|Send to printer »|
Feature Articles : Jan 15, 2010 ( )
Cancer Biomarker Chase Accelerates
Researchers Are Rapidly Gaining Ground in the Identification of These Early Warning Signals!--h2>
Expanding faster than the largest supernova, cancer biomarker development has become the priority of every single public and private entity performing cancer research nowadays. Evidence of this expansion comes in the form of multiple scientific conferences covering the topic. One of these conferences, IQPC’s “Oncology Biomarkers,” took place in Philadelphia late last year. Attendees discussed topics ranging from biomarker discovery to companion diagnostics.
Ann Kapoun, Ph.D., director of translational medicine at OncoMed Pharmaceuticals, described OncoMed’s approach to targeting cancer stem cells using the first drug in its pipeline, an antibody raised against DLL4—a ligand of Notch signaling that signals stem cell self-renewal, proliferation, and differentiation.
Dr. Kapoun also revealed OncoMed’s cancer biomarker strategy, specifically how “we use these models to develop pharmacodynamic markers, predictive biomarkers, as well as gene signatures that potentially can be used as cancer stem cell biomarkers.”
OncoMed essentially performs preclinical studies in its own cancer xenograft tumor models to identify and develop biomarkers that could be integrated and translated into early clinical trials. Dr. Kapoun shared data that followed pharmacodynamic markers in xenograft tumors as well as in different surrogate tissue types, such as blood, plasma, and hair follicles. “The markers that we are investigating represent the underlying response to the antibody against DLL4 activity, in other words, the markers are mediators of that signaling pathway.”
The xenograft models are part of OncoMed’s tumor bank, which is derived from primary patient specimens. “One of the main strengths of our tumor bank is that we can correlate data from preclinical studies with data from clinical studies and potentially identify patient stratification markers based on responders and nonresponders that we were seeing in these preclinical xenograft models,” Dr. Kapoun said.
She also presented preclinical data on the antitumor activity of OncoMed’s anti-DLL4 antibody, which is currently being tested in a Phase I trial. According to Dr. Kapoun, this antibody was active in a broad spectrum of human tumors in OncoMed’s xenograft models, including colon, breast, pancreatic, and lung cancers. The drug showed this activity as a single agent as well as in combination with different standard of care agents.
Joerg Heyer, Ph.D., director of genetic models at Aveo Pharmaceuticals, presented new data on the utilization of the company’s preclinical models to identify cancer biomarkers early in the drug development efforts. “There is a need to really get a head start on biomarker discovery, so that you can test them in Phase I and II and really make an informed Phase III trial,” reported Dr. Heyer. “We use preclinical models to mimic human cancer variation that we see in a patient population and to identify the biomarkers that allow us then to test the biomarkers in Phase II.”
Much of Dr. Heyer’s presentation focused on the development of these models. These genetically engineered models are developed with three elements of human cancer in mind: genetics, context, and variation. With the goal of developing a model with as much human context as possible, the entire population of genetically engineered tumors is grown in vivo, not in cell culture.
“What that does is that for every tumor that grows in our model, even though we’ve specified in genetic engineering some of the genetic alterations, we let the tumor decide which genetic alterations it needs to acquire to become a successful tumor,” said Murray Robinson, Ph.D., senior vp of oncology. There is significant tumor-to-tumor variation in the population, despite the fact that every tumor is derived from a single genetically engineered starting point. “And that, of course, is exactly what you see in human tumors.”
Dr. Robinson also presented results from a Phase II trial of Tivozanib, Aveo’s small molecule inhibitor of vascular endothelial growth factor receptor. “When this drug was put in xenograft models, which is the traditional model, it could work against virtually all of the xenograft tumors,” he explained.
However, like other receptor kinase inhibitors, when Tivozanib was tested in human trials drug response rates were significantly more modest. In Aveo’s population-based tumor model, Tivozanib replicated its behavior in human trials. “So what that has led to, and what Dr. Heyer presented, is a Tivozanib-responsive biomarker that we are currently testing in a Phase II trial,” Dr. Robinson explained.
A presentation given by Richard Kennedy, M.D., Ph.D., vp of experimental medicine at Almac Diagnostics, focused on the use of customized microarray-analyzed fresh frozen paraffin-embedded (FFPE) cancer tissue biopsies to identify novel cancer biomarkers. “We’ve developed a microarray technology that is optimized to work with FFPE tissue,” he said. “Most clinical trial datasets include large FFPE databases, and we are now able to analyze these in great detail using our technology.”
Almac developed disease-specific array (DSA) technology for five major cancers (colon, breast, lung, prostate, and ovarian). It has used the technology to develop a pipeline of diagnostic tests, one of which—a test for stage 2 colon cancer—was discussed in great detail by Dr. Kennedy.
The goal of the test, which is expected to be submitted to the FDA next year, is to predict which patients are at risk of recurrence following surgery and may benefit from adjuvant chemotherapy. “Typically arrays only work with the fresh [nonfixed or nonpreserved] tissue like frozen or fresh extracted tissue, so these tests are only possible because we’ve been able to use archived tissue samples,” Dr. Kennedy explained.
To build DSA, Almac sequenced all five cancer transcriptomes and then designed appropriate probe sets on an Affymetrix array format. One special feature of DSA that enables analysis of FFPE tissue samples is that the probes used to detect the message are designed to the true 3´ end of the sample RNA transcripts. This region is amplified better than the 5´ end from archived tissue, Dr. Kennedy said.
Christopher S. Foster, M.D., Ph.D., D.Sc., George Holt professor of pathology at University of Liverpool, planned to discuss his biomarker research at the meeting. A major premise of his work is that “approximately two-thirds of the prostate cancers do not need urgent treatment at the time of diagnosis because they will progress slowly, whereas one-third of prostate cancers progress relatively quickly.”
By focusing on developing biomarkers of prostate cancer progression in order to begin phenotyping prostate cancers, Dr. Foster’s research has yielded a number of predictive biomarkers that have been published in peer-reviewed scientific literature. The most powerful of these predictive biomarkers is heat shock protein 2, which enables researchers to accurately segregate between slow-progressing and fast-progressing prostate cancers at the time of diagnosis.
Dr. Foster plans to identify tissue biomarkers by analyzing spliced RNA transcript levels in prostate cancer biopsies by microarray instead of the more traditional histochemical approach. “That alone is a major step forward. If one can take a patient’s biopsy, phenotype it, and accurately predict its behavior when left untreated, or to begin suggesting drugs or therapies that might or might not be appropriate biologically, then it will be a major leap forward in cancer management.”
© 2016 Genetic Engineering & Biotechnology News, All Rights Reserved