Comparative Analysis of in Vitro Chemosensitivity to Four Proteasome Inhibitors in Human Myeloma Cell Lines

Proteasome inhibitors (PI) are effective chemotherapeutic agents in the treatment of multiple myeloma (MM), used alone or in combination with other anti-cancer agents, such as alkylating agents, topoisomerase inhibitors, corticosteroids, histone deacetylase inhibitors (HDACis) and immunomodulatory drugs (IMiDs). Bortezomib (Velcade/Bz) was the first PI to be approved by US-FDA for the treatment of relapsed and refractory MM. Other second generation PIs include carfilzomib (Kyprolis/Cz), ixazomib/Iz and oprozomib (Opz). Wide inter-individual variation in response to treatment with PIs is a major limitation in achieving consistent therapeutic effect in MM. Yet few studies have compared the efficacy of all four PIs in a range of myeloma subtypes.

In our current study, we performed comprehensive in vitro chemosensitivity profiling of response to four (4) PIs (Bz, Cz, Ix and Opz) in a panel of forty-five (45) human myeloma cells lines (HMCLs) generated through the immortalization of primary multiple myeloma cells (MMCs) and representing the biological and genetic heterogeneity of MM with regards to chromosomal abnormalities, oncogene mutations (e.g. Ras), tumor suppressor variations (e.g. p53), cell surface phenotypes, or growth factor response. Cells were treated with increasing concentrations of Bz, Cz, Ix and Opz as single agents and cell viability assays were performed using CellTiter-Glo luminescent cell viability assay to generate survival curves and determine the half maximal inhibitory concentration (IC₅₀) values by calculating the nonlinear regression using sigmoidal dose-response equation (variable slope). Our results in comparing the cellular responses to PI treatment among HMCLs showed wide range of variability in IC₅₀ values identifying some lines which were highly sensitive and some lines relatively refractory to PI treatment. Pearson product-moment correlation (PPMC) test demonstrated statistically significant (adjusted p values < 0.001) positive correlation between IC₅₀ values of the following drug pairs: Bz vs Opz (r = 0.82); and Ix vs Opz (r = 0.88); Bz vs Ix (r = 0.65); Cz vs Opz (r = 0.69) and Cz vs Ix (r = 0.63). Subgroup analysis revealed significant correlation between carfizomib IC₅₀ and chromosome number (p < 0.05). Furthermore, it was interesting to note that although all 4 drugs belong to the same drug class (PI), not all cell lines responded the same across all PI treatments. This demonstrates tumor heterogeneity even in response to inhibitors of the same class, and further demonstrates tumors refractory to one PI may still respond to another. We are currently examining genetic characteristics that are associated with response among the four PIs, and analysis of these characteristics will be presented.