Neonatal Mouse Bone Supports Growth Of Primary Myeloma Cells Ex Vivo and Assessment Of Responses To Antimyeloma Drugs

A barrier to optimizing treatment for multiple myeloma (MM) is the lack of a convenient assay to determine drug-responsiveness in individual patients. Although primary MM cells are routinely obtained from patient bone marrow aspirates, CD138+ cells die too rapidly in culture to be analyzed. We hypothesized that primary MM cell survival requires the bone microenvironment to provide the necessary soil for the growth of the cancer cell seeds. Immunodeficient animal models contain a complete multicellular bone marrow microenvironment, including extracellular matrix, and will support engraftment of patient MM cells. However, intact animal models are too slow, complex and expensive to provide timely and clinically useful predictions of individual tumor behavior and responses to drugs.

We hypothesized that a well-established model for studying bone in culture could be adapted to support the growth of primary MM cells and to study myeloma bone disease. We developed an ex vivo organ coculture assay (EVOCA), with neonatal mouse calvarial bones providing the soil to support tumor cell growth. Human MM cell lines and primary MM cells were successfully cocultured with immune-naïve neonatal hemi-calvariae for up to 2 weeks. EVOCA was optimized to support bone and MM cell survival and analyzed by histology and by PCR with species-specific primers for changes in expression of tumor (human) and host (mouse) genes. Four human MM cell lines (8226, H929, U266 and JJN3) and primary CD138+ cells from 5 patients (1 freshly isolated and 4 from previously frozen cells) survived for > one week in calvarial cultures, using as few as 10⁴ cells. Both primary MM cells and CD138+ cells were detected by immunohistochemistry for human CD138 after 2 weeks and by species-specific RT-PCR for the human housekeeping gene RPL32. Appropriate responses of mouse bones to MM were observed at 2-4 days, including increased osteolysis by histology, RANK ligand mRNA by PCR and osteoclast numbers by TRAP staining.

To determine responsivness of MM to anti-myeloma drugs, cells were cultured with calvariae for 48 hours to establish engraftment, then the EVOCA was treated with 10 uM lenalidomide, 10 nM bortezomib, or 10nM carfilzomib. Anti-myeloma treatments attenuated the bone responses including reduction of RANKL. Primary MM cells from different patients varied in responsiveness to different drugs. Among 5 samples tested, primary cells obtained at the time of diagnosis were more responsive to anti-myeloma drugs in blunting induction of RANK ligand compared to MM cells obtained at the time of relapse.

EVOCA, based on neonatal mouse calvariae supporting human MM cells, provides a useful platform for studies of myeloma bone disease. The assay can be run and analzyed in two weeks with 10⁴CD138+ cells (freshly isolated or from frozen stock) per well, making it practical to determine individual patient sensitivity to a panel of currently available anti-myeloma drugs at initial diagnosis or at relapse. This could permit rational selection of personalized MM therapy. EVOCA could also be used to develop novel myeloma drugs that target tumor or bone by efficiently testing them in a physiologically relevant setting.