1 MOLECULAR PREDICTORS OF CLINICAL RESPONSE TO THE PROTEASOME INHIBITOR BORTEZOMIB: CLINICAL IMPLICATIONS FOR INDIVIDUALIZED PATIENT-SPECIFIC ANTI-CANCER THERAPEUTIC ALGORITHMS

The proteasome inhibitor bortezomib is an active agent for treatment of advanced multiple myeloma (MM). However, similarly to other novel anti-cancer therapies, drug resistance (de novo or acquired) of MM cells to bortezomib curtails its clinical benefit. Unfortunately, no currently available clinical parameters can definitively predict response to bortezomib on an individual patient basis. This creates the impetus to identify molecular markers to identify in advance patients with low probability of response to bortezomib. Utilizing oligonucleotide microarray chips (which simultaneously quantify the expression of ˜ 22,200 genes), we characterized the baseline gene expression profiles of a panel of 40 patient-derived human MM cell lines, as well as the gene expression profiles of MM tumor cells isolated at baseline (i.e. prior to bortezomib therapy) from 45 relapsed refractory MM patients enrolled into a phase II clinical trial of bortezomib. Using a compendium of bioinformatic analyses, including hierarchical/functional clustering, relevance networks and pattern recognition analyses, we identified putative predictor genes correlating with ex vivo MM tumor cell responsiveness to bortezomib and genes differentially expressed in bortezomib-treated responders vs. non-responders, leading to generation of a composite candidate molecular signature associated with high risk for early disease progression after bortezomib treatment (median progression-free survival of 10 months for patients with vs. without this high-risk molecular signature, p = 0.012, log-rank test). Additional cross-validation of this molecular signature is independently provided by analyses of tumor samples from bortezomib-treated patients of a separate phase III randomized trial comparing bortezomib vs. high-dose dexamethasone in advanced MM. Interestingly, genes predicting for response to bortezomib are distinct from genes predicting response to other conventional (e.g. cytotoxic chemotherapy) or novel therapies, consistent with the distinct mechanism of action of bortezomib, i.e. the inhibition of proteasome function in tumor cells. Our study offers insight into the mechanisms regulating the anti-tumor activity of bortezomib and builds a framework for rational design of future individualized treatment protocols tailored to each patient9s tumor molecular features, an important step towards improving patient outcome and unnecessary risk for potential treatment-related adverse events.