By Malorye A. Branca, Editor-in-Chief, PharmaWeek

While they have different methods for finding, applying, and even defining biomarkers, all large pharmaceutical companies are putting tremendous emphasis on these as tools for improving drug discovery and development. At this week's CHI Clinical Biomarker Summit, Pfizer scientist Samuel DePrimo described some approaches the company is using to help grow its nascent targeted oncology portfolio quickly.

"At Pfizer we like to think about biomarkers as relating to targets, mechanisms, or outcomes," DePrimo said.  All three types of biomarkers are particularly useful in oncology because signal transduction pathways are "complex" and the same molecules can play different roles in different cancers. To be able to move forward with confidence, it is critical to get early confirmation that the same effects seen in preclinical models are occurring in human subjects.

DePrimo told "Tales of three protein kinase inhibitors" to illustrate how the three types of biomarkers were developed.  In all three cases, the preclinical work has uncovered markers that could have clinical utility as well.

In the first example, Pfizer scientists saw a neat correlation between levels of phosphorylated ERK (pERK), drug dose, and cellular response during preclinical development of the company's MEK inhibitor, PD 0325901.  They hoped to see the same kind of relationship with a couple of downstream markers of cell proliferation – Ki67 and caspase-3.  While the drug did inhibit these markers, neither one seemed linked to outcome.  This somewhat perplexing result is a reflection of "tumor heterogeneity" DePrimo said. While a lot of the same molecular players are involved in most cancers, different ones take starring roles in particular types of tumors.

The second example is related to the company's CDK4/6 inhibitor (PD 0332991).  In this case, two very different kinds of biomarkers both proved to be useful preclinical guideposts. Phosphorylated RB protein and gene expression profiling both affirmed the drug was having the intended effect in xenografts.  For the gene expression analysis, the group used the Affymetrix HG-U133 chip to look for genes down-regulated by the drug. They found "Transcripts with overlapping down-regulation in all 3 xenografts studied," DePrimo said. Even better: "A large number of the down-regulated genes are associated with DNA replication."

Finally, biomarkers played a key role in the development of SU11248 (Sutent), Pfizer's new treatment for kidney cancer and GIST. This drug attacks both the tumor itself and its vasculature by inhibiting fms-like tyrosine kinase 3 (Flt3), Kit, VEGF, and PDGF.  While all of these proteins are useful markers of activity, Flt3 levels proved to be particularly useful.  "SU11248 target modulation of Flt3 has been characterized preclinically and clinically," said DePrimo. 

Getting early confirmation that the target is being hit, and that that makes a difference, is a huge bonus for drug developers. But it may require "innovative clinical design" and extra steps during preclinical studies to get those results, DePrimo said.  Hopefully such steps will lead to even more drugs like Sutent – the first compound ever approved simultaneously for two cancers.

©  Copyright 2006, Cambridge Healthtech Institute. All Rights Reserved