The Lab

12 Dec 1980

Washington, 1980

The government paid for research and held the patents, and few were licensed: fewer than 5 percent of 28,000, GAO later reported, against 25 to 30 percent where a company had been allowed to keep title. In December 1980 Congress passed the Bayh-Dole Act. A university may keep a patent on federally funded work, and it must share royalties with the inventor, so a professor’s share of the profit is part of the design.

Fewer than 5%Of 28,000 federal patents had been licensed when Congress acted in 1980, as the GAO later reported

2005

Stanford

The law did not start this. Stanford’s first Cohen-Boyer patent on cutting and splicing DNA was filed in 1974 and issued in December 1980; ten of its 468 licensees paid 77 percent of the licensing income, and Stanford and the University of California earned $255 million by 2001. In 2005 Stanford’s licensing office reported turning the equity from its license to Google into $336 million. Both are universities’ income, set out in a scholars’ account and Stanford’s own report; neither source involves a relative.

25 Aug 2011

The rule

Since 1995, federal health-research rules have counted an investigator’s spouse and dependent children when they measure outside financial interests. The 2011 rewrite set a $5,000 floor, and asked institutions to make the conflicts of key staff public on a website or in writing within five business days. It declined to reach more distant relatives or friends. A father-in-law or a girlfriend would fall outside it.

13 Sep 1999

Philadelphia, 1999

Jesse Gelsinger, 18, joined a University of Pennsylvania gene therapy trial led by James Wilson. He died on 17 September 1999, four days after his dose, of a severe immune reaction to the viral vector, CMAJ reported. Wilson later said he believes a rare antibody effect was behind it, and that it cannot be proved. The form named Wilson and the company Genovo in one sentence; his parents said they should have been told more about the financial ties.

2000

The company

Wilson founded Genovo in 1992. It financed his institute, by sources’ counts more than $4 million over five years, or more than $4 million a year, or nearly a quarter of the budget, and held the rights to license patents from his lab. Penn held shares, 3.2 percent by one account, and a Harvard case study, citing Steinbrook and the Wall Street Journal, says Wilson and his family held 30 percent, non-voting. We could not open those two. This is the nearest documented case of a lab’s work going to a company its professor founded.

11 Feb 2002

Their answers

The FDA wrote to Wilson on 30 November 2000 that he “failed to adequately protect the safety and welfare of subjects,” and in 2002 rejected his explanation. Penn’s institute had answered in February 2000 that every patient gave consent, that data on two patients came late but were with the FDA for months before the sixth group, and that no causal link to Gelsinger’s ammonia level was shown. In 2000 Penn acknowledged weaknesses in monitoring but said they “did not contribute to Jesse’s death.” Wilson denied that money drove his research. Gelsinger’s father called Wilson “their scapegoat,” and the family’s suit was settled for an undisclosed sum.

9 Feb 2005

No admission

In February 2005 the U.S. Attorney announced civil settlements. It alleged the study produced toxicities that should have ended it, that reports to the FDA, the NIH and review boards misstated the findings, and that consent did not disclose all anticipated toxicities. Penn paid $517,496 and Children’s National Medical Center $514,622. No one admitted the allegations, and Wilson agreed not to sponsor an FDA-regulated trial for five years.

1992

Cambridge to Moscow

This case is not about selling research. In 1992 a Harvard institute signed a USAID agreement to support Russia’s reforms. The economist Andrei Shleifer was Project Director, and from 1995 the lawyer Jonathan Hay ran the Moscow office as General Director. USAID suspended the project in May 1997 over conflict-of-interest concerns and ended it on 6 June.

28 Jun 2004

The ruling

In June 2004, on motions and without a trial, Judge Douglas Woodlock held Hay liable on three False Claims Act counts, and Shleifer on one, conspiracy. Shleifer won another count, a third stayed open, and every defendant won on common-law fraud. The judge did not decide whether the rules bound Shleifer himself. Harvard was held liable for breach of contract only. The court found the men’s own money in Russian stocks, some in the name of a father-in-law, and $200,000 loans each to a fund company run by Hay’s then-girlfriend. The judge held the project rules did not bar investment in Russian government debt, and all claims against Shleifer’s wife and the girlfriend had been dismissed.

Aug 2005

Settled

The government also said FFIA, an investment firm partly owned and run by Shleifer’s wife, used project resources to found a Russian bond-trading company. FFIA denied it, called the claims “completely without merit,” and paid $1.5 million in 2004 as a “purely business decision.” In August 2005 Harvard agreed to pay $26.5 million, saying the court had found no institutional wrongdoing, and Shleifer $2 million; Hay paid by his income, $500,000 to $2 million by one count. No defendant admitted liability. Shleifer said his legal fees would have passed what he paid, and Hay’s lawyer that nothing argued questioned the value of the work.

4 Jun 2008

Boston, 2008

Senator Charles Grassley asked Harvard and Mass General for the conflict-of-interest forms of three psychiatrists. In June he told the Senate that, after the hospital asked them in March to look again, Biederman and Timothy Wilens each reported over $1.6 million from drug companies and Thomas Spencer over $1 million. Eli Lilly told him it had paid Biederman tens of thousands of dollars he had not accounted for; Grassley said in most cases the doctors reported less than companies did, and in a few more. Grassley said Harvard and Mass General had been extremely cooperative.

$1.6 millionWhat Joseph Biederman reported from drug companies after Harvard asked him to look again. The three doctors’ first forms, Grassley said, suggested a couple of hundred thousand dollars in seven years

Jul 2011

Harvard’s answer

In a letter to colleagues in July 2011, the three said Harvard Medical School and Mass General had completed a review and found they had violated conflict-of-interest policies. Each was to stay out of paid industry activities for a year, then seek approval for two more, and face a delay in promotion. The three called their errors “honest” mistakes made in good faith. Biederman died in January 2023.

22 Dec 2008

Atlanta, 2008

Emory found that Charles Nemeroff took in more than $800,000 from GlaxoSmithKline for over 250 talks, from 2000 to 2006, and did not report it. He had believed educational talks were exempt, Emory’s policy said otherwise, and he stepped down as chair after 17 years. He said he regretted the failure of full disclosure and had acted in good faith, and Emory’s review of his slides supported his claim the talks were not product-specific. Nature reported in May 2012 that a Justice Department inquiry was still open; we found no later report of how it ended, or of any charge.

13 Sep 2018

New York, 2018

José Baselga, chief medical officer of Memorial Sloan Kettering, resigned after ProPublica and The New York Times reported he had left company payments out of journal articles. He said the lapses were unintentional and that he would correct 17 articles. In December a cancer-research society’s panel concluded the omissions were inadvertent and let him keep publishing, yet asked him to leave his editor’s post for not meeting its standards; the New England Journal called it a “breach of trust.” He later led cancer research at AstraZeneca and died in March 2021.

60%Share of nearly 180 papers by José Baselga since 2013 with no industry ties reported, by ProPublica’s and the Times’ count

21 Jun 2023

A husband and wife, 2023

We searched for the founder’s case directly, a professor and a spouse with a company. In Oklahoma, a chemistry professor and his wife formed and controlled a firm that received Energy Department small-business technology grants and spent the money on unrelated things, including personal expenses, the U.S. Attorney said. Each pleaded guilty to a false-statement charge and was sentenced to 27 and 14 months. The release says nothing of a university lab’s research being sold to the firm.

$2.1 millionRestitution ordered to the U.S. Department of Energy from a University of Oklahoma professor and his wife, 2023

4 Oct 2013

Two that ended without a finding

At MD Anderson in Houston the president’s wife co-directed a project that won a $20 million state grant outside the usual approval; a University of Texas System review found no conflicts of interest related to it. The two had also co-founded a drug company, and he left its board and put his investments in a blind trust. At Missouri a colleague’s suit said two professors paid their wives from federal grants; the Justice Department did not pursue it and the university called it unfounded. The AP reported that soon after, a campus office moved nearly $60,000 in state funds to replace federal grants paid to one of the women.

2003

What studies count

Eight cases are anecdotes, not a rate. In 2003 a Yale team reviewed 37 studies. About a fourth of investigators had industry affiliations. Across eight articles covering 1,140 studies, industry sponsorship went with pro-industry conclusions, with an odds ratio of 3.6. The authors concluded such ties “can influence biomedical research in important ways,” which is an association across studies, not a finding about any one lab.

About 1 in 4Investigators with industry affiliations, in a 2003 review of 37 studies. Roughly two thirds of institutions held equity in start-ups that sponsored research at the same place

2017

A larger count

A Cochrane review of 2017 drew on 75 papers. Industry-sponsored studies reported favorable conclusions 1.34 times as often, on low-quality evidence, while harms results were similar, on very low-quality evidence. The authors said standard checks on study design did not explain the gap. It compares who sponsored studies. Its abstract does not report on stakes held by individuals or their relatives.

1.27How much more often industry-sponsored drug and device studies reported favorable efficacy results than other studies (95% confidence 1.17 to 1.37)

2011

The other side

Scholars count over US$35 billion in sales of an estimated 2,442 products made with recombinant DNA. A 2011 study found 153 approved drugs, vaccines and new uses traced to public-sector labs. A former Harvard Medical School dean, Jeffrey Flier, said industry ties are essential to developing new drugs, ProPublica reported. These counts show where discoveries came from, not whether the 1980 law changed it.

153New FDA-approved drugs, vaccines or new uses discovered through public-sector research over 40 years, in a 2011 study

2023

The test

Do labs sell research to relatives’ companies? In what we read, the nearest case is Genovo, a company a professor founded, which held rights to his lab’s patents and in which, by one account, his family held stock. Penn denied money played any part, and the government’s 2005 allegations do not mention the stake. The Russia case involved family investments in an aid project, and the Oklahoma couple’s own firm took federal grants. Two spouse cases ended without a finding. We found none in which a court or university found a lab selling research to a relative’s firm. That is the limit of what we read, not a count; no study here counts how often it happens.