HDAC Inhibitor AR-42 Decreases CD44 Expression and Sensitizes Myeloma Cells to Lenalidomide

Introduction: The first FDA-approved deacetylase inhibitor (HDACi), suberoylanilide hydroxamic acid (SAHA, Vorinostat), was shown to be effective in vitro by a number of anti-neoplastic mechanisms. Despite minimal single-agent activity in multiple myeloma (MM), phase 1b studies combining HDACi’s with bortezomib salvaged some relapsed patients and prolonged progression free survival (PFS) from days (Vorinostat) to months (Panobinostat), albeit at the cost of significant side effects including fatigue, nausea, and vomiting. Phase 1/2 studies in combination with lenalidomide have demonstrated tolerability and activity in lenalidomide-refractory patients, but randomized trials are lacking. In MM, the anti-neoplastic mechanism of action for HDACi’s is unknown, but at biologically achievable concentrations, it has been theorized that they sensitize MM cells to other drugs by interfering with cell adhesion mediated drug resistance (CAM-DR), primarily involving the integrins CD44, CD49d (VLA-4), CD54 (ICAM-1), and/or CD184 (CXCR4). AR-42 (ARNO Therapeutics) is a novel orally bioavailable phenylbutyrate-based class I/II HDAC inhibitor that has greater anti-proliferative effects compared to Vorinostat in vitro and in vivo. It has been previously shown that in MM cell lines, AR-42 down-regulates the expression of gp130, and inhibits IL-6 induced activation of STAT3 and downstream targets including BCL-XL and Cyclin-D1, with minimal effects on the PI3K/AKT and MAPK pathways. CD44 is a type I transmembrane glycoprotein, which is directly transcribed by β-catenin, and its role in cell adhesion-mediated drug resistance (CAM-DR) for MM as well as other cancers has been largely investigated. Recent published data have shown that CD44 forms a complex with STAT3 and p300 (acetyltransferase) causing STAT3 activation in a cytokine- and growth factor-independent manner, and that CD44 over-expression is one of the main molecular mechanisms that contributes to Lenalidomide resistance in MM cells.

Hypothesis: We hypothesize that CD44 down-regulation, both surface expression on MM cells as well as the soluble form in the blood, is the primary effect of AR-42 at concentrations achievable in humans, explaining its weak single agent effect and its improvement when combined with other therapeutic agents in the clinic.

Methods: As part of a single center, dose-escalating, first-in-man phase 1 trial of single agent oral AR-42 administered orally three times weekly in 28-day cycles (3 weeks of treatment followed by a 7-day off treatment period), patients were accrued at 20-70 mg TIW in a standard 3+3 cohort design. Using peripheral blood obtained during cycle 1 of therapy, nCounter® GX Human Immunology assays and nCounter miRNA expression profile was performed to assess differentially expressed genes after AR-42 treatment. Enzyme-linked immunoabsorbent assay (ELISA), qRT-PCR and luciferase assays were also performed.To examine whether AR-42 treatment could sensitize the cells to Lenalidomide in vivo, we used GFP+/Luc+ MM.1S cells engrafted in NOD-SCID mice.

Results: AR-42 in relapsed MM showed no confirmed partial responses, but did result in marginal responses in 3 out of 13 MM patients. We found that AR-42 in MM cells modulates the expression of many genes coding for surface receptors including CD44, CD48, CD46 and TRAF5 and affects the expression of several miRNAs. Our data show a decrease of CD44 mRNA expression in the CD138+ MM plasma cells and of the soluble CD-44 in the serum of AR-42 treated patients. We also show that in MM cells CD44 down-regulation upon AR-42 treatment is associated with impairment of STAT-3 signaling pathways and direct targeting of its regulatory RNA binding protein, Insulin grow factor 3 binding protein 3 (IGF2BP3), by miR-9-5p. We found that miR-9-5p is up-regulated in vitro and in the cancer cells of MM patients after AR-42 treatment. Moreover we show that AR-42 in combination with Lenalidomide show synergistic apoptotic effect on MM cells, enhancing Lenalidomide anti-myeloma activity invivo.

Conclusions: These findings show that CD44 is a therapeutic target for the HDACi AR-42 in MM patients, providing the rationale to support further clinical investigation of AR-42 in combination with IMiDs in patient cohorts with high pretreatment CD44 expression in the serum and on the surface of MM cells.