To the Editor:

The articles by DeFrancesco in the June issue (Nat. Biotechnol. 22, 663–664, 2004) and Miller in the August issue (Nat. Biotechnol. 22, 944, 2004) discuss Merrill Goozner's book entitled The $800 Million Pill: The Truth Behind the Cost of New Drugs. At every partnering or financing meeting, at least a dozen biotech presentations or conversations will refer to the Tufts Center for the Study of Drug Development figure for the average cost of developing a new drug. The latest Tufts study, dating back to November 2001, puts this figure at $802 million. With subsequent inflation and the natural tendency of biotech executives to round upward, this usually comes out as “just under a billion dollars.” Shareholders should remove, with immediate effect, any CEO of a biotech company that is working on a drug that costs that much to develop.

The $802 million figure applies to the kinds of drugs that were surveyed in the Tufts Center study. That study looked at 68 randomly selected new drugs from ten companies. Few biotech companies could muster an average of over six drugs each, so we conclude that the numbers did not come from typical biotech companies.

Most of the money that goes to make this average figure is spent running long or large late-stage clinical trials. The billion-dollar drugs will either be 'me-too' compounds, where a large patient cohort is required to reveal the marginal advantage of the compound over the current market incumbent, or they will be novel drugs for chronic conditions where long-term trials across a broad section of the population are required to show prolonged safety and efficacy.

For pharma companies, it is a valid commercial strategy—if not an entirely commendable one—to develop 'me-too' compounds to defend their franchises in particular indications. Pharma companies can also afford to commit huge resources in search of treatments for chronic conditions.

This is not for biotech companies, however. Investors will never give them enough money to allow them properly to leverage the value of compounds that require extended trials. Instead, biotech companies ought to concentrate on drugs that are both differentiated from current market offering and address acute medical needs.

The good news is that most biotech companies recognize this. They do focus on novel compounds, and they do work in indications like cancer and infectious disease. They often work in niche indications where the medical needs are more pronounced. Any longer term projects, such as therapies for metabolic and autoimmune disease, are out-licensed early in development.

Consequently, the drug development cost for biotech companies would be much lower than the Tuft estimate, perhaps by almost an order of magnitude. But this doesn't seem to stop biotech CEOs quoting the $802 million figure.