Drug Trials With a Dose of Doubt

But after hours of speeches and slides, the committee members, appointed by the U.S. Food and Drug Administration, had yet to vote. The members were focused on the quality of Merck's case for the new drug, which rested on the treatment of only 69 patients.

Merck summoned to the microphone one of its announced consultants, a man whose government job was nearby, at the National Institutes of Health. Dr. Thomas J. Walsh assured the committee that Merck's data describing the patients was "extremely robust and very, very rigorous." He said his government staff had assisted in vetting the company's data. About 30% of the patients were helped by the drug, he said.

The advisory committee voted unanimously to endorse the drug, called Cancidas. Sixteen days later, the FDA approved it. Doctors would later prescribe it for patients whose immune systems had been ravaged by chemotherapy and who were presumed to have a potentially deadly, invasive fungal infection. In its first five years on the U.S. market, Cancidas would generate $859 million in sales for Merck.

U.S. law generally prohibits a federal employee from representing an outside party before a government agency.

In building a career as an influential government scientist, Walsh has served as both a paid and unpaid advisor to pharmaceutical companies and has helped lead clinical trials that tested the effectiveness of their products. With his help, the companies have brought new antifungal drugs to market, but controversy has flared over whether results from two of the studies were misleading and whether some of the participating patients received adequate treatment.

In written comments for this article, Walsh said his advice to industry did not conflict with his position at the NIH's National Cancer Institute, or affect his scientific judgment.

"I am not and have never been a representative of, or advocate for, any pharmaceutical company," Walsh said.