On GSK, Ghostwriting, and the AHA: Grassley Sidesteps Confusing Evidence, Can't See Forest for Trees, Is Probably a Hypocrite, Etc.
The reason to care about medical ghostwriting is that it may be used to convey drug-industry messages through credited academic authors. In other words, there is an attempt by industry to influence a physician-reader's prescribing practices under the guise of academic objectiveness (although nobody, to my knowledge, has demonstrated this outcome).
But I would argue that the issue of medical ghostwriting is irrelevant when pharma-employed scientists are among the credited authors of articles (which are usually reports of pharma-supported studies). In these cases, the potential for commercial influence in the published material is already obvious, regardless of who actually wrote the article.
Nevertheless, the possibility that there are situations in which medical ghostwriting is immaterial is evidently lost on some (nor will it get in the way of a mission to root out ghostwriting, no matter how immaterial). Case in point:
In a 2-part effort to impugn the marketing of the antidiabetes drug Avandia (rosiglitazone), the Senate Committee on Finance sent a letter to FDA Commissioner Margaret Hamburg on July 12, alleging that manufacturer GSK 1) failed to publish negative studies of the drug "in a timely manner" and 2) hired ghostwriters to pen medical articles that were favorable to Avandia. The former charge, the more serious of the two, has received considerable press attention—most notably in a recent NYT article by Gardiner Harris. However, the ghostwriting charge has not, by any means, been ignored.
In a confusing post at the generally anti-pharma (and often sensationalist, IMO) blog Pharmalot, Ed Silverman asked his readers last Thursday, "Did the American Heart Association's Circulation journal publish a ghostwritten article about Avandia?" Silverman's answer, yes, lies in an exchange of letters among the Senate Committee, the FDA Commissioner, and the AHA. But a closer examination of the letters and the published literature reveals a more nuanced answer than what Silverman concluded (and what his blog readers are likely to be receptive to). A closer examination also reveals how deftly and stubbornly the Committee maintained its ghostwriting allegations against the AHA while failing to acknowledge 1) its confusing and/or mistaken evidence and 2) when medical ghostwriting is an important issue.
In its July 12 letter, the Finance Committee provided 2 attachments to substantiate allegations of Avandia-related ghostwriting by GSK. The first attachment, H (pp 58-109 of 158), contained an e-mail from the Avandia Publication Strategy Manager, who presented a draft manuscript to his GSK colleagues. The manuscript, which reviewed 5 modifiable risk factors for cardiovascular disease, was to be credited to University of Texas cardiologist Steven M. Haffner, MD, and "targeted" to the American Journal of Cardiology, an independent peer-reviewed medical journal that is published by Elsevier. The journal is not affiliated with the AHA.
The text of the redacted and "VERY poorly written"* manuscript mentioned rosiglitazone exactly twice (at the end of the article), both times in conjunction with its chief competitor, pioglitazone (Actos). A notable, suggested edit was to use "Avandia references" in the construction of a table detailing the effects of thiazolidinediones on cardiovascular risks.
The Committee's follow-up attachment I (pp 110-158) contained Haffner's "reworked" and highly condensed manuscript, entitled, "Modifying cardiovascular risk in the type 2 diabetes patient." The title page of the review now indicated the authorship of Haffner and nurse Holly B. Cassells, PhD, who is on faculty at the University of the Incarnate Word [?!]. The title page also acknowledged that the paper was "[s]upported by an unrestricted, educational grant from GlaxoSmithKline" (thereby acknowledging the potential for commercial influence in the manuscript). The heavily redacted text in the attachment mentioned rosiglitazone 4 times (by my count) and highlighted a study of the cardiovascular effects of combined therapy with atorvastatin (Lipitor; Pfizer) and rosiglitazone (Cohen et al, 2001). A GSK e-mail suggested that the review article should now be targeted to another non-AHA journal, Clinical Cardiology (published by Wiley).
In its attachment I to Hamburg, the Committee chased the Haffner review paper with another Haffner-authored article, which was evidently accepted into the journal Circulation, a bona-fide AHA journal. This article, "Effect of rosiglitazone treatment on nontraditional markers of cardiovascular disease in patients with type 2 diabetes mellitus," described the results of an actual study; in other words, it was not a review article (and it was clearly not the review manuscript previously cited by the Committee in its attachments). The article, published in 2002, had 6 credited authors, 3 of whom (including Martin I. Freed, MD) were acknowledged employees of GSK, and the study's GSK funding was made clear.
On August 23, the AHA responded to the Committee's charges by sending a letter to Hamburg. Donna K. Arnett, PhD, the chair of the AHA's Scientific Publishing Committee, attempted to clarify the distinction between the 2 articles cited by the Committee in its attachments. Arnett noted correctly that the review article was never published in Circulation and was targeted to at least 2 non-AHA journals.** Arnett concluded,
Since Circulation has never published a review article by Haffner and Cassel[sic] entitled, "Modifying Cardiovascular Risk in the Type 2 Diabetes Patient," as claimed in the letter to you, the letter and its appendices are misleading. We have requested that [the Committee] provide a correction for the record.
But not to let details get in the way of a mission, Ranking Member of the Committee, Senator Chuck Grassley, responded to Hamburg with yet another letter, dated September 1. Grassley disingenuously wrote (or one of his staffers disingenuously ghostwrote),
At the time, it was unclear if the emails about ghostwriting concerned the article that appeared in Circulation, or if they were discussing a separate manuscript.
To which I say: Rubbish. The featured, originally ghostwritten review article by Haffner (and then Haffner and Cassells) was the clear subject of the GSK e-mails.
Then Grassley (or his staffer) added, while completely ignoring the cited review article,
Regardless, my staff has consulted with GSK, and the company confirmed that the manuscript which appeared in Circulation was written for GSK by a medical education company.
To which I say, more or less, to the Committee (and Pharmalot): So what? Given that 3 of the 6 authors of the Circulation article were acknowledged GSK employees and that the study was transparently funded by GSK, the potential for commercial influence is obvious to any physician-reader, regardless of who actually wrote the article.
Now as far as the originally ghostwritten review article is concerned, which Senator Grassley (or his ghostwriting staffer) failed to address in the September 1 rebuttal letter, a quick search of the literature reveals that Cassells and Haffner (credited in that order) finally published, “The metabolic syndrome: risk factors and management,” in the Journal of Cardiovascular Nursing in 2006. This journal is not an AHA journal. Because access to the article requires a subscription, I am (at this time) unable to assess the objectiveness of the article, how closely the article adheres to the attachment I text, or whether commercial support is acknowledged.
* As it was described by GSK's Julia Eastgate, Head of Diabetes Communications, Europe, in her e-mail.
** Silverman called Arnett's letter "sharply worded"; on the contrary, I find Arnett's wording moderate in tone and matter-of-fact.
A comment from Ed Silverman of Pharmalot (who had some unidentified trouble with my blog's commenting system):
Hi Barbara,
Thanks for the mention. And if I may, I’d like to note a couple of things…
First, I understand your point that if three Glaxo employees co-author a study, then perhaps ghostwriting is not the appropriate categorization and, therefore, not an issue. I’ll leave it up to the readers – yours and mine - to decide based on the info available. With that in mind, let’s remember that we both noted that Grassley wrote to Hamburg in his latest letter that Glaxo “confirmed that the manuscript which appeared in Circulation was written for GSK by a medical education company.”
In other words, assuming Grassley and his staff got it right this time (and we all know that Senators and staffers do make mistakes), the study published in Circulation was written entirely – or largely - by outsiders, not the folks listed on the paper. On that basis, it would appear that ghostwriting is the correct description. I agree that having Glaxo employees as co-authors and Glaxo funding should make it clear that the paper was, for all intents and purposes, a commercial effort. But ghostwriting is a separate activity and that moves the ball further down the promotional court. And that, in my view, is a distinction worth noting.
One other thing: I appreciate that you stop by Pharmalot, but I think the notion that the site is anti-pharma and sensational is off base. Here’s why: The pharma biz has been undergoing an enormous amount of change in recent years and change creates tension. And examining those moments of tension is one way that journalists chronicle change. Those tension points, of course, have played out in all sorts of ways – patent expirations; battles between brand-name and generic makers; pricing issues; employment loss; the utility of sales reps; the debate over disclosing clinical trial data; mergers and acquisitions, including proxy fights and disputes with dissident shareholders; the wisdom of FDA policies, such as the accelerated approval program and experimental meds; R&D struggles won and lost; and legal rulings that hold significance to patients who claim harm from a medication, among other things.
Clearly, there is much to choose from when looking back to all of the posts that have appeared since the site was launched in early 2007. So I’m sure it’s easy to cherry pick and say there is the sort of slant you mention. But I would argue there have been many posts that examine a wide range of issues and, when taken together, offer a more nuanced, thoughtful and well-rounded picture of an industry that is trying to transform itself. And if in the process, there is discussion about business practices and how these can be improved, then so much the better. After all, if improvements can occur, that’s not a negative.
On a personal note, I live and work in suburban New Jersey, and so the plight of some drugmakers and their contribution to various state and local taxes is felt directly and indirectly in my own town. I have family and friends who work in pharma. My kids have befriended other kids whose parents work with or for pharma. And my kids have been vaccinated (I hope I haven’t broken any HIPAA laws). And yes, we take prescription meds when needed. So the vagaries of the industry are all around me and this permeates my own life in ways that may not seem evident in my coverage.
I don’t claim to be perfect, but I do hope this helps and perhaps we can continue the discussion on a more positive note.
Cheers,
Ed