Sequential Treatment with Rigosertib Followed By Azacitidine Maximizes the Effects on the Interferon Signaling Pathway in Hematopoietic Cells in Myelodysplastic Syndrome (MDS)

Background: MDS is a clonal stem cell disorder characterized by abnormal maturation and differentiation of hematopoietic cells. Azacitidine (AZA), a hypomethylating agent (HMA), was approved by the FDA for patients (pts) with MDS and is the standard of care as 1^st line therapy for higher- risk disease. About 50% of MDS pts respond to AZA, with a median duration of response of 14 - 24 month. For those pts responding to an HMA, most either relapse or progress with worsening bone marrow failure (BMF) and have a median survival of 4 to 6 months after treatment failure. Both primary and secondary resistance remains significant clinical problems and result in poor survival. A recent study reveals that AZA activates several immunomodulatory pathways leading to activation of the antiviral defense pathway and upregulation of interferon signaling (IFN) (Chiappinelli et al, 2017). INF has multiple effects on hematopoiesis and appears highly regulated. Interferon can stimulate hematopoietic stem cells (HSC) which may be countered through activation of p38 MAPK; inhibition of the latter may improve hematopoiesis. Short-term effects of IFN stimulate myeloid and erythroid hematopoiesis.

Rigosertib (RIG) is a Ras-mimetic that blocks the activation of Ras effector proteins. It has been identified as a novel anti-cancer drug which inhibits cell cycle progression and induces apoptosis of cancer cells (Athuluri-Divakar et al, 2016). It activates apoptosis related pathways by ameliorating several signaling pathways like Akt, p38, MAPK/ JNK, STAT3, β-catenin, GSK3α/β and PLK1 that can be dysregulated in MDS pts (Xu et al, 2014). In vitro, the combination of RIG with AZA (Skidan et al, AACR 2006) showed synergistic effect in inducing cell death in a sequence dependent manner requiring RIG priming. We have reported that it acts as a histone deacetylase inhibitor with chromatin modifying activity (Chaurasia et al, ASCO 2016). In a Phase I/II study the combination of RIG and AZA (NCT 01926587) produced an ORR of 85% in pts who were HMA naïve and importantly 62% in HMA failures (Navada et al, EHA 2017). The ability to reverse the clinical resistance phenotype is a novel observation with important clinical implications and understanding the mechanisms is critical to develop ways of reversing resistance.

Methods: We investigated the in vitro effects of RIG combined with AZA on two cell lines one resistant and one sensitive to AZA: BW90, an AZA resistant line and MDS-L, AZA sensitive and in specimens from 2 pts treated on a clinical trial of RIG and AZA in combination, to determine effects on hematopoietic cells and IN signaling.

Results: QT-PCR studies demonstrated that individual treatment of MDS-L and BW90 with RIG or combined with AZA in sequential treatment (AZA/RIG or RIG/AZA) altered chromatin remodeler (KDM2a, SET1, JMJD3 and LRWD1) transcript levels in a cell line specific context. Exposure of cells from a bone marrow from a patient prior to treatment with RIG alone or RIG/AZA failed to induce expansion of CD34⁺ cells and yielded maximum aldehyde dehydrogenase (ALDH) activity, a marker of primitive hematopoietic stem and progenitor cells (HSCs) (ANOVA, p=0.006). AZA alone yielded a 3.8 fold expansion of CD34⁺, with marked decrease in ALDH activity that was inversely proportional to the expansion of CD34⁺ cells. Expansion of CD34⁺ cells led to ≥2 fold increase in pluripotent genes (SOX2, OCT4, NANOG and ZIC3) expression levels. QT PCR studies of the effects of AZA or RIG alone or in combination on INF yielded differential effects on INFa and INFb expression as follows compared to control reported as percent fold increase for INFa - AZA 2x; RIG 0.75x, AZA/RIG 1.5x; RIG/AZA 2.5x; INFb AZA 1.3x; RIG 0.6x, AZA/RIG 1.2x; RIG/AZA 0.7x. There was a significant increase in INFa (p=<0.0001) but not INFb, with RIG priming compared to AZA alone or AZA/RIG.

Conclusion: Treatments with RIG either alone or combined with AZA results in epigenetic reprogramming of pluripotency genes, and expansion of primitive HSPC. This may serve to regulate HSPC pluripotency and expression of a maturation program. A significant increase in expression of IFNa seen after exposure to RIG/AZA may lead to enhanced hematopoietic function and may explain the differences in the alternative sequence of AZA/RIG. Further studies are underway to define more specific effects on these pathways to explain the impact on the resistance phenotype.