ALBERT

Description: Background: AZA monotherapy has demonstrated improvement in OS in HR MDS, clinically meaningful and durable responses continue to be limited to a subset of patients (Silverman 2006). One obvious strategy is to develop doublet therapy with drugs that are effective monotherapy in HR MDS and can be administered effectively and safely in combination with AZA. Several agents have been combined with AZA in first line therapy for pts with HR-MDS achieving ORR and CR/PR rates that are comparable to AZA monotherapy (ORR 38% and CR/PR of 24%) (Ades ASH 2018; Sekeres JCO 2017). RAS and other signaling molecules in the Ras pathway are frequently mutated in HR MDS and are proposed to drive leukemic transformation (Takahashi 2013). Given that rigosertib interferes with the RAS-binding domain of RAF kinases and inhibits the RAS-RAF-MEK & the PI3Ks pathways (Athuluri-Divakar, Cell 2016), it is an attractive candidate for combination with AZA. Furthermore, in vitro combo of rigosertib with AZA synergistically inhibits growth and induces apoptosis of leukemic cells in a sequence-dependent fashion (Skiddan AACR 2006). We report updated results from a Phase II study in a subset of patients receiving oral rigosertib in combination with standard dose AZA as first line therapy for HR MDS. Methods: In the Phase II study 09-08, a total of 39 treatment-naïve HR MDS/RAEB-t patients received oral rigosertib in combination with standard dose AZA. Rigosertib was given at 840mg/day (560 mg in the morning & 280mg in the afternoon); or 1120mg/day (560 mg in the morning & 560mg in the afternoon or 840 mg in the morning and 280 mg in the afternoon) (Maniar ASH 2018). Responses were determined by 2006 IWG criteria including transfusion independence (56 days without PBC or PLT transfusions). Oral rigosertib was administered on D1-D21 of a 28D cycle. Parenteral AZA 75mg/m2/day was administered for 7days from D8. Patients were evaluable for response if they received 3 cycles of therapy. Results: Median age of the pts was 64 years (42-90). IPSS-R score at study entry was 9 were Intermediate, 8 were High and 17 were Very High Risk (VHR). In total 20 pts were transfusion-dependent at study entry. CR/PR responses were observed across all IPSS-R cytogenetic prognostic subgroups: Very Poor 60%, Poor 25%, Intermediate 37.5% & Good 25%. CR/PR responses were also observed across all higher IPSS-R prognostic risk categories: Very High 42%, High 17% and Intermediate 25%. A summary of clinical benefit/risk is provided in Table 1 below. Median duration of response was 12.2 mos (0.1-24.2+) and median duration of treatment was 8 mos (1.3-27.3). Time to first and best responses was 1/4 mos. Four pts continue to respond to therapy. 10 AEs led to D/C: urinary tract pain (2), and 1 each for urinary retention, hematuria, hydronephrosis, osteolysis, cerebral haemorrhage, WBC count decreased, neutrophil count decreased, & abd pain. The most frequent AE in table 1, are similar to AEs reported for both rigosertib and AZA as monotherapies, & the GU toxicities were mitigated using specific management guidelines. The majority of MDS pts who experienced AEs ≥Grade 2 were successfully managed and continued to receive the doublet on study. Conclusion: The efficacy (90% ORR & 34% CR) and safety of oral rigosertib and AZA in combination is favorable as first line therapy in pts with HMA naïve HR MDS and are comparable to historical results for standard dose AZA monotherapy (ORR 38% and

Description: 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.

Description: Background: Patients with HR-MDS have a dismal prognosis after failure of hypomethylating agents (HMAs) (Zeidan 2014), with median overall survival (OS) of less than 6 months (Prebet 2011) and currently no approved second-line therapy (Garcia-Manero 2016). Targeted therapies with novel mechanisms of action, combination strategies, as well as innovative study designs, are all needed to expedite and address the high unmet medical need in patients with HMA refractory HR-MDS. Rigosertib is a unique targeted therapy that inhibits PI3K and PLK signaling pathways by binding directly to the Ras-binding domain (Athuluri-Divakar 2016) and in vitro cytotoxicity studies have demonstrated synergy with azacitidine (Cosenza 2015). INSPIRE study is an example of a study with a novel compound and unique mechanism of action as well as an innovative study design. Methods: INSPIRE (NCT02562443) is a global randomized Phase 3 trial in patients with HR-MDS after HMA failure. Patients are randomized in a 2:1 fashion to rigosertib or physician's choice of treatment. Key inclusion criteria: age