Children of the Cure: Missing Data, Lost Lives and Antidepressants
David Healy, Joanna Le Noury and Julie Wood:
Toronto: Samizdat-Health; 2020 (279 pages) Toronto, Ontario.
International Network for the History of Neuropsychopharmacology Review
INFORMATION ON CONTENTS:
This book covers a series of events leading to a study of paroxetine, versus imipramine and placebo in adolescents – Study 329, followed by the controversies that accumulated around this study, leading to a Black Box Warning on antidepressants, a fraud charge, the then biggest fine in corporate history, efforts to get the paper retracted that went nowhere and finally a “restoration” effort that now means there are two versions of this study in premier journals offering diametrically opposite conclusions. The least known part of the story lies in the difficulties the team trying to restore the study had – more of which came from BMJ than GSK. Readers are given a brief background history of depression, the origin of the SSRIs and controversies about suicidality linked to antidepressants (chapters 3-5), the Black Box story (chapters 6-7), the calls for retraction (chapters 7-8), the Restoration process (chapters 10-13) and a series of possibly pertinent subsequent events (chapter 15).
This book is a history rather than an argument. It follows events, some of which may be known to insiders but almost all of which will be new to a wider readership. It seeks to engage – actively asking readers to help fill in the blanks.
Children of the Cure offers either a fairy tale or an epic take (pay your money and make your own call) on Study 329 – the most famous clinical trial in medicine. There are other takes on what happened in the pipeline from the more academic The Illusion of Evidence-Based Medicine by Jureidini and McHenry to Paul Scott’s Malcharist, which may be satire or may be all too real, difficult to tell, and a compelling screenplay. These books complement Jim Gottstein’s The Zyprexa Papers.
Children of the Cure ends on a note suggesting there is scope to view the role of medical academics, at least the child psychiatrists, in the original trial and subsequent Keller et al article in Rip Van Winkle terms. Martin Keller, Neal Ryan and others of the merry band who struggled up the mountain were decent people. A number of them were women, so they weren’t escaping from their wives. When Study 329 began, they fell asleep in one kind of a world only to wake up in a different one. Where once they had run trials, now they supplied patients. A world in which some of them were deposed under oath. After their long sleep, despite other evidence of preserved cognitive faculties, Keller and Ryan at least had marked amnesia. The one person who seemed to have a good recall of what happened was Sally Laden – the author of the original paper, whose name doesn’t appear on the authorship line of the paper in JAACAP – the Journal of the American Academy of Child and Adolescent Psychiatry.
Three years later the Attorney General of New York, Eliot Spitzer, lodged a fraud charge against GSK. Several years after that again, GSK resolved an action taken by the Department of Justice for $3 billion, then the largest settlement in corporate history. In 2004, Spitzer claimed his action would transform the relationship between medicine and the pharmaceutical industry – no longer would trial data be sequestered. In 2012, many thought $3 billion would make some difference. But showing a mastery of the art of turning a problem into an opportunity, in 2004 GSK gained kudos and badly discombobulated Spitzer by trumpeting their new clinical trial register and by 2012 they were positioned to jump on the AllTrials bandwagon and when they did Andrew Witty, their CEO, featured on the front cover of the BMJ as the acceptable face of the pharmaceutical industry.
Oh, how much fun it must be to work in a pharmaceutical company marketing department with the prospect of a regular supply of Tom Wolfe moments – Wolfe famously said that he was hoping Bush would win the 2004 election so he could head out to JFK and watch all the departing liberals, who had sworn to leave the country if Bush won.
In the case of Study 329, its largely the liberals who seem to have set up the problem through the unintended consequences of regulations put in place to curb industry and liberal championing of health as a trump card – something to counterbalance industry’s dangerous practices in other domains. They seem unaware that their Access to Medicines campaigns feed straight into pharma’s hands or that creating rule books typically play into the hands of those with the deepest pockets – all but forcing corporate development. Outwitting such a band of self-righteous Ichabod Cranes is just too easy.
That said, there is a problem. There is barely a trial of medicines now that can legitimately be described as science – if we are conforming to the norms of science that is. The problem is that the apparent trials, let’s say trans-trials to contrast them with cis-trials, are now so numerous, especially in psychopharmacology, that it is difficult to see a way back.
The road to Study 329 began in the 1970s when industry began to outsource trials, outsource the reporting of trials, and to run multicentred trials in order to have enough patients to get snake oils on the market – the multiple centres opened up the way to a sequestration of trial data on the pretext of collecting the data centrally.
By the time we get to the mid-1990s and Study 329, clinical trials had become an industrial enterprise where, as with the Prozac and Paxil and Zoloft trials done in children, primary endpoints could be switched, trials negative on their primary endpoints could be written up as positive and serious adverse effects like suicidality could be written out of the script. Study 329 demonstrates all these features but so do all other trials of psychotropic drugs in pediatric depression. Despite the fraud change and $3 billion settlement, Study 329 was the industry norm not an aberration. And Study 329 and other pediatric depression trials run then were exemplary in terms of their conduct compared to trials done now – run in India or Africa or US foster homes or correctional facilities, but with Western “authors” for African trials and academics from the most distinguished US centres in the case of trials recruiting from correctional facilities and foster homes and in both cases likely a higher proportion of non-existent patients than we had in the 1990s.
The trials of SSRI and related drugs for pediatric depression remain a gold standard symbol that will be hard to surpass. They are universally negative with Prozac having more negative trials than any other drug and probably more suicidal events. The number of suicidal events in these negative studies may never be truly known. A trick, not noticed when Study 329 was being Restored, has since shown up in material emerging from the Dallas centre run by Graham Emslie, the lead author on the early Prozac trials, as well as a 329 author. No suicidal events showed up in Dallas – but children it seems could become disinhibited and homicidal without the treatment being viewed as ineffective or the events being seen as adverse events. The children were designated instead, as having an intercurrent illness. This designation makes the patient vanish without a trace of what went wrong. This trick is now widely used in both drug and vaccine trials.
What happens when there are 30 negative trials and not a single positive trial – the greatest concentration of negative trials for any indication ever? Why the treatments end up as the second commonest pharmaceuticals taken by teenage girls.
None of this is necessarily industry’s fault. If let get away with it, its difficult to blame companies for fooling doctors who are all but begging to be fooled. MAGA is an acronym that is unlikely to catch on, but it captures perfectly the spirit of the age – Medically Assisted Garbage Adoration.
MAGA is wonderfully illustrated in a tardive dysgnostic feature of the 329 story. Along with Keller and Ryan. Karen Wagner was one of the 329 authors. She was later the first ghost author on Pfizer’s studies of Zoloft in childhood depression and a Forest citalopram trial (CIT-MD-18). Many College students would be able to spot why two negative Zoloft studies jammed together and published as one should never have been published in any journal, never mind JAMA. CIT-18-MD has other serious flaws, and Forest were later fined by the Department of Justice, just as GSK had been.
Just after the Restored Study 329 was published in the BMJ in 2015, putting Wagner’s role in the original study in focus for members of the American Academy for Child and Adolescent Psychiatry (AACAP), including testimony that most articles featuring her name were ghost-written, AACAP members voted her in as AACAP President.
Children of the Cure opens with an invite to all readers to help us author Study 329 and the story of Study 329. We present a great deal of the data along with our take on it and make it clear that part of the point of the exercise was not to establish the truth about the study but to show that claims that science is authoritative are part of an authoring process that works best in the absence of data. Once the data becomes available, it becomes clear that claims can only be provisional and there may be many legitimate takes on the same data.
We also ask readers to assist in authoring the story of Study 329. There are holes in the narrative such as why have the data from the supposedly independent NIH sponsored Treatment of Adolescent Depression Study (TADS) – a study that likely had the highest concentration of suicidal events of any study – been destroyed.
Few of our readers are likely to be able to contribute much additional authoring apart from personal experience of grief on SSRIs or in dealing with health sevices afterwards – as happened Stephen O’Neill’s family (see DavidHealy.org) but academics, journal editors and especially people with experience working in or to the pharmaceutical industry are well placed to assist with both exercises in authorship.
Rip van Winkle, Sleepy Hollow, or some other Washington Irving story? Discounting industry input, over a third of the authors on the Keller et al paper, as well as Eliot Spitzer and many of the other players in the drama came from or had strong links to New York State. Spitzer – now there’s a twist; who set up his escort service downfall?