Friday, February 15, 2008

Spinning and Spinning for Cymbalta

I was browsing the January issue of the Journal of Clinical Psychiatry in order to get ideas for the "Research Updates" section of The Carlat Psychiatry Report, and I happened upon two articles, one for those intrepid counter-detailers out there, and the other for Lilly drug reps.

The first one, "Time to Rehospitalization in Patients with Major Depressive Disorder Taking Venlafaxine or Fluoxetine," is an article important for evaluating the long term usefulness of Effexor, but one which, I can guarantee you, will not show up in the hands of your Effexor rep. In this study, depressed patients in a Taiwan psychiatric hospital were discharged with instructions to continue either Effexor or Prozac, depending on which medicine made them better in the hospital. The two groups were compared to see whether Effexor would better protect patients from rehospitalization than Prozac. The results? About 45% of patients in both group were eventually reshospitalized. For patients on Effexor, the average time to rehospitalization ("survival time") was 223 days, and for patients on Prozac, it was 222 days. What does this prove? Not much, since patients were not randomly assigned to treatments. However, it's certainly no endorsement of Effexor's vaunted superiority over SSRIs. As the authors (who have no pharma relationships) state: "Our findings do not support the notion that venlafaxine, a dual reuptake inhibitor, is associated with less relapse/tachyphylaxis."

The next article, on the other hand, will be parlayed into a "teaching point" by your Eli Lilly rep, although there's precious little to teach. Entitled "Switching to Duloxetine from Selective Serotonin Reuptake Inhibitor Antidepressants: A Multicenter Trial Comparing 2 Switching Techniques," this Lilly-funded study identified 368 depressed patients who had not responded to at least 6 weeks of SSRI treatment. These patients were then randomly assigned to either abruptly switching to Cymbalta, or gradually switching to Cymbalta. They were then maintained on Cymbalta for 10 more weeks to see if they would respond. Before we get to the results, think carefully about the design of this study, and what questions one might reasonably be able to answer, or to not answer.

This was not a double-blind study: in fact, both patients and doctors knew that the participants were being switched to Cymbalta. Furthermore, there was neither a placebo arm, nor an SSRI continuation arm. This is crucial, because patients who don't respond to a medication after 6 weeks may well respond if continued on the same medication for an additional 10 weeks. Thus, if these patients improve after a switch to Cymbalta, we have no idea how to interpret this. It might mean that Cymbalta is better. But it might also mean that the 10 extra weeks of being in treatment worked its nonspecific placebo-related magic. After all, with enough time, many people become less depressed, no matter what treatment they receive. Thus, the design of the study allows no meaningful evaluation of Cymbalta's effectiveness. The only potentially useful information here relates to the technical issue of how to conduct a switch from an SSRI to Cymbalta. As it turns out, an abrupt switch works fine and is well-tolerated. And that's all you can conclude from the study.

However, if you read only the abstract (which is as far as most readers will venture) you'll get the uncanny sensation of having been teleported to Eli Lilly's website:

"Conclusion: Switch to duloxetine was associated with significant improvements in both emotional and painful physical symptoms of depression and was well tolerated and safe, regardless of which of the switch methods was used."

If it reads like copy written by a Lilly employee, it's because it was: Dr. Perahia, the first author, works for Lilly in England. One might have hoped that the editors of the Journal of Clinical Psychiatry would have caught this bit of blatant promotionalism before it went to press. Because of this awful oversight, now Lilly will likely end up paying the journal thousands of dollars to purchase article reprints for its drug reps--someone's head will roll!

13 comments:

Anonymous said...

Our profession's collusion with the drug companies has gotten so bad (as you have demonstrated so well with your website) that it is really not shocking anymore. I look forward to further posts on this blog, but, thanks in part to this blog, have little outrage left. What a bunch of suckers we are.

Anonymous said...

Now why are you bothering to read J of Clinical Psych? Everyone knows its a front for pharma

Anonymous said...

Why is this so much more prevalent in psychiatry? Is it just because drug mechanisms and clinical endpoints are harder to measure? It happens to some extent in other areas (DM), but it's nowhere near this prevalent.

And to be fair, you don't need to randomize to say something is safe/efficacious. You need to in order to say it's MORE safe/efficacious, but I've written papers where randomization was inappropriate (for example, in a rare condition) but the absolute numbers suggest it's OK.

Supremacy Claus said...

Dan: Is there a course required for all Harvard trainees, "Hatred of America 101?" You're obsessed with an anti-corporation ideology as all Harvard education victims are. Harvard teaches all students on pain of shunning, only central government may have any power.

What psychiatrist anywhere in the world, outside of hapless Harvard indoctrination victims, does not know that all anti-depressants have similar efficacy and discontinuation rates, with none better than imipramine?

Having a longer list of medications help clinicians match one of them to the patient that only responds to that one.

Is there any medical student in the world who does not know that after the first week in a psychiatry rotation?

Jim Sabin said...

Danny -

As always, thanks for the excellent post. Since busy clinicians often just read the abstract, the conclusion of the Cymbalta abstract is going to persuade trusting readers to a false and potentially harmful conclusion.

Best

Jim

RhetRx said...

Well, heads may not roll at the Journal if they're making thousands of dollars off the deal -- classify it in a "special supplement" to the journal and you'd see a lot of editors willing to consider the arrangement, if not outright encouraging it.

As for Supremacy Claus, the reason the infighting over minor differences among drugs happens in the first place is _because_ the differences are minor and each company is jockeying for the best position before losing patent protection. You'll see similar comparisons among oncology meds [If I go on to compare this to NASCAR racing behavior as the green flag drops, does it sound any more pro-American? Just wondering...]

Anonymous said...

obviously you find the journal respectable enough to browse in order to borrow ideas for your research update

therapyfirst said...

I hope people will consider reading Charles Barber's book "Comfortably Numb" about the overprescribing of psychiatric meds. He may not be a psychiatrist, but he was a mental health care provider in past work.

As for Clinical Psych, I stopped getting it years ago after the editors wrote one of the more obscene support editorials for pharma sponsored studies at the time. I have my doubts about Current Psychiatry now, so read that journal cautiously.

Again, if you don't like what you read in Dr Carlat's blog, why are you commenting on it negatively. I guess unbiased and objective aren't vocabulary terms in Pharma talk these days!

Daniel Carlat, M.D. said...

Why do I keep reading Journal of Clinical Psychiatry? Because, in general, it's a pretty good source of practical clinical research. Unfortunately, a big part of the Journal's business model is helping drug companies market their products, not only through legitimate (because transparent) advertisements, but also through their deceptive supplements. In addition, it seems that in every regular issue, the editors allow a couple of articles to sneak through that are Phase 4 trials designed to promote some marketing point. This was precisely my problem with the Cymbalta "study." As a demonstration that abrupt switching is well-tolerated, the article has some real (albeit minor) clinical value. But the promotional wordsmiths got their hands on the abstract and turned it into a mini-advertisement for Cymbalta. Who's to blame? Unfortunately, the blame lies squarely on the Journal's editors, who should have insisted that the abstract be rewritten to accurately reflect the content of the article. I assume (but cannot prove) that they allowed this to be printed because of pressure from the publisher, who knows that reprint orders are a huge source of profit. And as printed, with the deceptive abstact, they are much more likely to get a lucrative reprint order from Eli Lilly.

Supremacy Claus said...

Dan: You had the best training in psychiatry at Harvard. We would all like to hear more about it.

Did it include indoctrination sessions in left wing propaganda? Every utterance out of its alumni and faculty seems to have come from the same cookie cutter.

Perhaps, a separate blog entry would do it justice. I am sure, many of its grads will love to chime in. Tell us about the best psychiatric training in the country. What made it so good?

Anonymous said...

Suicidal ideation associated with duloxetine use: a case series.
J Clin Psychopharmacol. 2008 Feb;28(1):101-2.

Michael Benjamin said...

Hey, Dr. Carlat, maybe it wasn't a "blatant oversight" that led to the final sentence in the abstract! Perhaps, as you allude to, the journal purposely allowed the line so as to ensure massive reprint sales.

There is a doc at my hospital who is fond of saying, "most of what you read in journals is garbage." I think he might be right in this case.

Feel free to check out my independent online community at InteractMD.com

Anonymous said...

Supremacy Claus has an obvious problem with Dr. Carlat's Harvard education (hey, there are some great meds out there for obsession!), but where did HE learn that questioning the status quo and advocating for positive change equate with hatred of America or left wing propaganda? Exactly the opposite. Read the founding documents lately?