Takahashi, Koichi

Abstract The dismal outcome of acute leukemia undergoing immunophenotype-switch, compounded by the increased incidence in the era of immunotherapy, highlights the need for in-depth systemic studies. We investigated the clinicopathological, cytogenetic, and molecular features of 33 patients experiencing switch between acute lymphoblastic leukemia (ALL) and acute myeloid leukemia (AML). The median duration from the primary diagnosis to lineage switch was 7.8 months, with the subsequent median survival being < 3 months. The majority experienced a transition from B-ALL to AML. The AML cases demonstrated a spectrum of differentiation, ranging from minimal to erythroid, with monocytic being the most common. Cytogenetic analysis revealed 11q23/KMT2A fusions as the most recurrent abnormalities (n = 18); also observed were complex karyotype, 7/7q-, 9p-/CDKN2A deletion, t(5;14)/TCLX3::BCL11b, t(8;9)/PCM1::JAK2, t(9;22)/BCR::ABL1, and t(12;19)/TCF3::ZNF384. Remarkably, in 60% of patients lacking KMT2A fusions, a chronic myeloid neoplasm was observed preceding the first leukemia. In contrast, no prior hematological malignancies were observed in KMT2A cases. Deep mutational profiling reinforced the inherent clonal relationship between the initial and subsequent leukemias, evident from the overlapping mutational signatures. The presence of distinct additional mutations across the two phases supports the long-standing theory of lineage switch emerging via divergent evolution of a multipotential progenitor.

Abstract Background: Hypomethylating agents (HMA) are well established standard of care for patients (pts) with higher-risk MDS (HR-MDS). However, approximately half of the pts do not respond to HMA therapy and most of the responders eventually lose response (HMA failure). There is no standard of care for pts after HMA failure and median overall survival (OS) post HMA failure is around 6 months (Jabbour et al. 2015). While the mutational landscapes and their role in prognosis are increasingly becoming apparent in pts with HR-MDS at diagnosis, mutational profiles at the time of HMA failure and their impact on clinical outcomes are not well understood. Here, using samples collected from a global Phase 3 trial randomizing HR-MDS pts post HMA failure to I.V. rigosertib (RGS) or physician's choice (PC) (INSPIRE: NCT02562443), we analyzed the landscape of driver mutations in HR-MDS after HMA failure and investigated the association with the clinical outcomes. Since RGS is a non-ATP-competitive small molecule RAS mimetic (Athuluri-Divakar 2016), the study also offered an opportunity to test the hypothesis whether HR-MDS pts with oncogenic RAS pathway mutations benefit from RGS. Methods: HR-MDS pts after HMA failure were randomized 2:1 to RGS or PC. All pts failed to respond to or progressed on prior HMA therapy. Bone marrow samples or peripheral blood samples were collected at the time of trial screening. Genomic DNA was sequenced by the targeted capture deep sequencing of 295 genes (median 500x). Results: 372 pts were enrolled in INSPIRE trial (248 to RGS and 124 to PC). The median age of the trial participants was 73 (range: 40-85). All pts were previously treated and with an HMA with the median duration of prior HMA therapy of 6.7 months. 64% and 28% of the pts were classified as IPSS-R very high risk or high risk, respectively, at the time of randomization. Among the 372 participants, DNA sequencing of pre-treatment samples was performed in 188 pts (51% of the participants, N = 122 in RGS arm, N = 66 in PC arm). The most frequently identified driver mutations involved ASXL1 (36%) followed by RUNX1 (24%), TET2 (23%), STAG2 (22%) and TP53 (21%). Mutations in splicing pathway genes were found in 36% of the pts. Oncogenic RAS pathway mutations were detected in 15% of the pts (NRAS = 3 %, KRAS = 2%, CBL = 4%, NF1 = 5%, PTPN11 = 3%, and 1% had multi-hit mutations). Compared to the previously untreated MDS pts (N = 446, Papaemmanuil et al. Blood 2013), mutations in ASXL1 (36% vs. 18%, P &lt; 0.0001), BCOR (9% vs. 3%, P = 0.002), CEBPA (4% vs. 0.2%, P = 0.001), NF1 (5% vs. 0.9%, P = 0.003), RUNX1 (25% vs. 11%, P &lt; 0.0001), STAG2 (22% vs. 5%, P &lt; 0.0001), TP53 (21% vs. 6%, P &lt; 0.0001), and IDH1/2 (13% vs. 7%, P =0.016) were significantly more enriched in pts with HMA failure, whereas mutations in SF3B1 (7% vs. 37%, P &lt; 0.0001), TET2 (23% vs. 35%, P = 0.006), and splicing pathway genes (36% vs. 68%, P &lt; 0.0001) were significantly less frequent in HMA failure pts. These results are consistent with the high-risk profiles of HMA failure pts. Frequency of oncogenic RAS pathway mutations were similar between HMA failure and previously untreated MDS pts (15% vs. 13%, P = 0.612). Correlation analysis between the types of HMA failure and gene mutations showed that TP53 mutations were significantly enriched in pts who relapsed after initial response to HMA (P = 0.001), whereas oncogenic RAS pathway mutations were significantly enriched in pts who progressed during the HMA therapy (P = 0.03). Overall, RGS did not significantly improve the overall survival (OS) of HMA failure pts compared to PC. Survival difference between RGS arm and PC arm was not observed in any subgroups stratified by the gene mutations. The only subgroup that showed improved OS with RGS compared to PC was pts with RAEB-t (median OS 7.5 vs. 3.9 months, P = 0.049). Of note, among pts with oncogenic RAS pathway mutations, no survival difference was observed between RGS and PC arms. Conclusion: High-risk gene mutations, such as TP53, ASXL1, RUNX1, and STAG2 (Ogawa. Blood 2019). were significantly enriched in MDS pts with HMA failure, suggesting their role in HMA resistance and disease progression. Identifying the mutations present de novo and with HMA failure offers the opportunity to determine prognosis based on these mutations as well as potential strategies to target these mutations with new medical entities. Disclosures Takahashi: GSK: Consultancy; Celgene/BMS: Consultancy; Novartis: Consultancy; Symbio Pharmaceuticals: Consultancy, Membership on an entity's Board of Directors or advisory committees. Luger: Syros: Honoraria; Agios: Honoraria; Daiichi Sankyo: Honoraria; Jazz Pharmaceuticals: Honoraria; Brystol Myers Squibb: Honoraria; Acceleron: Honoraria; Astellas: Honoraria; Pfizer: Honoraria; Onconova: Research Funding; Celgene: Research Funding; Biosight: Research Funding; Hoffman LaRoche: Research Funding; Kura: Research Funding. Al-Kali: Novartis: Research Funding; Astex: Other: Research support to institution. Diez-Campelo: BMS: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Takeda Oncology: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau; Novartis: Consultancy, Honoraria, Membership on an entity's Board of Directors or advisory committees, Research Funding, Speakers Bureau. OffLabel Disclosure: Rigosertib for MDS patients after HMA failure

Background: More than 45 mutations have been identified in association with HR-MDS. In the majority of patients with MDS (80%) co-mutations are present and the prognostic contribution of each individual mutation remains elusive, especially after adjusting for clinical variables such as IPSS-R score. N-RAS and K-RAS mutations as well as regulators of the Ras pathway (e.g. PTPN11, NF1) are frequently observed (15-20%) in HR-MDS, however their clinical impact is unclear, especially in de novo MDS. Rigosertib (RGS) is a non-ATP-competitive small molecule RAS mimetic that has the potential to block RAS-RAF-MEK-ERK and PI3K-AKT-mTOR signaling pathways (Athuluri-Divakar 2016). Rigosertib has the potential to also inhibit wildtype upregulation of RAS. We report here genomic profiling at the time of study entry in the ongoing phase 3 randomized global study (known as INSPIRE) in patients with HR-MDS after failure of HMA therapy. Methods: INSPIRE (NCT02562443) is a global randomized Ph3 trial in pts with HR-MDS after HMA failure with an overall target enrollment of 360 pts with currently 298 pts randomized. Pts are randomized 2:1 to rigosertib or physician's choice of treatment. The primary endpoint is overall survival (OS). All pts failed to respond or progressed on HMA therapy. Key inclusion criteria includes: age < 82 years, RAEB-1, RAEB-2 or RAEB-t; intermediate risk (IR), high risk (HR) and very high risk (VHR) per IPSS-R; ≥ 1 cytopenia; duration of prior HMA ≤ 9 cycles within 12 months; last dose of HMA ≤ 6 months before enrollment; baseline blast counts between 5-29% and one of the following: progression any time after initiation of HMA treatment, intolerance to HMA, failure to achieve complete remission (CR), partial remission (PR), or hematologic improvement (HI) after six 4-week cycles of AZA or either four 4-week or four 6-week cycles of DAC, or relapse after initial CR, PR or HI. Bone marrow samples were collected at study baseline and throughout the study for mutational analysis as an exploratory endpoint. Baseline blast counts are described as % reported in bone marrow aspirate at screening. Genomic DNA was extracted from diagnostic bone marrow or peripheral blood samples and targeted capture deep sequencing of 295 genes were performed (median sequencing depth 500x) using Agilent's SureSelect custom panel. Modified Mutect and Pindel were used to identify high-confidence somatic mutations. Results: All data is presented as blinded aggregate results for both arms of the study. Baseline mutations are presented for 114 pts; 92 pts were randomized and 22 pts were screen failures. Median age is 72 years (59-81). The IPSS-R scores for the pts randomized were: Intermediate 5(5%), High 30 (32%) and VHR 56 (60%). In total 48 different mutations were identified at baseline prior to pts receiving study treatment. The most common mutations identified in pts were ASXL1 40%, STAG2 26%, TP53 24%, RUNX1 22%, SRSF2 and TET2 each 18%. Average number of mutations per pt was 3.06. At baseline, 4 pts (4%) had no mutations, 17 pts (18%) had only 1 mutation while 17 pts (18%) had between 6-8 mutations. N-RAS and K-RAS mutations occurred in 4 pts each (7% of pts) and all in the presence of other mutations. Mutations involving regulators of the Ras pathway (N-RAS, K-RAS, PTPN11, NF1) occurred in 12 (13%) of patients. Of note, 23 pts (25%) had IDH1 (6 pts) or IDH2 (17 pts) and the majority of these pts (68%) were High/VHR with an average blast count of 15% (range 5-27%) at study entry. The high proportion of IDH1/2 mutations observed is most likely due to inclusion of patients with RAEB-t. These results will be updated at the meeting with blinded baseline mutational analyses for all pts randomized into the study. Conclusion: The baseline mutational analyses from the INSPIRE study provides important initial insights into the genomic profile of pts with HMA failure, especially for the subset with VHR. Following analysis of the primary endpoint, it is anticipated that correlation of overall survival and clinical response with mutational status will be possible, including changes in mutations following therapy. Given the number of mutations involving the Ras pathway the efficacy of rigosertib in patients with this group of mutations will also be examined. Figure Disclosures Garcia-Manero: Amphivena: Consultancy, Research Funding; Helsinn: Research Funding; Novartis: Research Funding; AbbVie: Research Funding; Celgene: Consultancy, Research Funding; Astex: Consultancy, Research Funding; Onconova: Research Funding; H3 Biomedicine: Research Funding; Merck: Research Funding. Luger:Agios: Honoraria; Jazz: Honoraria; Kura: Research Funding; Onconova: Research Funding; Pfizer: Honoraria; Seattle Genetics: Research Funding; Daichi Sankyo: Honoraria; Genetech: Research Funding; Cyslacel: Research Funding; Ariad: Research Funding; Biosight: Research Funding; Celgene: Research Funding. Al-Kali:Astex Pharmaceuticals, Inc.: Research Funding. Jedrzejczak:Amgen: Consultancy, Other: travel support for hematology meetings (ASH, EBMT, EHA) ; Celgene: Other: travel support for hematology meetings (ASH, EBMT, EHA) ; Takeda: Consultancy; Novartis: Research Funding; Roche: Other: travel support for hematology meetings (ASH, EBMT, EHA) . Díez-Campelo:Novartis: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Celgene Corporation: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding. Zbyszewski:Onconova Therapeutics, Inc.: Employment. Cavanaugh:Onconova Therapeutics, Inc.: Employment. Woodman:Onconova Therapeutics, Inc.: Employment. Fruchtman:Onconova Therapeutics, Inc.: Employment. Takahashi:Symbio Pharmaceuticals: Consultancy.

Abstract Background: Blasts and leukemic stem cells of acute myeloid leukemia (AML) as well as several other hematologic malignancies express CD123, potentially providing a target for novel therapies. XmAb14045 (also known as SQZ622) is a potent bispecific antibody targeting both CD123 and CD3 that stimulates targeted T cell-mediated killing of CD123-expressing cells, regardless of T cell antigen specificity. It is a full-length immunoglobulin molecule designed to be dosed intermittently, in contrast to smaller constructs that are referred to as "DART" or "BiTE" bispecific antibodies that require a continuous infusion. Methods: Patients with relapsed or refractory AML, B cell acute lymphoblastic leukemia (B-ALL), blast phase chronic myelogenous leukemia, or blastic plasmacytoid dendritic cell neoplasm were eligible to enroll on this first-in-human, multicenter, open-label phase 1 dose-escalation study (XmAb14045-01) with standard 3+3 design. The primary objective was to estimate the maximum tolerated dose (MTD) or recommended dose and schedule of XmAb14045. Secondary objectives included safety, preliminary antileukemic activity, and pharmacokinetics/pharmacodynamics of XmAb14045. Treatment was administered weekly in 28 day cycles, using a weight-based dose with a single dose level in Part A and, in Part B, an initial priming dose on Day 1 followed by an escalated dose on subsequent weeks. Premedication to prevent cytokine release syndrome (CRS) was instituted as needed and included a steroid, acetaminophen, and diphenhydramine. Patients were premedicated at all XmAb14045 doses ≥0.075 µg/kg. Therapy was continued for as long as tolerated and there was continuing evidence of therapeutic benefit in the opinion of the investigator. Treatment response was assessed by the 2017 European LeukemiaNet (ELN) criteria after Cycle 1 and after each odd-numbered cycle. CRS was graded using the CRS revised grading system (Lee et al., Blood, 2014). Results: At data cut-off, 64 patients have been treated to date, 63 with relapsed/refractory AML and 1 with B-ALL. Patients had a median age of 61 years and were heavily pretreated (median of 3 prior therapies [range 1-8], including 19 [30%] who had undergone prior allogeneic stem cell transplantation). The recommended dose for Part A was 1.3 µg/kg after a single dose-limiting toxicity of Grade 4 CRS at 2.3 µg/kg; no MTD has been identified. CRS or its component symptoms was the most common treatment-emergent adverse event (TEAE). CRS episodes began within approximately 1-4 hours of the start of drug infusion and occurred in 49 of 64 patients (77%). Seven patients (11%) developed Grade ≥3 CRS, the majority of these on the first dose. There were no CRS-related deaths. Excluding CRS-related events, additional TEAEs occurring in >10% of patients included fatigue (31%), febrile neutropenia (30%), peripheral edema (30%), cough (23%), elevated hepatic transaminases (19%; all recovered without sequelae), pneumonia (17%), stomatitis (14%), hyperglycemia (13%), and sepsis (11%). No myelosuppression requiring dose modification or evidence of tumor lysis syndrome was seen. In Part A, single agent antileukemic activity was documented with a best response of CR (2) or CRi (1) in 3/13 AML patients (CR/CRi rate 23%) treated at the two highest dose levels studied to date (1.3 and 2.3 µg/kg weekly); no CR, CRi, or morphologic leukemia-free state (MLFS) responses were seen at lower doses. Antileukemic activity occurred quickly; all responders had achieved at least an MLFS response after 4 doses (1 cycle). Two responders were bridged to stem cell transplantation, and the third was ineligible for medical reasons but remains in remission at 14+ weeks after initiating therapy. Conclusions: XmAb14045 demonstrated evidence of antileukemic activity in heavily pretreated patients with relapsed/refractory AML treated in Part A at the 1.3 and 2.3 µg/kg doses administered once weekly, with a 23% CR/CRi rate. CRS was the most common toxicity but was generally manageable with premedications. The study is ongoing with further optimization of dose, schedule, and premedication regimen for CRS anticipated during accrual to dose escalation cohorts in Part B. ClinicalTrials.gov Identifier: NCT02730312 Disclosures Ravandi: Xencor: Research Funding; Abbvie: Research Funding; Amgen: Honoraria, Research Funding, Speakers Bureau; Sunesis: Honoraria; Astellas Pharmaceuticals: Consultancy, Honoraria; Abbvie: Research Funding; Seattle Genetics: Research Funding; Macrogenix: Honoraria, Research Funding; Jazz: Honoraria; Sunesis: Honoraria; Seattle Genetics: Research Funding; Bristol-Myers Squibb: Research Funding; Astellas Pharmaceuticals: Consultancy, Honoraria; Orsenix: Honoraria; Xencor: Research Funding; Orsenix: Honoraria; Jazz: Honoraria; Bristol-Myers Squibb: Research Funding; Amgen: Honoraria, Research Funding, Speakers Bureau; Macrogenix: Honoraria, Research Funding. Foran:Xencor, Inc.: Research Funding; Agios: Research Funding. Stock:Jazz Pharmaceuticals: Consultancy. Mawad:Swedish Cancer Institute: Employment. Blum:Astellas: Consultancy; Pfizer: Consultancy; Boehringer Ingelheim: Research Funding; Forma: Research Funding; Xencor: Research Funding; Tolero: Research Funding. Saville:Xencor, Inc.: Employment, Equity Ownership. Johnson:Xencor, Inc.: Employment, Equity Ownership. Vanasse:Novartis: Employment, Equity Ownership. Ly:Xencor, Inc.: Employment, Equity Ownership. Mims:Novartis: Consultancy; Agios Pharmaceuticals: Consultancy, Membership on an entity's Board of Directors or advisory committees; Abbvie Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.

Abstract Pracinostat, a potent oral pan-histone deacetylase inhibitor with modest single-agent activity in acute myeloid leukemia (AML), has shown synergistic antitumor activity when combined with azacitidine. This single-group, multicenter phase 2 study assessed the safety and efficacy of pracinostat combined with azacitidine in patients who were at least 65 years old with newly diagnosed AML and who were ineligible for standard induction chemotherapy. Patients received pracinostat 60 mg/d, 3 d/wk, for 3 consecutive weeks, plus azacitidine 75 mg/m2 daily for 7 days in a 28-day cycle. Primary endpoints were complete remission (CR), CR with incomplete count recovery (CRi), and morphologic leukemia-free state (MLFS) rates of the combination. Secondary endpoints included safety, progression-free survival (PFS), and overall survival (OS) of the regimen. Fifty patients (33 de novo, 12 secondary, and 5 therapy-related AML) were enrolled. Twenty-six patients (52%) achieved the primary endpoint of CR (42%), CRi (4%), and MLFS (6%). Median OS and PFS were 19.1 months (95% confidence interval [CI], 10-26.5 months) and 12.6 months (95% CI, 10-17.7 months), respectively, with a 1-year OS rate of 62%. Forty-three patients (86%) experienced at least 1 grade 3 or worse treatment-emergent adverse event with the combination, with infections (52%), thrombocytopenia (46%), and febrile neutropenia (44%) reported as the most common toxicities. The 30- and 60-day all-cause mortality rates were 2% and 10%, respectively. DNA sequencing revealed somatic mutations at baseline, and clearance rates correlated with response to treatment. Pracinostat plus azacitidine is a well-tolerated and active regimen in the frontline treatment of older patients with AML unfit for intensive therapy. A larger controlled trial is ongoing. This trial was registered at www.clinicaltrials.gov as #NCT01912274.

7034 Background: In a phase II study in 50 elderly patients (pts) with AML who were not eligible for intensive chemotherapy, treatment with pracinostat + azacitidine (AZA) was well tolerated, led to 42% complete remission (CR) rate and a median overall survival (OS) of 19.1 months (Blood 2016: 128;100). Here we investigate the impact of somatic mutations and their clearance on disease response and patient outcomes. Methods: 88 samples from 41 study pts were sequenced. All 41 pts were analyzed pre-treatment, and a median of 3 longitudinal samples were analyzed from 19 pts between Cycle 2 and 9. Mutations were assayed by SureSelect targeted capture exon sequencing (Agilent) of 295 genes that are recurrently mutated in hematologic malignancies. Longitudinal mutation clearance was analyzed by plotting variant allele frequency (VAF). Results: At baseline, 96 mutations in 28 genes were detected in 38 pts, with the most frequent being in SRSF2 (27%), DNMT3A (20%), IDH2 (17%), RUNX1 (17%), and TET2(17%). Mutations associated with CR rate and OS are indicated in the Table. A CR was achieved in 10/19 pts that had longitudinal sequencing analysis and at the time of CR, 9 (90%) had persistently detectable mutations in their bone marrow. In 7 of them, continued exposure to pracinostat + AZA lowered the VAF or cleared residual mutations. In 2 pts, relapsed samples showed re-expansion of the founder clone. Conclusions: Mutations in NPM1, and DNA methylation pathway were associated with a better response to pracinostat + AZA, while TP53 mutation was associated with a poor response. Persistent mutation at the time of CR suggests residual preleukemic clonal hematopoiesis in this elderly population. Benefit of prolonged exposure to pracinostat + AZA was also confirmed at molecular level where continued decline of mutation VAF was seen after achieving CR. Clinical trial information: NCT01912274. [Table: see text]