AJR:198, April 2012
Vicente Gilsanz Houchun H. Hu Michelle L. Smith Fariba Goodarzian Sherri L. Carcich Nicole M. Warburton Marcio Malogolowkin
OBJECTIVE. The objective of our study was to determine whether the depiction of brown adipose tissue (RAT) in PET/CT studies of pediatric patients with lymphoma is related to disease status.
MATERIALS AND METHODS. The PET/CT studies of 31 pediatric patients (17 boys and 14 girls) with Hodgkin or non-Hodgkin lymphoma were reviewed, and the prevalence of metabolically active BAT at diagnosis and the prevalence of BAT when there was no evidence of disease were compared.
RESULTS. The percentage of PET/CT studies depicting BAT was greater when there was no evidence of disease than at diagnosis (10% vs 77%, respectively; p < 0.001). The McNemar test indicated a strong inverse correlation between the presence of disease and the presence of BAT (p < 0.001). This correlation was noted when all subjects were examined together and when subjects with Hodgkin lymphoma and those with non-Hodgkin lymphoma were analyzed separately (p < 0.001 and < 0.05, respectively). When baseline and follow-up PET/CT scans for all patients were analyzed for the presence of BAT using conditional logistic regression, both the season when the study was performed and disease status independently predicted BAT: The winter months positively predicted BAT and the presence of lymphoma was negatively correlated with the depiction of BAT on PET/CT. Age, sex, treatment, and weight did not provide additional information when added to the model.
CONCLUSION. The knowledge that BAT is a predictor of disease status should contribute to the correct analysis of PET/CT studies in children with lymphoma.