Prognostic significance of bone marrow 2-[18F]-fluoro-2-deoxy-d-glucose uptake in diffuse large B-cell lymphoma: relation to iliac crest biopsy results.

ABSTRACT

AIM: To investigate the prognostic significance of bone marrow (BM) 2-[18F]-fluoro-2-deoxy-d-glucose (FDG) uptake in relation to posterior iliac crest BM biopsy (BMB) results in diffuse large B-cell lymphoma (DLBCL). MATERIALS AND METHODS: Pretreatment integrated positron-emission tomography(PET)/computed tomography (CT) images of 512 DLBCL patients who underwent BMB and received rituximab combined with cyclophosphamide, doxorubicin, vincristine, and prednisone (R-CHOP) chemotherapy were analysed retrospectively. BM uptake was assessed visually and by maximum standard uptake value (SUVmax). Associations with lymphoma-specific survival (LSS) were assessed using Kaplan-Meier and Cox regression analyses. RESULTS: FDG(+) BM was observed in 64 cases (41 focal, 12 heterogeneous, 11 diffuse). This finding distinguished iliac crest involvement (positive in 59 and negative in 453) with 89.6% accuracy (459/512) and 93.6% specificity (424/453). In BMB(+) patients, BM-to-liver SUVmax ratio >1.8 concurred perfectly with FDG(+) BM. During 52 months of follow-up, there were 156 lymphoma-related deaths. In the entire population, multivariate analysis revealed high International Prognostic Index (IPI; p<0.001), old age (p=0.003), bulky disease (p=0.011), BMB(+) (p=0.028), and FDG(+) BM (p=0.019) as independent predictors of worse LSS. In the BMB(+) subgroup, high National Comprehensive Cancer Network-revised IPI (NCCN-IPI; p=0.029) and FDG(+) BM (p=0.008) were significant independent predictors. Among BMB(+) patients with low to low-intermediate NCCN-IPI, FDG(+) BM was associated with significantly worse 2-year LSS (33.3% versus 100%; p=0.017). The same was true among those with high-intermediate NCCN-IPI (34.7% versus 76.9%.; p=0.026). CONCLUSION: Increased BM FDG in DLBCL is a predictor of worse LSS independent of BMB results and other prognostic variables including IPI/NCCN-IPI.