Traditional diagnostic tests for Her2 (the target of the drug Herceptin) have well-documented failings. In July, Monogram Biosciences launched a new Her2 test, and two months later, Genomic Health added Her2 status information to its breast cancer test. They believe that more comprehensive, multianalyte tests may better guide therapy and yield a pharmacoeconomic benefit, even at eight to 10 times the cost. These developments point to the need for a more nuanced approach to the management of cancer as a chronic disease and a more nuanced view of personalized medicine.

Despite the success of Gleevec, developing drugs that inhibit complex signaling pathways, while simultaneously trying to understand the biology around a drug's target, remains a challenge. Without biomarkers to help establish dosing and identify likely responding patients, clinical development of targeted cancer drugs will remain challenging. AstraZeneca's recent experience with Iressa bears this out. It's likely that single markers will not be sufficient to stratify patients by their tumor types; rather, patterns of gene and protein expression will be required. The technology to take these measurements is making its way from academia to industry, but the process is slow, and needs encouragement and better coordination between academia, regulators, and industry. Meanwhile, clinical trials themselves remain the best target validation tools. When all is said and done, efficacy is the best biomarker.

Tiny Vysis scored big this spring, winning the backing of two major therapeutic and diagnostic companies for its products. But is its core technology ready for prime time? Vysis sees an opportunity to expand use of its FISH diagnostics technology beyond niche applications in the clinical market. But skeptics say the technology is too complex and expensive for routine clinical use.