PHILADELPHIA (GenomeWeb News) — As drug makers increasingly use biomarkers to help them understand disease and drug activity, a passel of seemingly simple human oversights threatens to muddy results, according to a senior pharma official.
Poor communication, misaligned research protocols, careless vendor selection, fragmented research, and medical nescience will hurt biomarker-based drug-discovery efforts, Yu Chen Barrett, associate director of the cardiovascular and metabolics biomarker team at Bristol-Myers Squibb, said at a biomarker conference yesterday.
Drug makers in recent years “have set high expectations for biomarkers to provide critical information relating to drug mechanism of action, dose selection, efficacy, and safety so that candidate drugs can be developed with confidence,” Barrett said during her presentation at Cambridge Healthtech Institute’s Biomarker World Congress, held here this week.
She said pharmas use biomarkers to study the pharmacodynamics and pharmacokinetics of drug candidates; to ascertain disease or adverse event risk; to help predict efficacy; and to stratify patients in clinical trials.
But “the diverse nature” of biomarkers has also “generated considerable confusion” when it comes time to implement and interpret them in clinical trials, she said.
To avoid this, it is “imperative” for researchers to “understand the biological and physiological relevance of biomarkers,” Barrett said.
According to Barrett, if an assay developed to identify a biomarker fails, suspicion can run up and down the discovery continuum. The lab responsible for running the assay may blame the drug maker’s clinical team, which provided the samples, by questioning its ability to collect the samples in the first place.
“They’d say, ‘Do they know what they’re doing, because what they sent us is crappy,’” said Barrett.
Meantime, the clinical team may suspect the phlebotomists who drew the blood on which the samples are based.
“Many times labs and clinical teams don’t communicate,” she said.
In fact, Barrett said, bad assays can come from vendors that happen to calibrate them differently from the sponsor. Barrett stressed that pharmas should “make sure [a vendor’s] assay protocol agrees with your protocol.”
Failing to understand basic biology can also hurt biomarker research, she said. For instance, drug makers that use the C-reactive protein biomarker to track disease progress or drug efficacy often don’t account for the fact that CRP levels in some individuals fluctuate throughout the day and during seasonal changes, Barrett said.
Another opening for shoddy research is failing to ask basic medical questions: An assay that works in a Phase I study won’t necessarily play along in a Phase II trial. The earlier trial is populated with a smaller number of healthy people not on medication, while patients enrolled in larger Phase II studies are typically diagnosed with the disease a candidate aims to treat, and are therefore likely to be taking medication that can throw off assay results.
Barrett is not alone. John Bloom, executive director of diagnostic and experimental medicine at Eli Lilly, agreed. Speaking during his own presentation at the conference, Bloom said that “one of the biggest challenges” of biomarker-based drug development “is integrating all of the external and internal elements” of research.
Bloom, who called the task “daunting” but a “fun challenge,” said “most of the novel tools, disciplines, and strategies for optimizing drug development … require an organizational construct with functional components and partnerships that are integrated and fungible.”
“The traditional segmentation of pharmaceutical R&D, and the resulting functional ‘silos,’ will increasingly impede our ability to develop and implement phase-appropriate biomarker strategies and fully leverage enabling technologies and disciplines,” he added.