Using metabolic FDG PET criteria for the choice of target lesions in onco-haematological imaging

Introduction. - Malignant lymphoma is a heterogeneous and widespread disease. The morphological response assessment is currently based on the choice of target lesions at baseline, defined by size criteria on enhanced CT (eCT). FDG PET/CT is now commonly used at staging and the aim of this study was to evaluate the relevance of using metabolic criteria rather than size criteria to define the target lesions. Patients and methods. - Fifty-nine patients with aggressive lymphoma were retrospectively included. Target lesions were chosen by two radiologists on eCT and by two nuclear physicians on PET/CT. Response assessment, based on the sum of the products of the greatest diameters of up to six target lesions chosen either by size or metabolic criteria, was computed and compared to a clinical gold standard (GS) for each patient. Interobserver agreement and comparison to the GS were assessed with kappa (κ) and intraclass correlation (ICC) statistics. Results. - The spatial distribution of target nodal areas was equivalent among eCT and PET/CT readers with a maximum of target lesions in cervical and mediastinal areas. Choosing with PET/CT led to significant heterogeneity in the size of target lesions when compared with eCT alone (P = 0.03). Interobserver agreement for quantifying the response rate was equivalent in both groups. However, there was a greater correlation to the response of the GS when using PET/CT to target the patient (k = 0.64 vs. 0.47) and an increased rate of complete responses. Conclusions. - Metabolic criteria can replace current size criteria to define target lesions on FDG PET/CT. The morphological response rate appears accurate with an increased rate of complete responses after therapy and a better correlation to the haematological standard of reference. (authors)