Millennium Pharmaceuticals has discovered biomarkers that may identify likely responders to its multiple myeloma drug Velcade, and hopes that additional pharmacogenomic research will help improve outcomes and expand use of the drug into a broader multiple myeloma patient population.
Genomic analyses on tumor samples prospectively collected from previously treated multiple myeloma patients revealed that patients with an increased expression of genes in the NF-κB and adhesion molecule pathways may be more sensitive to the drug.
Additionally, Millennium identified gene-based predictive and survival classifiers that need further validation. The results from Millennium’s analyses appear in the April 15 issue of Blood.
The US Food and Drug Administration approved Velcade in 2003 to treat patients whose myeloma has relapsed after two prior therapies. It is the first proteasome inhibitor to win FDA approval.
George Mulligan, associate director of oncology clinical development at Millennium, told Pharmacogenomics Reporter in an e-mail this week that the company is currently focused on conducting genome-based clinical trials to identify patients who will respond to Velcade as a single agent or in combination with other therapies. But he said the research is too preliminary to determine if a molecular diagnostic would optimize treatment.
Millennium conducted the research in collaboration with the Dana-Farber Cancer Institute, the Myeloma Institute of Research and Therapy at the University of Arkansas for Medical Services, the Mayo Clinic, and the Department of Hematology at Erasmus Medical Center.
For the genomics research Millennium prospectively collected tissue samples from several studies. Among them was the APEX Phase III study, which showed Velcade to be superior to the standard of care, dexamethasone, in relapsed multiple myeloma patients. Dexamethasone is a steroid given to cancer patients to manage the side effects of their chemotherapy treatment.
APEX enrolled 669 multiple myeloma patients. Half were treated with dexamethasone and half with Velcade as a monotherapy. Millennium’s analysis “showed that the biomarkers did not predict sensitivity to dexamethasone therapy, indicating that the biomarkers may be specific to Velcade,” the company said in a statement accompanying the Blood paper last week.
The data “add to the body of evidence that supports the role of Velcade in the treatment of a broad range of multiple myeloma patients and highlight the applicability of genomics research,” Millennium said.
Velcade co-developers Millennium and Johnson & Johnson will conduct genomic analyses in combination with other cancer agents to explore optimal treatment strategies.
Genomics Classifiers Need Further Study
In the past, retrospective gene-expression analysis has been able to pinpoint tumor sensitivity to certain cancer therapies. However, Millennium’s pharmacogenomic goal with Velcade was to collect genetic samples prospectively and link gene expression with the proteasome inhibitor’s activity.
“Doing prospective work as part of clinical trials is necessary for this research to reach its full potential,” Mulligan said.
According to Mulligan, prospectively designed studies “begin to deal with analysis of patients rather than historical samples” and yield results that are “relevant for current drugs or drug combinations rather than a treatment that was standard five to 10 years ago.”
Millennium’s genomic research experienced a few setbacks. Researchers were unable to study approximately 50 percent of collected samples due to insufficient tumor sampling or RNA degradation, according to the Blood paper. Additionally, since data was gathered from multiple clinical trials, the differences between studies made comparisons challenging, the article states.
Still, utilizing the samples, the study authors identified a pretreatment gene-expression pattern and predictive classifier “subtly associated” with Velcade response, but not to dexamethasone.
Studies showed “Velcade was active in various molecular subtypes of myeloma and that a compilation of genes – a ‘classifier’ – could predict response to this new single-agent drug when examined in a separate dataset,” Mulligan said.
Researchers also found a distinct genomic classifier based on overall survival and pathways dictating NF-B activity and cell adhesion associated with Velcade activity.
According to the authors, many of the pathways associated with progressive disease regulate protein biosynthesis and mitochondrial function. These functions could be connected to protein load in secretory myeloma cells or the status of mitochondrial apoptotic pathways, the authors note in the paper.
“Consistent with the response prediction, some of these pathways (e.g. adhesion- and cytokine-related gene sets) appeared to be [Velcade]-specific when data for [Velcade] versus dexamethasone were compared,” the article states. “It will be important to induce and/or inhibit these pathways in model systems to test whether their activity confers sensitivity or resistance to [Velcade], dexamethasone, and/or other anticancer agents.”
According to Mulligan, researchers found that these pathways were elevated in between 50 percent to 70 percent of the relapsed multiple myeloma tumor samples. The researchers are investigating why this elevated presence doesn’t always confer sensitivity to Velcade.
“This is a very interesting area that involves but goes beyond bioinformatics,” Mulligan said. “We need to manipulate pathways to understand how their relationships confer resistance or sensitivity.”
“Will clinicians need a diagnostic to treat relapsed myeloma patients … with single-agent Velcade and what sort of accuracy would be necessary for them to use that tool? These remain unanswered questions, but it is expected and we hope that such questions will continue to come up and help improve the therapeutic strategies for various cancers.”
With regard to the survival classifier, the study authors recommended additional analyses and comparisons with other myeloma pharmacogenomic datasets to validate this finding.
“At this time, we cannot determine whether the association of the classifier to survival is specific for Velcade treatment because the majority of patients in the dexamethasone arm were subsequently treated with Velcade,” the article states.
“Although the classifier described here is promising, further refinement is necessary before it can be considered for clinical use in predicting patient response to single-agent [Velcade] in the relapsed setting.”
“The 75 percent overall accuracy might be improved with more patient samples, or it may be that there is not adequate information in the RNA levels of pretreatment, purified myeloma samples to make a significantly more accurate prediction.”
According to the Blood paper, the research is still too early to elucidate plans for a drug-response diagnostic. “The predictive accuracy required of a clinical diagnostic for myeloma treatment has not yet been defined,” the study authors acknowledged.
The development of a diagnostic may be contingent upon disease stage, if Velcade is used as a single agent or in combination with other therapies, and whether Velcade will achieve disease control or cure.
“The knowledge about less sensitive tumors will drive fresh rounds of new drug development,” Mulligan said. “If the accuracy of this predictive research is adequate, a diagnostic tool will need to be addressed.”
According to Millennium, J&J is continuing genomics research on Velcade in additional trials. A Phase III trial by the Dutch and German co-operative groups HOVON and GMMG is investigating a Velcade-based combination treatment in newly diagnosed multiple myeloma patients. Additionally, Millennium is conducting a Phase III genomic study looking at Velcade-based combination treatment in indolent non-Hodgkin's lymphoma.
Mulligan acknowledged that there may be technical hurdles to identifying genomic classifiers, but the most immediate challenges with Velcade are clinical. “That is, will clinicians need a diagnostic to treat relapsed myeloma patients … with single-agent Velcade, and what sort of accuracy would be necessary for them to use that tool?” Mulligan posited.
“These remain unanswered questions, but it is expected and we hope that such questions will continue to come up and help improve the therapeutic strategies for various cancers,” he added.