One of the things that fascinates me most about the slow-paced business of biotech is how utterly mismatched it is against the demands of Wall Street.
Asking for quarterly earnings reports from companies that are in the red, and likely to remain so for years, while they test drug candidates in early-stage clinical trials (which may or may not translate to success in a phase III study) is a little bit like requesting a Big Mac in a French restaurant. Investors want as-promised, fast-food results when biotech can only deliver a time-consuming coq au vin of variable success.
I was thinking about that earlier this week, when I saw a really interesting article about the tangible impact genomics and other “omics” are beginning to have in drug development. Those of you who have been around this business for a while remember the days of big promises by the likes of Human Genome Sciences, DeCode Genetics, and Millennium Pharmaceuticals. That was a little over ten years ago, and genomics was being touted as the field that would revolutionize medicine. It would lead to new drug targets and a better understanding of human disease.
And now, given the dearth of drug candidates to materialize from that promise as of yet (and the hundreds of millions of investor dollars lost in the process, since many of the companies from the genomics boom went bust or nearly bust), naysayers love to point their finger at it as an example of another bubble burst.
There is some truth to these lamentations, but similar things were said of monoclonal antibodies, and now these therapies are making a real difference in patients’ lives. Life cycles in biotech and pharma may be too long to agree with many investors’ patience spans, but sometimes it pays to just wait.
Which brings me back to that paper I mentioned before: it just came out in Nature Reviews Drug Discovery (June 2010 issue) and it gives some pretty interesting examples of how genomics and other “omics” (proteomics, metabolomics) are starting to impact drug development behind the scenes.
One of the goals of genomics is to find certain signatures that could be correlated with better response to treatment, or with a certain stages of disease (for example, an early form of cancer) to better select drug candidates or patient populations in clinical trials. But the problem is that this kind of information—usually a readout from massive gene-expression or other array-type experiments—is vast and hard to interpret.
For example, let’s say a company discovers after a failed phase II trial that higher expression of a set of 10 genes correlates, retrospectively, with response to therapy. A company might want to pursue the failed drug candidate within this specific subset of patients, but it would be too risky
