Six Weeks to Recover a Grin

I am among those who can be blamed for the disastrous MHRA supported drug label and messaging for antidepressants, along with Guideline recommendations that tell doctors and patients that it can take 6 weeks for antidepressant induced recovery to take place.

This idea has been interpreted by MHRA and lots of psychiatrists and family doctors, who have no background in pharmacology, to mean that people who are becoming badly agitated or actively suicidal in the days after starting an antidepressant should continue on their treatment regardless – because it can take 6 weeks to recover.

At a time like this, when they are feeling worse and their life is obviously at risk from this terrible disease, they more than ever need an effective treatment.

Here’s how this lethally crazy idea took hold.

SSRIs

Back in 1970 when SSRIs were just a glimmer in Arvid Carlsson’s eye, the focus was entirely on catecholamines – Restoring the Magic to Healthcare. Noradrenaline was low in depression and tricyclic antidepressants supposedly restored it to normal. Julius Axelrod got a Nobel Prize for this work. But there was one big problem. Tricyclics rapidly increase noradrenaline levels, but recovery didn’t immediately take place.

Unlike the muscle relaxation from benzodiazepines, the increased energy on stimulants, or the reduction in tension on antipsychotics, which are clear effects that start within minutes or hours of taking a first pill and remain the obvious therapeutic effect throughout treatment – if that is they are present to start with, there was no immediate onset obviously good thing that tricyclics did. Recovery took about 2 weeks or so to appear.

It was easy to think that rather than having a useful effect, tricyclic antidepressants must be correcting a defect. Everyone at this point still agreed the defect was in the noradrenaline system.

The goal was to make pure noradrenaline reuptake inhibitors in the expectation that these would be more effective and free from the nasty side effects caused by tricyclic actions on serotonergic, histamine and cholinergic systems.

Fridolin Sulser claimed the key lay in changing noradrenaline receptors. Chunks of receptor protein felt like they might take a little longer to change than simple neurotransmitter levels, perhaps even two weeks. Everyone jumped on the Sulser bandwagon.

Meanwhile, Arvid Carlsson came up with the idea of an SSRI. He based this on an almost immediate onset useful effect that some, but not all tricyclics, had. Listening to doctors and patients, Carlsson heard that clomipramine, imipramine and amitriptyline produce an early onset anxiolytic or serenic effect that made them useful for obsessive-compulsive disorder (OCD) and other anxiety states in a way that desipramine and nortriptyline weren’t.

Desipramine and nortriptyline are also tricyclic antidepressants. They inhibit noradrenaline reuptake but not serotonin. They are ‘purer’ versions of imipramine and amitriptyline, which inhibit both serotonin and noradrenaline reuptake.  They don’t do this something that imipramine for instance does – and it was becoming clear they are in some way weaker than clomipramine, imipramine and amitriptyline.

Rather than follow the noradrenaline inhibitor crowd, Carlsson set about making a pure serotonin reuptake inhibitor. He created zimelidine, Zelmid, and then alaproclate, both of which got withdrawn because of problems. But not before clinical trials and early practice showed that Zelmid worked. Carlsson inspired Rhone Poulenc’s indalpine, which worked but was removed because of liver problems. Under his influence, Lundbeck converted talopram, a noradrenaline inhibitor, into citalopram and later escitalopram. Finally, before Prozac, there was fluvoxamine – Luvox – later linked to the Columbine massacre.  LTEP

Prozac was about the sixth SSRI, made after Lilly figured noradrenaline reuptake inhibitors, like nisoxetine and atomoxetine were not the way forward. Desipramine and nortriptyline were purer but also weaker than their parents, imipramine and amitriptyline, and not especially anxiolytic.

Zelmid, Prozac and all the SSRIs were also pretty weak antidepressants, but this didn’t worry companies. They were serenic – anxiolytic – a bigger market than antidepressants.

SSRIs were/are immediately serenic even in much lower doses than the standard Prozac, Paxil and Zoloft doses. However, selling doctors the idea that, unlike benzos, these anxiolytics would not cause dependence was a problem. This problem could be camouflaged by rebranding them as antidepressants. No-one then thought antidepressants caused dependence.

An added advantage was that unlike anxiety, everyone at this point bought into the idea there was a lesion, a defect, in depression, which antidepressants correct, cure, remedy but it takes time.  They don’t sap our moral fiber by providing some early onset tranquilizer type crutch.

Even though both noradrenaline and serotonin inhibitors could now treat depression, even if weakly, Sulser’s Receptor Hypothesis survived. It transitioned from being a Catecholamine Receptor to a Final Common Pathway Receptor.

We held onto the idea of a chunk of moody protein that treatment took a few weeks to chisel back into the right shape.  But this Grin now got in the way of seeing the Cat – the immediate onset benefit SSRIs can produce

The weak antidepressant effects of SSRIs made it difficult to show they worked in the conventional antidepressant trial format. Clomipramine and imipramine walloped them in trials of severely depressed patients. In mild depression trials, SSRIs struggled to beat placebo at the 6 week point which was needed to get a license.

But beat placebo they did in roughly half the trials done and companies got licenses to claim SSRIs are antidepressants. Whence came the idea you must keep taking an antidepressant for up to 6 weeks – it used to be 2 weeks for tricyclics – even if you are having a disastrous short-term effect.

Companies could have run one week trials of their SSRI against placebo showing that 50% of us get an almost immediate serenic effect which placebo takers don’t get. Critics of the drugs, like Joanna Moncrieff, would not then have been able to say the effects are all placebo based or in the mind.  It would have been clear what the drugs do.

But clarity about what their drugs does not necessarily suit company marketing. Instead, doctors were sold the line that nothing useful was happening in the short-term. There was no challenge to the old idea that antidepressants don’t do much for weeks until some chunk of receptor started behaving itself.

In time, in marketing hands, the chunk of moody receptor protein transitioned into a chemical imbalance.

The problem with this idea is in mild depressions, natural recovery kicks in by 6 weeks even on placebo.

The best way to interpret the trial data is that some of us – no more than 50% of us – get a serenic benefit early on – within a day or two. Family and friends can often spot this. This can settle us down, leave us bothering a little bit less, until natural recovery kicks in.

If you aren’t more serenic after 48-72 hours you need to get off SSRIs/SNRIs. It is possible to damp down the agitation you may be feeling with a benzodiazepine or beta-blocker or other drug – this is what companies did in their trials – and you may limp through to 6 weeks and even appear to recover but this recovering will not be brought about by the SSRI/SNRI.

You may appear to be better and be advised to remain on treatment but you will be unstable and all it might take to trigger severe suicidality may be a tweak of the dose – up or down.

So how did I contribute to this confusion?

The first edition of Psychiatric Drugs Explained came out in 1993. Depression was still rare and SSRIs not much used. It offered the standard narrative about antidepressants.  Rather than do something useful, they correct a problem and this takes time, which we think has something to do with a final common pathway rather than low transmitter levels.

When revising PDE later it wasn’t absolutely clear this bit needed changing and later editions continued to take this line – with a nuance. As ever milder conditions were being diagnosed as depression and treated with SSRIs, there was an emphasis on the data that these conditions recovered naturally in a few weeks. There was also a new emphasis on evidence that those of us who recover without meds or therapy are more resilient in the longer run.

Therapeutic Principles

In mitigation, I also argued in a 1997 article that rather than being Magic Bullets, it was looking like serotonin and noradrenaline reuptake inhibitors offered different Therapeutic Principles. Some like noradrenergic drugs could be energy and vigilance enhancing, while serotonergic drugs were more serenic or anxiolytic.

The reason tricyclic are more potent that SSRIs or pure noradrenaline reuptake inhibitors is that they embody a number of beneficial effects.  They can enhance energy, make serene, as well as sedate by acting on histamine along with being mildly euphoriant through their anticholinergic effect.

I repeated this point in a 1998 article responding to Joanna Moncrieff and Simon Wessely (writing together) claiming the effects of antidepressants (SSRIs) were in the mind.  I don’t think either of them understood it.

I got the idea of Therapeutic Principles from Arvid Carlsson. For more on Therapeutic Principles and why it makes sense to illustrate the point with a slide on Constipation and Laxatives see Madness, Normality and Antidepressant Dysregulation – at Slide 45.

Retrospectively, it is easy to see that the times needed a few unambiguous statements.

1. SSRIs are serenic. If this effect is not clear early on, you should stop the treatment and change to something else – as Arvid Carlsson showed even nicotine might be better than an SSRI if it is not making you serenic almost immediately.
2. Because they are serenic for some of us SSRIs, if they suit us, can be helpful across a range of nervous states we now call depression but are better called anxiety states – many of which recover naturally in a few weeks.
3. Much lower doses of SSRIs should have been available from the start – some people might have done best on doses between 1-5 mg of fluoxetine, paroxetine or citalopram.
4. Nobody should ever have been put on doses more that 20 mg of the above 3 drugs or their equivalent for other SSRI or SNRI drugs. There is no point – 20 mg is the top of the dose range – things can only go wrong.
5. SSRIs are ineffective for severe depression (melancholia).
6. Tertiary Amine Tricyclic Antidepressants (clomipramine, amitriptyline and imipramine) are more effective for melancholia than SSRIs. These TCAs can do useful but subtle things in the short term and typically bring about a notable change within two weeks – although melancholia can also recover naturally in 5-6 months.

Claims that SSRIs don’t work, or their effects are all in the mind are wrong and almost as bad as the idea that someone should persist with treatment even if getting worse. SSRIs clearly do useful things for some people that are real enough and physical enough for family and friends to spot.

Many of those who are helped should probably not persist with treatment beyond a few weeks. Some folk who are more introverted and prone to rumination may benefit from longer periods on treatment, but they need to be aware that long-term treatment risks compromising health and even shortening lives.

Suicide Warnings

Louis Appleby, Britain’s Suicide Czar, and others linked to National Suicide Prevention Programs in the US, Europe and elsewhere skirt around the risks of antidepressant induced suicide – See An Appleby a Day Does Not Keep Suicide Away.

As with the role of antidepressants in chiseling away at a bit of moody receptor protein, outlined above, there is a background here that perhaps contributes to a policy dictated approach that is doing more harm than good.

Aspirin and Tylenol (paracetamol/acetaminophen) have been and likely still are among the commonest drugs implicated in overdose deaths, intentional or accidental.  For decades, the companies making them successfully campaigned against suicide warnings. The argument was that letting people know you could kill yourself with these drugs would increase rather than lower deaths by overdose.  Warnings might kill more people than they saved. It took a long time to get warnings in place.

The same factors likely applied to the original tricyclic antidepressants, some but not all of which can be lethal in overdose.  There is a twist to this story. The SSRIs when they came on the market were sold as safe in overdose compared with those dangerous tricyclic drugs. although this is not entirely true.  Companies were dishing out backhanded warnings – the labels of the older drugs didn’t tell you an overdose of tricyclics could kill you, but companies made absolutely sure every doctor got the message.

We now avoid mentioning overdose deaths or suicide in drug labels if we can.

The SSRIs trigger suicide problem has been buried in our ‘traditional approach’ to this tricky issue.  The problem with the traditional approach in this case is that no-one is accusing SSRIs of being dangerous in overdose. The problem that they and antipsychotics pose is that they are vastly more likely to trigger suicidal (and homicidal) behavior than anything else out there – including stimulants, cocaine, opioids and all other prescription drugs.

We are increasing not decreasing suicide rates by failing to warn about this.

Suicide and Withdrawal

Faced with their golden geese laying suicidal events in their SSRI trials, companies took suicidal events from the run-in phase of their trials – where people had been whipped off  prior treatment with an antidepressant – and moved these events into the trial placebo column.

This breached regulations and should not have been done. In so far as it could not have been done by accident – all companies could not accidentally have had the same ‘accident’ – it looks like fraud and FDA have tacitly conceded this.

If you can call it a silver lining, the silver lining in this case is that it drew attention to – or could have drawn attention to – the consequences of whipping people off prior treatments. Abruptly stopping causes suicidal events – perhaps even more so than at any other point in treatment exposure.  This became clear in company healthy volunteer trials also – most notably with the death of Traci Johnston.

Study 329 is famous for showing a huge gap between the rate of paroxetine suicidal events compared to placebo. GSK never published the continuation phase of this trial which showed the Taper phase to be the riskiest phase of all.  Study 329 Continuation.

Their healthy volunteer trials enabled companies to design their trials to hide the problem of withdrawal – any anxiety, depressive or suicidal symptoms were viewed as relapse of the original condition rather than withdrawal symptoms that had also occurred in healthy volunteers.

This applied to supposedly independent studies like the Antler or Kendrick studies also. Because the teams behind these studies don’t have access to prior healthy volunteer data they assume that anxiety, depressive and suicidal symptoms in patients withdrawal are evidence of relapse.

In addition, there is a regular stream of comments in response to articles in lay media on antidepressants and withdrawal hazards claiming that these drugs have saved the life of the person commenting – which they probably have but by treating a withdrawal issue rather than a mood disorder.

Amidst all the noise, it is difficult to detect the voice of anyone who knows anything about pharmacology or how clinical trials operate or can be made to operate. The best the critics seem able to do is squirt the octopus ink of ‘conflict of interest’.

Footnote

When the SSRIs became more than a glimmer in Arvid Carlsson’s eye, Lilly who had been developing nisoxetine and later atomoxetine pivoted to fluoxetine.

In the 1994 Wesbecker lawsuit, John Heiligenstein, one of Lilly’s honchoes when interviewed under oath seemed to have a terrible memory of the details of fluoxetine’s development.  He said yes he had, this was because he had Adult ADHD.  How did he know this – well he said he had found atomoxetine helpful for it.

Based, it seems, on his experience or whatever. Lilly were then in the process of bringing atomoxetine on the market not as an antidepressant but as Strattera. a treatment for ADHD, and their marketing department had a wonderful idea – they needed to persuade people that ADHD was not something you grew out of – you had it for life.

Since then millions of adults have developed ADHD – who never had it in childhood.