Memory Hole (21 October): Vioxx and a quacking duck

8 years ago today: A story of ducks and a single death in a Vioxx clinical trial

On 21 October 1999 a 73-year-old woman died. That death and our response to it tells volumes about the health of our profession.

In medicine we sell our wares under the banner of science. The technique of science (the "scientific method") consists of a number of steps. For a double blind randomized clinical drug trial those steps might be:

1. State a meaningful hypothesis
2. Design a study and statistical methods to test it
3. Do the study in a double blind randomized way
4. Unblind the study AFTER deciding what happened to each patient and locking the database
5. Study authors analyze the results
6. Share the results with everyone

That is the usual way of science. This method is not used by the practitioners of any system of quackery.

So that takes us to the Merck Advantage trial of Vioxx, and a single 73-year-old female participant in that trial who died on this day - Oct. 21, 1999. She died a few minutes after calling her son to tell him she felt short of breath.

The Advantage trial was a very short (12-week) clinical trial in 5,500 patients. If you want to avoid showing up long term side effects, you keep your clinical trials short and scientifically irrelevant (that's about the meaningful hypothesis part).

Now eight people taking Vioxx suffered heart attacks or sudden cardiac death in the Advantage trial, compared with just one taking naproxen, according to data released by the F.D.A. in 2005. The difference was statistically significant, but Merck never showed the data that way.

In the published study, Dr. Lisse (University of Arizona) reported that five patients taking Vioxx had suffered heart attacks during the trial, compared with one taking naproxen, a difference that did not reach statistical significance. But the paper never mentioned the other cardiac deaths of patients taking Vioxx, including the 73-year-old woman.

Dr. Lisse said that while he was listed as the paper's first author, Merck actually wrote the report (that's the authoring bit). "Merck designed the trial, paid for the trial, ran the trial," Dr. Lisse said. "Merck came to me after the study was completed and said, 'We want your help to work on the paper.' The initial paper was written at Merck, and then it was sent to me for editing."

Now, the other thing in usual science is not to allow company technocrats with a vested interest to decide themselves what happened to individual 73 year old women after the study is unblinded.

Neither patients nor their doctors were aware whether patients were receiving Vioxx or naproxen. Results from the trials were also supposed to be blinded when they were examined, so that Merck researchers could not bias the results. After examining the case, Dr. Eliav Barr, a Merck employee judged that the woman had probably died of a heart attack."

"Common things being common, the clinical scenario is likely to be MI (heart attack)" Dr. Barr wrote in an E-mail message in November 2000 to Dr. Reicin, the Merck clinical research executive.

Dr. Reicin quickly shot back: "I think this should be called an unknown cause of death." A few hours later, she wrote, "I would prefer unknown cause of death so we don't raise concerns."

"A fatal event not being treated as objectively as possible surprises me," said Dr. Myerburg, a prominent cardiologist. "It looks like they're playing around with this death."

In November 2001, an F.D.A. reviewer who had examined all the company's data concluded in a report to other agency officials that Merck had misclassified the death. But the F.D.A. never told anyone either.

Dr. Lisse, the study "author" said he had never heard of the case of the woman who died, until told of it by a reporter.

So can we really believe anything at all, and how to distinguish this from the worst type of quackery? As the old saying goes, ...If it looks like a duck, and it quacks like a duck...it's a bloody duck.

What is even more amazing is that the leadership of medicine says nothing -- nothing at all. Medical students learn none of this in any of their pseudo-ethics training.

Sources:

1. Correspondence in the Annals of Internal Medicine on the problem: Deeply disturbing evidence of medicine run amok
2. New York Times, April 24, 2005: "Evidence in Vioxx Suits Shows Intervention by Merck Officials"

Earlier|Later|Main Page

Memory Hole (16 October): Merck probed over "agreement" with FDA and data analysis

3 years ago today: Probe of Merck "arrangement" with the FDA and data analysis

On 16 October 2004 the Financial Times of London and the Washington Post reported on a probe of the FDA and its relationship with Merck. Vioxx had been withdrawn two weeks previously. The FT discussed a secret deal with US regulators that made Merck privy to government studies on its drugs. An FDA E-mail, suggested regulators discussed a notification agreement with Merck. In a May 14 2002 e-mail, Ann Trontell, FDA deputy drug safety director, warned colleagues that a Merck official had reminded her that "there had been an agreement that Merck would be informed prior to any FDA publication about one of their drug products".

The Washington Post reported that "Merck officials faced a difficult decision about how to handle the catastrophe". "The whole saga, industry experts said, raises unsettling questions about aggressive consumer marketing of drugs before their long-term safety has been proven."

"researchers had been warning about the drug's possible cardiovascular risks since 2000". "Data from a company study found then that users had four times as many heart attacks and strokes". Merck "repeatedly reassured the medical and financial communities that Vioxx was safe". Eric Topol said "Why didn't they stop the DTC [direct to consumer] marketing?". "That's the tragedy here."

"The FDA safety officer in charge of the report, David Graham, concluded that Vioxx posed much greater cardiovascular risk than the other major drug in its class" In an E-mail "a Merck official complained that the FDA had not lived up to a prior agreement to alert the company before releasing any negative information about its products."

Clinical trials include independent safety monitors. "The four monitors, along with a Merck statistician were the only ones with full access to information about the trial, including which patients were getting Vioxx and which were getting the placebo."

(For a glimpse into the way in which data and analyses can get combobulated under such incestuous high-stake conditions a single example will suffice. Therein lies a great deal of the problem.)



Merck is I believe one of the more ethical pharmaceutical companies. Who sets the standards?

Sources: Financial Times, 16 Oct 04, Merck under spotlight for 'deal' with US drug regulators; Washington Post, 16 Oct 04, Painful Withdrawal for Makers of Vioxx

Earlier|Later|Main Page

What have they got to hide? The Famciclovir case

Pending further description of methodology in Sheffield Actonel studies, my attention was drawn to a similar scandal in another parallel scientific universe. It involves the drug Famciclovir (Famvir) and is discussed at Pharmawatch and Healthcare Renewal.

Famvir is used for treatment of herpes zoster, and genital/labial herpes. The case involves a nine year delay before partial publication of findings which had adverse implications for the sponsor. As in the Actonel case, it also involved protracted denial of raw data to academic authors. The study and editorial were finally published a few weeks ago (1,2).

The authors relate that (1):

"....companies are understandably wary of devoting resources to a report that is not in favor of their product. Although we received data tables and listings for this study, we were not able to obtain data tapes to verify the analyses and to conduct what we consider the most clinically relevant analyses"

Dr Wald, the first author, notes in correspondence that:

"We asked for raw data and we were not given it. Initially, we were told that the company has other priorities and that they have not looked at the data. Then company has changed hands and we were told that they no longer had it. But we were given the summary tables and report from which i wrote the paper. The whole process took several years."

Should we as clinicians really be prescribing Famciclovir given this corporate approach to science, and given that we don't really know what was hidden from the authors?

The response of the medical profession and medical ethicists to such problems has been feeble. EBM specialists who collect Clinical Evidence scrupulously fail to document attempts to control, suppress or delay that evidence. It is also puzzling that the drug regulators have no interest in contamination of the scientific literature upon which doctors and patients rely. The malaise is reflected in the recent puzzling assertion in the UK that the drug regulatory body (MHRA) has no remit to investigate or to be concerned about allegations of scientific misconduct or contamination of the scientific literature after a drug is licensed. What then are they for?

We would do well to remind ourselves of the story of Betty Dong and the "thyroid storm" (3, 1996 Editorial in Science, another Editorial)

In 1990, Dong, a researcher at UCSF was funded by Boots to carry out research on a widely use thyroid treatment (Synthroid). She discovered the Boots drug was no more effective than three much cheaper competitors. When she tried to publish, Boots threatened to sue. The publication was withdrawn. She received no institutional support. Company executives attempted successfully to publish an inaccurate version of the findings while excluding Dong and threatening legal action. Nine years later the sordid details were exposed in the press and Dong’s paper was published. In 1999/2000 the company paid $170 million to settle class action lawsuits. However it is estimated that the company made a profit of $3billion in inflated costs during the nine year delay. No company executives were prosecuted. The regulators turned a blind eye.

1. Wald A, Selke S, Warren T et al. Comparative efficacy of famciclovir and valacyclovir for suppression of recurrent genital herpes and viral shedding. Sex Transmitted Dis 2006; 33: 529-533.
2. Fife KH. Are the antiherpes nucleosides really all the same? Sex Transmitted Dis 2006; 33: 534-535
3. Rennie D, "Thyroid storm" JAMA, Apr 1997; 277: 1238 - 1243

Earlier|Later|Main Page

Procter research shenanigans 5: Research misconduct summarized in one graph

This is the key graph from Richard Eastell et al. (2003) J. Bone. Miner. Res. 18:1051-6.

The solid lines supposedly show the relationship between the change in bone turnover (NTX) and new spine fractures in patients taking Actonel (Risedronate) at 1 year and at 3 years. The dashed lines show patients taking placebo. The purpose of this posting is simply to show that the scale of graphs in ghost-written papers by P&G had been drawn so that a proportion of the data simply "fell off" the left hand end of plots. I was supposed to have signed off on two further publications containing similar graphs and opaque statistical analysis ...... all while being refused access to underlying data codes held by P&G.

The first step of any statistical analysis is to PLOT THE DATA. Following press exposure P&G eventually released the raw data to authors in April 2006. The distribution of the actual data for NTX change (the X axis variable) in the above paper (Eastell et al 2003) for patients taking Actonel is:
A third of the data would not have appeared within the range of the Procter and Gamble graph as plotted by them (P&G's graph ends at -60%). Strange that.

The relevance of the missing end of the plot will be discussed in later postings. Suffice to say that these patients show responses more typical of the drug produced by P&G's competitor (Merck, Fosamax).

Earlier|Later|Main Page

Procter research shenanigans part 2: On graphs and honeybees

Misleading graphs form part of the attempt by Procter and Gamble Pharmaceuticals (P&G) to interfere with scientific evidence about their osteoporosis drug Actonel (backstory here, here and here). Several similar graphs (with supposedly supportive statistics) appeared in reports prepared by P&G based on data generated in Sheffield. They also appeared in a P&G funded publication (Eastell et al. 2003 J.Bone.Miner.Res. 18:1051-6), underpinned meeting abstracts, and two draft ghostwritten P&G publications with myself as intended first author -- all while P&G refused authors access to randomization codes and other underlying raw data.

Here is an example (from Eastell et al (2003) J. Bone. Miner. Res. 18:1051-6).



Much more of this later. However I was reminded of some correspondence in Science 30 years ago in response to a graph and statistics in a paper by Roubik (Science 1978: 201;1030).




Which prompted Robert Hazen to respond (Science 1978 202:823):



Earlier|Later|Main Page

Why bother with clinical trials?

Useful statistical methodology for big pharma here:
http://www.datanumerics.com/

See also: The freedom to say that 2 plus 2 make 4

Earlier|Later|Main Page