Genomics in myeloma: the philosopher's stone?

In this issue of Blood, Mulligan and colleagues describe advantages and challenges of gene expression profiling experiments in the context of prospective multicenter clinical trials in relapsed or refractory multiple myeloma patients. They identify a gene expression signature that is significantly associated with the response to the novel agent bortezomib.

Gene expression profiling has long evolved from a “mythical” technology, used by a small number of scientists in their search for alchemical gold, to a broadly applied tool aiming at the improvement of cancer classification, survival prediction, and treatment optimization.

In hematologic malignancies, recent gene expression profiling studies have laid the foundation for the precise molecular definition of already established, but also novel, biologically relevant subtypes of lymphomas, leukemias, and multiple myelomas.¹  Moreover, gene expression data have been used to construct powerful survival predictors that may be helpful in guiding future treatment decisions. However, most of these studies were retrospective in nature and, in many instances, included heterogeneous patient cohorts that had been treated with diverse and sometimes outdated therapeutic regimens. The availability of numerous new anticancer agents now calls for the definition of patient subsets that are most likely to benefit from these drugs.² 

In this issue, Mulligan and colleagues report on the first successful steps toward this goal. Gene expression data derived from bone marrow aspirates of more than 200 multiple myeloma patients enrolled in multicenter phase 2 and phase 3 clinical trials of bortezomib were searched for expression-based response and survival signatures.

Multiple myeloma, a neoplasm of bone marrow plasma cells with diverse underlying molecular alterations, is currently not curable and, not surprisingly, the treatment options have become increasingly complex. While traditional laboratory parameters such as serum albumin and β-2 microglobulin,³  but also genetic alterations (eg, the presence of the chromosomal translocations t(11;14) or t(4;14)), provide useful prognostic information in the context of high-dose therapy (HDT),⁴  there is an increasing need to identify patient subgroups that may benefit from novel therapeutic agents or therapies. One such agent, the proteasome inhibitor bortezomib, was tested in relapsed or refractory multiple myeloma patients in the just-mentioned phase 2 and phase 3 trials. The specific design of these trials coupled with the generation of gene expression profiles before treatment initiation enabled the authors to develop a genomic classifier that is specifically associated with the response to bortezomib. These results may lead to the rapid development of a diagnostic test that predicts the response to bortezomib using a technical platform that is suitable for broad clinical application.

The study clearly demonstrates that large-scale gene expression profiling studies as an adjunct to prospective multicenter clinical trials are feasible. Careful quality control steps, as performed by the authors (eg, for sample handling, purification, and processing), need to be undertaken in order to derive prognostic information of high accuracy. A major challenge in the present study was the comparison between different trials with some variation in patient characteristics, such as the number of prior therapies and responsiveness to previous treatments. A future prerequisite for developing therapeutically relevant genomic classifiers in the context of prospective trials is a careful trial design and the definition of prespecified gene expression signatures that are going to be tested on truly independent data.²  In multiple myeloma, but also in other hematologic malignancies, such efforts are being planned or are already under way—these trials deserve our special attention.