Positron Emission Tomography Reliably Distinguishes Aggressive but Not Low Grade Lymphomas from Reactive Lymph Nodes
Gang Zheng, Milena Vuica-Ross, Kathleen H Burns, Christopher D Gocke, Michael J Borowitz, Amy S Duffield. Johns Hopkins Hospital, Baltimore, MD
Background: 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) is widely utilized in clinical management of lymphoma, including pretreatment staging and therapeutic monitoring. In PET imaging, semi-quantitative analyses using standardized uptake values (SUVs) allow for an objective assessment of glucose uptake and metabolic activity. Significant FDG uptake can also be seen in benign conditions, creating diagnostic dilemmas. The aim of this study was to evaluate whether SUVs in FDG-PET studies can be reliably utilized to distinguish reactive conditions, low-grade B cell lymphomas and aggressive B cell lymphomas from one other.
Design: From January 2009 to September 2012, patients with a diagnosis of reactive lymphoid hyperplasia, low-grade B cell lymphoma (follicular lymphoma, mantle cell lymphoma, marginal zone lymphoma, small lymphocytic lymphoma, lymphoplasmacytic lymphoma, and low-grade lymphoma, not otherwise specified), or aggressive lymphoma (diffuse large B cell lymphomas and Burkitt lymphoma), who had received a corresponding FDG-PET scan for initial staging, were identified. SUV values were only utilized from locations where the tumor was biopsy-proven, and the FDG-PET scans were performed within two months of the original biopsy.
Results: 27 patients with reactive lymph node(s) were identified, as well as 32 patients with a low-grade lymphoma and 33 patients with an aggressive lymphoma. Average SUVs are 5.6 (1.2-18.4) for reactive lymph nodes, 5.0 (0-14.3) for low-grade lymphoma, and 14.7 (5.4-31.5) for high-grade lymphoma. High SUV is an excellent predictor for high-grade lymphoma (vs reactive conditions: AUC=0.91 by ROC analysis, p<0.001; vs low-grade lymphomas: AUC=0.93 by ROC analysis, p<0.001). At a SUV cutoff of 7.2, there is a sensitivity of 94% and a specificity of 85% in differentiating a high-grade lymphoma from reactive lymph nodes. However, SUV is not useful in differentiating reactive lymph nodes from low-grade lymphomas (AUC=0.54 by ROC analysis, p=0.59).
Conclusions: Although a high SUV (>7.2) strongly predicts a high-grade lymphoma, the significance of a relatively low SUV is uncertain as this can be seen in both reactive lymph nodes and low-grade lymphomas. In those situations, clinical correlation and possible biopsy are needed.
Tuesday, March 5, 2013 9:30 AM
Poster Session III # 161, Tuesday Morning