The tricyclic antidepressants TCAs are a group of drugs of similar structure (hence their tag of TCA, which refers to their arrangement in three-rings), but they are markedly heterogeneous (I.e. very variable and dissent) in terms of their pharmacological actions. Some of them should be considered as mis-classified in that they are not antidepressants. They therefore illustrate that although the structural similarity of drugs may be close, their pharmacological actions can be very different. For instance, clomipramine is a close structural analogue of chlorpromazine, but their effects could hardly be more different. See clomipramine:

http://www.drugbank.ca/structures/structures/small_molecule_drugs/DB01242

and chlorpromazine:

http://www.drugbank.ca/structures/structures/small_molecule_drugs/DB00477

Any discussion that generalises by saying ‘the tricyclics …’ is certain to be unhelpful and equally certain to be inaccurate (e.g. see the appalling APA textbook chapter on TCAs about which I have commented.) See also my seminal review paper about the TCAs [1].

When they were introduced in the early 1960s these drugs were labelled as either anti-histamines, or anti-depressants (according to how their properties were then perceived), some were mis-classified (e.g. doxepin should have been classified as an anti-histamine, and chlorpheniramine as an SNRI anti-depressant — indeed, the bromine analogue, brompheniramine was the molecule from which the first SSRI was actually synthesised). Some readers may know than doxepin has recently been ‘rebadged’ as a sleeping tablet (only 50 years too late! But that is psychiatrists’ knowledge of pharmacology for you!). I have been using it as a hypnotic for many years, and I explained the rationale of that in my TCA review in 2006, table 6: nice to see people catching up at last.

It is logical to select a few of the TCAs for antidepressant treatment and familiarise oneself with the use of those. The notes on ‘receptor affinities’ and ‘toxicity’ clarify and substantiate the rationale for the following preferences, again, see my TCA review paper.

Clomipramine: is probably still be the best and cheapest serotonin and noradrenalin reuptake inhibitor SNRI (But note recent ‘price gouging’).

Nortriptyline: potent as a noradrenalin reuptake inhibitor, modest but useful sedative effect, low anti-muscarinic effect, linear pharmacokinetics, no active metabolites, no significant inhibition of cytochrome P450 enzymes, cheap. It can be safely combined with both an SRI (preferably sertraline) and with an MAOI: so is a good ‘bridging’ drug.

General: I would usually start with nortriptyline, and then progress, if needed to clomipramine. But there is a place for these other TCAs mentioned here.

Amitriptyline: evidence for superior efficacy vs SRIs in severe depression. Many regard clomipramine as an even more potent AD — in my view there is no doubt about the fact that clomipramine is superior.

Imipramine/Desipramine: best side effect vs potency profile (Desipramine), particularly the lowest propensity to cause postural hypotension or anti-muscarinic side effects (note lofepramine [2] is similar and is metabolised into desipramine [3]). But Desipramine is very potent and the usually recommended dose range is too high.

A Brief word of explanation is appropriate here: because Desipramine is not a very effective antidepressant by itself, it was used in very high doses in the 1960s and the recommended dose was often given us 150-300 mg daily. That will undoubtedly produce, in some people, blood levels that are in the toxic range. The toxicity of drugs taken in overdose is worked out according to the so-called defined daily dose DDD, which determines the number of tablets in a packet, which is usually a month’s supply. Therefore, when people took an overdose of a packet of desipramine they were far more likely to come to harm, that is because the total number of milligrams of the drug in a packet was higher than other tricyclics. However, if it’s used in sensible doses of 50-100 mg daily, which produces a fully effective blockade of the NRI, then it is very safe and has a low toxicity index.

Note that just as amitriptyline is metabolised into nortriptyline, so also is imipramine (and lofepramine — which is a prodrug) metabolised into desipramine.

TCAs vs the rest

Selected tricyclic antidepressants remain important and effective first choice (‘first-line’) treatments for major depressive disorder [4].

The relatively higher toxicity of some of them in overdose (particularly dothiepin, by far the most toxic), and other side effects, persuade some authorities to over-generalise, over-simplify and say they are no-longer suitable for ‘first-line’ treatment in primary care. Nonsense, in the sense of ‘non-sense’. The discontinuation rates due to SEs are not meaningfully different for selected TCAs and newer ADs. Indeed, in my experience patients are less likely to need to cease NTP than SSRIs.

The evidence on toxicity does not logically support the widely advocated, but simplistic, approach of avoiding TCAs because they might have higher toxicity if someone takes an OD. First, some TCAs are less toxic than some ‘newer’ drugs (e.g. nortriptyline is less toxic than venlafaxine). Also, note that some 95% of deaths by suicide are not caused by the antidepressant drug prescribed, so prescribing a less toxic drug does not address the main problem of death from suicide attempts (unless it is clearly demonstrated to be more effective). That is especially so if the shiny new drug is, in reality, less effective (which some, perhaps many, of the newer drugs probably are). Indeed, the evidence that some old drugs are more effective than most of the new drugs remains strong — especially for amitriptyline, imipramine and clomipramine and tranylcypromine, phenelzine isocarboxazid (see other notes).

The risk of suicide, even if that is by over-dose with the prescribed TCA, is minimised by ‘good clinical management’ (e.g. among other measures, patients should be attending frequently (or be in hospital), only a small (sub-lethal) supply may be issued at each visit, if no responsible person is available to supervise the supply. Remember, 95% of those who die by suicide use means other than the drug they are prescribed.

Misinformed views about SEs and toxicity have caused most doctors to make SSRIs their ‘default’ choice (follow the ‘guidelines’, then you do not have to think for yourself).

There are sometimes good reasons not to use SRIs as ‘first-line’ treatment (see below), but most doctors are ignorant of what these are, to their patients’ detriment.

When to consider avoiding SRIs* as first choice

We can predict which patients are likely to find a TCA to be a more satisfactory option as a first-choice treatment:

Typical ‘endogenous’ symptom profile, or more severe, depressive illness.

Patients with previous episodes of severe depression unresponsive, or partially responsive, to an SSRI).

Those whose current episode is severe or who show psychomotor retardation or lack of energy (optimal choices, imipramine, amitriptyline and clomipramine, my preference is definitely for clomipramine because it is an ‘SNRI”).

Those with pre-existing history of, or present symptoms of, marked insomnia. Also, patients with restless leg syndrome, which is precipitated and exacerbated by SRIs.

Those with previous or present symptoms of gastrointestinal disturbance, reflux, dyspepsia, GI ‘overactivity, GI bleed, or any bleeding risk or tendency or propensity for bruising.

Those with anorexia and weight loss.

Those with present, or past, history of significant suicidal features or attempts. Evidence has been getting stronger and stronger over the last 10 years or so that such patients may be made worse by SSRIs [5, 6].

Those with high anxiety or agitation.

If concern over cytochrome P450 enzyme drug interactions is relevant. TCAs cause less problems in this area than many new drugs.

A history of hyponatremia.

The aged, because of > SEs compared to TCAs [7]

Any concern regarding discontinuation or withdrawal syndromes should be a cause for caution and careful consideration of alternative strategies.

*By which I mean both SSRIs and SNRIs

The above factors, and others, will indicate a TCA may be the optimal first choice.

Gillman’s maxim No. 2

The longer a new drug is in use the smaller its side effect advantages turn out to be, compared with previously existing drugs.

Other comments on the relative merit of different antidepressants

I am astonished that others took so long to appreciate the problems caused by SRIs in relation to bleeding, serious GI problems and frequent, almost unavoidable, sexual dysfunction, not to mention interactions caused by CYP450 inhibition.

That demonstrates the power of indoctrination by advertising.

There is evidence from both clinical experience and research that SSRIs noticeably increase the risk of gastric disturbance and bleeds. This is probably for two reasons; they are ‘pro-kinetic’ in the GI tract (hence the side effect of diarrhoea) and they increase the bleeding time (by depleting platelet 5-HT). It is astonishing that everyone seems to have forgotten so quickly that depletion of platelet 5-HT was the assay initially used indicate whether a drug was an SRI.

A quick point here about ‘side effects’, which is a mis-used term. The main so-called side effects of many drugs, particularly SSRIs, are an inevitable consequence of their main intended mechanism of action, as illustrated by increased bleeding time caused by depletion of platelet 5-HT, gastric upset, sexual dysfunction. Hence, these effects occur in all subjects, to a greater or lesser degree.

It behoves us all to be sceptical about SEs and alert to the emergence of long-term problems — such problems are still emerging with SSRIs after 30 years! E.g. see [8-10], and most recently the spectre of autism possibly being exacerbated or even caused by SRIs in pregnancy [11, 12]. Post marketing surveillance in most countries is a relatively neglected area; we must not expect that side effects will always be obvious.

Sertraline, like nortriptyline, has a particularly favourable pharmacokinetic and side effect profile with a lesser propensity for interactions and is a 1st choice [13]. It is also a weak dopamine re-uptake inhibitor and that may possibly confer advantage, particularly in relation to the insidious amotivational syndrome that I think is under-appreciated as an effect of SSRIs, which deplete frontal dopamine (see note on bupropion).

Three of the other selective serotonin reuptake inhibitors have significant and sometimes dangerous interactions, mostly via cytochrome P450 enzymes; these are fluoxetine (2D6 and 3A4), paroxetine (2D6) and fluvoxamine (1A2 and 2C9 /19). For this reason, it is hard to justify their routine use by primary care doctors who are likely to encounter difficulties keeping up with the multiple possible interactions. Some of these interactions have precipitated serious morbidity and have led to expensive litigation.

SSRIs are usually an inadequate treatment for more severe depressive illness.

I advise great caution if treating “endogenous” depression with these drugs, one often encounters degrees of improvement that fall well short of full remission of illness. That is unacceptable, because full remission of illness is the goal.

Less effective drugs probably include; moclobemide, citalopram, paroxetine, fluvoxamine, trimipramine, doxepin, dothiepin, trazodone, nefazodone and mianserin and its twin, 6-azamianserin marketed as mirtazapine.

Although the old monoamine oxidase inhibitors (MAOIs), tranylcypromine and phenelzine, are no longer advertised (how many young doctors do not know they even exist?) it is well to remember that they are not just very effective, but are the treatment of choice for many patients See section on MAOIs on this website and my various review papers on this subject [14-19].

The therapeutic effectiveness of many new drugs for serious ‘endogenous’ depression is not yet satisfactorily established (and probably never will be), yet a large proportion of doctors rush to use the most recently approved drug on the flimsiest of evidence.