Absence of correlation between amobarbital distribution as assessed with SPECT brain perfusion imaging and behavioral manifestations during the intracarotid amobarbital procedure (Wada test).

1. The IAP is used presurgically in patients with temporal lobe epilepsy to predict the effects on LTM and language of the planned temporal lobectomy. This prognosis presumes that a similar pattern of perfusion will result in anesthesia of the same cerebral regions in most patients. 2. Coinjection of Tc-99m HMPAO with the barbiturate during the IAP has been used to ascertain whether this actually is true, with variable results. Moreover, most studies document only unilateral IAPs and do not report on behavioral performance. 3. The authors coinjected Tc-99m HMPAO and amobarbital in 33 IAPs from 18 patients (15 injected bilaterally, 3 unilaterally) to clarify this and to evaluate the relationship of the perfusion pattern to behavioral performance; SPECT results were also compared to angiographic evaluation obtained at the time of catheter placement. 4. SPECT perfusion data was rated for presence/absence and intensity of perfusion to the ACA, MCA, PCA territories and to H, i or c to the injection site. V, STM and LTM were graded according to a standardized protocol. 5. MCAi was perfused in 100% of cases, ACAi in 91%, PCAi in 21% and Hi in only 6%. Cross-over flow was shown in 9 studies; 50% of the patients in whom both sides were injected (on different days) had crossover, involving the ACAc territory in 80% of cases. As expected, injection on the non-ES was associated with a significantly worse LTM performance than on the ES (p = 0.006). There was no relationship between the perfusion pattern and the V level of the patients (a potential confounding variable in memory/language evaluation) during IAP, nor between perfusion pattern and LTM. STM was significantly adversely affected by the presence of crossover perfusion. Angiography in general overestimated the extent of cerebral perfusion demonstrated by SPECT, most probably because of the markedly different injection conditions. 6. Despite the best efforts to standardize injections, the perfusion pattern has been mostly unpredictable in the patients. Moreover, it has little bearing on their behavioral performance, except for the prediction of poor STM performance (the clinical implications of this remaining dubious). Marked LTM alterations after non-ES injections confirm remote hippocampal effects in the presence of only rare direct perfusion of that region. Tc-99m HMPAO/Amobarbital coinjection was unhelpful from a clinical perspective, most probably because a large part of the effects of amobarbital arise from deafferentation of regions not directly perfused by the anesthetic agent.