Sunday, February 23, 2014

Pradaxa: More Evidence of Drug Firms Manipulating Science

Thanks to Dr. Roy Poses at Health Care Renewal:

--for calling attention to two articles in the New York Times by Katie Thomas:

Let me give some medical background and then review the contents of Thomas’s articles, relay Dr. Poses’ take on the matter, and then end with a reflection of my own.

For many years, the most commonly used oral anti-clotting drug (commonly referred to mistakenly and misleadingly as “blood thinners”) has been warfarin. Warfarin has a number of disadvantages. First, it has a very narrow safety range—let blood levels get too low and the person is at risk for forming potentially fatal clots; let them get too high and the person is at risk for a potentially fatal bleeding episode. That leads to the second disadvantage—that patients taking warfarin have to report regularly for blood tests to measure the levels and to adjust the dose if needed. So if anyone succeeds in developing a new-generation replacement for warfarin, but without these problems, then it would qualify as a better mousetrap par excellence.

Recently, several firms have proclaimed the advent of the better mousetrap; one version is Pradaxa (dabigatran) made by the German firm Boehringer Ingelheim. Pradaxa has been marketed for preventing clots in one of the most common conditions for which such drugs are prescribed, the irregular heartbeat known as atrial fibrillation, and the company claims it is at least as good as warfarin without requiring any inconvenient blood testing. Warfarin works by blocking the step in the clotting process that’s controlled by Vitamin K, so that if a patient taking warfarin starts bleeding dangerously, an intravenous dose of Vitamin K will immediately reverse the drug’s effects. Pradaxa and the other new drugs work at a different point in the clotting cycle and there’s no available antidote if a patient bleeds—leading some experts to advise that the drug is too dangerous to use for that reason alone. The FDA has received reports of 1000 deaths attributed to Pradaxa (out of some 850,000 patients prescribed the drug, netting Boehringer the hefty revenues of $2B).  

It’s patients suing the drug firm over Pradaxa-related harm that led to the articles in the Times. The judge ordered release of some documents related to the case that include internal company e-mails about a research study coordinated by a company scientist, Paul A. Reilly. Reilly’s study showed that a part of the safety problem with Pradaxa was that some patients had too high a blood level and some patients too low (sound familiar?). He concluded that a blood test that measured drug levels could be helpful for at least some patients in avoiding dangerous reactions. (Such a blood test is available now in Europe but not the U.S.)

A draft of Reilly’s paper that included these findings unleashed a storm of e-mails from other company scientists and officials. They argued that publishing a paper with these conclusions would undermine the company’s primary marketing point in favor of Pradaxa, the result of a fine-tuned marketing effort going back a decade. Moreover, some feared that if the paper were published, it would be that much harder to get the government regulators to hold off demanding blood tests. The end result was that the paper was published recently but with many of the offending details removed.

Boehringer Ingelheim insists that this was a simple matter of scientific review and refinement. A draft was circulated, others chimed in with appropriate criticisms, and in the end the final paper was suitably modified to better present the actual facts. Nobody here but us scientists, boss.

Other experts aren’t buying that, and Dr. Poses appropriately asks how often such censoring of scientific findings in the name of marketing occurs with no friendly judge to force the release of the secret company documents.

One of the common complaints from pharmapologists is that pharmascolds quite unfairly would have physicians distrust a scientific paper merely because it’s sponsored by a drug company, when ideally, they should read the paper carefully, review the methods, and believe the paper or not based solely on its merits. Dr. Poses comes out with a powerful contrary statement: “I strongly advocate that those who author authoritative systematic reviews, meta-analyses, and clinical practice guidelines… assume the likelihood that all commercially sponsored published clinical research has been manipulated…”

Now for my own comment—this sounds very much like a replay of the case of the so-called atypical antipsychotics that we previously discussed:

Recall there that a new class of drugs was introduced with great fanfare and said to be far superior to the original class of drugs, and only after years of use did it become clear in hindsight that the supposedly new drugs were in fact hardly different at all from the old ones, except in the eyes of the company marketers, who managed to bamboozle the entire medical profession very neatly, much to the harm of patients.

Reilly PA, Lehr T, Haertter S, et al. The effect of dabigatran plasma concentrations and patient characteristics on the frequency of ischemic stroke and major bleeding in atrial fibrillation patients: The REL-Y Trial (Randomized Evaluation of Long-term Anticoagulation Therapy). Journal of the American College of Cardiology 63:321-328, 2014.

Saturday, February 15, 2014

"You're Another" Should Be an Admission of Guilt

Charles Ornstein, investigative reporter at ProPublica, recently commented on a flap involving (to put it bluntly) university docs shilling for a device company:

He cited a blog by Paul Levy, a former CEO of New England Deaconess Medical Center in Boston. Levy noted an ad for the da Vinci surgical robot that featured a group photo of surgeons from the University of Illinois (Chicago) Hospital. Mr. Levy did a little digging around and discovered that such an appearance in an ad seemed to violate U-Ill. official policy. So he sounded off in several blog posts and the University eventually took note and said they'd look into the matter, while insisting that these surgeons did all this out of the kindness of their hearts and were not paid for appearing in the ad.

You might wonder why the flap over a surgical robot. These devices can cost a cool $2M or more each, and of course when introduced were touted as the greatest new medical breakthrough. Sadly, to the extent that evidence is becoming available (and I admit I have not followed this debate as closely as one might), the bulk of data seem to question whether robotic surgery produces outcomes that are superior to the old garden variety surgery, and indeed in some cases there may be worse complications--leading for instance the American College of Obstetrics and Gynecology (ACOG) to state (as Ornstein noted) that robotic hysterectomy has not even been shown to be as good as, much less better than, the old-fashioned and vastly cheaper procedure.

This has left a lot of hospitals, who rushed to obtain "the latest" out of fear that both their surgeons and their patients would desert them for the "better" hospital up the street, with expensive robots that now threaten to become white elephants. So you can see why the maker of the robot would want to run ads reassuring everyone that these devices are indeed worth their weight in gold.

What I wanted to comment on regarding this incident are two things.

First: Intuitive, the company that makes the robot in the ad, responded to the discussion by saying: "In the past year, there has been much misinformation about robotic-assisted surgery, spread largely by plaintiffs’ lawyers as well as segments of the health-care community threatened by our groundbreaking technology. ...  The University of Illinois, which uses our technology, and the people featured in the advertisement agreed to appear without compensation. Those who use our technology see first-hand the outcomes resulting from its use. Their unpaid testimonials of da Vinci surgery are credible and sincere."

So, first, only evil folks with an axe to grind say anything bad about robotic surgery (take that, ACOG); and second, no one paid these nice Illinois surgeons a dime to be in the ad. Now that latter may be true; I can't disprove it. But knowing what we know about both the largesse of device makers toward docile docs who tout their wares, and the extent the makers sometimes go to to launder the money as it changes hands, I'd frankly be quite surprised if there's no quid pro quo someplace.

Second: Ornstein spoke with Thomas Hardy, executive director of university relations at U-Ill. Chicago. Mr. Hardy wanted to be sure that Ornstein knew all about a shady event in blogger Levy's past, which if you want to read the dirt, is detailed here:

So Ornstein asked Mr. Hardy exactly how this previous, and admitted lapse of Mr. Levy was relevant to the discussion of the ad with the U-Ill. surgeons that appears to violate U-Ill. official policy. Hardy's reply was: “I believe if you’re attributing claims and accusations to the blogster, your readers deserve to know his reported background so they can make an informed decision about his credibility…Wanted to make sure you have the pertinent information.”

Right. So know we know what sort of person we're dealing with.

I have a proposal for a new rule: If a person accuses you of conflict of interest or a similar lapse of professional responsibility, and the best defense you can come up with is an ad hominem attack against the person regarding something that's irrelevant to the present controversy, then your ad-hominem attack ought to be viewed as in itself an admission of wrongdoing.

Wednesday, February 12, 2014

Update: Detailing in India and Pakistan

The very nice people at the Centre for Biomedical Ethics and Culture, Sindh Institute of Urology and Transplantation, Karachi, Pakistan, publish a newsletter. I was attracted by the cover story, "The Drug Industry and Doctors: An Unholy Alliance," by Sandhya Srinivasan, a journalist and researcher from India:

In many ways the issues in that part of the world are the same as here, but a few highlights of interest:
  • Presumably due to the general lower income enjoyed by physicians in poorer nations, it's become more acceptable and commonplace for drug reps to bribe physicians with less subtle "gifts" such as air conditioners, washing machines, microwaves, TVs, etc., compared to the pretense kept up in richer countries that the "gifts" relate solely to one's professional role.
  • While efforts have been made in both India and Pakistan to declare such bribes unethical, in India, especially, enforcement has been marred by numerous roadblocks--both the influence exerted over government by drug company interests; and also the dubious position of some of the folks issuing the "ethics" guidelines, for example: "The [Medical Council of India, who promulgated one Code of Ethics on the subject] is not viewed as a guardian of medical ethics. Less than a year after the amendment, its president was arrested on charges of corruption."
  • A study in Pakistan likewise found widespread acceptance of drug reps and their largesse, with doctors meeting with an average of seven reps daily.

It may be hard to imagine a part of the world where freeing medicine from the grip of industry influence is even more of an uphill struggle than here, but that seems to be the case in South Asia.

The NIH Accelerating Medicines Partnership—Francis Collins Sings a Few Bars

I had wanted to respond to a news release some days ago about the Accelerating Medicines Partnership announced by the NIH and 10 drug companies, but before I got around to it, I caught a tiny bit of this interview on The Diane Rehm Show on my car radio:
--in which not only did Dr. Francis Collins, head of the NIH, talk about the AMP, but also sang about it self-accompanied on his guitar (which, he explained, suitable for the former head of the Human Genome Project, has a double helix inlaid in mother-of-pearl on its fretboard).

I’ll let you enjoy the music if you wish on the audio and turn straight to the written transcript.

Here’s how Dr. Collins explains the AMP:

So more than 1,000 new drug targets have emerged in the last five years from that kind of study, but it's very hard to sift through them and pick out which ones are going to be the real home runs that we're all looking for…. this is full 50/50 skin-in-the-game kind of collaboration, $230 million committed to this over five years, half of that coming from NIH, half of that coming from the 10 companies that are participating. The scientists from both sectors will sit around the same table and work together to make this happen in a fully open access atmosphere…. We already have detailed research plans for these disease areas that have been worked out over the past year with very clear milestones that have to be met. This is not one of those where everybody goes off and plays in the lab. We are really serious here about making real progress.

Well, in this phase of trying to identify the next generation of drug targets, everybody agrees that this is precompetitive, and that is that all the information has to be openly shared both within this consortium and for anybody else who's watching. The competitiveness, Diane, kicks in once you've identified, oh wow, that particular molecule is a really promising target for the next generation of Alzheimer's therapy…. Then every company will run off and do what they do really well… And we want them to do that. That's good competition. That means that things move quickly and there's lots of ways that you can get to yes.

Let me see if I can make sense of this from the standpoint of the view of current pharmaceutical science that I proposed in HOOKED. I suggested then that the current drug pipelines were so dry because there are only a limited number of molecules that do useful things in the human body without killing us. Not that there would never again be a “golden age” of major advances in drug therapy; but the new golden age would not come from drug companies doing cranked-out research on their assembly line—it would require breakthroughs in our basic understanding of disease, discoveries that the NIH and academic centers are better able to make than industrial labs. So Take Home Message #1, I hear Dr. Collins saying that we’re poised to enter another age of advancement, that basic science research has now identified a bunch of exciting new possibilities—locks that we now need to discover the keys to.

Take Home Message #2 seems to be that the “translational science” model, which NIH has embraced in the last decade, tells them that we’re now ready for a more accelerated and focused phase of discovery, when NIH scientists and Pharma scientists can both participate in a program that will quickly move (as the translational mantra has it) from (lab) bench to bedside. At this point, as a complete non-expert, I would enter a note of caution. I generally support the translational science model (full disclosure: a small portion of my salary is currently funded under a translational science grant, to address the ethical issues). However, there are times in the life cycle of a discovery when you can safely go on the fast track and other times when you’d be better off “playing in the lab” just a little bit longer. I hope the NIH gurus are right that their particular list of diseases—Alzheimers, diabetes, and autoimmune disease—is truly ready for this hurry-up approach.

OK, so that’s the science part; what about the potential-conflict-of-interest part, the rocks on which so many previous scientific ships have broken up and sank? Says Dr. Collins:

With regard to the drug companies, I know that people are concerned about what their motives are. Again, they make pills. NIH doesn't. … We believe, by working together, we can speed up the process of getting the right answers. … You know, five years ago, I don't think this [collaboration] would have been possible. I think it's a combination of scientific opportunity that is so exciting but so overwhelming that no company can tackle it on their own, plus their own anxieties about the failures that are all too common, even now, in drug trials where you've spent hundreds of millions of dollars and you get to the end of that Phase III trial and it didn't work. They've had enough of that, and they're anxious to try something different.

So two things here—first, with a demand for open access, NIH figures they have a safeguard against one of the main dangers of industry collaborations of the past; and second, they believe that the drug industry today is in a different place and more willing to play nicely in the sandbox. Interestingly, Dr. Collins made a comment similar to one I have made previously in this blog—

[Rehm] I'm interested that Johnson & Johnson hired Yale University to oversee the sharing of clinical trial data. What can you tell us about that? [Collins] I think that's fascinating. So the person at Yale University, Mr. Harlan Krumholz, who's a cardiologist, who is a wonderful leader in this whole idea about getting information out there where everybody can see it -- and he has a lot of credibility so people will believe that if Harlan is involved that what Johnson & Johnson proposes to do is the real deal and not just some window dressing. I think it's great.

Final conclusion I’d offer: NIH seems to be going into the AMP with decent reasons and with eyes open, but so far, Pharma has an excellent track record of managing to have such folks for lunch nonetheless. I hope Dr. Collins is right and this time it will be different, and if so, we may have yet another good model of positive collaboration to work with. Stay tuned.

Sunday, February 9, 2014

AAUP Chimes In on COI

The American Association of University Professors has now published its 356-page report, Recommended Principles to Guide Academy-Industry Relationships:

What the AAUP essentially appears to have done is to review reports issued previously by such organizations as AAMC and IOM, and select the provisions their committee most approved of. In general, the AAUP has taken a strong stand in favor of what they consider to be academic values and academic freedom, and opposing any intrusion of financial conflicts of interest. For example, among the 56 guiding principles that they offer, Principle 3 demands the right for faculty to publish their research with at most a minimal delay (30-60 days) to protect patent and other so-called "intellectual property" interests. Principle 4 condemns any and all instances of ghostwriting. The report states clearly that disclosing COI is usually not enough and that avoiding COI is by far preferable.

While subsets of principles relate directly to academic medicine and biomedical research, the AAUP here attempts to address the broadest set of academic-industry relationships, and not to restrict their purview solely to medicine and the life sciences. They acknowledge, however, that the medical area has been especially problematic (no surprise to any readers of this blog).

One area that draws special attention in the report is what they call Strategic Corporate Alliances, major multi-year commitments between a university and a corporate sponsor to cover all faculty working in a specific field. The original poster child for such an alliance, discussed at some length in HOOKED, was the 1998 agreement between UC-Berkeley and Novartis for rights to all research conducted in its Plant Biology department. Since then, says the AAUP, such agreements have only grown and expanded. The principles they propose to govern such agreements strongly tips the balance in favor of an academic-governance model and away from giving too much control to the corporate sponsor--so much so that one wonders what corporations would agree to major alliances under the conditions specified.

I suppose in the name of full disclosure I should mention that among the 770 endnotes, at least one cites HOOKED-- I cannot state that I have read all 770 of them, however.

Sunday, February 2, 2014

Reps on Safety Issues, or, the Sounds of Silence

When writing in HOOKED about drug reps, I asserted that they seem to soft-pedal safety concerns, seldom mentioning any adverse reactions to their drugs. Unfortunately, the data at the time were restricted to one major study in France, so one could claim that we actually knew very little about what reps did or did not say to physicians on this subject.

We can therefore welcome the more recent study of Dr. Barbara Mintzes and her colleagues (subscription required), who studied drug rep-physician interactions in four cities in three countries, Montreal, Vancouver, Sacramento, and Toulouse. A major weakness immediately arises as they relied on physician recall of the visit rather than directly observing the visit. However, it's nearly impossible to imagine drug reps agreeing to have these visits observed, or acting "normal" if they knew the visit was being recorded, so I gather this is the best evidence that is ever likely to be available. The folks ended up with 255 physicians reporting on 1692 detail visits.

Essentially, this study confirmed the earlier French study. They used as a primary outcome "minimally adequate safety information," which they defined as "mention of [at least] 1 approved indication, [at least one serious adverse event (SAE), [at least] 1 common non-serious adverse event (AE), [at least] 1 contraindication (CI) and no unapproved indications or unqualified safety claims". They thought they were setting the bar pretty low with these requirements, so it's impressive that this desired endpoint was reached overall in only 1.7% of all visits, ranging from 0.9% in Sacramento to 3.0% in Toulouse.

As an example of just how far off the mark was the information provided in most details, in terms of safety, the physicians recorded what they took to be the "key message" of the detail. For details for Avandia, taken off the market in Europe and restricted in Canada and the U.S. in 2010 due to concerns about excessive heart risks, key messages included, "'Avandia is safe even in patients with heart disease, as long as they don't have heart failure' (Montreal); 'Avandia is not as dangerous as the public makes it out to be' (Sacramento); 'New studies indicate safety' (Vancouver)."

Despite the near-total absence of valid information about the safety of drugs, these physicians (who of course have to be willing to see reps to be included in the study; a number of candidates were excluded because they never saw reps) rated 57% of the details as "good" or "excellent," and nearly two-thirds of the time stated they were "somewhat" or "very" likely to start prescribing the drug being detailed.

The authors noted that details that were conducted in the way that the vast majority of these apparently were are actually illegal according to the official criteria for the FDA and equivalent agencies in Canada and France, but of course there's virtually no enforcement at the level of the private conversation between rep and doc in the office.

If anyone still thinks that visits from reps to physicians are "not marketing but education," this study seems to put one more nail in that coffin. (Hat tip to Primary Care Medical Abstracts for clueing me in to this article.)

Mintzes B, Lexchin J, Sutherland JM, et al. "Pharmaceutical Sales Representatives and Patient Safety: A Comparative Prospective Study of Information Quality in Canada, France and the United States." Journal of General Internal Medicine 28:1368-75, 2013.