The feasibility, safety, and preliminary effects of chronic vagal stimulation were studied in an aluminagel monkey model. Pilot studies to perfect the equipment, determine stimulation thresholds, and insure the comfort and safety of the animals preceded this study. Four monkeys were equipped with an indwelling, 2-electrode cuff (titanium bands spaced 7 mm apart; silicone encased; 1.5 cm total length) in contact around the right vagus nerve; avoidance of the cardiac branch was confirmed by electrocardiograms. After postsurgical recovery, the intact and awake animals received constant-current stimulation (5 mA; 83 Hz, 143 Hz, or 50-250 Hz randomly; 0.5-ms pulse width) at the onset of every spontaneous seizure for the duration of the seizure or every 3 h for 40 s if stimulation had not occurred in the preceding hour. Stimulation periods of 2-6 weeks, with differing levels of stimulation, were preceded and followed by at least a 2-week baseline period of no stimulation. During the stimulation periods, the seizure rate decreased to zero in two monkeys and the interseizure intervals became invariable in the remaining two monkeys. These effects carried over temporarily into the poststimulation baseline periods. Vagal stimulation had no consistent effects on seizure severity or EEG interictal spikes. Histological studies of six vagus nerves were unable to separate electrode cuff damage from any direct effects stimulation may have had on the nerves. Although it appears that chronic vagal stimulation is feasible and that epileptogenic processes are influenced, the safety and efficacy of the procedure are still in question.

The efficacy of cerebellar stimulation was addressed in a chronic monkey model (N = 12) of spontaneous focal motor and secondarily generalized seizures using 24 hr seizure frequency monitoring and all-night EEG recording. The anterior cerebellar vermis was stimulated employing parameters similar to those used in man, 10 Hz, 1 msec pulses, 10 min on, 10 off, at an average current of 2.0 mA. Six weeks pre- and post-base-line periods were compared to a stimulation period of the same length. The results contribute to a clarification of conflicting findings of previous researchers by revealing an inverse relationship between seizure frequency and interictal EEG bursts during the weeks of stimulation. Seizure frequency increased significantly and interictal bursts decreased. Both of these effects (especially the former) were evident in the post-stimulation period, but for different reasons than hypothesized for the period of stimulation. Whereas the therapeutic value of cerebellar stimulation on seizures may be in question, its utilization in the study of mechanisms of epilepsy may be warranted.

Summary The efficacy of valproic acid (VPA) compared t o ethosuximide (ESM) was tested in an alumina‐gel monkey model (N= 12) in‐ strumented with indwelling catheters (for drug infusion and blood sampling) and an EEG plug (for sleep staging and spike recording). The results indicated that VPA was effective in the plasma level range 50‐ 150μg/ml in terms of seizure frequency, duration, and severity, as well as EEG sharp activity. However, focal seizures were not attenuated until the higher plasma concentrations were reached. At the lower plasma levels there was an immediate but transitory effect of VPA on seizure frequency which lasted 2 days only, and a more permanent decrease in seizure frequency at the higher plasma levels. Even with a constant‐rate intravenous infusion of VPA, the mean diurnal fluctuations in plasma drug levels was 36% and considerably more in some animals. EEG bursts per minute tended to correlate inversely with VPA circadian, plasma concentration increases at night and positively with ESM concentration in several monkeys. ESM exacerbated seizure frequency to some extent. The changes in seizure frequency during administration of either drug appeared not to change the basic patterns of time of occurreixe of the seizures for each animal, but rather augmented o r attenuated epileptic activity within those patterns. The findings indicate the ability of this model to quantify correlative processes in efficacy evaluation which lead to a greater understanding of mechanisms of action of antiepileptic drugs. RÉSUMÉ l'efficacité de l'acide valproique (VPA) fut comparée à celle de l'ethosuximide (ESM) chez la singe rendu épileptique par gel d'alumine (N = 12) et muni de cathéters permanents (perfusion des drogues et obtention d'échantillous de sang) ainsi que d'une prise permanente pour l'EEG (enrégistrement des niveaux de sommeil et des pointes). Les réisultats out montré que le VPA était efficace, à des taux plasmatiques de 50 à 150 μg/ml, en ce qui concerne la fréquence, la durée et la severité des crises, ainsi que la diminution des points dans l'EEG. Cependant, les crises focales ne furent attenuées qu'avec les concentrations plasmatiques elevees. Le VPA diminua la fréquence des crises d'une facon immédiate mais transitoire aux concentrations plasmatiques basses; cependant, un effet permanent sur la frequence des crises fut obtenu aux concentrations élevées. Malgré que le VPA fut administrt par perfusion intraveineuse continue à vitesse constante, les concentrations plasmatiques à“l'édquilibre” ont montré des oscillations dues au cycle veille‐sommeil. Cellesci étaient de 36% en moyenne et atteignaient des valeurs beaucoup plus élevées chez certains animaux. Les résultats montraient une tendence de correlation inverse entre les “burst” de l'EEG et les augmentations des concentrations sanguines du VPA durant la période de sommeil. lly avait également une tendance de correlation directe entre les “burst” de l'EEG et les concentrations plasmatiques de l'ESM. l'ESM aggravait (à un certain degré) la fréq...