FDA Squelches an Article Raising Doubts on Safety Of Device to Repair Artery – WSJ
Fri, 9 Jul 2004
A front page article in The Wall Street Journal exposes another case of safety data suppression by the FDA. The case involves a device to repair the heart artery which FDA’s own experts found to be less safe – i.e, increased mortality when compared to patients who underwent open surgery.
The WSJ reports: “The story of Medtronic’s vigorous lobbying behind closed doors to modify the conclusions of midlevel FDA specialists offers a rare look at the agency handling a complaint by a big company.”
The WSJ reports the company threatened the editors of the Journal of Vascular Surgery with “criminal and civil sanctions” if they didn’t pull an article from the web site. The article was co-authored by Dr. Dale Tavris, head of epidemiology in FDA’s device center. “In the end, the FDA asked the journal to remove the paper” from the site and from publication.
As WSJ reporters note: “The episode also shines a spotlight on a topic of intense debate: What right does the public have to see the information companies collect about their drugs and medical devices?”
Clearly, concealment of data from pediatric SSRI clinical trials is not an isolated case. Senator Grassley and the Finance Committee which is currently investigating FDA’s suppression of Dr. Andrew Mosholder’s report and undisclosed antidepressant safety data, will, no doubt, find this case of interest, as it provides added confirmation that FDA is failing to protect the public from hazardous drugs and devices. The agency has become an agent for the suppression of safety data that confounds marketing goals–its recommendations about drugs and devices are not evidence-based, but are determined by pressure from industry.
Contact: Vera Hassner Sharav
THE WALL STREET JOURNAL
Invasive Procedure After Medtronic Lobbying Push, The FDA Had Change of Heart
Agency Squelches an Article Raising Doubts on Safety Of Device to Repair Artery Threat of ‘Criminal Sanction’
By ANNA WILDE MATHEWS and THOMAS M. BURTON
July 9, 2004; Page A1
In May, an article posted on a prominent medical journal’s Web site raised concerns about the long-term safety record of a device to treat bulges in the main artery leading from the heart. Now, all that remains on the site is the title.
The article, written by Food and Drug Administration researchers, focused on the AneuRx stent graft, a 5-inch metal-and-fabric cylinder designed to prevent deadly ruptures of the arterial bulges, called aneurysms. It suggested that, over time, a rival treatment — invasive surgery — was safer.
The device’s maker, Medtronic Inc. of Minneapolis, sharply disputed the study’s conclusion, reprising arguments from 2003 when it quietly persuaded the FDA to tone down a public notice on AneuRx’s safety. Medtronic objected to the FDA that the authors used confidential data without permission. Its lawyer also threatened the editors of the Journal of Vascular Surgery with “criminal and civil sanctions” if they didn’t pull the article from the Web site. In the end, the FDA asked the journal to remove the paper from the site, and in late June the agency officially withdrew it from publication.
The story of Medtronic’s vigorous lobbying behind closed doors to modify the conclusions of midlevel FDA specialists offers a rare look at the agency handling a complaint by a big company.
The episode also shines a spotlight on a topic of intense debate: What right does the public have to see the information companies collect about their drugs and medical devices? Today, companies reveal important data to the FDA, but the agency typically releases only the information that it thinks is essential for doctors to know. The idea is to avoid confusion or overreactions and protect companies’ proprietary data, which the FDA is generally prohibited from disclosing. Now some critics are calling the system inadequate, citing the example of industry findings about possible dangers of antidepressants to children that weren’t published.
“You can’t make improvements in patient care by hiding information,” says Lazar Greenfield, a vascular-surgery professor at the University of Michigan who co-authored the squelched article during a stint as a visiting scholar at the FDA.
The FDA says in a statement that its decision to withdraw the paper “rested on substantial legal concerns” including “that the paper may have drawn on commercial confidential information, which was submitted to the agency and which the FDA must protect under the law.” The FDA said it will “further review” its 2003 public health notice.
In a statement, Medtronic says it “has a proven and unequaled record” in releasing data on its aneurysm products. The company added that evidence shows the AneuRx is a “safe and effective device.”
The AneuRx (pronounced “an-you-rex”) has annual sales of nearly $200 million. It treats aneurysms in the abdominal section of the aorta, the big pipe that carries blood from the heart. Every year, more than 200,000 Americans are diagnosed with these bulges, and more than 16,000 die when the bulges lead to a rupture of the aorta, experts estimate.
For years, the only treatment option was open surgery, in which doctors open up the patient’s abdomen and sew a synthetic-fabric tube inside the aorta in the area of the bulge. Blood then flows through the tube, which takes pressure off the bulge and makes the artery less likely to burst. But recovery from the surgery can take several days or weeks.
In the late 1990s, medical-device makers developed another approach — small devices called stent grafts that, like the synthetic tubes used in surgery, form a new channel for blood and take pressure off the bulge. They can be slid into the body through a small incision in the thigh, a far easier procedure for frail patients. Medtronic sells about 10,000 AneuRx stent grafts a year; other companies sell another 10,000 stent grafts. About 55,000 people have open surgery for abdominal aneurysms every year.
WORDING OF A WARNING
A squelched article by FDA officials was more specific about the risks of the AneuRx device than the agency’s public notice. . Article by FDA officials: “mortality in patients with an AneuRx … probably exceeds that in patients who undergo an open procedure, within three years or more.”
. FDA public notice: “mortality from the AneuRx Stent Graft is likely to crossover and exceed the … mortality from open surgeryat some point in time.”
In April 2001, two years after AneuRx was approved, the FDA put out a health notice warning of possible problems with the AneuRx and a competing device. One concern was “migration,” meaning that the AneuRx might move from the spot where it was implanted. The notice also cited possible leaks in which blood flows outside the device into the aneurysm, thus leaving a continued risk of rupture.
The notice didn’t answer a crucial question: Do these problems make the AneuRx riskier overall than surgery?
The FDA turned to that question in 2002, preparing a public notice to doctors. Studies and interviews with vascular surgeons indicate that at top hospitals, the short-term mortality rate for surgery is very low. Typically, about 1% to 2% of patients die in such hospitals after the operations. At two highly regarded programs, the University of Michigan and the Cleveland Clinic, the death rate is about 1%. But at community hospitals, the rate is thought to be in the 4% to 6% range.
In an internal 2002 draft of its notice to doctors, FDA officials cited the lower 1% to 2% figure as the most suitable comparison for the AneuRx data being supplied by Medtronic. The hospitals that treated the AneuRx patients included in Medtronic’s data were the same type of high-volume hospitals that had good results in surgery, so the authors of the draft considered this an apples-to-apples comparison.
AneuRx fared badly in the face-off. Its aneurysm-related mortality rate in the initial period after implantation surgery — the period when surgical complications usually arise — was around 1.5%. That was no better than the open-surgery rate cited in the draft. And the draft said long-term mortality from aneurysm-related causes was worse for the AneuRx than for open surgery.
The FDA draft concluded by recommending that AneuRx not be used when a patient’s chance of dying during or shortly after surgery was less than 2%. That would effectively rule out the device for some patients.
A Medtronic consultant quickly fired back. Christopher Zarins, a principal investigator in the original AneuRx trial and the chief of vascular surgery at Stanford Medical School, wrote to the agency on July 29, 2002, questioning the low surgical death rates and warning that the FDA risked “giving patients misleading information.” To get such low rates, the FDA “selectively picked the references that suited their argument,” he says.
Medtronic asked Dr. Zarins and other well-known surgeons to make their arguments to the agency in person. On Aug. 1, 2002, the surgeons met with agency officials at the suburban Maryland offices of the FDA device center. Dr. Zarins offered a 47-slide PowerPoint presentation.
Another prominent doctor, Ron Fairman, chief of the vascular-surgery division at the University of Pennsylvania, was even more blunt. Dr. Fairman, a longtime stent-graft researcher who says he wasn’t a Medtronic consultant or paid to attend the meeting, recalls criticizing the draft’s approach as an “embarrassment” and “unprofessional.” He says the sources cited in the draft, which included a medical textbook and an article in an Italian medical journal, were too narrow or unreliable to produce a trustworthy figure for the general mortality rate from open surgery.
Those arguments didn’t sway the FDA, and a later draft still contained the points that drew the company’s objections. Medtronic brought another group of leading vascular surgeons, including the then-president of the Society for Vascular Surgery, to an FDA meeting on Oct. 30, 2003. Medtronic had helped fund the society’s aneurysm-screening program. Of the other two surgeons present, one was on Medtronic’s scientific advisory board, but the other wasn’t a Medtronic consultant. These experts agreed that the FDA draft presented an unfairly negative picture of AneuRx by exaggerating the safety of open surgery.
An FDA official fought back. Dale Tavris, the head of epidemiology in the FDA device center’s office of postmarket surveillance and a key player in drafting the public notice, asked pointed questions of the experts, according to people who were present. Higher-ranking FDA officials appeared more conciliatory and agreed to re-examine the surgical statistics, participants say. The FDA declined to make Dr. Tavris available for an interview or comment on the meeting, citing its policy of not discussing internal deliberations about regulatory actions. Janet Woodcock, the agency’s acting deputy commissioner, says that it is routine to consult with concerned parties about pending safety notices.
The FDA finally issued its public-health notice on Dec. 17. As in the drafts, it pointed out the long-term death rate from aneurysm-related causes for AneuRx recipients was higher than that of open-surgery patients — about 0.40% per year vs. an average 0.18% per year.
But the final notice no longer highlighted the very low short-term surgery mortality rate of 1% to 2%. Instead, the notice mentioned a broad range of rates. The key recommendation from the drafts — that the devices shouldn’t be used on patients whose risk of dying from the surgery was less than 2% — was gone. In its place, the FDA recommended use of AneuRx in patients who met “the appropriate risk-benefit profile” and listed factors for doctors to consider in making the decision.
The tug of war seemed to have been resolved. Then on May 7 of this year, an article called “Aneurysm-related mortality rates in the US AneuRx clinical trial” appeared in the Journal of Vascular Surgery’s online preview section. Its authors were Dr. Tavris and two FDA colleagues, as well as Dr. Greenfield, the University of Michigan vascular-surgery professor.
It isn’t uncommon for FDA officials to publish papers in academic journals. The agency’s official policy encourages the practice, although it says that articles written as part of a staffer’s official duties must be reviewed for “policy implications” by superiors. Such articles are the property of the FDA, the policy says.
The paper relied on the same Medtronic data used by the FDA in preparing its public-health notice. The paper revived the 1% to 2% open-surgery death rate, saying a low figure was most relevant for comparison because the AneuRx recipients covered in the data were at institutions that treated lots of aneurysms.
The paper concluded that the mortality rate for patients getting the AneuRx probably exceeded that for surgical patients by three years or more after the treatment, because the AneuRx had little advantage in preventing immediate post-surgical deaths and caused or allowed more problems down the road.
Medtronic responded in a letter to the FDA dated May 20, asserting that the paper’s authors were using confidential company data without permission. The letter argued again that the article was underestimating the surgery death rates. Medtronic said it had reached agreement with the FDA about the December public-health notice and the article went “well beyond” the notice, “in some cases reverting to the position[s] in the initial draft … which we … believed were wrong and which were ultimately eliminated in the final version” of the notice.
Medtronic was particularly upset about a remark in the paper’s abstract that the AneuRx death rates calculated by the authors were a “conservative estimate” because of potentially incomplete data. The company’s letter said the remark showed Dr. Tavris and his colleagues were “biased.”
Five days later, an outside lawyer for Medtronic, Edward Basile of King & Spalding, wrote to the editors of the Journal of Vascular Surgery demanding that the journal remove the paper from its Web site. Mr. Basile’s letter said the journal’s “disclosure of … proprietary information and breach of confidentiality protections are subject to both criminal and civil sanctions which will be pursued vigorously if this situation is not remedied.”
On June 8, the journal removed the paper. The decision came because of a request from the authors and was made before the editors received the Medtronic lawyer’s letter, says Jack Cronenwett, a professor at Dartmouth Medical School who is an editor of the peer-reviewed journal.
But e-mails sent by the FDA’s Dr. Tavris suggest the authors didn’t make that decision on their own. In a June 14 e-mail sent to Dr. Cronenwett and others after the journal agreed to remove the paper from its Web site, Dr. Tavris said he didn’t “believe that I have the authority to make any decisions on the matter without FDA permission.” Dr. Tavris also said in an e-mail to a colleague on June 16 he was “not directly privy to the conversations being held on the matter” at high levels of the FDA.
Dr. Tavris also disputed Medtronic’s accusation of bias. “These conclusions were peer reviewed by a panel of subject area experts who are in a position to be much less ‘biased’ than Medtronic,” he wrote in a June 15 e-mail to a colleague. He also argued that the article didn’t contradict the FDA’s public health notice but “expands on the details. … I am prepared to argue any instance where they claim a contradiction between the two.” In an e-mail the next day, Dr. Tavris said his supervisor believed “the Commissioner will decide to pull the article permanently.”
That turned out to be prescient. On June 22, the acting director of the FDA’s device center, Daniel Schultz, called to tell Dr. Cronenwett, the journal editor, that the agency was formally withdrawing the paper. In a letter to Dr. Cronenwett dated June 25, Dr. Schultz wrote that “because the conclusions drawn in the article submitted for publication … go beyond the information provided in the Public Health Notification, the article does not reflect FDA’s current position regarding AneuRx-related mortality.”
Dr. Greenfield, the paper’s co-author at the University of Michigan, says the FDA didn’t inform him that it planned to formally withdraw the article, and he disagreed with the decision.
As of yesterday, the Journal of Vascular Surgery’s Web site still gave the paper’s title on a list of upcoming articles. Dr. Cronenwett says it’s the journal’s policy to leave the title of withdrawn articles on the site for 30 days, along with a note available to subscribers explaining that the article has been removed.
FAIR USE NOTICE: This may contain copyrighted (C ) material the use of which has not always been specifically authorized by the copyright owner. Such material is made available for educational purposes, to advance understanding of human rights, democracy, scientific, moral, ethical, and social justice issues, etc. It is believed that this constitutes a ‘fair use’ of any such copyrighted material as provided for in Title 17 U.S.C. section 107 of the US Copyright Law. This material is distributed without profit.