Toward a quantitative multivariate analysis of the efficacy of antiseizure therapies.

Seizure frequency is the only variable used for assessing therapeutic efficacy. Advances in quantitative analyses allow measurement of intensity, duration, spread, and time between seizures (TBS). These variables are used here to investigate the efficacy of brain electrical stimulation in humans with pharmacoresistant seizures. The results of a trial of contingent high-frequency electrical stimulation (HFES) for abatement of clinical and subclinical seizures are examined using principal component analysis (PCA) and regression models. HFES significantly: (1) Decreased seizure severity in two of eight and increased TBS in one of eight subjects; (2) decreased seizure severity in the primary epileptogenic zone of one subject but increased it in the secondary zones; (3) had both a beneficial and detrimental effect on severity and/or TBS (increase). These effects were immediate and also outlasted the duration of stimulation ("carryover"). Contingent HFES has multifarious and complex effects, intra- and interindividually, on seizure severity and TBS. Two inferences, at once promising and sobering may be drawn from these results: one, that contingent electrical stimulation deserves a place in the armamentarium of therapies for pharmacoresistant seizures, and the other, that its apparently narrow therapeutic ratio calls for careful implementation and multivariate quantification of its effects.