Metabolic tumour volumes measured at staging in lymphoma: methodological evaluation on phantom experiments and patients.
Membres identifiés du Cancéropôle Est :
Dr CASASNOVAS Olivier
Tous les auteurs :
Meignan M, Sasanelli M, Casasnovas RO, Luminari S, Fioroni F, Coriani C, Masset H, Itti E, Gobbi PG, Merli F, Versari A
PURPOSE: The presence of a bulky tumour at staging on CT is an independent prognostic factor in malignant lymphomas. However, its prognostic value is limited in diffuse disease. Total metabolic tumour volume (TMTV) determined on (18)F-FDG PET/CT could give a better evaluation of the total tumour burden and may help patient stratification. Different methods of TMTV measurement established in phantoms simulating lymphoma tumours were investigated and validated in 40 patients with Hodgkin lymphoma and diffuse large B-cell lymphoma. METHODS: Data were processed by two nuclear medicine physicians in Reggio Emilia and Creteil. Nineteen phantoms filled with (18)F-saline were scanned; these comprised spherical or irregular volumes from 0.5 to 650 cm(3) with tumour-to-background ratios from 1.65 to 40. Volumes were measured with different SUVmax thresholds. In patients, TMTV was measured on PET at staging by two methods: volumes of individual lesions were measured using a fixed 41% SUVmax threshold (TMTV41) and a variable visually adjusted SUVmax threshold (TMTVvar). RESULTS: In phantoms, the 41% threshold gave the best concordance between measured and actual volumes. Interobserver agreement was almost perfect. In patients, the agreement between the reviewers for TMTV41 measurement was substantial (rho c = 0.986, CI 0.97 - 0.99) and the difference between the means was not significant (212 +/- 218 cm(3) for Creteil vs. 206 +/- 219 cm(3) for Reggio Emilia, P = 0.65). By contrast the agreement was poor for TMTVvar. There was a significant direct correlation between TMTV41 and normalized LDH (r = 0.652, CI 0.42 - 0.8, P
Eur J Nucl Med Mol Imaging. 2014 Jun;41(6):1113-22