cerebral blood flow may lead to anaerobic cerebral glycolysis and increased cerebral lactateproduction. Lactate levels are elevated both in arterial blood and in cerebrospinal fluid (CSF). Since lactate in the blood does not cross the blood-brain barrier into the CSF (Posner and Plum, 1967), this finding suggests that lactate is being independently generated in both fluid systems. The magnitude of CSF lactate elevation is higher in fatal than in nonfatal cases of cerebral malaria (White et al., 1985). Lactate is also a building block utilized by the liver in the synthesis of glucose. Hepatic gluconeogenesis may be impaired in severe malaria, and this could contribute to elevated plasma levels of lactate (Taylor et al., 1988). Another potential source of high plasma lactate levels is the metabolism of the parasites: P. falciparum consumes glucose and generates lactate as a byproduct. A mass of sequestered parasites may cause localized hypoglycemia (low blood sugar) and elevated lactate concentrations in the brain.

Another potential contributor to the pathogenesis of cerebral malaria is tumor necrosis factor (TNF), a cytokine. Plasma levels of TNF are elevated in adults with severe malaria (Kern et al., 1989) and in children with acute P. falciparum infections (Grau et al., 1989; Kwiatkowski et al., 1990). TNF is produced by monocytes and macrophages in response to a number of stimuli (Carswell et al., 1975; Cuturi et al., 1987) and has been shown to cause a wide variety of physiological effects in humans (Tracey et al., 1986). Some of these, such as fever and hypoglycemia, are common in children with severe malaria. Others, like low blood pressure, kidney failure, and a disruption of the blood clotting mechanism, occur frequently in adults with severe malaria. The extent to which TNF contributes to the pathogenesis of human clinical illness is unclear, but in a study of mice infected with a mouse malaria parasite, treatment with anti-TNF antibodies protected the animals from the cerebral complications of the disease (Grau et al., 1987), and high plasma levels of TNF in children with malaria are associated with an increased risk of dying (Grau et al., 1989; Kwiatkowski et al., 1990).

It has also been suggested that immune mechanisms may play a role in cerebral malaria. In some rodent models of malaria, cerebral lesions contain local collections of inflammatory cells. These cerebral lesions and associated neurological symptoms can be prevented by pretreatment with corticosteroids (potent anti-inflammatory agents), cyclosporin A (an inhibitor of T-lymphocyte function), or antibodies to TNF. Inflammatory lesions are not found in human cerebral malaria, however, and neither cyclosporin A nor corticosteroids have proven to be an effective treatment for human cerebral malaria, although pretreatment with these drugs has not been possible. In this respect, human P. falciparum malaria differs significantly from certain animal malarias, where accumulations of mononuclear cells are found in cerebral vessels (Grau et al.,