FIRM THAT PAID FOR UCSF STUDY SEEKS DAMAGES

Furious at report that AIDS drug came up short

By Sabin Russell

San Francisco Chronicle 1 Nov. '00

A Southern California drugmaker is seeking millions in damages from the University of California at San Francisco after university researchers insisted on publishing a report showing that the company's experimental AIDS drug did not work as planned.

The dispute has spilled over into a national debate about the inherent conflicts of interest found in the financial web tying private business to independent university research.

Immune Response Corp. of Carlsbad, San Diego County, initially attempted to block UCSF researchers from publishing data in the Journal of the American Medical Association showing that the company's AIDS drug, Remune, failed to slow the progression of AIDS in a test involving 2,500 patients.

The report is being published intact, however, in today's issue of the prestigious medical journal, along with a series of articles commenting on the report and looking at unrelated conflict of interest issues.

"Our decision to publish this study is based on the belief that the integrity of the research process must be protected and preserved," wrote the journal's editor, Dr. Catherine DeAngelis, in an editorial accompanying the piece.

Dr. James Kahn, a UCSF AIDS clinician who was principal investigator for the drug trial, said the drug company was attempting to force its own spin on some bad news. "The question we posed in the study is, 'Does it work?' and the answer is, 'No, it does not,'" he said. "There is no ambiguity here."

Immune Response maintains that UCSF improperly used its data and published its report based on incomplete information. It is seeking between $7 million and $10 million through an arbitration proceeding that could be resolved in several months.

Company executives concede that their drug did not work in the two-year clinical trial, but they contend that Kahn is ignoring findings within that same study suggesting that Remune was having a beneficial effect that could be proven over time.

Remune was developed originally by the late Dr. Jonas Salk, inventor of the polio vaccine. Made from killed copies of a strain of the AIDS virus, the drug was designed to stimulate the immune system of patients already infected with the virus.

At issue is an Immune Response analysis of data on 10 percent of the patients whose blood was tested more frequently. The company contends those results show the drug reduced the level of virus in patients' blood -- which other research depicts as closely correlated with better health.

"The crux of the matter is that we wanted this 250 patient cohort included," said Dennis Carlo, Immune Response's president and chief executive.

Kahn said that the data from the 250 patients was included in his report, but his analysis of it showed there was no significant effect. "Here is a company manipulating data to try to have a positive outcome," Kahn said. He described it as a well-known tactic known in research circles as "data dredging."

The major findings of the study -- that Remune did not halt the progression of HIV infection to AIDS or reduce AIDS deaths -- have been known since a review board halted the clinical trial in May 1999, more than two years after it began. A peek at interim data showed the drug was not working.

The dispute with Kahn broke out shortly afterward, when the company put out a press release which stressed the findings of the smaller cohort. Kahn strongly disagreed with the company's optimistic spin, and relations with Immune Response deteriorated to the point where he said the company would no longer share information with him.

"Like a bully in a sandbox, they pulled their data and said, 'We're outta here,'" he said.

Some investigators who worked on the Immune Response trial say, however, that it is the UCSF researcher who is being heavy-handed.

"I am stunned. I was floored when I found out they were coming out with a paper about which I knew nothing," said Dr. John Turner, a Philadelphia AIDS clinician who said he enrolled the largest number of patients in the trial. Turner said Kahn never shared his manuscript with him.

Turner said he believes the data on the smaller cohort clearly show some important effects.

"You don't selectively report something to make a drug look ineffective," Turner said. "I'd hate to feel that was done here, but there's not much of any other way to interpret it."

Editors at the medical journal are siding with UCSF, contending that such interference violates the independence of academic research. "We thought it was good work, and very important to have published," said Dr. Drummond Rennie, a professor at the UCSF Institute for Health Policy Studies and also a deputy editor of the medical journal.

Rennie said it is the "duty" of academics to publish the results of their research, whether they are positive or negative. "If things come out badly, you've got to publish it. That's just life," he said.

Kahn's study is accompanied by several articles in today's issue of the medical journal that focus on the question of conflict of interest in academic research. Rennie himself is the co-author of one report, which examined written conflict of interest policies at 89 institutions which receive the highest amount of funding from the National Institutes of Health.

Rennie's study found "wide variation" in the policies of those institutions. "Most lack specificity about the kinds of relationships with industry that are permitted or prohibited," he said. "When you just say academics should do good, that isn't good enough."

Only 55 percent of the institutions studied required conflict of interest disclosures of all faculty, while the rest required disclosures only of principal investigators or those conducting research.

Another report in the journal, authored by UCSF Institute of Health Policy researchers Elizabeth Boyd and Lisa Bero, made a case study of UCSF's own conflict of interest policies. It found that in 1999, 7.6 percent of faculty investigators reported personal financial ties with sponsors of their research, compared to 2.6 percent in 1985.

The financial ties varied widely. One third of those disclosing ties during a 20-year period (1980-1999) involved income from speaking fees, which ranged from $250 to $20,000 per year. Another third had consulting contracts that paid between $1,000 and $120,000 per year. The remaining third held positions on the board of directors, or scientific advisory boards, of the sponsoring companies.

One in seven researchers reported holding stock in the sponsoring companies, and 12 percent had more than one financial relationship with the sponsors.

Author Bero said that 26 percent of conflicts of interest at UCSF were deemed significant enough to require university intervention, such as requiring the investigator to sell stock, refuse additional payment for speeches, resign from a board, or require that the sponsor name a different principal investigator.

---

TIES THAT BIND

An analysis of financial disclosures from 225 UCSF researchers over a 20-year period (1980 to 1999) found a wide variety of financial ties to sponsoring companies. The brakdown below shows the nature of those ties:

Speaking fees: 34% ($250 - $20,000 a year)
Consulting fees: 33% ($1,000 - 120,000 annually)
Paid board position: 32% (board of directors or advisory board)
Stock ownership: 14% (average value under $100,000)
Multiple interests: 12% (e.g., stock ownership and board seat)
Patent ownership: 2%
Outstanding loans: 1%

Data are based on 488 reports from 225 individuals. Percentages are greater than 100 percent because data are not mutually exclusive. Source: Journal of the American Medical Association.