Lenalidomide In Combination with the Activin Receptor Type II Murine Fc Protein RAP-011: Preclinical Rationale for a Novel Anti-Myeloma Strategy

Abstract 4075

The introduction of novel treatment strategies targeting tumor cells within their microenvironment have resulted in prolonged survival for Multiple Myeloma (MM) patients. Lenalidomide belongs to this category of agents working via tumoricidal, anti-osteoclast and immunomodulatory activities. However, lenalidomide lacks bone anabolic effects. We have recently reported that activin A mediates osteoblast inhibition in MM and neutralizing activin A via a soluble receptor, RAP-011 (murinized form of ACE-011) (Acceleron Pharma, Cambridge, MA), restores bone architecture and reduces tumor burden in vivo. We therefore hypothesized that the combination with RAP-011 may potentiate lenalidomide effects and vice versa as they act via complementary mechanisms.

Our previous data demonstrates that 50% of MM patients have increased bone marrow (BM) levels of activin A that correlate with osteolytic burden. Here, we observed that 2 and 10 μ M lenalidomide (Selleck Chemicals, Houston, TX) upregulated activin A in 3 out 6 bone marrow stromal cell (BMSC) samples by 1.9 and 2.8 fold (average ± st.dev. 1052 ± 190 and 1667 ± 732 pg/ml respectively, compared to 734 ± 553 in control, p=0.2). There was no time-dependent upregulation of activin A. Of note, no augmentation of activin A was noted in BMSC which already expressed high baseline levels of the cytokine (average ± st.dev 3638 ± 3755 pg/ml in control vs 3074 ± 2997 after lenalidomide 10 μ M).

Previous data suggest that high concentrations of activin A induce growth arrest and apoptosis in myeloma and breast cancer cells and may therefore mediate lenalidomide cytotoxicity. To ensure that inhibition of activin would not antagonize lenalidomide anti-tumor effects, we investigated whether activin A inhibition affected the cytotoxic and anti-proliferative effects of lenalidomide on MM cells alone and in co-culture with BMSC. As previously demonstrated, RAP-011 did not exert any direct anti-tumor effects. Lenalidomide 10 μ M induced between 20 and 40% of apoptosis in several myeloma cell lines, such as MM1.S, LR5, DOX40 and RPMI, independent of activin A inhibition. Similarly, lenalidomide almost completely reversed the proliferative advantage conferred by BMSC to tumor cells. Combining lenalidomide and RAP-011 was not antagonistic to the inhibition of the proliferative advantage conferred by BMSC to myeloma cells, suggesting that lenalidomide's direct anti-tumor activity is not mediated through activin A.

Finally we assessed OB differentiation in the presence of both RAP-011 and lenalidomide. We have previously reported that activin A inhibitory effects on OB differentiation are reversed by RAP-011 treatment. Here, we noted diminished alkaline phosphatase (ALP, a marker of osteoblast activity) expression during osteoblastogenesis in the presence of increasing concentrations of lenalidomide (17% ± 3 decrease by 2 μ M and 26% ± 11 by 10 μ M compared to control, p=0.01 and 0.06 respectively). In contrast, combination with RAP-011 restored the osteogenic potential by increasing ALP expression close to control levels in healthy donor-derived BMSC and above control levels in MM-derived BMSC.

These preliminary data suggest that lenalidomide results in upregulation of activin A expression in MM BMSCs which have low baseline levels. Combining lenalidomide with RAP-011 results in restored osteogenesis presumably by inhibiting activin signaling. Importantly, we observed no antagonistic effect of RAP-011 on lenalidomide's anti-tumor activity, confirming that activin A does not mediate the anti-tumor activity of lenalidomide. Ongoing in vivo studies using a murine model of myeloma will confirm the efficacy of this promising combination. Our results provide a preclinical rationale for combining lenalidomide with ACE-011 to target myeloma by manipulating the microenvironmental compartment, specifically the bone compartment.