WE ARE producing too few new drugs. America’s Food and Drug Administration (FDA) approved 53 new drugs in 1996 but only 21 in 2003. Moreover, of the drugs that are launched, too many are recalled on safety grounds: Merck’s recent withdrawal of the painkiller Vioxx, for example, threatens the future of the whole class of Cox-2 inhibitors.
The pharmaceutical industry is suffering: after 20 years as the most profitable out of the 47 industries listed in the Fortune 500, it is now ranked third. Moreover, after a decade of annual growth in sales of around 14 per cent, pharmaceuticals are now looking at only 8 per cent.
The problem lies in unsuccessful research. In recent decades, researchers have searched for new drugs by the traditional route of exploring new sciences (such as DNA). Now that that search is failing, researchers are inverting the conventional model, and they are looking for new uses for old drugs.
Jeffrey Rothstein, a neuroscientist from Johns Hopkins University, Baltimore, organised a consortium of 27 research laboratories, each of which studies some aspect of neurobiology. He then selected, randomly, 1,000 drugs that the FDA had already approved for the treatment of a variety of diseases and asked the 27 labs to test them on experiments they were already undertaking.
To everybody’s surprise, Rothstein found that penicillin is a potentially useful treatment for amyotrophic lateral sclerosis, the neurodegenerative disease from which Stephen Hawking suffers. If trials confirm that it might indeed be a useful neurological drug, it could be introduced clinically with little delay or cost. New uses for other established drugs may emerge similarly over the next decades.
Although Rothstein has conducted a powerful experiment, life for the poor bloody infantry at the bench could not get much more boring than screening 1,000 randomly chosen chemicals on a repetitive assay. There’s an irony here, because the first antibiotic was found in a similar way. In 1910 the German scientist Paul Ehrlich discovered 606 as a “magic bullet” treatment for syphilis, and he named it 606 because it was the 606th randomly chosen chemical that he injected into syphilitic rabbits.
PERHAPS the poor bloody infantry at the bench deserve to be bored because they stick at the level of experiments. In 1662 the Royal Society was founded to “improve the knowledge of natural things, and all useful arts, Manufactures, Mechanic practices, Engynes and Inventions by experiment”. But experiments are a chore, and few of the really important scientists do many of them.
Consider physics. This year we celebrate the centenary of Einstein’s trio of great papers, but Einstein did no experiments. He wasn’t even an academic — he worked as a patents clerk in Bern, and he ruminated over other people’s experiments.
Einstein’s great predecessor, Newton, did, of course, perform experiments, particularly in optics (and alchemy) but he explained that “if I have made any valuable discoveries it has been owing more to patient attention than to any other talent”. He was primarily an observer and thinker.
In biology, the story is clear. The four great biologists of the modern era were Darwin and Wallace (of natural selection) and Watson and Crick (of DNA). Not one of them was an experimenter, each merely observed Nature and other people’s experiments.
The infantry soldier of the DNA story was, of course, Rosalind Franklin, but she was soon forgotten (although she is now being reinvented as a victim and feminist icon) because, although she did the key experiments, she failed to understand their significance. Yet the story has a happy ending because Franklin knew herself not to be a thinker, and she was happy for others to explain her data.
In chemistry, too, the story is clear but less happy. The greatest of chemists was the French aristocrat Lavoisier, who inaugurated modern atomic theory, not by his own experiments but by reinterpreting those of the English prole Priestley. Priestley, who died still believing in phlogiston (that fire is a substance), never forgave Lavoisier.
But we do forgive Lavoisier, because experiments are only one basis of science, not its essence.