Nicotinamide phosphoribosyltransferase (NAMPT) catalyzes the rate limiting step in the nicotinamide adenine dinucleotide (NAD +) salvage pathway: the conversion of nicotinamide and 5-phosphoribosyl-1-pyrophosphate into nicotinamide mononucleotide. Because NAMPT plays a central role in cancer metabolism (e.g., Warburg effect) and is upregulated in many hematologic malignancies, it is a high-value target for oncology applications. Despite promising preclinical data showing the potential of complete NAMPT inhibitors (NAMPTi) as effective oncology agents, early clinical trials yielded limited benefit due to the small therapeutic window for the agents tested. Complete NAMPT deactivation disrupts essential biological pathways in both cancer and healthy cells, and hence could not induce efficacy without causing dose-limiting toxicities. Here we introduce a novel partial NAMPTi, RPT1G, which was optimized to allow residual activity to meet the metabolic needs of healthy cells and to overcome the challenge of on-target toxicity. RPT1G's 1 st-in-class mechanism of action (MOA) simultaneously inhibits and stabilizes NAMPT in a catalytically active state, disrupting NAMPT enzymatic activity without ever turning it off. RPT1G achieves significant efficacy in a B-cell acute lymphoblastic leukemia (B-ALL) mouse xenograft model and avoids the severe on-target toxicities seen with complete NAMPTi. To demonstrate the importance of NAMPT residual activity in preserving cellular metabolism in healthy cells, we determined the mitochondrial respiration rate for RS4;11 (B-ALL) and human peripheral blood mononuclear cells (huPBMCs) treated with RPT1G. The oxygen consumption rate (OCR) markedly decreased in treated RS4;11 cells, consistent with RPT1G-induced inhibition of cellular bioenergetics. In contrast, in huPBMC, increasing concentrations of RPT1G elicited a dose-dependent decrease of OCR, up to a concentration equal to 5-fold the IC50 in RS4;11, beyond which, further reductions in OCR were not observed. This indicates that RPT1G's influence on mitochondrial respiration had reached its maximum effect, consistent with huPBMCs being able to use their reserve capacity to maintain cell survival. To further validate the novel MOA of RPT1G in vivo, we tested the viability of >100 cancer cell lines as well as huPBMCs upon RPT1G treatment. Leukemias, specifically ALL cells, displayed high sensitivity to RPT1G. Also, the concentrations required to achieve >90% NAD + reduction in RS4;11 cells do not cause NAD + reduction, nor significant decrease in viability in huPBMCs, highlighting the specificity of RPT1G against cancer versus healthy cells. In contrast, complete inhibitors led to a decrease in cell health for both leukemia and healthy cells, which likely explains their poor tolerability in previous clinical trials. To narrow down the sensitivity profile of RPT1G, we performed multiomic analyses across >200 genomically diverse cancer cell lines using a pharmacologically related analog of RPT1G and identified TET2 mutations as prospective prognostic biomarkers. We then conducted efficacy and non-GLP tolerability proof of concept studies to assess the therapeutic index (TI) of RPT1G. The efficacy was evaluated in a human B-ALL RS4;11 xenograft mouse model at 25 mg/kg BID and 150mg/kg QD, while the tolerability was assessed at doses up to 200 mg/kg QD. Comparison of the exposure levels from both studies revealed that RPT1G has an enhanced therapeutic window, with a TI >7.9 (previous NAMPTi TI < 3). RPT1G exhibited significant efficacy at 25 mg/kg BID while maintaining an acceptable level of tolerability at 200mg/kg QD, showing no meaningful toxicological change in hematology or clinical chemistry, and no gross organ lesions, thereby providing crucial insights on dose selection for its clinical development. Our results demonstrate that the 1 st-in-class partial NAMPTi RPT1G modulates NAMPT activity and exploits the different sensitivities and metabolic demands of tumor and normal cells. This results in selective killing of leukemic cells, without on-target toxicity seen with complete NAMPTi. RPT1G emerges as an ideal clinical candidate - its favorable efficacy and tolerability target profile in preclinical models provides robust support for its utility in treating leukemias, warranting the initiation of our phase I clinical study to assess both safety and efficacy in patients.