'Cerebral malaria'is a clinical syndrome consisting of coma not attributable to convulsions, hypoglycemia, or meningitis in patients with P. falciparum parasitemia. The deaths of 1-2 million African children each year are attributed to cerebral malaria, and this clinical case definition is used to diagnose and treat individual patients, to enroll groups of patients in studies to discern various associations (with clinical and laboratory features, with genetic traits associated with resistance or susceptibility to severe disease), and to assess the impact of various prevention strategies (insecticide-treated nets, intermittent presumptive therapy, malaria vaccine candidates). Data from our autopsy-based study suggest that this standard clinical case definition is wrong 23% of the time (62). Seven of 31 patients who met the clinical case definition of cerebral malaria and died had no histological evidence of cerebral malaria;other, non-malarial causes of death were identified at autopsy. The only clinical feature which differentiated the patients with 'true'cerebral malaria from those with coma and incidental parasitemia was ocular fundus findings known collectively as the 'malarial retinopathy'. These data are intriguing, but before a new definition of cerebral malaria can be considered, the findings must be confirmed in multiple sites across the epidemiological spectrum of malaria in Africa, and the sample size must be large enough to evaluate the association of eye findings with cerebral histology (in fatal cases) and to assess the prognostic significance of the 'malarial retinopathy'in all cases satisfying the standard clinical definition. Our primary hypothesis is that the clinical case definition of cerebral malaria would be significantly improved by including the results of a bedside eye examination. Histological evidence of the cerebral sequestration of parasitized red cells will be the 'gold standard'diagnosis. A secondary hypothesis is that 'brain smears'(prepared and stained as thin blood films are, and thus within the scope of most district hospitals in Africa) can identify 'true'cerebral malaria as reliably as brain histology. These hypotheses will be tested within the SMAC (Severe Malaria in African Children) network, which includes five sites in East and West Africa. We expect to be able to enroll enough patients (n=1,964) to complete this study within 60 months. Patients who satisfy the clinical case definition of cerebral malaria will be eligible for enrollment. SMAC clinicians will be trained in direct and indirect opthalmoscopy, and consent to obtain brain tissue via the supra-orbital plate will be sought from family members in the event of a death. Eye findings will be compared to established prognostic factors in cerebral malaria (in all patients) and to brain histology (in fatal cases). If this eye examination does prove useful, a new definition of cerebral malaria would greatly improve the efficiency of pivotal studies of malaria treatment, pathogenesis, and prevention.
