For one, the company defines pharmacogenetics and pharmacogenomics as basically the same thing. "It falls under the same umbrella of understanding the variability in response to drugs or disease that can be due to genetic differences," said Albert Seymour, PhD, head of discovery pharmacogenomics for Pfizer Global Research & Development, New London, Conn.
Pfizer incorporates pharmacogenomic data early in the discovery process. "We take our disease targets and add human-relevant data preclinically, as early as we can, to help build confidence that the gene plays a role in the disease mechanism," says Seymour.
Last November, the FDA issued a guidance on the use of pharmacogenomic data that encouraged drug developers to conduct pharmacogenomic tests during drug development and voluntarily submit that data with their applications to the FDA.
Seymour asks, "How can we use our knowledge of the human genome to understand and predict human response in the clinic? We are all 99.9% identical at the DNA level, but that 0.1% makes all the difference," he said. "We expect that this same type of variation can be understood at the disease level and in drug response."
These genetic differences could include inherited variation, differences in the DNA or RNA state, which can be elucidated through expression profiling, and even differences at the proteome level, he said.
The goal, said Seymour, "is to take several of these DNA variants and use them to predict response and the clinical phenotype or trait that we are interested in."
Without well-characterized clinical phenotype data, the studies are poorly designed. "The better clinical phenotype data you have access to, the better the study," said Seymour. Having the statistical expertise as well as powerful, secure data management is also very important, he said.
"We apply pharmacogenomics across the pipeline from early discovery through disease genetics to better understand some of the drivers of a disease," Seymour said. "Out of that can come target identification and validation, confidence in rationale around those targets, and genetic biomarkers that may not be druggable themselves but can be used to distinguish or differentiate a population."
One of the uncertainties for the company, said Seymour, is being able to drive efforts to differentiate compounds on the market and prove population stratification.
"When we take a set of exploratory targets and run these human genetic association studies against them, we can identify the disease-relevant models with a function for these particular alleles at the biological or pharmaceutical level," he said. "By the time we get to proof-of-concept, we have an optimal indication and an optimal compound."