Tailoring antidepressants to patients
Evidence about antidepressant use has not always reflected GP experience Professor Philip Cowen examines
When I was training in psychiatry the prescription of antidepressant medication was straightforward. The only widely used drugs were tricyclic antidepressants (TCAs) and the main issue, apart from their toxicity, was the perennial spat between GPs and psychiatrists (in my family between my father and myself) as to whether doses of 75mg or less of TCA really worked. To my dismay a recent meta-analysis suggested the GPs had been right1.
There are now many more kinds of antidepressant and prescription has become more complicated. More choice is better for patients and increases the chances of finding a preparation both beneficial and well-tolerated. Newer antidepressants are certainly safer in overdose than TCAs2.
At the same time the use of antidepressants has come under increased scrutiny from the media and pressure groups. This is partly a result of covert stigma against psychiatry and its treatments, and partly a general issue of loss of trust in institutions with pharmaceutical companies being viewed with particular suspicion. Practitioners need to know the kind of fears patients may have about antidepressants if they intend to recommend them in treatment (table 1, see page 44).
Despite some protestations to the contrary there is excellent evidence that antidepressants work in clinical depression. Summarising placebo-controlled trials, the British Association of Psychopharmacology concluded the number needed to treat (NNT) is between three and four and that in the broad range of depressed patients, currently available antidepressants are about equally effective3. But it does seem that in patients with more severe depression, amitriptyline and venlafaxine have the edge over selective serotonin
re-uptake inhibitors (SSRIs).3,4,5
If antidepressants are about equally effective in most patients it's helpful when using them to have some scheme of classification that describes their clinical profile, particularly their adverse effects. An easy way to do this is to group agents as sedating or non-sedating; this distinguishes most TCAs and receptor antagonists from selective amine re-uptake inhibitors (SSRIs, venlafaxine and reboxetine). Receptor antagonists (mirtazapine and trazodone) can be distinguished from TCAs because while sedating they lack anticholinergic and cardiotoxic effects (table 2, see page 44).
Are there differences between SSRIs?
There are differences between SSRIs, some of which are clinically important6 (table 3, see page 44). Paroxetine has attracted much media opprobrium because of its prominent withdrawal symptoms and studies have consistently shown abstinence symptoms after paroxetine cessation are worse than after other SSRIs.
On the other hand paroxetine has a large database showing efficacy in many different kinds of anxiety disorder, although it seems unlikely that in the real world it is superior to sertraline and citalopram in this respect. Fluoxetine has a long half-life, which makes it easier to withdraw and it is also licensed for the treatment of bulimia nervosa. But fluoxetine tends to be more activating than the other SSRIs and some patients find this uncomfortable.
Citalopram is least likely of the SSRIs to cause significant drug interactions through CYP450 inhibition, but it may be somewhat more toxic in overdose6. The claim that the recently marketed isomer of citalopram, escitalopram, has superior efficacy to its parent compound has been greeted with scepticism in some quarters7.
SSRIs are undoubtedly useful drugs. Their dosing is simple, many patients tolerate them well and they have helpful therapeutic effects in a range of anxiety disorders. But it needs to be remembered that some people can become agitated, restless and sleepless at the beginning of SSRI treatment. My own practice is to warn patients about this side-effect and to begin treatment with a half dose for about a week (fluoxetine can be given as 20mg every other day).
The main problem during longer-term SSRI use is persistent sexual dysfunction, particularly anorgasmia. A number of remedies have been suggested but none are well-established and careful lowering of the dose, if possible, can be the most helpful solution.
Venlafaxine and reboxetine
At starting doses (75mg daily of the XL preparation) venlafaxine acts as an SSRI. At higher doses a degree of noradrenergic activity conveys some additional efficacy4 but also a risk of hypertension. Otherwise the adverse effects of venlafaxine are the same as those of SSRIs. As with paroxetine, withdrawal symptoms can be prominent, at least in the short-term.
Reboxetine is a selective noradrenaline re-uptake inhibitor. It has a fairly activating profile and can cause insomnia, but otherwise its common adverse effects such as dry mouth and constipation resemble those of TCAs. Reboxetine might be helpful where depression is accompanied by prominent symptoms of lethargy and fatigue.
Under this heading are two sedating agents, mirtazapine and trazodone. Mirtazapine is a derivative of a venerable atypical antidepressant drug, mianserin, and like it can cause drowsiness, increased appetite and weight gain. Unlike mianserin there is no need for blood count monitoring at the beginning of treatment.
Trazodone often finds use at low doses (50-100mg) as a hypnotic but at doses of 150-300mg has antidepressant and anxiolytic properties. Careful dose titration is advisable to improve tolerability. Mirtazapine and trazodone are much less likely to cause sexual dysfunction than SSRIs and venlafaxine. Indeed trazodone can rarely cause troublesome priapism.
Don't forget tricyclics
Low-dose TCAs (50-75mg) probably are helpful in some patients with moderate depression1, particularly those with anxiety and sleep disturbance, and where they are reasonably tolerated, titration to 125-150mg can give efficacy that has yet to be surpassed5.
Such doses, however, can cause troublesome anticholinergic side-effects and should be avoided in patients with significant cardiovascular disease. Also at low doses TCAs can have useful additional therapeutic effects on neuropathic pain and irritable bowel syndrome which are, of course, common in primary care.
Careful dose titration is the key to successful use of TCAs and in clinical trials dothiepin appears well tolerated though the evidence base for its efficacy is less good than that of amitriptyline. Toxicity in overdose (accidental or otherwise) remains the greatest problem. In this connection it is worth remembering that lofepramine is much safer in overdose than other TCAs, but its profile is more activating than dothiepin or amitriptyline and it is not well tolerated by some.
In the longer-term
In many patients clinical depression seems to manifest as a chronic disorder requiring long-term medical and psychological management8. In this situation maintenance treatment with antidepressants may be appropriate. Meta-analyses suggest that in patients prone to depression (two or more previous episodes) longer-term prescription of antidepressants reduces the risk of relapse by over 50 per cent9.
In terms of preparation and dosage for longer-term treatment, the old adage 'what gets you well, keeps you well' is worth following. This suggests that in patients who have been well on long-term antidepressant treatment, attempts at dose lowering and treatment withdrawal should be carried out in small steps over several months where at all possible.
1 Furukawa TA et al. Meta-analysis of effects and side-effects of low-dosage tricyclic antidepressants in depression: systematic review.
2 Henry JA. Epidemiology and relative toxicity of antidepressant drugs in overdose. Drug Saf 1997;16:374-90
3 Anderson IM et al. Evidence-based guidelines for treating depressive disorders with antidepressants:
a revision of the 1993 British Association for Psychopharmacology guidelines. J Psychopharmacol 2000;14:3-20
4 Smith D et al. Efficacy and tolerability of venlafaxine
compared with selective serotonin reuptake inhibitors and other antidepressants: a meta-analysis.
Br J Psychiatry 2002;180:396-404
5 Barbui C, Hotopf M. Amitriptyline v. the rest: still the leading antidepressant after 40 years of randomised controlled trials.
Br J Psychiatry 2001; 178: 129-44
6 Edwards JG, Anderson I. Systematic review and guide to selection of selective serotonin reuptake inhibitors. Drugs 1999;57:507-33
7 Dyer O. Lundbeck broke advertising rules. BMJ 2003;326:1004
8 Andrews G. Should depression be managed as a chronic disease?
BMJ 2001; 322: 419-21
9 Geddes JR et al. Relapse prevention with antidepressant drug treatment in depressive disorders: a systematic review. Lancet 2003;361:653-61
10 Khan A et al. Suicide rates in clinical trials of SSRIs, other antidepressants, and placebo: analysis of FDA reports.
Am J Psychiatry 2003;160:790-2
Some common beliefs about antidepressant medication
Some common beliefs What we know
They don't work Controlled trials show conclusively that antidepressants do
it's all placebo work; they are as effective as most general medical treatments.
They are addictive They are not addictive in the proper meaning of the term that
people crave them and try to get more. Stopping medication
suddenly can cause abstinence syndromes but this is not
addiction. Slow withdrawal is tolerated by most people.
They cause suicide Clinical depression is certainly associated with suicide. Meta-
analyses indicate antidepressants are not10.
However, some patients on activating drugs (such as SSRIs) can
become more agitated and distressed early in treatment
perhaps increasing the risk of self-harm.
They don't deal with Antidepressants facilitate activity in brain pathways known
the problem, just to be important in coping with stress. They help you deal with
stop your complaining problems. Being clinically depressed makes problems seem insoluble.
Drug Sedating Weight gain Sexual Cardiotoxicity
TCAs +++ +++ + +++
SSRIs 0 0 +++ 0
venlafaxine 0 0 +++ +