A great deal of nonsense has been written about molecular medicine. This ranges on the one hand from puff pieces claiming that targeted therapeutics will provide the perfect drugs to cure every patient without side effects, all the way to warnings of doom for the industry as markets get sliced into smaller and smaller pieces. Like most such notions based on group-think rather than the hard analytical work, both ideas are simply wrong.
Some of the first targeted cancer therapies, such as Herceptin and Gleevec, have shown complete cures with a single drug will be the exception, even in cancers driven by a specific target, and initial impressive responses may not be durable.
In contrast to Herceptin, which is given only to patients who overexpress HER2, Avastin and Tarceva have no restrictions related to patients' expression of their targets, demonstrating that a targeted drug does not automatically equate to a narrow label.
In fact, deconstructing the short history of targeted therapies shows that a more complex picture is emerging.
For starters, in cancer and other diseases, the targeted therapies themselves have created new markets in patients whose diseases have become resistant to these drugs.
At the same time, researchers are pursuing targeting strategies to restore normal biological processes that clearly will have utility in very broad populations.
The most important lesson is that targeting strategies present clinical researchers with key choices. One is whether to develop molecules that are exquisitely specific or relatively promiscuous. Another is whether - and when - to enrich clinical trials. The choices will be influenced by the knowledge of the biology, the ability to assay for the target, and the regulatory strategy.
The upshot is that targeted therapies will be neither magic bullets, nor the end of blockbuster drugs. Instead, they will help define large, new, biology-driven markets at the same time that they create new niches.
When the idea of targeted therapeutics gained widespread currency in the mid-1990s, it was viewed in many ways like gene therapy: researchers would find and understand a genetic defect, there would be one target, and the resulting therapy would cure the disease.
The reality has been quite different. With the possible exception so far of Gleevec imatinib from Novartis AG (NVS; SWX:NOVN, Basel, Switzerland), the silver bullets look like they will be rare. Moreover, at least in cancer, even those bullets will provide only a temporary reprieve.
In retrospect, this shouldn't be surprising. Biological systems, including disease systems, generally have multiple workarounds to ensure achievement of critical path goals. In practice, this reality has played out in several ways.
The early lessons came from Herceptin trastuzumab, an anti-HER2 monoclonal antibody from Genentech Inc. (DNA, South San Francisco, Calif.). It was approved in 1998 to treat patients with metastatic breast cancer whose tumors overexpress HER2. Initial approval was as a first-line therapy in combination with paclitaxel, and as a single agent in second- and third-line therapy.
The target population had the characteristics of one that would respond to a silver bullet, as the 20-25% of women with metastatic disease who overexpress HER2 have poor prognoses. Yet despite the fact that only patients who overexpress the target are treated, not all of them respond.
"Sometimes targeted therapies will be a magic bullet, but in general cancers are complicated and have sophisticated systems for mutating, " said Hal Barron, SVP of development and CMO at DNA. "And our ability to differentiate tumors into different pathways that are causing the problem hasn't been perfected yet. Even in HER2 overexpressing patients, other pathways can overcome it."
Over the years, DNA has learned more about the molecular mechanisms that are important in driving tumor growth in patients who progress despite Herceptin therapy.
Several hypotheses have been proposed, according to Kenneth Hillan, vice president of development sciences at DNA. These include activation of the PI3 (phosphoinositide 3) kinase pathway (e.g., via inactivation of