Properties of after-discharges from cortical electrical stimulation in focal epilepsies

This study sought to determine whether certain aspects of after-discharges (ADs) obtained during cortical functional mapping provide better correlations between stimulus site and that of spontaneous seizures. Secondly, we wished to determine the percentage of stimulations evoking ADs and, of these, the percentage which clearly involves more than one electrode position, potentially inaccurately localizing cortical function. Thirdly, we wished to quantify the incidences of the several AD morphologies described by Jasper [in: Epilepsy and the functional anatomy of the human brain, 1954, p. 183; p. 692] and to assess whether certain morphologies had a greater tendency to evolve in frequency or morphology.

In these 29 patients requiring invasive recordings to determine principal epileptogenic areas, only subdural strips were placed in 19 patients, only grids in 2 patients, and both in 8 patients. A median of 21 electrodes per patient was stimulated of a median of 63 electrodes placed, with the following parameters: biphasic, monopolar, 50 Hz, 0.3 ms pulse duration, 1.5-18 mA. Coverage involved the frontal and parietal lobes (9 patients), frontal parietal temporal lobes (8), frontal temporal (3), temporal (2), occipital (2) and occipital temporal (2) with other combinations in 3 additional patients. Classification of AD morphologies was determined by a pilot study using IFSECN definitions [Electroenceph clin Neurophysiol 1974;37:538] and descriptions by Jasper [in: Epilepsy and the functional anatomy of the human brain, 1954, p. 183] and Gloor [in: Advances in neurology, vol. 8, 1975; p. 59].

Four hundred and two ADs (12%) were elicited by 3358 trains of electrical stimuli of which 260 (65%) clearly involved more than only the stimulated electrode position. Thus, 260 (8%) of 3358 stimulations evoked an AD that could mislocalize cortical function. The proportion of stimulating electrodes eliciting ADs ranged among patients from 4 to 83% (median 33%). Polyspike bursts and sequential spikes were the most common AD morphologies. Ten percent of ADs evolved in morphology, frequency or both. Evolution occurred more commonly (44%) with rhythmic waves than with other AD morphologies (7%). Neither evolving ADs, ADs producing clinical seizures, or ADs exceeding 10 s correlated topologically with spontaneous seizure origins.

Although occurring in a minority of cortical electrical stimuli, ADs may involve more than the stimulus site and therefore may inaccurately localize cortical function. Our material failed to disclose any consistent relationship between the site of stimulus eliciting ADs and that of spontaneously appearing seizures, even when certain aspects of such ADs were analysed.

These data illustrate the need for scrutiny of the post-stimulus electrocorticogram for ADs and particularly those involving extra stimulus sites. Whether varying the stimulus parameters prospectively will disclose a better seizure localizing value for ADs remains to be determined.