Study Objective. To delineate possible explanations for a nonmonotone
hematopoiesis, dose-response curve with filgrastim therapy after high-
dose chemotherapy. Design. Sequential two-phase study. Settings. Unive
rsity teaching hospital and basic pharmaceutical sciences laboratory.
Subjects. Thirty-nine patients with breast cancer or melanoma and 15 n
ormal CF-1 male mice. Interventions. Serial blood samples were obtaine
d from patients after high-dose chemotherapy to evaluate hematopoiesis
and tumor necrosis factor-alpha (TNF-alpha) concentrations. Murine he
matopoiesis was induced by filgrastim with or without coadministration
of lipopolysaccharide. Measurements and Main Results. Detection of pl
asma TNF-alpha in patients corresponded to substantially slower recove
ry of granulocytes, erythrocytes, and platelets, and was directly prop
ortional to the prescribed dosage of filgrastim. Lipopolysaccharide st
imulated the secretion of TNF-alpha in mice and totally aberrated filg
rastim-induced granulopoiesis. Conclusions. This in vivo evidence sugg
ests that regulatory pathways involving endogenous cytokines may overr
ide the effect of recombinant cytokines.