The pharmaceutical industry is addicted to developing new drugs. Its patents last only 20 years; after that, generics take over the market and companies start jonesing for their next moneymakers. (The patents for two of the most profitable drugs in the US, Amgen's Epogen and Johnson & Johnson's Procrit, expire this year.) But inventing a medication can take more than a decade, and much-hyped advances in robotics and automated chemistry haven't revolutionized drug discovery as promised. Anxious corporations are still chasing their next high. Here's a look at their search methods.
Serendipity
Find a new use for a failed drug or take advantage of lab "accidents" and unexpected results.
Pitfall: Dumb luck is not a reliable business plan.
Successes: Alexander Fleming discovered the first antibiotic - penicillin - in 1928 when mold contaminated petri dishes in his lab and killed bacteria. Viagra failed as a heart medication, but study subjects swelled with joy at its side effects.
Bioprospecting
Comb oceans, rain forests, and other habitats for organisms with medicinal properties.
Pitfall: Biologically active molecules often have negative side effects in humans.
Successes: Several cholesterol-lowering statins are derived from the fungus Aspergillus terreus. The cancer drug Taxol originally came from the bark of Pacific yew trees. (It's now synthetic.)
Combinatorial libraries
Mix a bunch of chemical building blocks into thousands of molecules; use robots to sort through the hodgepodge for compounds with potential.
Pitfall: Early efforts have produced molecules that can't be absorbed or processed by the body.
Successes: None yet; researchers are fine-tuning the approach and teaming it with other tools.
Rational design
Build drugs that attach to molecules associated with a given disease, using information about structure and function.
Pitfall: It takes lots of expensive computing power and can't predict whether the drug will work.
Successes: HIV protease inhibitors arose from studies of an enzyme used by the AIDS virus. The cancer fighter Gleevec blunts a signal specific to leukemia.
Biopharmaceuticals
Find enzymes, hormones, and antibodies that could act therapeutically.
Pitfall: The body destroys such molecules very quickly, often before they can be effective.
Successes: Erbitux (the ImClone/Martha Stewart hot potato) is an antibody that flags tumor cells for death. Vitravene, a snippet of genetic material, blocks a virus that causes vision problems in AIDS patients.
Genomics
Search newly decoded human genes and corresponding proteins for those that cause or prevent disease.
Pitfall: Just because you identify the gene doesn't mean you also know what it does or how to alter it.
Successes: None yet, but researchers are already using genome data to track rare side effects or identify which patients will benefit from which therapies.
- R. John Davenport
credit: Illustrations by Stephen Rountree
START
The Shoe With the 20-MHz Brain
Psssst. Get Microsoft Office. Cheap.
How Big Pharma Finds Its Next Fix
The Racer's Edge: Backseat Driving at 200 mph
