It's not that medical science has conquered all the challenges facing humanity. Of course, many diseases and conditions remain frustratingly resistant to treatment. Rather, the problem is that many of the most obvious targets -- the molecules, enzymes, or molecular processes that are involved in disease and can be affected by a drug -- have been exhausted. That leaves only newly discovered, less studied targets, or more complicated targets that have proved difficult to build drugs around.
Developing drugs for "known targets," as they are called, is tough enough. Look at the history of the cholesterol-reducing drugs known as statins, such as Pfizer's Lipitor and Merck's Zocor, which inhibit a molecule that raises cholesterol. "It took a decade-and-a-half of research for statins to become drugs," points out Bob Ruffolo, president of research and development at Wyeth Pharmaceuticals (WYE
). "I don't mean to trivialize the impact or importance, but that was the low-hanging fruit." Indeed, most drugs in the marketplace now interact with one or more of the some 500 basic enzymes and chemical reactions that scientists for decades have known to play roles in disease.
RISKIER PROPOSITION. The next wave of drug development will be much more challenging. While potential targets can be identified fairly easily, more basic research needs to be done before scientists can begin to develop drugs that interact with them. The current downside: a dry pipeline.
The upside potential, however, is great: Future blockbusters will likely be much more effective in treating disease than those now available. "We need to validate our knowledge of how these targets work in diseases," says Jeff Bird, partner at Sutter Hill Ventures. "It's hard work, notwithstanding an explosion in tools available to do this kind of work."
For the next several years, the industry likely will experience more failures than successes. In the first half of 2002 alone, a dozen or so companies -- from the ranks of both the biotech and pharmaceutical sectors -- announced canceled studies or rejections by regulators. "The risk has gone up, not down," Ruffolo says. "It's not surprising to see a higher number of late-stage failures."
A NEW APPROACH. Still, encouraging signs are around. Vertex Pharmaceutical (VRTX
) in Cambridge, Mass., is using an approach it calls chemogenomics. First, multiple potential drugs are developed to focus on a single target. Then, using information about molecular structure, Vertex experiments with more sets of potential drugs specific to similar targets in the same family. The company says the strategy allows it to expand its drug-development regime.
Vertex is using chemogenomics to design drugs that block production of proteases, a substance in the body that can help certain viruses replicate themselves. "These are not the newest kids on the block. They've been a focus of research for a long time, but the industry has only developed drugs against two proteases -- HIV protease and a protease involved in hypertension," says John Thomson, vice-president for research. Vertex has drugs in testing that interact with proteases involved with hepatitis C and rheumatoid arthritis.
Also promising are drugs that focus on a broad set of targets called kinases. These enzymes regulate communication both within cells and between cells. Drugs that interact with kinases can regulate such cellular signaling, allowing the body to fight infection, for example, but calming cellular side effects.
"GREAT TARGETS." Take Novartis' groundbreaking cancer drug Gleevec, which targets a kinase. The drug is one of the first cancer treatments that leaves patients feeling less sick than chemotherapy typically does. Unlike older cancer drugs that attack both healthy and cancerous cells, Gleevec calms only the signals involved in turning normal cells into cancerous ones.
"There are hundreds of enzymes that are [targets]. It's just that less is known about them," says Julian Adams, senior vice-president for drug discovery at Millennium Pharmaceutical (MLNM
) in Cambridge, Mass. He says more than 500 kinases exist in the body. "Not all of [our research on them] will turn out like Gleevec," he admits, "but I've got to believe that there are several dozen great targets in there."
In the meantime, companies will put a lot of resources toward technology that can help scientists work more efficiently to understand which targets are integral parts of disease and how drugs need to be designed around them if they're to be effective. Software and microchips that can cull and sort through databases of genes and proteins are already speeding up research. In fact, over the past decade, a myriad of tiny companies has sprung up to focus only on making sense of the genome.
SOARING DEVELOPMENT COSTS. Ultimately, advances in technology can take researchers only so far, says Joseph Zammit-Lucia of Cambridge Pharma Consultants. "The current impasse is really because we have so much information, and it's a new kind of information that requires a whole new way of analysis," explains Phyllis Gardner, associate professor of medicine and molecular pharmacology at Stanford University School of Medicine. Synthesizing that data doesn't come cheap: The $20 million companies typically spent a few years ago to develop a new drug has soared to $100 million today, Zammit-Lucia estimates.
No one doubts that knowledge from the genome and data-mining technologies will yield the wonder drugs of the future. But the transition will require patience and persistence on the part of researchers, marketers, and investors. "The pharmaceutical industry may need to stop thinking narrowly about blockbusters, and think more broadly about portfolios of products," says Ravera. That may be the healthiest approach of all. Tsao covers biotechnology issues for BusinessWeek Online. Follow The Biotech Beat every week, only on BusinessWeek Online