Sir

In response to your Editorial “A changing drug supply” (Nature 445, 460; 2007), W. Ross Tracey points out in Correspondence that niche drugs may be just as expensive as blockbuster drugs to bring to the market, but a pharmaceutical company's revenue from them is likely to be only a fraction of what it can earn from selling blockbusters (“Niche drugs aren't a cheap alternative to blockbusters” Nature 445, 818; doi:10.1038/445818c 2007).

This is not necessarily so, in my opinion. Many drugs fail in clinical development, and some drugs have to be withdrawn after causing severe side effects in some patients. Once it becomes possible to determine what causes these patients to show adverse reactions, it may become feasible to pre-select a smaller patient population that will only respond positively and not suffer any adverse events. Although this strategy may not deliver blockbuster drugs, it could considerably reduce the present high attrition rate in drug development and thus make niche drugs economically viable.

Another point is that the US Food and Drug Administration uses financial incentives and an accelerated review process to support the development of drugs for 'orphan diseases' — conditions so uncommon that drug companies would otherwise have no incentive to seek cures for them (see the Orphan Drug Act, http://www.fda.gov/orphan/oda.htm).