Comparative correlations of HMPAO SPECT indices, neurological score, and stroke subtypes with clinical outcome in acute carotid infarcts.

BACKGROUND AND PURPOSE: The prognostic value of single-photon emission computed tomography (SPECT) remains controversial. The aim of this study was to compare the prognostic value of stroke severity, stroke subtypes, and SPECT indices and to determine which predictive factors have an independent effect on clinical outcome. METHODS: We studied 55 consecutive patients with acute (< 12 hours) carotid infarct within 36 hours of symptom onset with SPECT. Clinical presentation was assessed by the Canadian Neurological Scale and stroke subtypes. SPECT indices were the degree and size of hypoperfusion and crossed cerebellar diaschisis as assessed by a semiquantitative analysis. Outcome was evaluated by the functional status and mortality (Rankin Scale score at 1 month). RESULTS: The Rankin Scale score correlated with the degree (r = .580; P < .00001) and size (r = .616; P < .00001) of hypoperfusion. The mean degree and size of hypoperfusion were significantly higher in patients with poor outcome. Crossed cerebellar diaschisis had no significant predictive value. Statistical analysis determined threshold values for the Canadian Neurological Scale score and the degree and size of hypoperfusion for the functional status and mortality. The degree and size of hypoperfusion had no higher performance than the Canadian Neurological Scale score. The negative predictive value was excellent for both clinical and SPECT indices. Multivariate analysis selected only the size of hypoperfusion as an independent predictor for the functional status (P = .004) and the Canadian Neurological Scale score for mortality (P = .009). CONCLUSIONS: SPECT performed within 36 hours of onset predicts clinical outcome, but different clinical and SPECT indices with threshold values should be chosen according to the relevant outcome end point.