Pay a visit to the website of The Cancer Genome Atlas, and you'll find this reassuring statement about the samples scientists hope to obtain for the project: “Tissue samples will be carefully catalogued, processed, checked for quality and stored, complete with important medical information about the patient.”

I can't imagine it's going to strengthen people's faith in the NIH or in science. Thomas Murray, Hastings Center

The information will have limits, the website hastens to add, and “samples will be coded to remove any descriptors that might connect a sample with the patient's private information.”

But will those assurances be enough to win donors' confidence? In June, Congressional investigators disclosed a National Institutes of Health (NIH) finding that Trey Sunderland, an NIH researcher, had committed “serious misconduct” when he shipped 3,200 samples of cerebrospinal fluid to the pharmaceutical company Pfizer. Sunderland did so without authorization and without reporting the more than $600,000 in consulting fees he received from the company—$285,000 of it for work related to the samples.

The Congressional report said that the case raises “serious questions” about inadequate oversight and control of human tissue samples at the NIH.

“I can't imagine it's going to strengthen people's faith in the NIH or in science,” says Thomas Murray, president of the Hastings Center, a bioethics think tank in Garrison, New York. But Murray, who in 2004 served on a panel that examined the NIH's conflict-of-interest policies, says the long-term impact may not be significant. The public, he says, may perceive it “as just one isolated, egregious case.”