Wednesday, February 25, 2009

Of Seroquel, Sex, and Secret Documents

Seroquel (generic name, quetiapine) is an atypical antipsychotic that made a tidy $4.4 billion for AstraZeneca last year. Seroquel’s golden goose status is based on a string of FDA approvals for schizophrenia, mania in bipolar disorder, and more recently, depression in bipolar disorder. In addition, it is commonly used off-label for insomnia, and the company is applying for FDA approval for the treatment of generalized anxiety disorder.

The drug’s approval for bipolar depression was particularly valuable, because it opened the door to off-label use in garden variety, unipolar depression—a much bigger market than bipolar depression. The problem is that the “BOLDER” studies, which formed the basis for FDA approval, had a glaring weakness. The patients who were enrolled were very different from the patients we see in clinical practice. The BOLDER studies excluded patients with:

--Current depression lasing longer than 12 months
--Substance use within 12 months
--Any other recently treated psychiatric disorders
--Suicidal ideation

In addition, any patients who had tried and not responded to two previous trials of antidepressants were not eligible for the study. This restriction is especially problematic, because by the time most real patients get the diagnosis of bipolar depression, they have already been on many rounds of antidepressants.

Thus, Seroquel is approved for a very special and unusual kind of depressed bipolar patient—someone who is who is unlikely to be seen by practicing psychiatrists. The patients I see are sicker and much more complex; prescribing Seroquel for them may or may not make them better, but is all but certain to cause significant weight gain, rendering them more vulnerable to heart disease and diabetes.

Which is all a very long-winded introduction to the latest scandal-in-the-making from AstraZeneca. As covered in BusinessWeek
, and the Furious Seasons blog, AstraZeneca is being sued by thousands of patients who allege that the company hid information about Seroquel’s risk of weight gain and diabetes. 6,000 claims have been consolidated into a single case before a federal court in Orlando Florida.

Reams of documents have been shipped in by AstraZeneca related to the case, but none have been disclosed yet, because no actual trial has commenced. It is possible that in order to avoid disclosure of embarrassing documents, the company will settle the case and prevent a trial. But the news agency Bloomberg has filed a motion to force these documents to be unsealed, citing “the public's right of access to judicial documents."

Presumably, these documents will be just as damaging as the leaked Zyprexa documents which showed
that Lilly’s “Viva Zyprexa” campaign deliberately encouraged primary care doctors to prescribe Zyprexa for unapproved uses.

Maybe more damaging, actually, because at least the Lilly documents did not involve a senior company scientist having sex with a company funded researcher and a ghostwriter. That’s right, folks—according to one publically available court document, the company’s lead research director, Wayne Macfadden, was getting awfully busy with women who did Seroquel research and wrote up the BOLDER results.

The document alleges that Macfadden's relationships with the women were "relevant and highly probative evidence of one high level AstraZeneca employee's determination to exploit his sexual relationships with these women in order to elevate Seroquel's status in the prescribing medical community through supposedly 'independent' publications of Seroquel safety and efficacy data….Moreover, the mere existence of these relationships calls into question the integrity of the scientific work product of those involved."

We don’t know if all these allegations are true, although Macfadden has already admitted to the dalliances.

At this point, AstraZeneca needs to do the right thing, which is to unseal the secret documents. Only then will the company be able to clear its name.

24 comments:

Anonymous said...

Agreed. Excellent Post.

Anonymous said...

unseal to clear it's name, or, seal it's doom.

I have been saying here for a year now this push for multiple applications for novel antipsychotics has gone beyond out of hand. If you are starting to hear cackling, you know who is home to roost.

I forget the name of the writer of a column in Psychiatric Times back in the mid to late '90's, but he was a psychologist/lawyer who wrote about the struggles with managed care, and in the last, or near last one he wrote, it was about the unfortunate collusion a physician had to engage to get insurers indicted in malfeasance/neglegence. The same seems to go for pharma's insidious role with meds. In the end, MDs will have to go down with the bigger players. Shame we live by the adage, 'the road to hell is paved with good intentions.'

I won't. Glad you have returned from your 2 week hiatus.

therapyfirst

PS: clinicians need to read Furious Seasons and if willing to contribute to his ongoing blog efforts AND ignore the commentary, I feel it is worth the investment to send some funds. Mr Dawdy has a good intent with his work, the shady followers notwithstanding.

Just an opinion.

Anonymous said...

Dr. Carlat, I know you don't like to publish posts that stray from the topic but I hope you make an exception in this case.

Allowing TF to post and slander readers of the Furious Seasons Blog by calling them shady unintentionally makes it seems like you condone that type of practice. Would you have allowed that if the post had been from someone who is not a psychiatrist? I don't think so.

I have greatly enjoyed your blog but if you allow TF to continue to slander people in this manner, then you are essentially no better than he is.

AA

Daniel Carlat said...

Anon--My policy on comments is to err on the side of permissiveness. Accepting a comment certainly does not imply that I agree with it. Slander means making a statement of fact you know to be false, and TF made a statement of opinion. I discourage rank insults and name-calling.His comment that Furious Season's followers are "shady" certainly treads a fine line in terms of insults, but I chose to publish it and allow others to question that opinion so that he can explain himself. Which is evidently what it happening.

Anonymous said...

Dr Carlat, I would hope you will publish this comment as it addresses both yours and Mr Dawdy's blog's value to the mental health community who frequent the internet.

Your efforts to write your blog are most likely maintained in a different manner financially than what Mr Dawdy depends on from his sources. He is in the midst of a fund raising drive and has published a post that you have referred to in this posting now. I am concerned that Mr Dawdy provides a service that, while might be exasperating or annoying to clinicians, he is more often on the mark and needs to be heard and considered, but might be lost if not financially supported as he sees fit. As I have told him both as a commenter (I asked and was forced to be banned last year due to my hassles with the crowd I spoke of in the above comment prior, and my addictive nature at the site in general) and as an emailer to him, I feel his commentary group that frequents the site diminishes it's value, which is my opinion and he does not agree with me. That said, I am noting to YOUR readers here to check out the site and be wary of the commentary section, as I am directing readers to the meat of the site, i.e. his postings, and take the commentary as it is worth.

Shady is not a term that is slanderous or libelous, which I appreciate your agreement, and I just want to go on the record that those who frequent his site use those terms inappropriately to try to control and diminish opposing views. I am glad you moderate your site, as you have not printed a few comments by me in the past, which in reflection was correct in the end. I just want to support Mr Dawdy, who you reference, as do I to my colleagues. His "supporters" I hope are doing so with their wallets, not just their mouths/keyboards.

Thank you for considering this for publication. By the way, I went back to your March 08 post about how reps feed on reciprocity, as I am at a new clinic which had their first drug lunch yesterday in a long while and were surprised I did NOT attend. I hope they read the post I copied for each clinician's and front office staff's mailboxes. Still a great read to me.

therapyfirst

soulful sepulcher said...
This comment has been removed by the author.
James M. La Rossa Jr. said...

While the recent posting, "Seroquel, Sex, and Secret Documents," is highly interesting for its gymnastic ability to weave together such disparate items as Seroquel’s BOLDER study, the “Viva Zyprexa” campaign, and a lustful Irishman, I respectfully submit that some cautious perspective is warranted.

Upon a very brief first reading of the BOLDER studies’ exclusions, my first thought was that they seemed reasonable to garner approvable status for GAD, so why is DC objecting? Then I realized that this was the original study criterion for the bipolar depression indication and I understood. So the big issue here is NOT with AstraZeneca’s GAD application, but with the FDA’s acceptance of the original criteria. Perhaps it is time for the DSM-crazed world of psychiatry to spend some time customizing the CONSORT guidelines (See: Consolidated Standards of Reporting Trials -- http://www.consort-statement.org/?o=1011) so that researchers can better connect diagnosis and treatment to real patients. Obviously, the FDA could use the help in its assessment of future trials and I know of at least one journal which would publish such customized CONSORT guidelines. It seems axiomatic that ANY trial, which recruits patients reporting suicidal ideation, for example, be given “bonus points” as BLACK BOX WARNINGS continue to sweep through the world of psychoactive drugs.

Secondly, we need to be equally cautious in quoting from “publicly available court documents.” ESPECIALLY about sexual allegations. These “court papers” are nothing but motion papers written by the opposing side. Of course they are going to claim that Mr. Macfadden’s actions were “relevant and highly probative evidence of one high level AstraZeneca employee's determination to exploit his sexual relationships with these women in order to elevate Seroquel's status…” To repeat this nonsense is just that! Does it occur to anyone that these “women” are already on the payroll, so what more could MacFadden’s dalliances buy? Would it matter if one of these ladies was now Mrs. MacFadden, suggesting that love and not influence was the motive? Impeaching AstraZeneca over a sexual matter exposed in the heat of a lawsuit where the parties are throwing the proverbial kitchen sinks at one another has no place in serious medical debate. Thank you. -j

Daniel Carlat said...

James,

I certainly agree that we don't want to perpetrate wrongful character assassination. In this case, Dr. Macfadden has already admitted to the relationships. One can argue, just as is true when discussing relationships between doctors and patients, that there is something inherently ethically unacceptable about about choosing to have a sexual relationship with somebody over whom you have professional control, especially when the outcome of that control can have effects that ripple out into the public health.

Of course, it would be much more egregious if it turned out that there is evidence that his relationships were willfully exploited—but at this point this is simply an allegation, and I hope any reader can understand the crucial difference between an allegation and a fact.

James M. La Rossa Jr. said...

Danny,

You make ample sense, as always, but quoting from one side's motion papers without representing the other POV puts the reader in the position of accepting the allegation AS fact. Furthermore, the language used is exploitive, much like The Starr Report exploited the sexual content to impeach President Clinton. In this case, repeating these allegations is unworthy of the professional debate customary on The Carlat Psychiatry Blog. Forgive me if I'm being a prude. Best, j.

soulful sepulcher said...

Today's court proceedings are producing some internal documents, according to Bloomberg at 12PST.

James, as always an interesting discussion between yourself and Carlat. I always appreciate reading.

Doug Bremner said...

I posted a comment on FS that I will pledge ten dollars for every psychiatrist that makes a donation to the site. I too think the use of atypical antipsychotics for depression is way over marketed and overused. It is a sad reflection on our field that we let the pharmaceutical industry bring us down this path.

Anonymous said...

Therapy First, the issue is that you are very unfairly painting FS readers with a broad brush and it needs to stop.

You would be on me in a heartbeat if I said that all psychiatrists are pharma lores, which obviously, they aren't. So why is it ok for you to paint people with a broad brush insult?

To the topic on hand - I know someone has developed serious medical problems as the result of antipsychotics and I am greatly worried This person is now on Risperdal but previously took Seroquel. I

Anyway, I will be watching these events with alot of interest.

AA

Anonymous said...

As fascinating as this group sharing is, to Danny's original posting:
I'm curious, in what way are those exclusion criteria different from nearly every FDA registration trial in mood disorders in the past decade? The inconsistency between efficacy trials and clinical effectiveness is discussed to death (including by those same awful clinical researchers who your readers would like to run out of town with pitchforks). But as long as the objective is to get a drug approved, the companies will do what the FDA asks them to, and only that: show separation from placebo. As with CME, you get what you pay for.

Doug Bremner said...

To get back on topic, why would anyone take a patient with new onset depression, who had not had a couple of antidepressant trials, and put them on a drug that could cause diabetes and/or akathisia?

Anonymous said...

I agree with localdoc about the selection/exclusion criteria for clinical trials for most psych drugs. So many people are excluded that the trials are not representative of the intended population and not generalizable to the community. Yesterday, I read the NIH on the required elements of ethical clinical trials and generalizability is one of the most important. I believe that a case can be made that these trials which are so exclusive are not scientific and not ethical.

soulful sepulcher said...
This comment has been removed by the author.
Anonymous said...

Um, a little reality check here folks. When you market a drug that brings in double to quadruple the profit margin when it gets indications for illnesses that have a larger incidence in the population, greed outweighs responsibility. Haven't we seen this over and over again, not just in the pharma industry, but with business in general? So, to the above comments that seem to ask if Dr Carlat is jumping to conclusions, I just want to know how much subterfuge, deflection, and projection do we as responsible clinicians have to wade through before the adage of "if it walks like a duck and quacks like one..." is applicable.

Sorry, while I respect the adage of innocent before guilty, do responsible and ethical people/organizations ask for evidence to be suppressed "for the good of the public"? Doesn't that line alone just want to make you want to slap the commenter real hard!?

These AZ reps have been disgusting in minimizing and rationalizing the side effect issues these past 4 plus years I have been challenging them about them. First I watch Lilly bs me about Zyprexa, then AZ with Seroquel, and BMS that makes Abilify is right behind them both.

This is bad stuff folks, and just go back to posts I offered early last year at this and other blogs that asked specifically what is coming down the pike with Seroquel. I am sorry to say this, but I hope some clinicians end up culpable in these suits for pushing these drugs to indications that were just wrong. And you as colleagues who just sat back and said nothing ain't so innocent if you knew better to begin with.

Harsh, but needs put out here!!!

Anonymous said...

Been an interesting last couple of days in reading psych journals, the most interesting was what I read in American Psychiatry News, at www.americanpsychiatrynews.com .

I like Glen Treisman's columns as he usually attacks poor judgments by intervening sources who have no business intervening, but I was a bit taken aback in reading the CME piece inside, that, in my opinion, was a bit too supportive of using Seroquel, albeit in the refractory patient determination. The article in my opinion was not as balanced when it spent 2 paragraphs at a time on antidepressants, buspirone, benzo's, and then was about as long for use of antipsychotics per the total content for the others, the second half of it(and twice as much as mention of Zyprexa) mentioning Seroquel. No mention of AZ as a sponsor that I could find, and if I missed it in the fine print, my bad in that, BUT, not what I would have expected from this publication.

I emailed Dr Treisman in abbreviated form these similar comments and look forward to his reply. I say it here to encourage readers to look at the articles and form your own opinions.

Hope March is better than February was to me. And, as a hopefully final goodbye to Supremacy Claus, if your last post was meant as humor, you bombed, and if meant to demean, then I hope the author finally realizes you are a detriment to this site and bans your sorry ass away.

Sorry for ending that way, but I do not like these people who believe that insulting and demeaning others directly should be tolerated. Maybe I am guilty by the last paragraph, but mincing words with an intent to be kind with people like SC is lost in the wind.

Have a nice weekend.

Cara T. Hoepner, MS, RN, CNS, PMHNP said...

Hello, all, and thank you James for introducing me to this blog, and esp this post on trial (and other) methodology. I've been reading the Carlat Report in hard copy for several yrs now, but am new to blogging, so pls enjoy my first attempt!

EXCLUSION CRITERIA
Dr. Carlat makes an excellent point about the exclusion of many people who attempt to take part in efficacy trials. It's unfortunate that, the more highly experimental (i.e. scientific) a trial is, the less generalizable it is. Less scientific trials such as retrospective analyses and quasi-experimental studies (open-label and/or not randomized, etc.) generalize more easily to real-world patients, but lack the rigor required to demonstrate efficacy. Trials such as STAR*D, with its equipoise design and few exclusion criteria, attempt to bridge the gap.

---------------------
The Code of Federal Regulations requires only that patients recruited for a study be properly diagnosed for the medical condition being studied; there is no requirement or recommendation for exclusionary factors (21 C.F.R. 314.126). In fact, exclusion criteria may fail their goal of optimizing drug-placebo differences.

Some criteria may be judged more useful or ethically important than others. For example, psychotically depressed or suicidal patients should not be treated with placebo, bipolar patients should not risk drug-induced mania and may respond differently from unipolar patients to drug therapy, current or recent substance abuse may obscure course of illness or preclude response, and treatment-refractory patients may also fail to respond while very mildly ill patients may not require medication. That said, a HAM-D 17 cutoff of 20 like that used in BOLDER excluded—as a single criterion—54.3% of one clinical treatment-seeking sample (Zimmerman et al's) from antidepressant efficacy trials. Duloxetine efficacy trials utilized a cutoff of 15, and hopefully this will become more common?

Questionable criteria, not widely employed, include brief or lengthy episodes, selected non-“severe” Axis I disorders including dysthymia, and medical illnesses. Where these are concerned, would including more patients to increase statistical power offset any possible increase in placebo response? Antidepressant medications may benefit some depressed patients with comorbid anxiety, medical illnesses, mild depression or chronic depression.

The above is loosely quoted from a paper a friend and I wrote (J Clin Psychopharmacol, 2007), in which we managed to exclude 91.4% of our treatment-seeking community outpatient sample from an antidepressant trial utilizing the same 11 criteria Zimmerman and Posternak had used in a similar study. We then went a step further and, employing only the five most common and most arguable criteria, we still knocked out 74.7%. Had we added the treatment non-response criterion, we'd have eliminated 79.3%.

So I was tempted to analyze my data to see how many clients I could knock out using what criteria I have in my dataset that are similar to those in BOLDER—anyone kinda curious, please ask.

BOLDER was far more lenient than many trials, BTW, in terms of how some of the criteria were defined. The recently treated psychiatric disorders listed are not just any of those which have been treated in past 6 mos, but only those which have been of primary focus. Most trials exclude for comorbid diagnoses on Axis I, and a large number exclude for borderline personality disorder. The most important distinction is the operationalization of treatment resistance: this study used the definition proposed by numerous authors. The trial did not exclude for failure of two or more antidepressants, but rather of adequate trials of antidepressants from three or more classes, during the current episode... an episode lasting no longer than 12 months. This means a person may have had to try an SSRI, an SNRI, and Wellbutrin or Remeron within the current episode in order to be excluded on this count. Many other trials are less lenient in that they use >2 trials within same episode, but 1.) few trials exclude for chronicity (episode ay be far longer) and 2.) most don't require that the drugs be from different classes. Other criteria were more stringent than is typical, including a HAM-D17 cutoff of 20 (18 is more often used), and substance misuse w/in 12 mos (6 mos almost always used).

To clarify the information on suicidal ideation, BTW, only "serious" SI or homicidal ideation were excluded for, as "decided by investigators" (?) Typically a scale such as the SADS-SI is used, w/ a cut-off of > 2.

-------------------------
As regards the GAD studies, comparison w/ 6 RCTs of SNRI, SSRI, agomelatine, and other atypicals revealed some of which had truly lenient criteria and others which were stringent, and those w/ quetiapine were abt midline among them. The GAD criteria were actually more restrictive than were those employed in BOLDER, with any comorbid PTSD, OCD, and severe personality disorders (BPD and ASPD) excluded for, and ongoing psychotherapy. Typically trials only disallow initiation of psychotherapy, and ask that those in regular sessions maintain the type and frequency of their therapy during the trial. The GAD studies also excluded but pregnancy/lactation/absence of birth control (paroxetine category D pregnancy, but lactation safe). They also list concurrent use of psychotropic medications/herbs/supplements within 2 weeks of study entry. These items were likely used in BOLDER as well; they are standard, and the BOLDER reports listed their criteria as "main" criteria, so these may be in addition to those disclosed.

-------------------------
An aside: note that MADRS was primary efficacy measure in BOLDER, where HAM-D was used as inclusion criteria, supposedly to "minimize rating inflation"... in these trials, both the MADRS and the HAM-D are used both for assessment and follow-up ratings. Typically the HAM-D 17 is used for f/u, and often the HAM-D 21 is used for initial assessment. Choosing later, after data is analyzed, which outcomes are to be reported is playing dirty pool. First thing I thought upon seeing this was that it seems seems "rating inflation" may have worked to advantage in outcomes measurement... I mean, weren't they leaving themselves open for that one? So I went hunting and... where I didn't find corroboration—and I didn't find any rants from David Healy (though he has a site!)—I did find a blog on which someone had reanalyzed the stats of BOLDER II, demonstrating the effect of some alternative methods used to calculate effect size (LOCF vs mixed effects models). Check out this clinpsyc blog....

-------------------------
SEX
As for the sex-for-reporting bias trade (this case gives the "cherry picking" of results a whole new spin), it may be variably related here... James makes a cogent point abt what sort of, umm, pull McFadden might have w/ established AZ employees. The instance in which improper use of a boardroom table might net profits for a writer is in the case that these women are freelance, which they indeed may be... drug companies often hire mktg companies to do this sort of thing, or pinch hitters as copy writers and the like. And then, of course, there's always those little blue boxes... (for male readers: Tiffany's).

Thanks for listening—Cara

Oh—PS—I have to add a little bit on the wt gain and diabetes situation... back in 2001-2002 of this century, before it was v well known (or at least before doctors were acknowledging) that the atypicals in use at the time may cause a few problems, a psychiatrist I knew—a well-respected researcher at UCLA (and no, not in Gitlin, Frye and Altshuler's group), and PI on phase II trials of various medications—all but ignored my concern about hyperlipidimia, hypercholesterolemia, and hyperglycemia in our clients. Poor eye contact, restricted affect, the lot. He did, in fact, say he didn't believe it was an issue. Until I shoved a thick print-out of peer-reviewed papers and abstracts under his nose, and in a less-than-casual tone he asked where I got them.
Medline.

Cara T. Hoepner, MS, RN, CNS, PMHNP said...

I was intrigued by the topic Stephany brought up re Seroquel (quetiapine; QTP) as an abusable in prison and on the streets... around 90% of the population I treat in my urgent care setting abuses substances, so I find this of particular interest and want to provide an 'urban' clinician's perspective:

Suggesting QTP as an abusable substance would generalize also to Wellbutrin—which enhances DA, and is also insufflated by some. Controlled substances are listed as such b/c they are highly abusable (to varying degrees, hence the hierarchical schedule). We could really pad schedule II if we got going.

QTP [a medication w/ 5HT2a, H1 and DA reverse agonist (the 'proper term' for blocking) properties] is only a drug of abuse secondary to other street drugs in a discrete population and doesn't have abuse/addiction potential on its own. It may be used on the street as "sleeping w/ Sara" or the "'quel"... histaminic drugs, when mixed with opiates, heighten and prolongs the effects of the opiate drug. See Kaplan & Saddock. QTP is the most histaminic of the atypicals. One could get the same effect using Benadryl (diphenhydramine) or Atarax (hydroxyzine)—maybe I should tip them off!

QTP will not, however, be found being abused by those to whom it is prescribed, unless they're involved in street drug culture. In fact, many of the people who abuse it on the street purchased it from a person who had it rx'd to them for legitimate symptoms. You won't find people forging rx's for it, either.
BTW—you mention that you experienced w/d from QTP; this is normal w/ a host of psychotropic drugs. It is worse with those that may cause downregulation of receptors (e.g. opiates and benzos). Not all abusables cause w/d and not all drugs that cause w/d are abusable.

-------------------------
On another front, Stephany mentions that clients "...for the most part are not told these are antipsychotics being re-marketed for a profit for AstraZeneca and BMS." This brings up something I like to tell my clients, which reduces stigma: what if depression had been the first indication? Drug classes are simply reflective of the first diagnosis they received indication for... drugs are not antipsychotics, antidepressants, or anxiolytics so much as molecules that target behaviors! Pass that one on, please...

Cheers—Cara ❦

Anonymous said...

Thank you, James, for bringing some helpful facts into the conversation.

As for freelance investigative journalists, these are typically newsroom veterans, well-seasoned over many years at first-rate papers and highly trained in vetting sources and the like.

Anonymous said...

Doug Bremner wrote:
"To get back on topic, why would anyone take a patient with new onset depression, who had not had a couple of antidepressant trials, and put them on a drug that could cause diabetes and/or akathisia?"
------

Good question, Dr. Bremner!

Personally, I don't know of such occurrences. I only know of Seroquel being prescribed when all else had failed, and it truly performed miracles (without the diabetes or weight gain).

But I'm sure that what you question happens, and here's a possible reason why:

Not only are some physicians reckless in prescribing, but they also get bored with the same ol' Rx. They have "scientist/researcher" envy and fool themselves into thinking they can do their own little clinical trials, despite ridiculous biases and sloppy follow-up. They also want to be seen as cutting-edge, I suppose. JMHO.

Anonymous said...

Seroquel is the most scary drug I have ever taken and I have taken quite a few. It helps in sleep but delivers panic attacks, muscle aches, headaches, gastrointenstinal upset and generally makes someone so sick that all they can do is lie in bed. If you try to have any interaction with peopl while on Seroquel or even days later, you appear to be on CNS depressants. I suppose I am as normal as the next person with a personality and simply suffer from the traumas of a tough bunch of bad luck, with being prescribed Seroquel at the end of a long line of other bad things that have happened to me. Now, I really want to kill myself with this drug running around my brain. Seroquel is great if you want to have no personality and want to be speech and thought disabled.

Anonymous said...

I have been on seroquel for 3 years (Major Depressive Disorder) and have gained 70lbs. I am concerned about my health & will be switching to Abilify soon. I loved Seroquel because it worked. I become exetremely emotional when I don't take it (3x day). I am worried about going off of it, but I have to lose this weight & get healthier. I'll will be going into the hospital soon to get weened off it.