Summary:Purpose: To evaluate the efficacy and risk of complications of selective posterior cerebral artery (PCA) amobarbital anesthesia in memory assessment of patients with epilepsy under consideration for temporal lobe resection.Methods: Thirty-two candidates for temporal lobectomy in whom conclusive memory assessment could not be obtained by the standard intracarotid amobarbital procedure were submitted to a selective PCA amobarbital test. A mean dose of 75 mg amobarbital was injected via microcatheter into the P2 segment of the PCA. Ten common objects were presented for naming and remembering while the anesthesia was judged efficient. After return to neurologic baseline, recall and recognition memory were assessed.Results: In all of the 32 patients, angiography and PCA anesthesia were successfully accomplished without serious adverse events. All but one of the patients remained alert and cooperative for memory testing under the anesthesia, and 28 of these patients showed adequate memory capacity of the hemisphere contralateral to the side targeted for surgery. So far, 19 patients have proceeded to surgery, and no case of global amnesia or serious, material-specific memory impairment has resulted. Three patients failed the PCA test (fewer than 67% items correctly recognized) and were excluded from surgery, partly on the basis of the PCA test results, but also supported by an overall evaluation of all the diagnostic procedures used.Conclusions: The selective PCA amobarbital test appears justifiable when performed by interventional neuroradiologists and may significantly reduce the risk of erroneously excluding patients with epilepsy from temporal resection. Further corroboration of the safety of the procedure seems warranted.

It is concluded that unilateral amobarbital anesthesia as such, irrespective of side, may trigger both euphoric and dysphoric responses. The relative frequency obtained is influenced importantly both by the emotional responsiveness of the subjects and the emotional climate of the test situation. Finally, it is suggested that brain regions supplied by the PCA contribute less to modulation of euphoric and dysphoric responses than those supplied by the MCA or the ICA.