In summary, these data suggest that widespread primary or secondary cerebral vasoconstriction is common in patients with neurological complications from cocaine. In most patients, SPECT showed wide-spread hypoperfusion in regions that had no clear clinical significance (e.g., the periventricular area). In many, the SPECT was performed more than 24 hours after the onset of neurological symptomatology. These findings raise several questions. It has been assumed that these SPECT changes in patients with acute neurological symptoms are temporary, although it will be important to determine whether these areas of hypoperfusion persist after symptoms have abated. Recently, Holman and colleagues (1991) found multifocal and deep areas of hypoperfusion with SPECT in 16 of 18 patients with a history of chronic cocaine abuse. Although most of the subjects tested positive for cocaine, several had abstained from cocaine use for weeks prior to the study. All 18 subjects had neuropsychological deficits, 13 mild and 5 moderate. Similarly, Pascual-Leone and colleagues (1991) have shown that CT scan atrophy strongly correlates with the duration of cocaine abuse, suggesting that brain injury may occur with continued use of cocaine. It is the authors' concern that cocaine abuse might produce permanent changes in cerebral perfusion. In conclusion, brain SPECT was found to be a useful procedure in the evaluation of acute cocaine intoxication. Brain SPECT revealed focal cortical lesions not seen on head CT or MRI, which corresponded to clinical deficits. In addition, [99mTc]HMPAO brain SPECT had a characteristic scalloped appearance, and this may be a marker for acute intoxication with cocaine. This study further supports the contention that cocaine causes neurological disease by its vasoconstrictive action.