncommon to see complete efficacy with a new drug when others have failed. In add-on and monotherapy trials of new drugs in patients with refractory seizures, the seizure-free rate varies between 0% and 36%.^9-12 There are also patients whose seizures actually worsen with the addition of an antiepileptic drug (AED). Although sometimes this has to do with an incorrect diagnosis and an inappropriate drug, it can occur unpredictably.¹³ However, to our knowledge, there has never been an investigation into whether certain patient characteristics are predictive of success or failure with a new drug. Because of this, we examined a population of patients who received one of the newer drugs, levetiracetam, and identified those who had either a dramatic improvement or a significant worsening of seizures.

All medical records of active patients at the Comprehensive Epilepsy Center, Department of Neurology, Columbia University, New York, NY, were examined to see whether patients had received levetiracetam therapy. The conditions of all had previously been proven refractory to other medications. All maintained seizure calendars. Three groups were identified based on seizure frequency before and after treatment: those who had a clear worsening of seizures (at least a doubling of baseline seizure frequency), those with clear improvement of seizures (>50% reduction in baseline seizure frequency), or those who became seizure free. Patients were excluded from the study if they did not meet any of these 3 criteria or if they had not been taking levetiracetam long enough to determine its effect. Timing for exclusion was less than 3 months or less than 3 times the baseline seizure frequency, whichever was greater. In other words, a patient with seizures every 2 months had to be seizure free for at least 6 months. Levetiracetam therapy was discontinued sooner for patients with a clear worsening of seizures depending on the clinical judgment of the treating physician. Improvement or seizure freedom was sometimes not achieved with the initial dose; however, if the aforementioned criteria were met following a dose increase, then these patients were included.

Characteristics of these patients were then tabulated, including medication (levetiracetam dose, concurrent medications, and previously failed medications), demographics (age and sex), seizure type, and epilepsy syndrome (based on all available data). These were then compared to see if there were differences between the groups.

Demographic and medication data are given in Table 1. By screening the medical records of more than 200 patients who received levetiracetam therapy, we identified 9 patients whose seizures were worse, 19 whose seizures were significantly improved, and 16 whose were seizure free. Patients who improved and who became seizure free were significantly older than those whose seizures worsened. Age of seizure onset was also significantly later for patients who became seizure free; in the improved group, this did not reach statistical significance. When the subset of patients with localization-related epilepsy (LRE) was examined (n = 31), a similar age relationship was seen although statistical significance was seen only for patient age in the seizure-free group compared with the group whose seizures worsened. There was no difference in sex. Patients who improved or became seizure free received a lower levetiracetam dose on average than did patients whose seizures worsened, although this difference was not statistically significant. Patients who became seizure free had lower baseline seizure rates, mainly because of the few patients in the improved and worse groups who had a large number of simple partial seizures. This difference was not significantly different than the worse group for patients with LRE.

Seizure type and syndrome are given in Table 2. A higher proportion of patients with generalized epilepsy worsened than became seizure free, although the numbers were small and this difference was not statistically significant. Among patients with LRE, a higher proportion in the group that worsened had frontal-lobe–onset, while more in the improved and seizure-free groups had temporal-lobe–onset. The comparison of temporal-lobe–onset in worsening vs improved vs worsening seizures was statistically significant.

The large number of AEDs available gives additional options for patients with incomplete efficacy with 1 or more drugs. Although clinical experience shows that 1 drug can be effective for a given seizure type when another has failed, differences across populations have not been demonstrated. If there are individual patient or seizure characteristics that make a drug more likely to succeed or fail, this would be useful clinically and avoid the present, somewhat arbitrary trial-and-error therapy of AEDs once patients have failed initial monotherapy.

Levetiracetam therapy has been shown to be highly effective for partial-onset seizures^14,15; there is evidence of possible efficacy in generalized seizure types as well.¹⁶ Our results suggest that levetiracetam therapy may be particularly effective in patients with late-onset epilepsy. This finding could be because of a higher number of generalized seizures in a younger population; however, the findings were similar when only patients with LRE were included. It was not caused by an inadequate dose, as patients who did well actually took a lower average dose (although it is possible that a higher levetiracetam dose actually caused worsening in some patients). Seizure-free patients had a lower baseline seizure frequency than the other 2 groups, mainly because of a few patients in the improved and worse groups who had a large number of simple partial seizures. Among patients with LRE, the improved group actually had more seizures at baseline than did the worse group. Because the results were consistent for the improved and seizure-free groups, it is likely that the better response in patients with late onset is not due simply to the baseline frequency. It is unlikely that improved and seizure-free patients represented a less refractory group, as all had previously tried approximately the same number of AEDs. A possible alternative explanation is cause: older patients with late onset of seizures would be more likely to have epilepsy resulting from degenerative or cerebrovascular disease; younger patients with earlier onset of seizures were more likely to have had trauma, congenital disease, or a central nervous system infection.¹⁷

Our results must be interpreted with considerable caution, as they represent a few selected patients and need to be confirmed in a larger, independent series. Use of comprehensive databases from clinical trials, while retrospective, may help to further define patient groups with particularly good responses (and would have the advantage of control groups for comparison). Because compliance was not measured, it is also possible that certain subgroups (those with later onset) were more compliant; this would require documentation of serum levels, which was not performed in this study. It is also possible that these data only show an overall improved drug responsiveness in patients with late onset, although this does not explain why all groups were previously unresponsive to a similar number of other drugs.

Perhaps more interesting is the observation that patients with frontal-lobe–onset LRE may respond differently than those with temporal-lobe–onset. In particular, our patients who did well were more likely to have temporal lobe epilepsy than those who did worse; the same was not true for patients with frontal lobe epilepsy. There is evidence of differences in seizure spread according to sleep state.¹⁸ Temporal lobe seizures are more likely to secondarily generalize during sleep than are frontal lobe seizures, suggesting different paths of spread that are modulated by sleep. If this is true, it would not be surprising that certain drugs (including levetiracetam), which presumably modulate neurotransmission in various ways, would also affect various pathways of seizure propagation differently. One open-label trial suggested that a combination therapy of divalproex sodium and lamotrigine may be particularly effective for frontal-lobe–onset seizures.¹⁹ Larger, controlled trials would also be necessary to confirm these observations.

Our results suggest that levetiracetam therapy may be most effective in patients with temporal lobe epilepsy who have late onset. More importantly, we suggest that similar investigations into the characteristics of patients whose seizures benefit greatly or whose seizures worsen from certain drugs may help in guiding the choice of AEDs for those patients who previously experienced failure in taking a drug, and perhaps even in the choice of an initial AED.

Accepted for publication June 18, 2002.

Author contributions: Study concept and design (Drs Bazil, Resor, and Hirsch); acquisition of data (Drs Bazil, Resor, and Yapicular); analysis and interpretation of data (Drs Bazil and Hirsch and Mr Rose); drafting of the manuscript (Dr Bazil); critical revision of the manuscript for important intellectual content (Mr Rose and Drs Resor, Hirsch, and Yapicular); administrative, technical, and material support (Mr Rose and Dr Yapicular); study supervision (Drs Bazil and Hirsch).

Corresponding author: Carl W. Bazil, MD, PhD, Comprehensive Epilepsy Center, Department of Neurology, Columbia University, 710 W 168th St, New York, NY 10032 (e-mail: cwb11@columbia.edu).