To determine if the elevated transferrin saturations found in some pat
ients with severe malaria are associated with an adverse outcome in ce
rebral malaria, we retrospectively measured baseline saturations in st
ored serum samples from 81 Zambian children with strictly defined cere
bral malaria. The children had been treated with quinine, sulfadoxine-
pyrimethamine, and intravenous infusions of either placebo (n = 39) or
the iron chelator, desferrioxamine B (n = 42), in a previously report
ed trial (Gordeuk et al, N Engl J Med 327:1473, 1992), More than one-t
hird of children in both the placebo- and iron chelator-treated groups
had transferrin saturations exceeding 43%, which is 3 standard deviat
ions above the expected mean for age. Among children receiving quinine
and placebo, those with elevated transferrin saturations had a delaye
d estimated median time to recover full consciousness (68.2 hours) com
pared with those with saturations less than or equal to 43% (25.4 hour
s; P = .006). The addition of iron chelation to quinine therapy in chi
ldren with high saturations appeared to hasten recovery (P = .046). We
conclude that increased transferrin saturations may be associated wit
h delayed recovery from coma during standard therapy for cerebral mala
ria and that serum iron and total iron binding capacity should be meas
ured in future studies. (C) 1995 by The American Society of Hematology
.