[Prognostic value of bone marrow (18)F-FDG uptake pattern of pretreatment PET-CT in extranodal NK/T cell lymphoma].

Objective: To investigate the prognostic value of bone marrow (BM) (18)F-fluorodeoxyglucose ((18)F-FDG) uptake pattern of pretreatment positron emission tomography/computed tomography (PET-CT) in extranodal NK/T cell lymphoma (ENKTL) patients. Methods: We retrospectively collected clinical data from a series of 63 ENKTL patients with stageⅡ~Ⅳ, who have received both (18)F-FDG PET-CT and bone marrow biopsy (BMB) prior to treatment. According the BM (18)F-FDG uptake pattern of PET-CT, the patients were divided into three groups: focal BM FDG uptake higher than liver (fPET+ ), diffuse BM uptake higher than liver (dPET+ ) and normal BM uptake (lower than liver) (nPET). The Kaplan-Meier method and Log-rank test were respectively used for survival analysis and univariate analysis, and COX proportional hazards model for multivariate analysis. Results: Among the 63 patients, 22 patients had nPET, 24 patients showed dPET+ , and 17 patients had fPET+ . BMB positive was found in 8 patients, and negative in 55 patients. Thirty-seven patients had disease progression or relapse, and 31 patients died. The 3-years progression free survival (3y-PFS) rates of fPET+ patients and nPET patients were 14.7% and 63.6% (P=0.006). The 3-years overall survival (3y-OS) rates were 18.8% and 64.8% (P=0.005). The 3y-PFS of dPET+ patients and nPET patients were 35.6% and 63.6% (P=0.161), 3y-OS were 47.9% and 64.8% (P=0.280). Univariate analysis showed that lactate dehydrogenase (LDH) level, Epstein-Barr virus DNA (EBV-DNA), Korean prognostic index (KPI) and BM (18)F-FDG were related with PFS and OS (all P<0.05). Multivariate analysis showed EBV-DNA and BM (18)F-FDG were independent predictors for PFS (P<0.05). EBV-DNA was also an independently predictor for OS (P<0.05). Conclusions: PET/CT-directed BM patterns are meaningful in predicting prognosis of newly diagnosed ENKTL patients. Focal BM (18)F-FDG uptake pattern is an independent predictor for PFS.