Clinical findings

Of the 650 patients, 96 (15%) had survived CSE, 49 with parenchymal lesions and 47 without (Table 14.1). In 27%, hippocampal sclerosis and an elective parenchymal necrosis of the neocortex were identified. This figure is in the same range as that for the entire study group of 650 patients, with 30.5% of cases exhibiting hippocampal sclerosis (9). Neocortical elective parenchymal necrosis was found in 3% of survivors of CSE, compared with 5.6% of the entire group. These figures indicate that the development of SE does not change the rate of elective parenchymal necrosis in the brain of patients with epilepsy.

Thirty-two patients had initial CSE and 64 had intercurrent CSE (Table 14.2). Parenchymal ischemic lesions were identified in 75% of the patients who had initial CSE but in only 40% of those who had intercurrent SE. This finding probably indicates that this type of pathologic change is related to the etiological event responsible for the initial CSE. The recurrence of SE did not significantly affect the proportion of cases without parenchymal lesions. No ischemic lesions were identified in 25% of patients with initial CSE as an isolated event and in 31% of patients with initial CSE and recurrent SE. No ischemic lesions were identified in 60% of patients with intercurrent CSE as an isolated event and in 62% of those with recurrent SE. Surprisingly, even recurrent SE did not change significantly the prevalence of ischemic lesions.

More extended sclerotic ischemic lesions (ulegyria) are much more involved in initial episodes of SE (intercurrent, 35%, isolated, 46%) than in intercurrent SE (intercurrent, 15%, recurrent, 20%). Cases with recurrent SE are also more related to these sclerotic lesions. In general these figures also support the hypothesis that the analyzed lesions are much more related to the causative factors of initial CSE than they are a consequence of the seizures themselves. This is also supported by the fact that recurrent SE does not modify the figures of the nonaffected specimen. In cases with intercurrent SE but recurrent events, there is the same rate of parenchymal necrosis in both groups.

The hypothesis that the ischemic lesions analyzed might be related to the causative factors is also supported by our observation that in the 49 patients with ischemic lesions, 80% did not have any other pathologic deviation (Table 14.3). In contrast, only 8% of the 47 patients with nonischemic lesions and SE did not have any other pathologic lesions. The detailed pathologic findings in this group of 47 patients with grand mal SE and no ischemic lesions are given in the table. Leading findings are migration disturbances, seen in almost 50% of the cases (Table 14.4).

The conclusion that nonictal factors are responsible for the pathologic changes is also supported by the analysis of elective parenchymal necrosis in patients with final (terminal) convulsive SE. Out of 59 patients with final SE, 29 had epilepsy and 30 had no epilepsy (Table 14.5). Of the patients with epilepsy, 76% had no elective parenchymal necrosis, whereas only 40% of patients without epilepsy had no elective parenchymal necrosis. This finding suggests that the pathologic changes seen are more related to the pathologic event that causes the pathologic lesion and initiates the epileptogenic functional disturbances. An analysis of etiologic factors responsible for final SE in patients without epilepsy shows that the etiology itself is responsible for the elective parenchymal necrosis (Table 14.6).

Morphometric studies of neuron density in the hippocampus clearly showed that the status epilepticus does not influence the neuron density of the pyramidal cell layer. We analyzed 29 control patients without epilepsy, 27 patients with temporal lobe epilepsy, and 12 patients with primary generalized idiopathic epilepsy. On morphometric analysis, neuron density in sector CA4 (H3) was influenced by the total number of grand mal seizures but not by the generation of SE (6).

We were able to analyze in detail the influence of SE in different clinical epilepsy syndromes—Lennox-Gastaut syndrome (n = 30), generalized idiopathic epilepsies (n = 15), and temporal lobe epilepsies (n = 27).

Five patients with Lennox-Gastaut syndrome had grand mal status. Hippocampal sclerosis in this subgroup with CSE was 20%, in the same range as for the whole group (Table 14.7). Also, cerebellar lesions, the predominant finding in this group (10), were seen in 60%, again in the same range as for the whole group (67%).

In the group of 15 patients with generalized idiopathic epilepsies, only one patient had grand mal SE (Table 14.8). In this patient we found a circumscribed elective parenchymal necrosis in the mesial thalamic nucleus. The distribution of this lesion was the same as demonstrated in our study of thalamic lesions after global ischemia (7). This patient sustained an intercurrent cardiac arrest after a suicide attempt.

Of the 27 patients with temporal lobe epilepsy, nine had SE (Table 14.9). Of these nine patients, only 44% had hippocampal sclerosis, compared with 56% in the whole temporal lobe group. Only the frequency of cerebellar lesions was increased, 66%, compared with 41% for the entire study group. Although this finding could not be demonstrated in other epilepsy syndromes, this slight increase in cerebellar involvement might not be a consequence of the status event. Thus, the cerebellar involvement in this syndrome might more frequently be related etiologically to the development of CSE.