Outcomes of synchronous systemic and central nervous system (CNS) involvement of diffuse large B-cell lymphoma are dictated by the CNS disease: a collaborative study of the Australasian Lymphoma Alliance.

De novo diffuse large B-cell lymphoma (DLBCL) presenting with synchronous central nervous system (CNS) and systemic disease (synDLBCL) is not well described and is excluded from clinical trials. We performed a retrospective analysis of 80 synDLBCL patients treated across 10 Australian and UK centres. Of these patients, 96% had extranodal systemic disease. CNS-directed treatment with combination intravenous cytarabine and high-dose methotrexate ("CNS-intensive") (n = 38) was associated with favourable survival outcomes compared with "CNS-conservative" strategies such as intravenous high-dose methotrexate monotherapy, intrathecal therapy and/or radiotherapy (2-year progression-free survival [PFS] 50% vs. 31%, P = 0·006; 2-year overall survival [OS] 54% vs. 44%, P = 0·037). Outcomes were primarily dictated by the ability to control the CNS disease, with 2-year cumulative CNS relapse incidence of 42% and non-CNS relapse 21%. Two-year OS for CNS-relapse patients was 13% vs. 36% for non-CNS relapses (P = 0·02). Autologous stem cell transplantation as consolidation (n = 14) was not observed to improve survival in those patients who received CNS-intensive induction when matched for induction outcomes (2-year PFS 69% vs. 56%, P = 0·99; 2-year OS 66% vs. 56%, P = 0·98). Hyperfractionated or infusional systemic treatment did not improve survival compared to R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine, prednisolone) (2-year OS 49% for both groups). Our study suggests that adequate control of the CNS disease is paramount and is best achieved by intensive CNS-directed induction.

Prognostication of diffuse large B-cell lymphoma in the molecular era: moving beyond the IPI.

Diffuse large B-cell lymphoma (DLBCL) is a heterogeneous disease with variable outcomes. Despite the majority of patients being cured with combination chemoimmunotherapy, up to 30% eventually succumb to the disease. Until recently, baseline prognostic assessment has centred on the International Prognostic Index (IPI), although this index is yet to impact strongly on treatment choice. Molecular features such as cell of origin, MYC and BCL-2 genetic alterations and protein overexpression were identified over a decade ago, yet their prognostic value is still not fully elucidated. Adding complexity are the plethora of new clinical, biological and molecular prognostic markers described in the recent literature, most of which lack independent validation, likely act as surrogate markers for those already in common use and have yet to substantially impact on therapeutic decision making. This review comprehensively assesses the value of individual prognostic markers in the clinical setting and their potential to predict response to novel agents, and ways to optimise their use in future research.