Abstract
Long-Term Outcome of Mild Mesial Temporal Lobe Epilepsy: A Prospective Longitudinal Cohort Study
Labate A, Aguglia U, Tripepi G, Mumoli L, Ferlazzo E, Baggetta R, Quattrone A, Gambardella A. Neurology 2016;86:1904–1910.
OBJECTIVE: To identify clinical and imaging features at presentation that might predict long-term outcome in patients with mild mesial temporal lobe epilepsy (mMTLE), which is defined by at least 24 seizure-free months with or without antiepileptic medication. METHODS: In the setting of a prospective, population-based, longitudinal cohort study, we followed up 101 patients, all with mMTLE at enrolment. By protocol, patients underwent clinical evaluation every 3–12 months. Independent t test, Mann-Whitney test, or χ2 test was used for comparing 2 groups. The incidence rate of refractory MTLE (rMTLE) was expressed as number of cases every 100 person-years. RESULTS: After a mean follow-up of 12.2 ± 3.7 years, 16 patients dropped out and 85/101 (mean age 46.5 ± 13.3 years) were available for the present analysis. Of these, 64/85 (75%) patients remained seizure-free and 21/85 (25%) became refractory (rMTLE), the latter corresponding to 2.0 cases per 100 persons per year. Patients with rMTLE showed a longer duration of epilepsy (p < 0.001), earlier age at epilepsy onset (p = 0.006), more frequent febrile convulsions (p = 0.02), and hippocampal sclerosis (HS) at MRI (p = 0.004) as compared to those with mMTLE. CONCLUSIONS: mMTLE is a syndrome representing the mildest form of the wide spectrum of MTLE. Earlier age at onset, history of febrile convulsions, longer duration of epilepsy, and the presence of HS on MRI predict a worse outcome.
Commentary
Mesial temporal lobe epilepsy, especially with pathologically or radiologically proven mesial temporal sclerosis is considered a highly refractory and drug-resistant type of epilepsy. It is the poster child syndrome to proceed to epilepsy surgery if there is drug resistance (1). Most case series of temporal lobe epilepsy are reported after surgery and therefore include only refractory patients. For that reason, the above-mentioned study by Labate et al. is of particular interest. The authors admirably followed a cohort of 85 patients with temporal lobe epilepsy, which was controlled with medication at the time of enrollment, for a median time of 11.4 years. One-quarter of the patients became refractory to medications, with the additional finding of a three times higher likelihood of becoming refractory if there was the presence of hippocampal sclerosis on MRI.
Their observed incidence of patients evolving from non-refractory to refractory epilepsy is estimated to be 2.0 cases per 100 person-years, which translates into a waiting period of 50 years to switch from nonrefractory to refractory, which is consistent with their other finding that later onset of epilepsy decreases the likelihood of becoming refractory. These findings overall mean good news for our patients. If they respond to medications, the likelihood that they remain that way is high, and we should be optimistic even in patients with the classic syndrome amendable to surgery.
Others have reported that in temporal lobe epilepsy, periods of seizure control are intermixed with periods of persistent seizures (2). One study estimated that the time for becoming drug resistant is, on average, 9 years (2). These findings, in addition to the Labate et al. study, would explain why it was difficult to enroll patients early in a large, multicenter study designed for surgical intervention within 2 years of disease onset in temporal lobe epilepsy (Early Randomized Surgical Epilepsy Trial [ERSET] study) (3). It also explains why it is difficult to perform epilepsy surgery early in the course of the disease. It seems intuitive that early intervention would prevent later detrimental consequences.
There is a silver lining to the above optimistic message. The study supports evidence that mesial temporal lobe epilepsy is a progressive disease. Although three-quarters of patients remained controlled, there was still one-quarter who went on to be refractory. Those patients were more likely to be patients with hippocampal sclerosis on MRI and to have a history of febrile seizures, earlier onset, and prolonged course of epilepsy. It is possible that the patients exhibiting hippocampal sclerosis on MRI represent a cohort with more significant neuronal cell loss and therefore are more likely to become refractory, as it merely reflects more extensive disease. There were 3 (7%) of 43 patients who did not have hippocampal sclerosis on initial MRI but developed hippocampal sclerosis on later imaging, which adds additional evidence that temporal lobe epilepsy is progressive. Curiously, all three of those patients remained seizure-free. Unfortunately, the authors did not add any longitudinal measurements of hippocampal volumes to the study, which would have been informative. Other studies have clearly shown that there is progressive atrophy of the hippocampus as well as of the lateral temporal neocortex over time with temporal lobe epilepsy, but mainly in refractory patients (4). A very recent study reported continuing gray matter loss even in seizure-free patients as compared with normal controls (5).
If temporal lobe epilepsy is progressive as evidenced by MRI, even if patients remain seizure-free, what are the consequences for cognition and memory? Memory problems can be more disabling than seizures. Cognition and memory has been studied extensively in the surgical cohort, and it seems that seizures and epileptiform activity have a negative influence on cognition (6). However, our knowledge about memory impairment in well-controlled epilepsy is somewhat limited. It is difficult to distinguish medication effects from the underlying pathology and developmental changes (7). Longitudinal studies are scarce (7, 8), but certainly worthwhile investigating in the future.
From intracranial EEG recordings, it is often observed that there is frequent subclinical, seizure-like EEG activity, in a sclerotic hippocampus, without any overt clinical manifestations (9). Whether this is also the case in mild temporal lobe epilepsy controlled with medications remains to be demonstrated. Ongoing epileptiform activity has an influence on memory and disease progression (6) and may also be an important determinant of overall cognitive outcome in mild temporal lobe epilepsy.
Labate et al. found a relationship between becoming refractory and a history of febrile seizures. The correlation of febrile seizures with intractable temporal lobe epilepsy is well known, but the interrelationship still remains much debated (10). The multicenter effort of the FEBSTAT studies tries to address such questions (10). Only about 10% of children with refractory febrile status developed hyperintense imaging abnormalities immediately after status. Most of these children developed hippocampal atrophy later. Whether this will be associated with epilepsy, mild or refractory remains to be awaited. The FEBSTAT study only includes patients with febrile status epilepticus, defined as seizures for more than 30 minutes, as compared with just a history of febrile seizure as reported in the Labate et al. study (10). A small but not insignificant difference.
In summary, it is laudable that the authors were able to follow this cohort for a prolonged period of time in a prospective manner. Having temporal lobe epilepsy that is initially controlled with medications, especially without any imaging findings, is good news for our patients, but we should not neglect that the disease may be progressive. We also need to put our attention to details beyond seizures such as cognition and memory, in particular in patients who are not necessarily refractory.