As an example, Reilly said Merck has used tumor microarrays in tests of pathway-specific biomarkers of response to its drug Vorinostat, which is a histone deacetylase inhibitor used to treat lymphoma.
At DDT, Two Big Drug Makers Shed Light on Their Genomic, Proteomic Biomarker Programs
BOSTON — Last week’s IBC Drug Discovery Technology meeting here offered some insight into how pharmaceutical companies are incorporating biomarkers into drug development for more efficient pipelines and, in some cases, for later pharmacogenomic diagnostics.
Besides Bristol-Myers Squibb (see article, this issue), Johnson & Johnson, Amgen, and Merck are taking advantage of genomic and proteomic biomarkers — in the form of markers of drug response or patient stratification — to help them usher drugs through clinical trials more efficiently.
For J&J, its focus on identifying molecular biomarkers calls for casting a wide net to evaluate many types of biomarkers, and then quickly cutting down the number of choices during biomarker development. Meantime, Amgen, whose concentration on biomarkers is somewhat narrower than J&J’s, focuses on protein biomarkers that can be easily obtained and efficiently tested. And Merck’s use of immunohistochemistry in testing response biomarkers has required the company to update its instrumentation.
Biomarkers for Decision-Making at J&J
Like many other pharmas, J&J uses genomic, proteomic, and other biomarkers to help stratify or eliminate patients in clinical trials, establish clinical trial dosages, or detect drug efficacy.
To find biomarkers for clinical decisions, the company begins with an extensive literature search that can yield a variety of leads, such as proteomic, metabolomic, methylation, imaging, comparative genomic hybridization, and gene-expression biomarkers, Hans Winkler, director of functional genomics, bioinformatics, and target validation at J&J, said during his DDT presentation last week. Efficient and thorough text mining can help the process, he said.
Following detection in samples using microarrays, positron emission tomography, mass spectrometry, or another technique, development teams use MS or nuclear magnetic resonance to winnow down the list, Winkler added.
A common problem in establishing response biomarkers for stratification during drug development is that samples are hard to obtain in the time available, said Winkler. He suggested taking advantage of the accessibility of serum and other bodily fluids to help speed the process. Because the samples are easy to obtain, pre- and post-dose analyses can identify response biomarkers that can be quickly attained during later trials.
Meantime, aiming for parallel, simultaneous drug and biomarker development can lead to problems of its own, even if the biomarker assay is not intended for use outside a clinical trial. It can be time-consuming and expensive to develop assays before clinical evaluation in phase 1 testing, and can lead to a time gap between drug and diagnostic development, said Winkler. The industry needs new assay-development techniques to solve the problem, he said.
Biomarkers for Early Clinical Development at Amgen
Scott Patterson, Amgen’s senior director of medical sciences, said the company’s early clinical development program is focused more on finding protein biomarkers that modulate the target rather than those that can be used as surrogate endpoints.
Amgen researchers are also interested in whether a biomarker can be used to help determine dose selection, stratify patients, or identify rapid responders, he said.
Like J&J, Amgen is particularly interested in analyzing blood and other bodily fluids that are easily obtained, said Patterson. From these, Amgen researchers measure cytokines, chemokines, and other proteins, as well as levels of free DNA.
Amgen uses multi-analyte profiling with bead-based and planar arrays to test body-fluid-based protein biomarkers and to detect signaling proteins in cell lysates, Patterson said.
Amgen also uses flow cytometry in permeabilized cells for signaling-protein assays, as well as for cell-surface protein assays, he added.
The company’s genomic assays involve tumor-based methods – profiles of specific RNA transcripts, gene sequencing, and gene copy-number detection.
Amgen has also had some luck using pathway analysis to choose stratification markers, said Patterson. The team developing AMG706, a phase 2 compound for non-small cell lung cancer, chose six protein analytes that would likely be affected by the inhibitor because they modulate angiogenesis, he said.
The researchers created two multiplexed ELISA assays for the analytes, grouping analytes with similar in vivo concentrations into the same assay to save time and materials.
Biomarkers of Response at Merck
Merck appears to be improving an older technology in order to develop biomarkers. John Reilly, a research fellow at the company, shed some light on the Merck’s development of responder-identification biomarkers using IHC as part of his DDT talk.
The old technique is well understood and samples are copious due to the availability of formalin-fixed paraffin embedded tissues, but IHC still suffers from a major problem: very low throughput.
However, the drug maker has been making use of automation, and the analysis step has recently had two “key” advances that have accelerated the technique, said Reilly.
First, the addition of image recognition and analysis capabilities has allowed machines to move into the one step of the protocol that has previously required a human eye, he said. Second, tissue microarrays have greatly increased the number of samples that can be processed and analyzed simultaneously, he added.
Outside of analysis, the use of pre-validated antibodies can reduce costs and time spent on research.