Cerebral hypoperfusion in stroke prognosis and brain recovery.

Stroke welcomes Letters to the Editor and will publish them, if suitable, as space permits. They should not exceed 1,000 words (excluding references) and may be subject to editing or abridgment. Please submit letters in duplicate, typed double-spaced. Include a fax number for the corresponding author. The interesting report by Davis et al 1 that correlated the results of '""Tc-labeled hexamethylpropyleneamine oxime/single-photon emission-computed tomography (HMPAO-SPECT) with stroke outcome was limited to predicting the outcome only in patients with hypoperfusion demonstrated by SPECT. In 250 consecutive patients we found that three types of SPECT patterns (normal, low, and absent perfusion) had different prognostic as well as diagnostic values (high and mixed patterns associated with cardi-oembolism and normal perfusion with lacunar stroke) when SPECT is performed at any time within the first 5 days after the stroke onset. 2-3 The "unexpected finding" of the significantly increased hypoperfusion between the acute and follow-up SPECT results at 3 months' could be explained by cystic changes that occur in chronic infarction. 2 It would be interesting to know the CT findings at that time. The predictive value of SPECT scanning decreases with time and has no predictive value 2 weeks after the ictus. 2 Furthermore, although the authors found that SPECT hypoperfusion independently predicted neurological outcome after allowing for other measures (eg, Canadian Neurological Scale [CNS] and Allen score), the increment in predictive value (ie, in R 2) was not clear. If the change is only slight, the clinical value of acute SPECT should be questioned. The question of whether SPECT is better than clinical assessment in predicting outcome therefore remains. The combination of SPECT with transcranial Doppler (TCD) appears more promising ,-' since ultrasound adds information on the degree of arterial patency. We combined semiquantitative visual SPECT patterns (normal, high, mixed, low, and absent) with TCD flow results (normal, collateral, stenotic, and occlusive) to give the cerebral perfusion index (CPI). 3 This index is calculated by multiplying the arbitrary given values of SPECT and TCD patterns: CPI=SPECT patternxTCD pattern. Compared with the CNS as a clinical predictor of recovery, the CPI performed better than clinical evaluation alone: the CPI predicted all degrees of the recovery at 2 weeks (poor, partial, good, and complete) whereas CNS differentiated only poor and partial from good recovery. 3 The CPI also differentiated the reversible deficits (transient ischemic attacks and minor strokes) from deteriorating strokes within the first 6 …