Diffuse large B-cell lymphomas (DLBCLs) are a heterogeneous group of diseases in terms of cell of origin, genetics and clinical outcome. About 30% of all DLBCL patients represent an unmet clinical need as they either do not respond to the standard first line chemo-immunotherapy or recur after initial remission. DLBCL classifications based on the NMF (Non-Negative Matrix Factorization) and LymphGen algorithms have led to the identification of genetic subtypes based on the co-occurrence of specific lesions. Repositioning drugs that are approved or in active clinical development for other indications can be a powerful strategy to match these newly uncovered DLBCL subtypes to targeted therapies.We have previously shown that the screening of large panels of DLBCL cell lines representative of the genetics of the disease can facilitate the repositioning of approved drugs for biomarker-selected populations of DLBCL patients. Repositioned drugs can then be rapidly translated to the clinical use, as they have already been extensively characterized for their safety profile. As a proof of concept, we have demonstrated that Dasatinib, a Src/Abl inhibitor approved for B-cell Acute Lymphoblastic Leukemia and Chronic Myelogenous Leukemia, is highly effective in PTEN-positive DLBCLs, irrespective of their COO class (Scuoppo et al., PNAS 2019).Here we present the results of a new screening that was performed on a panel of eight cell lines (4 ABC- and 4 GCB-DLBCLs) to test the activity of 212 drugs, either approved or in advanced clinical development, for repositioning in DLBCL, followed by validation in a larger panel of 34 genetically characterized cell lines. Our results point to inhibitors of the Nicotinamide Phosphoribosyl Transferase (NAMPT) as potently active in 63% of GCB-DLBCLs. NAMPT catalyzes the conversion of Nicotinamide (NAM) to Beta-Nicotinamide Mononucleotide (Beta-NMN). This reaction is the rate-limiting step of the Nicotinamide Adenine Dinucleotide (NAD) salvage pathway, the major metabolic route of NAD regeneration in mammalian cells. We validated the on-target activity of NAMPT inhibitors by multiple genetic and pharmacological approaches. First, we found that the activities of three chemically distinct drugs (FK-866, STF-118804 and KPT-9274) were highly correlated across the 34 lines DLBCL panel. Second, we were able to abolish the activity of all three NAMPT inhibitors by supplementing cells with Beta-NMN. Third, we showed that transduction of the drug-resistant mutant NAMPT H191R offsets the activity of the drugs.Dose-response assays on the full DLBCL cell line panel confirmed ABC-DLBCL resistance and also highlighted the presence of sensitive (GCB-S) and resistant (GCB-R) GCB subsets that can be separated by a subnanomolar IC50 threshold. To generate biomarkers capable of predicting GCB-S patients, we examined the RNA-seq profiles and genetic make-ups of the DLBCL cell lines collection and observed that the GCB-S subset was associated to the LymphGen EZB subtype, characterized by the presence of EZH2 mutations and BCL2 translocations. Conversely, the GCB-R subtype was linked to a 5-gene expression signature for the Kynurenine De Novo pathway, an alternative route for NAD synthesis. Accordingly, expression of each of the five De Novo Kynurenine pathway genes induced resistance to NAMPT inhibitors. These results were validated in xenotransplants of luciferized DLBCL lines and Patient Derived Xenotransplant models (PDXs) that were transcriptionally classified for the status of the Kynurenine De Novo signature. Together, these data support the repositioning of NAMPT inhibitors as a therapeutically relevant strategy for EZB-type GCB-DLBCLs. DisclosuresPasqualucci: Astra Zeneca: Research Funding; Sanofi: Research Funding.