Abstracts of 2nd Scientific Meeting on Bone Disease in Multiple Myeloma
Myeloma cell homing and reciprocal cultivation of the bone microenvironment in the SCID-hu system
J Epstein: Little Rock
In SCID-hu hosts, purified myeloma cells (>98.7% purity) grow in and interact with a human microenvironment. Myeloma cell growth is associated with neo-angiogenesis, osteoclastogenesis, and frequently with severe resorption of the human bones. Both the vascular endothelial cells and osteoclasts are human, as determined by their reactivity with monoclonal antibodies specific to human CD34 and Vitronectin receptor, respectively. Both cell types arise from the implanted human bones. Therefore, myeloma plasma cells are capable of sustained proliferation as attested to by four sequential transfers from one SCID-hu host to the next, and a non-myelomatous microenvironment can support sustained proliferation of myeloma cells and produce typical myeloma manifestations.
Whereas myeloma cells from patients with bone marrow restricted disease grew exclusively within the human bones of the SCID-hu hosts, cells from patients with extra medullar disease (pleural effusion, plasma cell leukemia) grew also on the outer surfaces of the human bones. This difference in growth patterns indicates that while dependent on a human microenvironment for survival, extramedullar myeloma is not dependent on the bone marrow microenvironment. In both cases myeloma cells disseminated from one human bone to another, remotely implanted human bone.
Myeloma dissemination could be passive, with circulating myeloma cells seeding and growing only in the bone marrow where they find a hospitable, supportive microenvironment. Alternatively, dissemination can reflect an active process of homing to the bone. Primary myeloma cells from practically all patients express the chemokine receptor CXCR4 and migrate ex vivo to a source of its ligand SDF-1. Femora and vertebrae of SCID mice injected intravenously with a myeloma cell line engineered to constitutively express high levels of CXCR4 had significantly higher infiltration of myeloma cells than those of mice injected with cells engineered to express very low levels of the receptor (12%±1 vs. 0.02%±0.02, p=0.01 for femora, and 25%±13 vs. 0.02%±0.02, p=0.01 for vertebrae, respectively). IL-6 down regulated CXCR4 expression and SDF-1 upregulated MMP2 and MMP9 activity in myeloma cells, compatible with an invasive phenotype. Thus, myeloma cell dissemination appears to be an active, SDF-1-mediated process of homing to the bone marrow.
To identify elements of the bone marrow microenvironment that are crucial for myeloma growth, the effects of agents that interfere with myeloma-induced changes in the human bone marrow were investigated.
Treatment with the bisphosphonates pamidronate and zoledronate effectively halted further resorption of the human bones. Human bones of zoledronate treated hosts had 7±6 osteoclasts/mm2 compared with 27±10 in untreated hosts (p=0.003). In contrast, pamidronate did not reduce the numbers of osteoclasts (28±9 and 22±10 in control and treated hosts, respectively) but inhibited their activity as indicated by stoppage of bone resorption. Both bisphosphonates inhibited the growth of medullar myeloma but had no effect on extramedullar myeloma. Treatment of SCID-hu hosts prior to introduction of myeloma cells prevented the growth of myeloma, even without further drug treatment, and inhibition of TRANCE by an OPG analog (TR-Fc) had anti-myeloma effects similar to the bisphosphonates, inhibiting medullar but not extra medullar myeloma. It thus appears that medullar myeloma requires active osteoclasts for sustained survival and growth in the SCID-hu host, a feature that distinguishes it from extramedullar disease.
To determine the role of neo-angiogenesis, myelomatous SCID-hu hosts were treated with thalidomide. In 5 of 6 hosts that also had human liver implants, thalidomide had clear anti-myeloma efficacy (contrasting with 1 response among 6 hosts without human liver implants). In contrast to the bisphosphonates, thalidomide was also effective in hosts with extramedullar myeloma. However, it was not clear if thalidomide's effect was related to the drug's anti-angiogenic activity, as ample microvessels were present in areas of residual myeloma and in areas of regenerating marrow in the human bones of treated hosts. Recombinant endostatin administered daily to myelomatous SCID-hu hosts had a clear anti-myeloma effect that was associated with strong inhibition of anti-human CD34 expressing microvessels. Similar effects were seen with Adenovirus-mediated endostatin gene therapy. It would thus appear that survival and growth of myeloma cells in the SCID-hu host require active neo-angiogenesis.
These observations suggest the following scenario for myeloma: Downregulation of CXCR4 by IL-6 and upregulation of MMPs activity by SDF-1, help dislodge myeloma cells from their nests in the bone marrow. The circulating myeloma cells are directed to reenter the bone marrow by SDF-1/CXCR4-mediated chemotaxis. Upon re-entry, the myeloma cells affect changes in the new bone marrow sites by inducing osteoclast formation, neo-angiogenesis, and cytokine production, resulting in a microenvironment with an altered cellular makeup and cytokine milieu that is supportive of their growth and is responsible for disease manifestation. Interference with this "cultivation" process deprives the myeloma cells of essential factors required for their survival and can halt and potentially reverse the disease process.