Are flash-evoked visual potentials useful for intraoperative monitoring of visual pathway function?

Flash-evoked visual potentials (VEPs) recorded from the scalp were used in a series of 35 patients with tumors along the visual pathway: 3 orbital tumors, 25 perisellar tumors, 4 intraventricular tumors, and 3 occipital lesions. Preoperatively, various combinations of impaired visual fields and visual acuity were observed in over 90% of the patients. A postoperative decrease in visual function was observed in 3 cases. Of the 25 perisellar lesions, 13 were operated through a standard frontotemporal craniotomy and 12 were operated through a transnasal-transsphenoidal approach. VEPs were highly susceptible to volatile anesthetics, and there was a significant incidence of spontaneous latency increases and amplitude decreases in a large number of patients. There was an unacceptably high number of cases with significant VEP alteration occurring without concomitant visual function change. During trepanation or the transnasal approach, a reversible potential loss was observed in 11 patients, a profoundly altered wave form was seen in 8 cases, and a loss of single peaks was observed in 15 patients. During dissection of the tumor, a reversible potential loss or a potential with unidentifiable peaks was found in 25 cases; however, the VEPs recovered during closure or in the recovery room. There was no correlation between intraoperative VEP changes and the postoperative changes in visual function. In only 1 patient with an insignificant postoperative decrease in visual acuity from 0.4 to 0.3 was there a concomitant intraoperative potential loss. The major conclusion of our findings is that light-emitting diode flash-evoked VEPs demonstrate intraoperative changes that appear too early and too prominently to be caused solely by manipulation of the optic pathways.(ABSTRACT TRUNCATED AT 250 WORDS)