Search

Supplementary and expanded descriptions of methods and figure legends for supplementary data

Supplementary Figure S1. Identification and refinement of cancer-specific super-enhancers in human AML. Supplementary Figure S2. Mapping of super-enhancers to genes. Supplementary Figure S3. AML SE profiles show pronounced variation in enhancers related to myeloid differentiation and enable de novo stratification into six distinct epigenomic subgroups. Supplementary Figure S4. Recapitulation of SE-defined clusters in AML data from the TCGA. Supplementary Figure S5 An SE is present at the RARA locus in a subset of AML samples. Supplementary Figure S6 SY-1425 is a selective and potent RARα agonist. Supplementary Figure S7 Comparison of SY-1425 response to RARA expression and enhancer score and the effect of an RARα antagonist. Supplementary Figure S8 SY-1425 in vivo pharmacokinetics. Supplementary Figure S9 SY-1425 survival and tissue effects on AML PDX models. Supplementary Figure S10 SY-1425 shows greater potency than ATRA in vitro and in vivo. Supplementary Figure S11 Induction of markers of myeloid differentiation in RARA-high AML cell lines upon SY-1425 treatment. Supplementary Figure S12 Comparison of SY-1425 expression and molecular response to APL. Supplementary Figure S13 SY-1425 expression and molecular response.

GSEA of transcriptional response to SY-1425 in RARA-high cell lines in perturbation gene sets & GSEA of transcriptional response to SY-1425 in RARA-high cell lines in hallmark/pathways and perturbation gene sets

Targeted-sequencing analysis of PDX and PairTree predicted mutational population clusters

Disease recurrence causes significant mortality in B-progenitor acute lymphoblastic leukemia (B-ALL). Genomic analysis of matched diagnosis and relapse samples shows relapse often arising from minor diagnosis subclones. However, why therapy eradicates some subclones while others survive and progress to relapse remains obscure. Elucidation of mechanisms underlying these differing fates requires functional analysis of isolated subclones. Here, large-scale limiting dilution xenografting of diagnosis and relapse samples, combined with targeted sequencing, identified and isolated minor diagnosis subclones that initiate an evolutionary trajectory toward relapse [termed diagnosis Relapse Initiating clones (dRI)]. Compared with other diagnosis subclones, dRIs were drug-tolerant with distinct engraftment and metabolic properties. Transcriptionally, dRIs displayed enrichment for chromatin remodeling, mitochondrial metabolism, proteostasis programs, and an increase in stemness pathways. The isolation and characterization of dRI subclones reveals new avenues for eradicating dRI cells by targeting their distinct metabolic and transcriptional pathways before further evolution renders them fully therapy-resistant.Significance:Isolation and characterization of subclones from diagnosis samples of patients with B-ALL who relapsed showed that relapse-fated subclones had increased drug tolerance and distinct metabolic and survival transcriptional programs compared with other diagnosis subclones. This study provides strategies to identify and target clinically relevant subclones before further evolution toward relapse.See related video: https://vimeo.com/442838617See related article by E. Waanders et al.

RNA-seq derived signatures and per-cluster weights for TCGA clustering

Table of median survival in TCGA data by predicted clusters

GREAT analysis of genes near H3K27ac peaks up-regulated by SY-1425 in RARA-high cell lines in perturbation gene sets

Patient sample features including cytogenetic alterations and mutations

Table of Enhancer, mRNA, and SY-1425 EC50 values in AML cell lines

Materials

Supplementary Figure 5 and legend

Characteristics of the primary AML samples

Mutant isocitrate dehydrogenase 1 (IDH1) and IDH2 block the differentiation of acute myeloid leukemia (AML) cells through production of R-2-hydroxyglutarate (R-2-HG). IDH inhibitors can induce differentiation of AML cells by lowering R-2-HG but have limited clinical efficacy as single agents. Here, we performed a genome-wide CRISPR knockout screen in an Idh1-mutated hematopoietic progenitor cell line to identify genes that increased the differentiation response to ivosidenib, an IDH1 inhibitor. The screen identified C-type lectin member 5a (Clec5a), which encodes a spleen tyrosine kinase (SYK)-coupled surface receptor, as one of the top hits. Knockout of Clec5a and Syk rendered cells more sensitive to ivosidenib-induced differentiation through a reduction in STAT5-dependent expression of stemness-related genes, including genes in the homeobox (HOX) family. Importantly, direct inhibition of STAT5 activity was sufficient to increase the differentiation response to IDH inhibitors in primary human IDH1- and IDH2-mutated AML cells, including those harboring mutations in receptor tyrosine kinase (RTK) and MAPK genes that have been linked to drug resistance. In patient-derived xenograft models of IDH1-mutated AML, combination treatment with ivosidenib and the STAT5 inhibitor pimozide was superior to each agent alone in inducing differentiation in leukemic cells without compromising normal hematopoiesis. These findings demonstrate that STAT5 is a critical mediator of resistance to IDH inhibitors and provide the rationale for combining STAT5 and IDH inhibitors in the treatment of IDH-mutated AML.Significance:A CRISPR knockout screen identifies a mechanism of resistance to IDH inhibitors in AML involving activated STAT5 signaling, suggesting a potential strategy to improve the clinical efficacy of IDH inhibitors.

Venetoclax requires a 75% dose reduction when coadministered with voriconazole. In a 10-year historical cohort of treatment with venetoclax, we did not observe a worse hematologic outcome in patients who received voriconazole prophylaxis versus those who did not. Subtherapeutic voriconazole levels and a triazole exposure history may contribute to breakthrough invasive fungal infection.

The impact of adverse risk genetic profiles on outcomes in acute myeloid leukemia (AML) patients following allogeneic hematopoietic stem cell transplantation (HCT) has not been fully elucidated. Accordingly, we have profiled somatic mutations at diagnosis using next-generation sequencing (NGS) in 178 AML patients who received allogeneic HCT. NGS revealed 598 somatic mutations in 165/178 patients (92.7%). Frequently mutated genes include DNMT3A, TET2, NPM1, RUNX1, IDH2, and FLT3. Commonly detected cytogenetic profiles include normal karyotype, trisomy 8, monosomal karyotype (MK), deletion 5, complex karyotype (CK), and monosomy 7. In univariate analyses, TP53 mutation, MK, CK, and monosomy 7 were associated with decreased overall survival (OS), relapse-free survival (RFS), and a higher relapse incidence (RI). We defined adverse molecular-genetic profile as harboring at least one of the molecular/genetic abnormalities of TP53 mutation, MK, CK, monosomy 7, and deletion 5. The patients harboring adverse molecular-genetic profile (n = 30) showed a lower 2-year OS (24.9% vs. 57.9%; p = 0.003), RFS (23.7% vs. 57.9%; p = 0.002), and higher RI (47.2% and 17.2%; p = 0.001) after HCT when compared to patients without those lesions. Multivariate analysis confirmed adverse molecular-genetic profile as an independent prognostic factor, associated with decreased OS (HR 2.19), RFS (HR 2.23), and higher RI (HR 2.94).

Background: venous thromboembolism (VTE) is a well-known complication in adults with acute lymphoblastic leukemia (ALL), especially in patients treated with asparaginase (ASNase)-including regiments. However, VTE risk in adult Philadelphia-positive ALL (Ph+ve ALL) patients treated with non-hyperCVAD chemotherapy is unclear. In this study, we examined VTE incidence in adult Ph+ve ALL patients treated with imatinib plus a pediatric-inspired asparaginase (ASNase)-free regimen modified from the Dana Farber Cancer Institute (DFCI) ALL protocol. Methods: a single centre retrospective review of Ph+ve ALL patients treated at Princess Margaret Cancer Center (PMCC) from 2008&ndash, 2019 with imatinib plus modified DFCI protocol was conducted. Results: of the 123 patients included, 30 (24.3%) had at least 1 radiology confirmed VTE event from diagnosis to the end of maintenance therapy. 86.7% (26/30) of the VTE events occurred during active treatment. Of all VTE events, the majority (53.3%) were DVT and/or PE while another significant portion were catheter-related (40.0%). Major bleeding was observed in 1 patient on VTE treatment with low molecular weight heparin (LMWH). Conclusion: a high VTE incidence (24.3%) was observed in adults Ph+ve ALL patients treated with imatinib plus an ASNase-free modified DFCI pediatric ALL protocol, suggesting prophylactic anticoagulation should be considered for all adult Ph+ve ALL patients including those treated with ASNase-free regimens.

Mutant isocitrate dehydrogenase 1 (IDH1) and IDH2 block the differentiation of acute myeloid leukemia (AML) cells through production of R-2-hydroxyglutarate (R-2-HG). IDH inhibitors can induce differentiation of AML cells by lowering R-2-HG but have limited clinical efficacy as single agents. Here, we performed a genome-wide CRISPR knockout screen in an Idh1-mutated hematopoietic progenitor cell line to identify genes that increased the differentiation response to ivosidenib, an IDH1 inhibitor. The screen identified C-type lectin member 5a (Clec5a), which encodes a spleen tyrosine kinase (SYK)-coupled surface receptor, as one of the top hits. Knockout of Clec5a and Syk rendered cells more sensitive to ivosidenib-induced differentiation through a reduction in STAT5-dependent expression of stemness-related genes, including genes in the homeobox (HOX) family. Importantly, direct inhibition of STAT5 activity was sufficient to increase the differentiation response to IDH inhibitors in primary human IDH1- and IDH2-mutated AML cells, including those harboring mutations in receptor tyrosine kinase (RTK) and MAPK genes that have been linked to drug resistance. In patient-derived xenograft models of IDH1-mutated AML, combination treatment with ivosidenib and the STAT5 inhibitor pimozide was superior to each agent alone in inducing differentiation in leukemic cells without compromising normal hematopoiesis. These findings demonstrate that STAT5 is a critical mediator of resistance to IDH inhibitors and provide the rationale for combining STAT5 and IDH inhibitors in the treatment of IDH-mutated AML. Significance: A CRISPR knockout screen identifies a mechanism of resistance to IDH inhibitors in AML involving activated STAT5 signaling, suggesting a potential strategy to improve the clinical efficacy of IDH inhibitors.

Techniques for analyzing circulating tumor DNA (ctDNA) to detect, characterize and monitor cancer have matured rapidly. An increasing body of clinical evidence is demonstrating the capabilities of this technology as a diagnostic test. The full potential of ctDNA liquid biopsy in the diagnosis, characterization and management of solid and hematological malignancies will be uncovered through interventional clinical trials evaluating clinical utility. In this Review, we discuss the current landscape of ctDNA liquid-biopsy applications across the cancer continuum and highlight opportunities for clinical investigation. Evaluation of circulating tumor DNA in blood has emerged as a powerful technology for oncology research. Lillian Siu and colleagues review the potential applications of liquid biopsy, highlighting clinical-trial designs to establish its clinical utility.

Key Points DNMT3A R882, TET2, ASXL1, and SRSF2 mutations identified at the time of diagnosis are associated with delayed count recovery. Persistence of preleukemic mutations in remission at high variant allele frequency is associated with delayed count recovery.

Optimization of post‐remission therapies to maintain complete remission and prevent relapse is a major challenge in treating patients with acute myeloid leukemia (AML). Monitoring patients for measurable residual disease (MRD) is helpful to identify those at risk for relapse. Hypomethylating agents are being investigated as post‐remission therapy. Identification of recurrent genetic alterations that drive disease progression has enabled the design of new, personalized approaches to therapy for patients with AML. Emerging data suggest that targeted post‐remission therapy, alone or in combination with chemotherapy, may improve outcomes. Results of ongoing clinical trials will further define potential clinical benefits.

Espresso efficiently distinguishes true mutations from background noise common in deep sequencing data., Sensitive mutation detection by next-generation sequencing is critical for early cancer detection, monitoring minimal/measurable residual disease (MRD), and guiding precision oncology. Nevertheless, because of artifacts introduced during library preparation and sequencing, the detection of low-frequency variants at high specificity is problematic. Here, we present Espresso, an error suppression method that considers local sequence features to accurately detect single-nucleotide variants (SNVs). Compared to other advanced error suppression techniques, Espresso consistently demonstrated lower numbers of false-positive mutation calls and greater sensitivity. We demonstrated Espresso’s superior performance in detecting MRD in the peripheral blood of patients with acute myeloid leukemia (AML) throughout their treatment course. Furthermore, we showed that accurate mutation calling in a small number of informative genomic loci might provide a cost-efficient strategy for pragmatic risk prediction of AML development in healthy individuals. More broadly, we aim for Espresso to aid with accurate mutation detection in many other research and clinical settings.

Core-binding factor acute myeloid leukemia is characterized by t(8;21) or inv(16) and the fusion proteins RUNX1-RUNX1T1 and CBFB-MYH11. International guidelines recommend monitoring for measurable residual disease every 3 months for 2 years after treatment. However, it is not known whether serial molecular monitoring can predict and prevent morphological relapse. We conducted a retrospective singlecenter study of 114 patients in complete remission who underwent molecular monitoring with real-time quantitative polymerase chain reaction analysis of RUNX1-RUNX1T1 or CBFB-MYH11 transcripts every 3 months. Morphological relapse was defined as re-emergence of >5% blasts and molecular relapse as ≥1 log increase in transcript level between two samples. Over a median follow-up time of 3.7 years (range, 0.2-14.3), remission persisted in 71 (62.3%) patients but 43 (37.7%) developed molecular or morphological relapse. Patients who achieved

Current treatments for acute myeloid leukemia (AML) are often ineffective in eliminating leukemic stem cells (LSCs), which perpetuate the disease. Here, we performed a metabolically focused drug screen to identify LSC-specific vulnerabilities and found that inhibitors of nicotinamide phosphoribosyltransferase (NAMPT), a rate-limiting enzyme in the NAD+ salvage pathway, selectively killed LSCs, while sparing normal hematopoietic stem and progenitor cells. Treatment with KPT-9274, a NAMPT inhibitor, suppressed the conversion of saturated fatty acids to monounsaturated fatty acids, a reaction catalyzed by the stearoyl-CoA desaturase (SCD) enzyme, resulting in lipotoxicity and apoptosis of AML cells. Transcriptomic analysis of LSCs treated with KPT-9274 revealed an upregulation of genes involved in lipid metabolism, including SCD . This transcriptional response conferred partial protection against NAMPT inhibitors, and its inhibition enhanced the cytotoxicity of KPT-9274 on LSCs in vivo. Our work demonstrates that altered lipid homeostasis plays a key role in NAMPT inhibitor-induced apoptosis and identifies NAMPT inhibition as a therapeutic strategy for targeting LSCs in AML.

Current treatments for acute myeloid leukemia (AML) are often ineffective in eliminating leukemic stem cells (LSCs), which perpetuate the disease. Novel therapies that target LSCs have the potential to improve clinical outcomes. An important step towards achieving this goal is identifying the fundamental processes that regulate cell fate decisions in LSCs. Perturbation of these processes may impair LSC activity and form the basis of novel therapies. Here, we investigated the role of mitochondrial quality control in the regulation of LSCs by inhibiting PTEN-induced kinase 1 (PINK1). PINK1 is a serine/threonine kinase that serves as a critical sensor of mitochondrial damage and is at the apex of multiple quality control pathways that maintain mitochondrial homeostasis. Damage to mitochondria results in activation of PINK1 on the outer mitochondrial membrane, which in turn triggers mitochondrial fission and mitophagy. We first determined if the overall quality of the mitochondrial pool differs between LSCs and their more differentiated progenies (henceforth termed "non-LSCs") in two human AML cultures known as OCI-AML-8227 and OCI-AML-21. In both cultures, LSCs assayed by xenotransplantation are found only in the CD34 +CD38 -fraction and not in the CD34 +CD38 + and CD34 - fractions. We monitored the turnover rate of the mitochondrial pool in the different fractions by expressing MitoTimer, a mitochondria-targeted fluorescent protein that matures from green to red fluorescence over 48 hours. A high ratio of green to red fluorescence is indicative of active mitochondrial protein turnover and high overall mitochondrial quality. In both cell lines, the fluorescence ratio was highest in the CD34 +CD38 - fraction compared with the other fractions, suggesting that LSCs maintain a higher quality pool of mitochondria than non-LSCs. Based on the above findings, we hypothesized that PINK1-dependent mitochondrial quality control mechanisms are involved in the regulation of LSC fate. To test this hypothesis, we silenced the expression of PINK1 in OCI-AML-8227 cells using lentiviral vectors expressing validated shRNAs under a doxycycline-inducible promoter. Depletion of PINK1 shifted the MitoTimer fluorescence from green to red, reduced oxygen consumption rate, and disrupted mitochondrial ultrastructure in all cell fractions, consistent with a reduction in mitochondrial quality. Unexpectedly, PINK1 downregulation resulted in a block in differentiation and cell cycle arrest at G1/G0 phase in CD34 +CD38 - cells. Re-expression of PINK1 was unable to reverse the cell cycle arrest, suggestive of a state of cellular senescence. Indeed, we observed other hallmarks of senescence including an increase in p21 WAF1 and p16 INK4a expression and SA-β-gal activity. To determine the mechanism by which PINK1 depletion causes senescence, we performed RNA sequencing analysis of sorted CD34 +CD38 - cells expressing PINK1 shRNAs. This analysis revealed a marked decrease in the expression of MYC target genes, with TERT (Telomerase Reverse Transcriptase) being the most downregulated gene. Consistent with the role of TERT in telomere maintenance and telomere shortening as a trigger of senescence, we found that PINK1 knockdown decreased telomere length in CD34 +CD38 - cells, and overexpression of TERT effectively rescued the senescence phenotype. These findings collectively indicate that inhibition of PINK1-dependent mitochondrial quality control pathways induces senescence of LSCs through telomere shortening. To determine whether these changes translated to a reduction in functional LSC activity, we performed colony forming unit assays and serial xenotransplantation assays in immunodeficient NSG mice. Depletion of PINK1reduced the colony forming capacity and engraftment potential of 3 primary AML samples in primary and secondary recipients. Importantly, PINK1 depletion had minimal impact on the colony forming capacity and engraftment potential of normal CD34 + hematopoietic stem and progenitor cells (HSPCs) derived from cord blood, suggestive of a therapeutic window in vivo. In summary, our results demonstrate that mitochondrial quality control pathways regulate cell fate decision in LSCs. Inhibition of PINK1 activity impairs LSC activity by inducing senescence, while sparing normal HSPCs. Our findings provide the basis for exploring PINK1 as a therapeutic target against LSCs in AML. Disclosures Chan: AbbVie: Research Funding; BMS: Research Funding.

BACKGROUND: Isocitrate dehydrogenase 2 (IDH2) mutations are found in about 10-15% of acute myeloid leukemia (AML) cases. Enasidenib (ENA) is a first-in-class mutant IDH2 inhibitor that induces differentiation of IDH2-mutated leukemic cells. However, the clinical efficacy of single agent ENA therapy in relapsed or refractory (R/R) AML is limited, underscoring the need for combination therapy. Preclinical studies have shown that IDH-mutated leukemic cells are particularly sensitive to BCL2 inhibition with venetoclax (VEN), a finding supported by clinical studies of VEN combination therapies. We previously demonstrated that the combination of ENA and VEN can be more effective than each drug alone, in reducing leukemic burden in patient-derived xenograft models of IDH2-mutated AML. Here, we report preliminary safety and efficacy results of an ongoing open-label, single-arm, phase Ib/II trial (NCT04092179) of ENA in combination with VEN in patients with IDH2-mutated myeloid malignancies. METHODS: Patients 18 years of age or above with IDH2-mutated R/R AML or high-risk MDS/MPN and an ECOG performance status of 0-2 were eligible. Patients previously treated with an IDH2 or BCL2 inhibitor were excluded. Participants received VEN continuously starting on cycle 1 day 1 with a 3-day ramp-up to a target dose of 400 mg daily (dose level 0) in cohorts 1 and 2. ENA was administered at 100 mg daily continuously starting on cycle 1 day 15. Each cycle was 28 days. Concurrent use of a moderate or strong CYP3A4 inhibitor was allowed with 50% dose reduction of VEN after completion of the first 2 cycles at 100% target dose. Interruptions of VEN and/or ENA were permitted for management of adverse events (AEs). Primary endpoints were 1) safety and tolerability and 2) overall response rate (ORR) defined as complete remission (CR) + CR with incomplete blood count recovery (CRi) + morphologic leukemia-free state (MLFS) + partial remission (PR) by revised IWG criteria. Secondary endpoints include pharmacokinetic (PK) profiles of VEN, duration of response, overall survival (OS), and IDH2 mutant allele burden in bone marrow by ddPCR. RESULTS: The study opened to recruitment in November of 2020. As of July 28, 2021 (data cutoff), 11 patients were enrolled on study; 10 with R/R AML and 1 with very-high risk MDS by IPSS-R. Six patients had a R140Q mutation, and 5 had a R172K mutation. Median age was 72 years (range: 32 - 80); 6 patients were male. Participants had received a median of 2 prior lines of therapies (range: 1 - 4). Six of 10 AML patients had primary refractory disease. The MDS patient experienced secondary azacitidine failure. Key treatment emergent grade ≥ 3 AEs regardless of attribution were: febrile neutropenia (n=3), intracranial hemorrhage (n=3), lung infection (n=2), other infection (n=2), elevated AST/ALT (n=2), sepsis (n=1), leukocytosis (n=1), TRALI (n=1), and small bowel obstruction (n=1). No cases of differentiation or tumor lysis syndrome were observed. No patients discontinued the study due to AEs. The addition of ENA, a known inhibitor of CYP3A4, did not significantly affect the PK profiles of VEN. Nine of the AML patients completed at least 1 cycle of treatment and are considered evaluable for efficacy. One AML patient died from intracranial hemorrhage prior to completion of cycle 1. Median duration of observation is 3.5 months. Of the evaluable patients, CR was achieved in 2 patients (22%) and CRi in 3 patients (33%) for an ORR of 55% (Fig. 1). Median number of cycles to response was 3. All responders remain in remission and on study with a median of 6 cycles received to date (range: 3 - 8). Of the remaining 4 patients, 2 patients (22%) remain on study with stable disease, and 2 patients (22%) experienced progressive disease and died (one after 7 cycles and the other after 1 cycle). Median OS for the entire cohort has not been reached. A sustained reduction in mutant IDH2 allele frequency in bone marrow correlated with response (Fig. 2). For the MDS patient, no response was observed after 1 cycle of treatment. CONCLUSIONS: VEN in combination with ENA is a well-tolerated regimen with no dose-limiting toxicities observed at the current dose level. The preliminary efficacy of this combination is encouraging with an ORR of 55% in evaluable R/R AML patients, with some responders achieving deep molecular remissions. Patient enrollment in dose-escalation cohorts is ongoing. Figure 1 Figure 1. Disclosures Chan: AbbVie: Research Funding; BMS: Research Funding. Gupta: Sierra Oncology: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; Constellation Pharma: Consultancy, Honoraria; Incyte: Honoraria, Research Funding; AbbVie: Consultancy, Honoraria; Roche: Consultancy; BMS-Celgene: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees; Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Pfizer: Consultancy. Maze: Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees; Celgene-BMS: Honoraria; Takeda: Research Funding; PharmaEssentia: Research Funding; Kronos Bio: Research Funding. Minden: Astellas: Consultancy. Schimmer: Takeda Pharmaceuticals: Consultancy, Research Funding; Medivir AB: Research Funding; Novartis: Consultancy, Honoraria; Jazz: Consultancy, Honoraria; Otsuka Pharmaceuticals: Consultancy, Honoraria; UHN: Patents & Royalties. Schuh: AbbVie: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Amgen: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Astellas: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Bristol-Myers Squibb: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Pfizer: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Agios: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Servier: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; GlycoMimetics: Research Funding; Kite/Gilead: Research Funding; Jazz: Membership on an entity's Board of Directors or advisory committees; Teva: Honoraria, Membership on an entity's Board of Directors or advisory committees. Yee: Pfizer: Membership on an entity's Board of Directors or advisory committees; Forma Therapeutics: Research Funding; F. Hoffmann La Roche: Membership on an entity's Board of Directors or advisory committees, Research Funding; Tolero: Research Funding; Genentech: Research Funding; Janssen: Research Funding; Novartis: Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding; Shattuck Labs: Membership on an entity's Board of Directors or advisory committees; Onconova: Research Funding; Bristol-Myers Squibb/Celgene: Membership on an entity's Board of Directors or advisory committees; Otsuka: Membership on an entity's Board of Directors or advisory committees; Paladin: Membership on an entity's Board of Directors or advisory committees; Geron: Research Funding; MedImmune: Research Funding; Jazz: Research Funding; TaiHo: Membership on an entity's Board of Directors or advisory committees; Takeda: Membership on an entity's Board of Directors or advisory committees; AbbVie: Honoraria; Astex: Membership on an entity's Board of Directors or advisory committees, Research Funding. DiNardo: Agios/Servier: Consultancy, Honoraria, Research Funding; Notable Labs: Current holder of stock options in a privately-held company, Membership on an entity's Board of Directors or advisory committees; Takeda: Honoraria; AbbVie: Consultancy, Research Funding; Novartis: Honoraria; Foghorn: Honoraria, Research Funding; Forma: Honoraria, Research Funding; GlaxoSmithKline: Membership on an entity's Board of Directors or advisory committees; Bristol Myers Squibb: Honoraria, Research Funding; ImmuneOnc: Honoraria, Research Funding; Celgene, a Bristol Myers Squibb company: Honoraria, Research Funding. OffLabel Disclosure: Enasidenib is approved for the treatment of relapsed or refractory AML as single agent. Venetoclax, in combination with azacitidine or low-dose cytarabine, is approved for the treatment of newly diagnosed AML patients who are 75 years or older, or who have comorbidities that preclude use of intensive induction chemotherapy.

Mutations in isocitrate dehydrogenase 1 (IDH1) promote leukemic transformation through the production of an oncometabolite, 2-hydroxyglutarate (2-HG). Ivosidenib is an inhibitor of mutant IDH1 approved for the treatment of IDH1-mutated AML. Treatment with ivosidenib can induce terminal differentiation of leukemic blasts via suppression of 2-HG. However, ivosidenib as a single agent has limited efficacy, highlighting the need to better understand the mechanisms of drug resistance. The study of drug resistance mechanisms has been hindered by the lack of IDH1/2-mutated AML cell lines models. To address this issue, we derived an Idh1-mutated AML cell line from bone marrow cells of a murine AML model generated by crossing mice expressing mutant Idh1R132H with mice expressing mutant Npm1 (Npm1c). This cell line (henceforth termed "OCI-mIDH1/N") undergoes partial myeloid differentiation in response to ivosidenib treatment in vitro. The establishment of OCI-mIDH1/N enabled us to perform a genome-wide CRISPR knockout screen to identify genes that upon inactivation, increased the differentiation response to ivosidenib. Through this screen, we identified C-type lectin member 5a (Clec5a) as one of the top hits. Clec5a encodes a cell surface receptor that signals through the intracellular spleen tyrosine kinase (SYK). To confirm this hit, we generated Clec5a knockout clones of OCI-mIDH1/N cells. Consistent with results of the screen, ivosidenib treatment induced higher levels of Gr-1, a myeloid differentiation marker, on Clec5a-/- cells than on parental Clec5a+/+ cells. Next, we investigated the role in SYK. Clec5a-/- cells had lower levels of pSYK compared with Clec5a+/+ cells, and overexpression of Syk in Clec5a-/- cells reversed their sensitization to ivosidenib. Furthermore, direct inhibition of SYK with fostamatinib was sufficient to sensitize Clec5a+/+ cells to ivosidenib. Our findings show that CLEC5A-SYK signaling promotes resistance to ivosidenib-induced differentiation. Mechanistically, we found that CLEC5A-SYK signaling drives ivosidenib resistance through STAT5 dependent expression of self-renewal genes in the HOX family. Clec5a-/- and Syk-/- OCI-mIDH1/N cells exhibited lower levels of STAT5 activation compared with wildtype cells. Furthermore, SYK inhibitor treatment downregulated pSTAT5 and decreased STAT5 occupancy at HOX gene clusters. Functionally, overexpression of constitutively active STAT5 in Clec5a-/- and Syk-/- cells reversed their heightened sensitivity to ivosidenib and increased the expression of key HOX self-renewal genes. To determine the clinical relevance of these findings, we analyzed two independent RNA-seq datasets from patients treated with IDH inhibitors (Quek et al., 2018; Wang et al., 2021). Gene set enrichment analysis revealed that poor responders expressed significantly higher levels of STAT5 target genes compared with responders in both datasets. Mutations in the receptor tyrosine kinase (RTK) pathway have previously been shown to be associated with resistance to IDH inhibitors. Given that these pathways signal through STAT5, we hypothesized that the mechanism by the RTK mutations confer ivosidenib resistance is through STAT5 activation. To test this hypothesis, we ectopically expressed KRASG12D or PTPN11E76G in OCI-mIDH1/N cells. Expression of the oncogenes increased pSTAT5 and conferred resistance to ivosidenib-induced differentiation. Importantly, knockdown of Stat5 completely restored their sensitivity to ivosidenib, in support for our hypothesis. Next, we tested if the STAT5 inhibitor pimozide synergizes with ivosidenib to treat patient-derived xenograft (PDX) models of human IDH1-mutated AML. In two PDX models where ivosidenib alone induced moderate differentiation, the addition of pimozide to ivosidenib significantly increased the expression of myeloid differentiation markers on AML cells. In two PDX models where ivosidenib alone induced no differentiation, pimozide alone was sufficient to induce profound differentiation of AML cells. Collectively, these results suggest that ivosidenib resistant cells shift their dependence from 2-HG to STAT5 to maintain their undifferentiated state. In summary, our findings demonstrate that STAT5 is a critical mediator of resistance to ivosidenib, and combination therapy with pimozide and ivosidenib is a promising therapeutic approach for IDH1-mutated AML. Disclosures Chan: BMS: Research Funding; AbbVie: Research Funding.

Clonal Hematopoiesis of Indeterminate Potential (CHIP) refers to the clonal expansion of hematopoietic stem and progenitor cells (HSPCs) carrying mutations associated with hematologic malignancies in individuals without evidence of a blood disorder. CHIP carriers are at a 10-fold higher risk of developing blood cancers relative to non-carriers. The time interval from acquisition of CHIP to overt neoplasia can span many years. This finding suggests a potential window of opportunity to intervene to prevent or delay malignant transformation; however, this goal is not yet possible due to the lack of known effective interventions. One of the most commonly mutated genes in CHIP is TET2, which encodes an epigenetic modifier involved in DNA demethylation. Tet2 knockout mouse models have an expansion of HSPCs and are at increased risk of development of myeloid malignancies, thus establishing TET2 mutation as a bona fide driver of pre-leukemia. To identify drugs that selectively target Tet2-deficient HSPCs, we generated isogenic murine cell lines by overexpressing HOXB4 in Sca-1 + HSPCs derived from a Tet2 knockout (Tet2-/-) mouse and a Tet2 wildtype (Tet2 +/+) littermate. HOXB4 overexpression has previously been shown to expand and immortalize HSPCs indefinitely without causing leukemogenesis. Using this isogenic cell line system termed HPC HOXB4, we conducted a drug screen of 37 small-molecule inhibitors of epigenetic signaling to find compounds that selectively reduced the proportion of Tet2-/- to Tet2+/+ HPC HOXB4 cells in competition assays. Through this screen, we discovered that DOT1L inhibitors, including SGC0946 and pinometostat, preferentially reduced the competitive advantage of Tet2 -/- HPC HOXB4 cells. Importantly, SGC0946 treatment also reduced the competitive advantage of unmodified Tet2 -/- murine HSPCs over wildtype cells in short-term competition assays. Mechanistic studies revealed that DOT1L inhibition induced higher levels of apoptosis in Tet2 -/- than in Tet2+/+ HPC HOXB4 cells and promoted the differentiation of Tet2-/- HPC HOXB4 cells. Genetic knockdown of Dot1l expression using shRNAs phenocopied the effects of pharmacologic DOT1L inhibition. DOT1L is a histone methyltransferase that catalyzes histone H3K79 methylation. Unexpectedly, we found that global H3K79me2 and H3K79me3 levels were higher in Tet2 -/- than in Tet2+/+ HPC HOXB4 cells. SGC0946 treatment effectively removed the H3K79me2/3 mark in Tet2 -/- cells. Next, we conducted RNA-seq analysis to gain insights into the transcriptomic changes in Tet2 -/- HPC HOXB4 after DOT1L inhibition. We found that expression of Mpl, which encodes the thrombopoietin receptor (TPOR), was over 20-fold higher in Tet2 -/- HPC HOXB4 than in wildtype cells at baseline and was strongly decreased after SGC0946 treatment. In contrast, SGC0946 had no effect on Mpl expression in Tet2 +/+ HPC HOXB4 cells. Mpl expression was also elevated in unmodified Tet2 -/- HSPCs compared with Tet2+/+ cells. Given that TPO signaling promotes HSC self-renewal and proliferation, we hypothesized that the effects of Dot1L inhibition are mediated through suppression of Mpl. Consistent with our hypothesis, enforced expression of Mpl in Tet2 -/- HPC HOXB4 cells completely rescued the effects of DOT1L inhibition, and downregulation of Mpl expression using shRNAs phenocopied the effects of DOT1L inhibition on Tet2 -/- HPC HOXB4 cells. Based on the findings above, we hypothesized that inhibition of TPOR signaling would selectively target Tet2 -/- over Tet2+/+ HSPCs. Given that Janus Kinase 2 (JAK2) is required to transduce signals downstream of the TPOR, we studied the effects of two potent JAK2 inhibitors, fedratinib and AZD1480, along with a JAK1/2 inhibitor, ruxolitinib, on Tet2 -/- and Tet2+/+ HPC HOXB4 cells in competition assays. In line with our hypothesis, all three compounds reduced the competitive advantage of Tet2 -/- HPC HOXB4 cells. Ruxolitinib also reduced the competitive advantage of unmodified Tet2 -/- HPSCs cells over wildtype cells in a dose-dependent manner. In summary, our data demonstrate that 1) the TPOR signaling pathway is upregulated in TET2-mutated HSPCs through epigenetic dysregulation by DOT1L, and 2) inhibition of DOT1L or the MPL/JAK2 signaling axis selectively targets TET2-mutated HSPCs. Our findings have important implications in the development of pharmacologic interventions against TET2-mutation driven CHIP. Disclosures Chan: AbbVie: Research Funding; BMS: Research Funding.