Glucose metabolic phenotyping and prediction of tumor grade in soft tissue and bone tumors by 18F-FDG PET imaging

Objectives The aims of this study were twofold: first we examined whether the glucose metabolic phenotype varies among different WHO sarcoma subtypes. Second, we determined whether the degree of FDG uptake correlates with tumor grade determined by a) a two tier grading system and b) a three tier grading system of the French Federation of Cancer Centers Sarcoma Group (FNCLCC).

Methods 124 patients (64 males, 60 females; 48±18 yrs) with 21 soft tissue and bone tumor subtypes underwent baseline FDG-PET/CT imaging. Fourteen patients (11%) had benign soft tissue tumors, 97 patients (78%) had soft tissue sarcomas, and 13 patients (10%) had bone sarcomas. Tumor SUVmax was compared amongst subtypes and with histopathologic grade.

Results Although some histologic subtypes showed a wider range in SUVmax compared to others, SUVmax did not differ significantly among the various high grade sarcoma subtypes (p=0.54). Tumor SUVmax differed significantly among grade 1 (4.36±1.86) grade 2 (9.94±7.21) and grade 3 (14.88±11.46) tumors (p<0.001). In addition, SUVmax was significantly higher in high grade compared to low grade sarcomas (11.70±8.95 vs. 4.94±4.16; p<0.001). The FNCLCC grading system and the 2 tier grading system predicted tumor grade in primary tumors with similar accuracy (AUC=0.84 and 0.83; p=0.85). The optimal separation of high and low grade tumors was achieved by a SUVmax threshold of 5.2 (Sensitivity, 79%; Specificity, 82%).

Conclusions FDG-PET cannot discriminate between different sarcoma subtypes. The degree of FDG tumor uptake correlates reasonably well with histological grade.