Correlates of arterial-filling patterns in the intracarotid amobarbital procedure

To determine behavioral correlates of anterior cerebral artery (ACA) and posterior cerebral artery (PCA) perfusion patterns in the intracarotid amobarbital sodium procedure.

Multivariate analysis of covariance and partial correlations of behavioral measures to ACA crossflow and PCA filling.

Angiography and the intracarotid amobarbital procedure at a comprehensive epilepsy center.

Forty-two patients with intractable epilepsy (right-hemisphere seizure focus [n-23]; left-hemisphere seizure focus [n = 19]).

Internal carotid angiography was performed both at a standard injection rate (8 mL of contrast per second) and at 1 mL/s, which matched the rate of the subsequent amobarbital injection. The degree of ipsilateral PCA and contralateral ACA filling were graded on a seven-point scale and compared with postinjection behavior, language, and memory measures.

The ACA crossflow did not correlate significantly with that of any measure. The degree of PCA-filling pattern correlated significantly only with the level of consciousness (r = .31, P < .004), but it was not significant after accounting for the effects of seizure laterality, injection side, and amobarbital dosage. Neither ACA crossflow nor PCA filling correlated significantly with memory. The degree of ACA and PCA filling was overestimated at standard angiography (8 mL of contrast medium per second) injection rates.

Although the degree of PCA filling correlates mildly with the level of consciousness postinjection, possibly by perfusion of thalamic or mesencephalic branches, it is not reliably predictive and is less contributory than the injection side and seizure laterality. The PCA filling is not required to produce valid memory assessment in the intracarotid amobarbital procedure, and ACA crossflow is not predictive of behavioral responses.