Published on Sunday, June 30, 2002 in the Washington Post
A Hospital's Conflict Of Interest
Patients Weren't Told Of Stake in Cancer Drug
by Justin Gillis
One of the nation's largest cancer centers enrolled 195 people in tests of an experimental drug without informing them that the institution's president held a financial interest in the product that stood to earn him millions.
The tests at M.D. Anderson Cancer Center in Houston involved Erbitux, the controversial cancer drug that is at the center of broad investigations in New York and Washington. Most of the patients, who were quite ill by the time they enrolled in the tests, have died.
The cancer center, a unit of the University of Texas system, has since acknowledged that it should have informed the patients of the conflict of interest involving its president, John Mendelsohn. It has recently adopted policies to ensure that patients are told ahead of time if Mendelsohn or the cancer center itself has a financial stake. Ethicists say that such conflicts of interest pose risks to patients and to the integrity of scientific studies.
Erbitux is at the center of an unfolding scandal involving stock trades by executives of the company sponsoring its development, ImClone Systems Inc. One of them was charged with illegal trading. Lifestyle guru Martha Stewart is under investigation for a stock sale she made the day before bad news came out about the drug.
The way Erbitux was tested at M.D. Anderson highlights a different issue: growing conflicts of interest in medical research that could affect any American touched by serious illness.
Genetic research is making possible a burst of medical innovation that promises new treatments, even cures, for serious illnesses such as cancer. But that same research has led to growing ties between academic doctors who invent new drugs and corporations that sponsor their development. What happened at M.D. Anderson, one of the country's most prestigious medical centers, is a case study in the ethical problems such links can pose.
Mendelsohn has been under scrutiny in recent months for his role in two simultaneous business disasters. One involved Enron Corp., the Houston energy trader now in bankruptcy protection -- Mendelsohn served as a director and member of Enron's audit committee. The other involved ImClone, a New York biotechnology company. Mendelsohn invented the company's leading drug, Erbitux, and served on its board.
ImClone is the lesser of the two debacles, but it has caused pain for numerous investors in recent months. ImClone stock, riding high most of last year on the basis of upbeat proclamations from the company, plunged in December after the Food and Drug Administration rejected ImClone's application for approval of the drug, known variously as C225, cetuximab or Erbitux. The FDA cited serious flaws in the company's tests at another facility.
Though Mendelsohn's financial ties to the company and the $6 million he made in an ImClone stock deal last year have been widely reported, his institution's failure to inform cancer patients of those tieshas not been.
Details come from hundreds of pages of M.D. Anderson records, released under Texas public-information law, and from interviews in Texas, New York and Washington.
On the basis of the available material, which does not include patient medical records, there is no reason to believe any patient at M.D. Anderson was harmed by taking the drug, which studies indicate is relatively safe.
At the same time, there is no way to know whether any of the 195 patients who enrolled in Erbitux tests from 1997 to 2001, had they been informed of Mendelsohn's financial stake, might have chosen to take a different drug -- or might have fared better as a result.
In other recent cases, human subjects in flawed research studies have died prematurely, and subsequent questions were raised about whether the doctors involved gave advice biased by their financial interests. But experts say they are unaware of any instance when a university president has held a large, personal financial stake in a drug being tested at his institution.
Mendelsohn, in an interview, acknowledged the potential for conflict between his roles as M.D. Anderson president and a board member of ImClone, but he said he had tried to balance them carefully.
He emphasized that he had played no direct role in caring for patients as they were deciding whether to take Erbitux. And he said his institution had enrolled patients in tests of competing drugs.
Mendelsohn said it was on his initiative, last year, that M.D. Anderson decided to toughen its policies on disclosure so that patients would be formally told of such a conflict. Mendelsohn said he was moved to act by public concern stemming from the cases elsewhere, including a situation involving a young man's death at the University of Pennsylvania.
"I'm not sure it's necessary even today," Mendelsohn said. "But I think you move with the times. I don't want to take any chances that a patient will feel they've been deceived at M.D. Anderson."A Need for Disclosure
Advocates for improved research ethics tend to think, by contrast, that disclosure in such cases is the bare minimum needed to protect patients from harm. Some go farther, saying that conflicts of interest in medical research should be prohibited.
"What's happening is a lot of talk about 'managing' conflicts of interest" -- in part by disclosing them to patients -- "and not prohibiting them," said Marcia Angell, a Harvard University lecturer who is a former editor of the New England Journal of Medicine and a leading voice on the issue. "I think disclosure is better than nothing, absolutely. But there should be a flat-out ban."
One of the first doctors to work on Erbitux at M.D. Anderson was an oncologist named Roman Perez-Soler. He has left the hospital for unrelated reasons and is chief of medical oncology at Albert Einstein College of Medicine in New York. His name has appeared with Mendelsohn's in two scientific reports about Erbitux.
In an interview, Perez-Soler said he believed Mendelsohn had conducted himself honorably in the Erbitux situation, but could nonetheless have handled it better, perhaps by emphasizing to doctors involved that there would be no repercussions if they declined to test the drug or reported negative findings about it.
Perez-Soler stressed that he knew of no specific misconduct in the Erbitux trials, but he said the results of any medical test conducted under such conditions must be viewed cautiously, since the financial interests involved would likely translate into pressure on faculty members to produce favorable results.
When he was asked to get involved in testing the president's drug, "I knew right away this was dangerous territory," Perez-Soler said. "You need a promotion. You need a salary increase. You need another lab. It distorts the normal conduct of things, because you go all the way to try to please the boss."
Perez-Soler said he believed that doctors, and even university presidents, must be allowed to have a financial stake in drugs they develop, but that careful ethical rules need to be worked out to insulate faculty members from pressure and ensure the integrity of research.
A doctor still working for Mendelsohn said he had felt no undue pressure in the case of Erbitux.
James Abbruzzese, chairman of gastrointestinal oncology at M.D. Anderson, tested the drug against pancreatic cancer and reported moderately favorable results, but then turned down a larger study in colon cancer because he didn't like the way ImClone planned to conduct the tests. ImClone executives brought the matter to Mendelsohn's attention, but Mendelsohn said he declined to intervene, and Abbruzzese said he felt no pressure to reverse his decision.
"Really, Dr. Mendelsohn was not involved in any of this discussion," Abbruzzese said. "If I had wanted to walk away from both of these studies, I don't think there would have been any issue raised with him."
As evidence of his even-handedness, Mendelsohn noted that his institution had enrolled more than 300 patients in tests of Iressa, a drug sponsored by another company that is the main potential competitor of Erbitux. Mendelsohn has no financial stake in that drug.
There's continuing controversy among doctors and investors about just how good a drug Erbitux is. The tests at M.D. Anderson found results similar to those at other institutions, namely that Erbitux is mildly effective.
The plunge in ImClone shares came after the FDA raised questions about tests led by another institution, the Memorial Sloan-Kettering Cancer Center, asking if they were skewed in a way that overestimated the benefits of Erbitux. ImClone has acknowledged to Congress that it could have done a better job managing the tests but said it still believes the drug works for some patients.Financial Incentives
The situation involving Erbitux is the latest chapter in a 20-year effort to develop a new type of cancer therapy.
Work in the late 1970s and early 1980s had suggested that many cancers proliferate in response to specific proteins called growth factors. Working at the University of California at San Diego, a young John Mendelsohn and colleague Gordon Sato wondered what would happen if one of the most important, the epidermal growth factor, were blocked by a drug.
With help from the National Cancer Institute, Mendelsohn and Sato devised a drug they thought might work. Other companies, picking up on their theory, did the same, and through many ups and downs, these drugs in recent years have entered the advanced stages of human testing.
Eli Lilly & Co., the drug giant, was the commercial sponsor of the Mendelsohn drug for a while but dropped it. A frustrated Mendelsohn eventually persuaded a small New York company, ImClone Systems, to pick up the project.
Mendelsohn's career flourished in the meantime as he took successive positions at the nation's two most prestigious cancer centers. He worked as chairman of the department of medicine at Memorial Sloan-Kettering, in New York, then left in 1996 to take the top job at M.D. Anderson.
That huge Texas institution, named for a Houston cotton broker, is a unit of the state university system that is of critical significance in the nation's struggle against cancer.
Every year it enrolls some 1 percent of all cancer patients, perpetually rivaling Sloan-Kettering as the largest center. It offers the sickest patients access to experimental treatments that represent their last hope, and it often defines new approaches to cancer care that spread across the country.
At the time Mendelsohn took over as president, he was an adviser to ImClone and held a financial stake in the success of the drug. He had not yet joined the ImClone board.
At M.D. Anderson, as at some other institutions, outside financial ties were once prohibited. But by the mid-1990s these strictures were being lifted across the country, including at M.D. Anderson, under pressure from faculty members who felt they should be able to benefit from valuable discoveries they made.
Mendelsohn said he knew from the outset he would need to keep his financial interests from coloring his judgment as president. At the same time, he said, he felt tests of the drug could go forward if he kept a proper distance. ImClone's representatives "negotiated directly with our contracts office and the investigators involved" to mount tests, Mendelsohn said. M.D. Anderson had no guidelines requiring patient notification of a conflict involving top officers.
It is clear some patients knew of Mendelsohn's role in inventing the drug -- they tracked him down to plead for access to it, and were referred to other M.D. Anderson doctors. But in the formal document patients signed consenting to experimental treatment, they were not informed of Mendelsohn's financial stake.
The theoretical risks in this situation are numerous, according to safety advocates. With hidden financial interests at stake, patients might be pushed toward particular experimental drugs so as to complete those trials quickly, rather than given objective options. "Human research subjects are in short supply, and what the sponsors want is to get as many enrolled as rapidly as possible," Angell said.
Additionally, according to a report by the General Accounting Office, the investigative arm of Congress, the financial interests might create an incentive to play down the risks of a particular drug. And they might also create a motive for doctors with stock options on the line to ignore side effects or massage test results.
"There's a temptation, in these circumstances, to under-report toxicity and over-report the activity of the drug," Perez-Soler said.Institutional Conflict
Serious questions about conflict of interest were raised in two recent cases of national import. Jesse Gelsinger, 18, of Tucson, died at the University of Pennsylvania in 1999 after volunteering to test an experimental treatment designed to help babies who shared his genetic ailment, ornithine transcarbamylase deficiency.
James Wilson, director of the center where Gelsinger died, had a direct financial interest. It later emerged that Wilson and other doctors had ignored serious danger signals when they proceeded with the test that killed Gelsinger.
Unlike in the M.D. Anderson instance, some of the financial interests were disclosed to the Gelsinger family. But still, serious questions were raised after Gelsinger's death about whether he got unbiased medical advice, and the University of Pennsylvania settled with his family for several million dollars. "Looking back, I can see that I was very naive to have been as trusting as I was," his father, Paul Gelsinger, told Congress.
Similar issues are being debated in Seattle, where the Seattle Times recently questioned tests that the newspaper alleged had killed patients prematurely at the Fred Hutchinson Cancer Research Center. Some of the doctors who designed those tests had a financial stake in the outcome.
Mendelsohn, in an interview at his office in Houston, said he had been aware for many years that financial interests can cloud a doctor's judgment. That's why he recently instituted a policy at M.D. Anderson,he said, that no doctor directly involved in clinical care of a patient can have a financial stake in an experimental drug being offered to that patient.
Policies like that have been spreading nationwide in light of the Gelsinger disaster, but they are far from universal.
Just as the academic world is groping toward policies for individual doctors, a new type of conflict -- the type that Mendelsohn exemplifies -- has gained increasing attention. This is "institutional conflict of interest," or a circumstance in which an institution or one of its senior officers holds a financial interest in a drug that institution plans to test.
More and more, universities are filing patents on key discoveries and licensing those to drug companies, giving the university a potentially lucrative stake. And, as the Mendelsohn case shows, discoveries researchers make early in their careers can create conflicts after they have advanced to the top ranks.
Professional bodies have only lately begun studying the issue. "On the institutional side, we're way, way farther back in our work nationally," said David Skorton, a vice president at the University of Iowa who is active in the issue.
The GAO has complained repeatedly that the federal government has been slow to improve standards. Such work is underway, said Greg Koski, director of human research protection in the Department of Health and Human Services, but no national consensus has emerged on exactly how far to go in restricting institutional conflicts.
"I don't think anyone at this point is entirely convinced that they have all the answers," Koski said.Separate Roles
After he had been at M.D. Anderson for two years, Mendelsohn was asked to join ImClone's board, deepening his financial involvement.
He sought legal advice from M.D. Anderson's attorney, who approved the arrangement based in part on Mendelsohn's claim that he had kept his two roles separate. "I understand that you have strictly avoided any participation in the research being conducted at the institution sponsored by ImClone," attorney Dan Fontaine wrote in a Feb. 6, 1998, memorandum.
There are questions, however, about whether Mendelsohn continued to do so after joining the ImClone board. He has appeared as an author on 11 scientific papers and abstracts reporting research on Erbitux that were published after the date of the memo, and four of those involved research on human subjects that was conducted partly or wholly at M.D. Anderson.
Mendelsohn said the papers in question reported studies that he helped design before moving from Sloan-Kettering to the presidency of M.D. Anderson. He said the work took considerable time to complete and write up for publication, which he said explains why some of it appeared as late as 2001. He said he has had no recent involvement in the research and did not feel he had violated the commitment implicit in the Fontaine memo.
"I just have gotten out of it," Mendelsohn said.
However modest the Erbitux results, as the drug came up for FDA consideration in 2001, glowing articles appeared about it in Business Week and other publications. ImClone shares rose 94 percent that year to peak at $73.83 in December.
In the midst of the optimism, ImClone cut the biggest deal of its type ever with drug giant Bristol-Myers Squibb Co., which agreed to invest more than $2 billion for a stake in ImClone and some rights to the drug. One beneficiary was Mendelsohn, who earned $6.3 million when he sold a stake to Bristol last fall in an offer open to all ImClone shareholders.
But then disaster struck. ImClone announced on Dec. 28 that the FDA was throwing out its application for early approval, citing flawed tests. Subsequent disclosures raised serious questions about whether company insiders, including board members, knew as far back as the time of the Bristol tender offer, in October, that bad news was coming. Unlike some company insiders, Mendelsohn reported no stock sales in December, and he retains a significant stake in ImClone.
ImClone shares have plummeted 88 percent, closing Friday at $8.69. Extensive investigations are underway in Washington and New York, and a slew of corporate executives and FDA administrators testified about the matter on Capitol Hill on June 13.
Samuel Waksal, who was chief executive of ImClone during the events of December, has been arrested on charges that he tipped family members to sell their ImClone shares just ahead of the bad news. A related investigation is focusing on whether he, or anyone else, tipped Stewart to sell her 4,000 shares in the nick of time.
Mendelsohn said he had never sold stock based on inside information, and he stressed that he had always been up front about his financial stake. Only after Erbitux had become embroiled in controversy, however, did most faculty members of the M.D. Anderson Cancer Center receive word that from now on, they would have to inform patients in writing of Mendelsohn's special connection to Erbitux.
Now the institution is wrestling with an additional issue: Even when patients are told of a financial interest that might influence the medical advice they get, they don't necessarily understand that what is being disclosed to them is an extra risk to weigh.
"When they find out their doctor is the person that invented something, they think that's just sliced bread," said Leonard Zwelling, Mendelsohn's vice president for research administration. "They say, 'I've come to the right place. This is the best I could hope for.' "
© 2002 The Washington Post Company