A Phase I/II Study of the Combination of Oral Rigosertib and Azacitidine in Patients with Myelodysplastic Syndrome (MDS) or Acute Myeloid Leukemia (AML)

Background:Rigosertib is a small molecule anti-cancer agent targeting PI3/polo-like kinase pathways that promotes G2/M arrest and has effects on the B-Raf and Ras pathways. It is currently being tested as a single agent with the intravenous (IV) formulation in patients (pts) who have relapsed or are refractory to hypomethylating agents (HMAs) as well as with the oral formulation in lower-risk, red-cell transfusion-dependent MDS patients. Azacitidine (AZA) is first-line therapy for pts with higher-risk MDS. In vitro, the combination of rigosertib with AZA acts synergistically to inhibit growth and induce apoptosis of leukemic cells (Skidan et al 2006). This effect appears to be sequence dependent, requiring exposure to rigosertib first, followed by AZA. These nonclinical results provided the rationale to combine the 2 agents in a phase I/II study in pts with MDS and AML.

Methods: Pts with MDS and non-proliferative AML, who were previously untreated or had failed or progressed on an HMA were included in the phase I component of the study. Oral rigosertib was administered twice daily from day 1 through day 21 of a 28-d cycle. AZA 75 mg/m²/d was administered for 7 days starting on day 8 of the 28-d cycle. Pts were entered in 3 escalating-dose cohorts of rigosertib in a classic 3+3 design: [1] 140 mg twice daily; [2] 280 mg twice daily; [3] 560 mg qAM and 280 mg qPM. A CBC was performed weekly and a bone marrow (BM) aspirate and/or biopsy was performed at baseline and every 4-8 weeks afterwards.

Results: Eighteen pts have been treated with the combination of oral rigosertib and AZA. Pts had diagnoses of intermediate-1 MDS (3), intermediate-2 MDS (6), high-risk MDS (2), CMML (1), and AML (6); median age was 70.5 years; 61% of pts were male. Pts have received 1-10+ cycles of treatment with the total number of cycles administered thus far being 58. Cytogenetic profiles by IPSS were good (8 pts), poor (8 pts), and intermediate (2 pts). 11of 18 patients were transfusion dependent at baseline [RBC (11), platelet (6)]. One patient became RBC transfusion independent after 3 cycles of treatment. 5 additional patients have had a reduction in their RBC and platelet transfusion requirements. 56% of patients received prior treatment with HMAs: AZA (6 pts), decitabine (4 pts). The most frequent adverse events (AEs) in Cycle 1 included constipation, diarrhea, nausea, fatigue, hypotension, and pneumonia. The AEs did not differ significantly among the 3 cohorts. Elevation in creatinine in 1 pt in cohort 1 was a possibly related grade 3 dose-limiting toxicity that required subsequent expansion of the cohort. Drug-related dysuria/cystitis was not reported in this pt population. Responses according to IWG 2006 criteria were observed in the BM and peripheral blood: Complete Response (CR) (1 pt), Cr_i (CR with incomplete blood count recovery) (4 pts), stable disease (2), hematologic improvement-erythroid (1). Six pts received fewer than 4 cycles of treatment and are too early to evaluate. Six pts came off study for the following reasons: progression of disease (1), pt request (1), death from pneumonia (2), received stem cell transplant (1), persistent fungal pneumonia (1). Two evaluable pts have responded to the combination after progression or failure on HMA alone.

Conclusions: The combination oforalrigosertib at 560/280 mg BID (recommended phase II dose) and standard-dose AZA can be safely administered and appears to be well tolerated in repetitive cycles in pts with MDS and non-proliferative AML. The AE profile does not differ significantly from that of AZA alone. Data from the Phase I component of this study suggest activity in patients with MDS after HMA failure. Additional data are required to evaluate this observation. The Phase II segment of this study is underway to further assess the response of the combination.