In psychiatry, we’re rarely afforded the definitive data enjoyed in a lot of science. It’s just the way things are. But that doesn’t mean we can’t be glad when some definitive graphs come along – particularly when they support clinical experience. The data in this 2004 paper is an example. I received this link from multiple directions when I wrote about the recent article claiming increased efficacy with increasing dose [an inconvenient truth…]. The graphs are of 5-HTT transporter occupancy at varying doses of five different SSRI/SNRIs [the minimal effective dose marked with the dotted vertical lines]:
by Jeffrey H. Meyer, Alan A. Wilson, Sandra Sagrati, Doug Hussey, Anna Carella, William Z. Potter, Nathalie Ginovart, Edgar P. Spencer, Andy Cheok, and Sylvain HouleAmerican Journal of Psychiatry 2004 161:826–835.
Objective: Minimum therapeutic doses of paroxetine and citalopram produce 80% occupancy for the serotonin [5-HT] transporter [5-HTT]. The authors used [11 C] DASB positron emission tomography to measure occupancies of three other selective serotonin reuptake inhibitors [SSRIs] at minimum therapeutic doses. The relationship between dose and occupancy was also investigated.Method: Striatal 5-HTT binding potential was measured in 77 subjects before and after 4 weeks of medication administration. Binding potential is proportional to the density of receptors not blocked by medication. Subjects received citalopram, fluoxetine, sertraline, paroxetine, or extended-release venlafaxine. Healthy subjects received subtherapeutic doses; subjects with mood and anxiety disorders received therapeutic doses. Percent reduction in 5-HTT binding potential for each medication and dose was calculated. To obtain test-retest data, binding potential was measured before and after 4 weeks in six additional healthy subjects.Results: Substantial occupancy occurred at subtherapeutic doses for all SSRIs. Compared to test-retest data, each drug at the minimum therapeutic dose had a significant effect on striatal 5-HTT binding potential. Mean occupancy at this dose was 76%–85%. At higher plasma SSRI concentrations, 5-HTT occupancy tended to in- crease above 80%. For each drug, as the dose [or plasma level] increased, occupancy increased nonlinearly, with a plateau for higher doses.Conclusions: At tolerable doses, SSRIs have increasing occupancy with increasing plasma concentration or dose. Occupancy of 80% across five SSRIs occurs at minimum therapeutic doses. This suggests that 80% 5-HTT blockade is important for therapeutic effect. Occupancy should be measured during development of antidepressant compounds targeting the 5-HTT.
hat tip to alto et al…
In spite of 25 years of dreams and schemes, if these drugs don’t help on the first shot, they’re unlikely to do much better with prestidigitation. Because of the escalating side effects with increased dose, these drugs, like many in clinical medicine, have a therapeutic window. I’m awed with the number of patient dosed well beyond the window. Getting them to come down slowly is often a major undertaking, though there’s rarely a complaint once on a more rational dose. More doesn’t seem to be better, at least in my hands…
This article is silent on the matter of benefit versus risk of increasing dosage. We need to see NNT benefit compared with NNH toxicity in relation to dose and blood level. I don’t recall seeing any such studies so far.
But increasing dose does make sense if all you are doing is harnessing a placebo effect. By increasing dose, you buy time. The same occurs when you switch to another SSRI. You are just buying time. Wait long enough and nearly everyone with depression will get better. Those who do not, you can call “treatment-resistant”. But if you increase dose and switch drugs for a long enough time, the vast majority of people will get better, and you can declare your treatment a success.
Bernard,
Now that would be some data of value. I couldn’t find it either. With all the Trials that have been done, it could be easily generated were the raw data “transparent”…
Were depressed subjects treatment-naive? My unanswered question is how does the brain adapt when a person has been prescribed 300mg/day sertraline for years, and what effect does that adaptation have during gradual withdrawal of sertraline? This in an elderly person whose renal function is sufficiently poor that not taking sertraline for three days due to a supply problem did not result in noticeable withdrawal symptoms, i.e., who knows what her actual blood titer of sertraline might be?
I don’t know the answer either. One of the confusing things about SSRIs is that even though I suggest taper to everyone coming off of them, some people notice no adverse symptoms and just stop, while others have a really hard time withdrawing. Likewise, it seems to me that people with difficult withrawal symptoms have been on the medication for a long time at high doeses, but again that’s an anecdotal comment.
I think Ryan makes an excellent point.
Me too…
I think I said this before, when in residency, reading how the placebo effect was 33-40%, and then respectful efficacy of antidepressants was 67-70%, I noted that meant medications only are effective for about a 1/3 of people taking them.
Did not go over too well with that attending teaching the class. But, I think it is true, take 10 people with true Major Depression, and do nothing, 3-4 will come back in 3-6 months and feel fairly much better. Now give them all meds at day one, and about 7 will come back feeling better.
The two points to those alleged statistics are simple to me: First, what can we do with those 3 who won’t be seeing any progress after either trial, and second, of those 7 coming back from the meds trial, what do we say to them if they have unpleasant, if not debilitating side effects from meds, depending on what are the parameter high end dosing provided of the med(s)?
“oops, if you were one of the 3 who would have gotten better without more invasive treatment, sorry…”?
Again, to patient readers out there, find the intestinal fortitude to ask the prescriber if he/she would eagerly and without pause take the medication regimen you are asked to swallow.
Especially if it is more that 2 psychotropics concurrently. Watch for two failed replies: the stammer, or, the curt “yes”. If you ask me, the reply is often “if this is the next step, then yes, but, what other less invasive options do I/you have?”
But, what percentage of patients really want to hear the truth, get better with the pill, or, if you want to get it right, face the truth of their status quo? Ask people around you how many really have maximized participating in honest effective psychotherapy, if even going at all!
Which is a separate argument, why are we tolerating charlatans and frauds who claim to be therapists these days? Unfortunately, I won’t be around for that debate, as I am signing out from blog activities in about 10 days.
Cue Mike Meyers in drag from the old SNL skit, “talk amongst yourselves”…
BACK THEN, the numbers were 1/3 placebo 2/3 medication group got better. That was when the NNT was 3-4. Since the definition MDD expanded faster than the universe and SSRIs replaced TCAs, the studies in the past 20 years are more like 50 percent placebo, 60 percent drug. Which of course means that people who were going to get better anyway got better. But why should we worry? The former Grand Poobah of the APA indicated in his recent book that psychiatric disorders were precisely defined, even though the preamble to the DSM says the opposite.
One reason there are so few relevant data is that NIMH entrusted Phase IV studies to Pharma. In turn, Pharma saw to it that lots of self-serving experimercials were conducted but few really cogent studies of significant questions like this one.
Discussion of adaptation to many years of antidepressant treatment is contained in the literature on downregulation, desensitization, or depopulation of receptors, as well as that on iatrogenic blunting, apathy, indifference, and tachyphylaxis. Examples:
Psychiatry (Edgmont). 2010 Oct;7(10):14-8.
SSRI-Induced Indifference.
Sansone RA, Sansone LA.
J Affect Disord. 2009 May;115(1-2):234-40. doi: 10.1016/j.jad.2008.07.007. Epub 2008 Aug 9.
Does tachyphylaxis occur after repeated antidepressant exposure in patients with Bipolar II major depressive episode?
Amsterdam JD1, Shults J.
Giovanni Fava has written many papers theorizing about the effect of antidepressants and withdrawal, such as
Prog Neuropsychopharmacol Biol Psychiatry. 2011 Aug 15;35(7):1593-602. doi: 10.1016/j.pnpbp.2010.07.026. Epub 2010 Aug 20.
The mechanisms of tolerance in antidepressant action.
Fava GA, Offidani E.
J Clin Psychiatry. 2003 Feb;64(2):123-33.
Can long-term treatment with antidepressant drugs worsen the course of depression?
Fava GA.
This is a paper looking at the neurobiology:
Biol Psychiatry. 2003 Nov 15;54(10):1105-17.
Neurobiology of antidepressant withdrawal: implications for the longitudinal outcome of depression.
Harvey BH, McEwen BS, Stein DJ.
There is next to nothing discussing how long it takes for downregulated receptors to normalize after discontinuation of an antidepressant. If anyone has references, please post them!!
It is possible that any given individual’s receptors are either resistant to downregulation (which would render the antidepressant ineffective at any dose) or instantly re-adapt, forestalling withdrawal syndrome. No one really knows.
The rate of antidepressant withdrawal syndrome is 20%-80% (in short, frequent), indicating quick re-adaptation is possible but not guaranteed.
Looking carefully at figure 6 it seems that the general dosing in practice is about double what is needed for 80% occupancy, except for citalopram which appears to be dosed accurately. I wonder, could anyone speak anecdotally about their experience using the lower dosing reflected in this study and treatment outcomes?
For reference, figure 6 lists 80% occupancy dosing as citalopram 20-40mg/d, fluoxetine 20mg/d, sertraline 50mg/d, paroxetine 20mg/d, venlafaxine XR 75mg/d.
If the half life is short (ie paroxetine and to a lesser degree sertraline), should we consider using slightly higher dosing than the 80% occupancy dosing to prevent valleys in concentration? An important note when thinking about this question is that the study says that the serum sampling occurred 6-13 hours after last dose (which seems to me to mitigate the concern of this concentration valley question).
Come on, medication is to some degree just a crapshoot. How does a psychiatrist honestly know what is the best antidepressant to use? Me, having the good fortune to use them all, see what the side effect profile REALLY is, not what the Company that makes it claims, and then realize that what the person who told me of the 4 P’s of pharmacology is true at the end of the day:
Promise (for the first indication it gets), Panacea (the arrogance and greed to think it treats almost everything), then Placebo (not always pure placebo but the prior drugs out there equivalent to the new one seems to be the more reliable known), and finally, the POISON.
Isn’t it just fascinating how these drugs get FDA approved, and then we find out later about that allegedly unknown side effect that almost takes the drug out of the marketplace, if the company who made the Brand name doesn’t have the FDA in it’s pocket, like Lilly as example A?!
Prozac and Zyprexa, gotta love how Lilly sold these drugs. Prozac, one dose fits all; Zyprexa, weight gain and diabetes, oh yeah, a true class phenomena and not just their own product. Oh, and Cymbalta, my bet is somatic MDs definitely out prescribe us now, and the dosages for pain management…
I have a patient who is becoming hypomanic as her pain MD doubled her dose from 60 to 120mg, without notifying me, and writing the script. Gotta love this s— pervasive in American medicine.
2 drugs need to be taken off the pharmacology marketplace, Paxil, and Xanax. Do you know there are still docs who put pregnant women on Paxil, a class D designation now for 4 years?
Anyway, I will miss the banter, but, I won’t miss the BS.
1.
Identification and Treatment of Antidepressant Tachyphylaxis
Steven D. Targum, MD
Innov Clin Neurosci. 2014 Mar-Apr; 11(3-4): 24–28.
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4008298/
Dr. Targum is bold. “It has been suggested that antidepressant tachyphylaxis is a form of relapse related to evolving drug tolerance…” (and discusses a possible role of neurogenesis.)
and
“It has been reported that depressed patients with “true” antidepressant tachyphylaxis may be less responsive to new treatment interventions.”
2.
How often do SSRIs and other new-generation antidepressants lose their effect during continuation treatment? Evidence suggesting the rate of true tachyphylaxis during continuation treatment is low.
Zimmerman M1, Thongy T.
J Clin Psychiatry. 2007 Aug;68(8):1271-6.
http://www.ncbi.nlm.nih.gov/pubmed/17854252
Not so bold. This meta-analysis somehow gave the authors the impression that if a patient stops responding to a drug that initially worked, the initial response was probably a placebo response. (24% of placebo responders and 7.4% of drug responders stopped responding.)
I don’t think they can say much of anything with the data they had, which must be why they describe the study as “evidence suggesting.”
Hedge hedge hedge: “the majority of relapses in patients taking antidepressants during continuation treatment could be attributed to…”
Neurogenesis should be listed as a side effect. Who would take a drug that will turn their neurons into Magic Rocks®?
http://www.wards.com/the-original-magic-rocks.pro
What great comments!
[I don’t know what to call non-Melancholic depression any more, but by whatever name, it’s what I’m talking about below]
We used to see depression as time limited. So the modern practice of antidepressant “maintenance” is hard for me. And it’s obvious that there are “depressive people” – people with characterologic problems in which depressive symptoms come and go. So complaints of “my antidepressant stopped working” have to be taken less literally. Was the earlier effectiveness the placebo effect of being cared for?, or medication effect? Is there a new stressor? Did you miss something last time? Did the husband stop drinking, then start again? Or the other things suggested above.
The complaint “my antidepressant stopped working” is not necessarily a call for a prescription pad with something else or something added, it’s time to take another history in my opinion – a new case. I doubt tachyphylaxis is a major factor, but even if it is, I’d say the same thing.
Neurogenesis? Show me…
Classically, melancholia was seen not only as time limited but also as recurrent. That is, it was seen as an episodic disorder. The natural history is one of shortening well intervals between the episodes. That was the basis for introducing preventative maintenance antidepressant treatment. The rule of thumb that I adopted came from Jules Angst in Switzerland – a patient who has had three lifetime episodes, two of which occurred in the last 5 years, will benefit from maintenance treatment. That was in the days of real antidepressants – the tricyclics. Lithium also has a preventive effect in unipolar melancholia, and it is the first choice in bipolar cases.
I certainly second Dr. Carroll’s point. It’s the reason for my [I don’t know what to call non-Melancholic depression any more, but by whatever name, it’s what I’m talking about below] quip.
Depression and depression are barely cousins…
I recommend reading the following review of “fixed dose” comparisons of modern antidepressants: http://www.dialogues-cns.com/publication/dose-response-relationship-of-recent-antidepressants-in-the-short-term-treatment-of-depression/
As expected by the binding curves in the article discussed by 1BOM, the dose-response curves are mostly flat.
What I find hard to understand is why in OCD higher doses do seem more effective when tested by the same method:
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2888928/?report=classic
dysphoria or anhedonia and…
four of eight other symptoms…
meaning two people can have the same “disease” and only have one of five symptoms in common…
Major Depression is hardly a cousin of itself. Meningitis and a cervical disc have more in common in terms of symptoms.