Comments on Beyond Seizure Control: A Case Series on Levetiracetam Associated Psychiatric Manifestations

Dear Editor,

We read with interest the article by Durairaj et al. ¹ on six patients—a 34-year-old woman (case 1), a 35-year-old woman (case 2), a 17-year-old man (case 3), a 38-year-old man (case 4), a 54-year-old man (case 5), a 37-year-old man (case 6)—who developed psychiatric symptoms including mania (n = 3), psychosis (n = 3), 3–14 days after an increase in levetiracetam (LEV) dose by 500–1000 mg/day (cases 1, 2, 3, and 5) or after starting LEV 1,000 mg/day (cases 4 and 6). ¹ The psychiatric symptoms subsided in all six patients either after discontinuation of LEV (case 6) or after switching to valproic acid (VPA) (cases 1, 2, 3, and 5) or to brivaracetam (case 4). ¹ The study is noteworthy, but some points require discussion.

Discussion

The first point is that the taxonomic hierarchy of seizures and epilepsies remained unclear in cases 3, 5, and 6. ¹ We should know whether the seizures in these three cases should be classified as focal, generalized, or unknown, with motor, non-motor, sequential, or unclassified presentations. It should also be mentioned how many patients had focal, generalized, combined focal and generalized, or unknown epilepsy.

The second point is that LEV serum levels were not measured in any of the six patients to determine whether the psychosis or mania was due to intoxication (above normal range) or therapeutic levels. Intoxication can be caused by incorrect overdose, reduced metabolism, reduced excretion, or adjustment of dosage to body weight. Therefore, we should know who was responsible for obtaining LEV, whether the daily dose was witnessed, whether kidney and liver function were normal at the time of the onset of psychiatric symptoms, and whether any of the six patients were underweight. Since the daily LEV doses exceeded 1,500 mg in only one patient (case 5), it is rather unlikely that any of the six patients had LEV serum levels above the upper reference limit. We should also know whether any of the six patients were regularly taking concomitant medications to rule out drug interactions as a cause of elevated serum LEV levels.

The third point is that none of the six patients had EEG recordings after the onset of symptoms. ¹ Therefore, it cannot be ruled out that the psychiatric symptoms in some patients were actually due to persistent paroxysmal activity manifesting as non-motor symptoms, or to postictal focal, or generalized slowing. The EEG could also help classify seizures and epilepsy.

The fourth point is that no cerebral imaging results were reported. ¹ Since two patients had a history of traumatic brain injury, it is crucial to know whether focal or generalized cerebral morphological changes could explain the psychiatric symptoms. Were there any signs of hemorrhage or ischemia in these two patients?

The fifth point is that it is not clear why four patients were switched to VPA (cases 1, 2, 3, and 5) at a dosage between 1,000 and 1,500 mg/day. ¹ Since VPA causes delirium, it would have been essential to rule out delirium as the cause of the cognitive and behavioral abnormalities in these four patients. VPA can also be responsible for irritability, restlessness, and abnormal behavior. ² There is also evidence that VPA can cause psychosis. ³

The sixth point is that case 6 suffered from a dissociative disorder before the onset of the dissociative episode that led to the initiation of LEV treatment. ¹ How can one rule out that the newly occurring episode was merely an exacerbation of a pre-existing disorder that was independent of LEV?

Final, the Naranjo score indicated probable causality in five cases and possible causality in one patient. ¹ However, the classification as probable or possible means that in none of the six cases could a definite causal relationship between LEV and the psychiatric symptoms be established.

Recommendations

Conclusions

The causal relationship between psychiatric symptoms and LEV remains unproven as long as alternative causes for the psychiatric symptoms have not been sufficiently ruled out.

Supplemental Material

Supplemental material for this article is available online.