Abstracts

Rationale: Wada test or intracarotid amobarbital procedure (IAP) is commonly used for memory and language lateralization prior to epilepsy surgery despite variation in the protocol among centers. Here, we investigated 2 IAP protocols and their correlation with neuropsychological and imaging tests.Methods: We retrospectively identified consecutive patients who underwent 2 protocols of IAP. IAP1 (object-based) was performed with a variable dose of sodium amytal tailored to contralateral arm weakness using 8 dually-encodable objects; recall was tested by forced recognition of the targets interspersed with 16 foils. IAP2 (material-specific) was performed with a fixed dose of amytal (120/90 mg) using 18 stimuli (9 pictures, 9 words); recall was tested by recognition of targets presented simultaneously with 7 foils. Ipsilateral EEG delta slowing was used as a measure of duration of hemi- anesthesia. Memory recall was tested after return of strength and language to baseline. Patients with baseline IAP memory scores of <50% were excluded. We defined 3 types of IAP memory lateralization based on total recall score and asymmetry: lateralized (LAT: asymmetry index (ipsilateral minus contralateral) >20%); nonlateralized intact (NLInt: >=50% recall for each injection and asymmetry index <=20%); nonlateralized impaired (NLImp: <50% recall for both injections regardless of asymmetry index). Neuropsychological testing included verbal memory composite score derived from WMS-4 logical memory and CVLT-2, and visual memory composite score derived from WMS-4 visual reproduction and Rey-Osterrieth figure. Results: Forty-six IAPs (from 40 patients) were analyzed: 20 IAP1; 26 IAP2. 6 patients underwent both protocols. Amytal dose was 40 130 mg for IAP1, 90 120 mg for IAP2. Mean dose was significantly lower for IAP1 than IAP2 (90 vs. 118 mg) (p<0.0001). Memory acquisition occurred during the period of delta slowing in 100% of IAP1 and 71% (58% of patients) of IAP2 injections. Language lateralization was mostly left hemispheric in both protocols. Memory classifications in IAP1 and IAP2 respectively were: LAT (50%, 70%); NLInt (15%, 7%); NLImp (35%, 23%). In 6 patients who had both protocols, language (100%) and memory (83%) lateralizations were concordant except for one patient who showed NLInt on IAP1 but NLImp on IAP2. In patients with left-sided seizure foci, verbal memory scores negatively correlated with an injection difference score for IAP1 ( =-0.58, p=0.08) and IAP2 ( =-0.87, p=0.01). Other correlations between verbal/visual scores and IAP recall scores were not significant. In patients with unilateral mesial temporal sclerosis (MTS, n=22), the IAP memory lateralization was contralateral in both IAP protocols. Conclusions: Despite significant methodological differences in the IAP protocols, the language lateralization, memory lateralization (in MTS), and verbal memory (in patients with left-sided seizure foci) tend to be similar. Lower amytal dose and fewer memory items facilitate quicker memory acquisition during IAP. Further studies should examine whether or not the two IAP protocols impact surgical decision making.