Language function following subdural grid-directed temporal lobectomy.

The purpose of the study was to determine the extent to which a temporal resection may be undertaken without producing risk to temporal language areas. Patients undergoing craniotomy and placement of a subdural electrode array (SEA) for evaluation of intractable epilepsy were studied to determine the variability of distance of temporal language cortex from the temporal pole. Hemisphere dominance was determined by intracarotid sodium amytal injection. Temporal lobe speech arrest (SA) was mapped with a 64 contact point SEA. Thirty-one patients had left dominant hemisphere SEAs. Thirty had SA 5 cm to 9 cm from the temporal pole (median 7 cm). One had SA at 3 cm. Twenty-one patients subsequently had temporal lobectomy (TL). Mean extent of resection was 5.7 cm (range 3 to 9 cm). In 18 TL patients who had neuropsychometric evaluation of language function pre- and post-surgery, there was no significant deterioration. Thirty-nine patients had right non-dominant SEAs placed. Eighteen had TL. Thirteen of these had pre- and post-surgery language evaluation and there was no significant change. Comparison of preoperative scores showed significant superiority of the right non-dominant group over the left dominant group for naming. TL up to 5 cm without stimulation mapping of language areas would be safe in the majority of cases, but one subject (3%) had SA mapped anterior to this and a small number of cases may therefore be at risk to language function following a 5 cm TL. Extensive lateral resections up to 9 cm are possible with preservation of language function with stimulation cortical mapping.