To evaluate the effects of hematopoietic cytokines on bone marrow glucose metabolism noninvasively, we studied serial quantitative FDG-PET images in 18 patients with metastatic melanoma and normal bone marrow who were undergoing granulocyte-macrophage colony-stimulating factor (GMCSF) or macrophage colony-stimulating factor (MCSF) administration as an adjunct to chemotherapy.
Methods: All patients received 14 days of cytokine therapy in three groups: four patients were treated with GMCSF (5 micrograms/kg/d SQ), eight patients were treated with GMCSF (5 micrograms/kg/d SQ) and monoclonal antibody (MAbR24) and six patients were treated with MCSF (80 micrograms/kg/d IVCI) and MAbR24. Dynamic FDG-PET imaging was performed over the lower thoracic or upper lumbar spine at four time points in each patient.
Results: Baseline glucose metabolic rates in the bone marrow of these three groups of patients were similar (5.2 +/- 0.7, 4.4 +/- 0.8 and 4.8 +/- 1.2 micrograms/min/g as mean value and standard deviations, respectively). In both GMCSF and GMCSF + R24 groups, rapid increases in bone marrow glucose metabolic rates were observed during therapy. After GMCSF was stopped, bone marrow glucose metabolic rates rapidly decreased in both groups. The glucose metabolic response in these two groups was not significantly different by pooled t-statistics (p = 0.105). In the MCSF+R24 group, the increase of glucose metabolic rate on Days 3 and 10 was 35% and 31% above baseline on the average, but was not significant.
Conclusion: The results support the use of parametric FDG-PET imaging for noninvasive quantitation of bone marrow glucose metabolic changes to hematopoietic cytokines in vivo.