SPECT IMAGING IN CEREBRAL VASOSPASM FOLLOWING SUBARACHNOID HEMORRHAGE

Cerebral vasospasm is a frequent complication after subarachnoid hemor rhage and contributes to overall morbidity and mortality. Arteriograph y is the standard test for determining the presence of vasospasm. A re trospective review of 16 patients with cerebral aneurysm was undertake n to assess the sensitivity and specificity of SPECT for diagnosis of vasospasm. Fourteen patients were hospitalized after subarachnoid hemo rrhage and 2 patients were hospitalized for elective aneurysmal clippi ng. The patients' condition on discharge was correlated to clinical an d SPECT evidence of vasospasm. Methods: Vasospasm was defined as the n ew onset of neurological signs and symptoms not explained by rebleed o r hydrocephalus. A total of 20 SPECT studies were performed for 16 pat ients during their admission and 14 of 16 patients had a single angiog raphic study. Results: Thirteen of 16 patients had 14 episodes of clin ical evidence of vasospasm and 14 SPECT studies were performed in thes e 13 patients. The sensitivity and specificity of SPECT in this retros pective study were 89% (8/9) and 71% (5/7), respectively. Our small sa mple of arteriograms yielded in comparison a sensitivity of 67% (2/3) and specificity of 100% (9/9). The one false-negative SPECT study occu rred in conjunction with the one false-negative arteriogram in the pre sence of clinical findings consistent with vasospasm, Three false-posi tive SPECT studies occurred in 2 patients who had perfusion abnormalit ies in areas of normal CT findings without clinical or arteriographic evidence of vasospasm. Five of 5 patients who died became unresponsive as a result of clinically presumed vasospasm and 4 of 5 of these pati ents had diffuse or hemispheric SPECT perfusion defects. Of the 11 pat ients who survived, none became unresponsive; 1 of 11 had positive dif fuse or hemispheric perfusion defects. Conclusion: SPECT is a sensitiv e and fairly specific lest for corroboration of clinical findings of v asospasm. A negative SPECT study may obviate the need for arteriograph y. Unresponsiveness is the best predictor of poor outcome; however, he mispheric SPECT perfusion deficits are also associated with poor outco me.