Abstract
The pharmacological treatment of depression has changed considerably over the last 10 years with the introduction of the second-generation anti-depressants. Four new classes of antidepressants are now available in Australia; the selective serotonin re-uptake inhibitors (SSRIs), a reversible inhibitor of monoamine (RIMA), a serotonin and noradrenaline re-uptake inhibitor (SNRIs) and 5-HT2 receptor antagonists. The use of these agents, especially the SSRIs, has exploded since their introduction in 1991. In 1991, the number of prescriptions written for the first SSRI, fluoxetine, was 81 661 at a cost to Government of $3.8 million. By 1997/1998, the number of prescriptions written for the SSRIs was 28.9 million, at a cost to Government of $93.1 million. Over the same period, the use of the TCAs has dropped slightly from 2.8 to 2.3 million prescriptions, with the cost falling from $10.5 million to $10.3 million. The newer agents now account for 61% of the cost of all the anti-depressants (data from the PBS). The use of these newer antidepressants is becoming commonplace [1] and is recommended by some as the first-line treatment for depression [2]. With this change in prescribing patterns, the role for the TCAs is becoming redundant, and it has been suggested that they should only be used as a second-line drug. There has now even been the suggestion that prescribing aTCAraises the potential for a professional negligence claim to be brought against the practitioner [3].
Three reasons are given to support the view that the new antidepressants should be the first-line treatment of depression: efficacy, tolerability and safety. It has been argued that there is little, if any, difference in the efficacy between the new antidepressants and the TCAs and that the newer agents are better tolerated as a result of their low side-effect profile. Finally, the new agents are safer, particularly when taken in overdose; a desirable attribute for a medication prescribed for persons at risk for suicide. While these are persuasive arguments in favour of the new antidepressants, they do not necessarily mean that TCAs should not be used as, we believe, they still have an important place as a first line treatment for melancholia.
Efficacy
Depression is a biopsychosocial disease and the optimal treatment involves both pharmacological and non-pharmacological approaches. The treatment approach depends upon the type of depression (broadly speaking melancholic vs non-melancholic depression [4, [5]]), the severity of the depression, causal factors and patient preference. For those patients with mild to moderate, non-melancholic (non-endogenous) depression, non-pharmacological treatments such as cognitive-behavioural therapy, interpersonal psychotherapy or problem-solving therapy are as effective as antidepressant medication [6, [7], [8], [9], [10]] although expensive in the short term. The two treatment modalities should not be considered as mutually exclusive as drug treatment in isolation is, in reality, only partial treatment given that medications do not deal with psychosocial antecedents and consequences of depression. In contrast, pharmacotherapy is the mainstay of treatment for melancholia (or endogenous depression), with non-pharmacological approaches being essential adjuncts to treatment [4, [5], [10], [11]].
There have now been many studies evaluating the efficacy of the second-generation antidepressants in comparison with the older agents. The findings of these studies have been subjected to meta-analyses [12, [13], [14], [15], [16]]. In the main, these studies show that the both groups of drugs have equivalent efficacy. This is hardly a surprising finding since, if both classes of drugs are more effective than placebos, finding a significant difference between classes of effective treatments requires laige sample sizes [17, [18]]. These studies have focused on persons suffering from mild to moderate depression, and there are few studies that have evaluated the efficacy of the different classes of antidepressants in the treatment of melancholia or delusional depression. Where they have done so, the TCAs appear to be more efficacious for the treatment of severe (melancholic, endogenous or psychotic) depression [19, [20], [21], [22], [23], [24]]. There are a number of studies which suggest that the classes of antidepressants are equivalent or at least that there is no statistical difference. A meta-analysis examining studies of SSRIs versus TCAs [15] found that while there was no difference in efficacy between SSRI and comparator antidepressants for SSRIs taken together or individually, when studies were classified into high and low depression scores, based on a median split of initial Hamilton Depression Rating Scale scores, there was a slight advantage to TCAs in the high HDRS group. In addition, SSRIs were slightly less effective than TCAs in inpatients. The latter is of interest as inpatients are often the more worrying or difficult-to-treat group, due often to factors other than severity alone such as suicide risk, personality problems and psychosis. However, in those studies there is a trend to be seen for the TCAs to be more efficacious (e.g. [25]).
Tolerability
The major advantage claimed for the newer agents is their different side-effect profile, especially the lack of anticholinergic and cardiovascular side effects. While they have been promoted as being well tolerated, they are not side effect free. They have a distinct but different side-effect profile; prominent effects are nausea, diarrhoea, anxiety, agitation and insomnia. This different adverse-effect profile is an advantage if it results in better tolerability, most often measured as the number of patients who drop out of treatment during a course of antidepressants. However, discontinuation is only one measure of tolerance; others are the number, impact and severity of side effects of the medication.
There are now several studies which have been subjected to meta-analyses [13, [16], [26], [27]], all of which seem to indicate that the new antidepressants are better tolerated than the tricyclic antidepressants. This holds for clinical trials as well as naturalistic studies in primary care [27]. However, a recent review from the Canadian Coordinating Office for Health Technology [28] reported that while study completion rates were equivalent, some side effects such as nausea, diarrhoea, insomnia, headache, nervousness, agitation and anxiety were more common for SSRIs. By contrast, the TCAs had different side effects, particularly the anticholinergic ones. The differences in discontinuation rates in the published studies are small [29] and may in part be the result of some of the studies using the older TCAs (with the most prominent side effects) rather than the newer TCAs [16]. The main indicator of tolerability has been discontinuation rates; other measures of tolerability such as the number, impact or severity of side effects have not been used. The differences in discontinuation rates are not great, although there do appear to be fewer persons discontinuing when prescribed the second generation antidepressants than those being treated with TCAs. The number of side effects reported by patients on the TCAs and the second-generation anti-depressants appear to be equivalent [28].
A number of points could be made about discontinuation studies.
(1) These studies were conducted in the first part of the 1990s whenthe second-generation anti-depressants were very new on the market. At that time, they were touted as having few, if any, side effects. The side effects of the tricyclics are well recognised, particularly the anticholinergic side effects. There may have been a preference for medical practitioners to discontinue the TCAs if these side effects were reported so that the patient could be switched to the new antidepressants with ‘few side-effects’. This may have led to higher rates of discontinuation with tricyclics than with the new second-generation antidepressants.
(2) The majority of the studies that have evaluated discontinuation have been conducted in treatment trials. Such treatment trials are generally short-term, up to 12 weeks. There is no doubt that the side effects of the TCAs are particularly troublesome during the early phase of treatment especially the anticholinergic side effects. Such side effects are disabling for patients who are, at the same time, struggling with their depressive symptoms, although drowsiness and sleepiness may be an advantage for some. In the short term, it is highly probable that persons will discontinue medication as these side effects are at their worse when the person is most depressed. By contrast, some of the side effects with the new antidepressants are not so troublesome when the person is at their most depressed. For example, sexual dysfunction and somnolence may not be a particular problem for the patient at the nadir of their depression but these side effects do persist and may become more of a problem after recovery. Long-term follow-up studies examining later rates of discontinuation would be of benefit. With the tricyclics, side effects diminish over time and when the person is well they tend to be minimally disabled by side effects, particularly the anticholinergic ones, although constipation may persist which is easily remedied by a high fibre diet or bulk laxatives. In a recent paper, Zajecka et al. [30] found treatment emergent sexual dysfunction in 60% of males and 57% of females studied who were treated with SSRIs, with 30% of both unable to reach orgasm. In our clinical experience, this side effect is particularly distressing and requires switching medication to a TCA, moclobemide, or nefazadone which do not lead to sexual dysfunction.
The serotonin syndrome is a potentially severe adverse drug reaction that may be associated with SSRI use. Causes of the syndrome include excess levels of or unusual sensitivity to SSRI agents, certain drug combinations, and the administration of excessive amounts of serotonin precursor agents [31].
One component of tolerability that has not been adequately dealt with in this literature is the impact of drug interactions. There has been a growing list of drug interactions with the SSRIs. These drug interactions can make prescribing these new agents problematic, particularly in the general medical setting among older patients who may be taking a range of drugs for coexisting medical illnesses [32]. By contrast, the TCAs do not have effects on the cytochrome P450 system and thus have few potentially harmful drug interactions. Contraindications of course do exist for the TCAs, particularly for those who have glaucoma, prostatic hypertrophy or cardiac disease.
Like the TCAs, the SSRIs have the potential to produce withdrawal symptoms if the drug is ceased abruptly. The most frequently reported symptoms are dizziness, paraesthesia, nausea, visual disturbances and headache. Symptoms have been reported to occur despite slowly tapered withdrawal and to persist for up to 21 days after drug withdrawal [33].
Safety
The major reasons for advocating the first-line use of the new generation antidepressants is that they are safe if taken in overdose and fatal outcomes do not generally occur following overdose of these agents, although it must be noted that fatal outcomes have been reported [34]. Suicide is one of the likely sequelae of depression, with up to 6% of persons suffering from depression going on to suicide [35], and about half of successful suicides were suffering from depression [36]. There is no doubt that the TCAs are potentially fatal if taken in overdose, with 81% of deaths caused by antidepressant overdoses attributable to the TCAs [34]. This would appear to have been a very powerful argument for not using the TCAs to treat depression. However, only about 4% of all suicides are due to overdoses of a single anti-depressant [37], and other methods of suicide, such as hanging or shooting are more common, especially among young people (the most vulnerable group) [38]. Among those who take an overdose as the means to end, medications such as dextro-propoxyphene and short-acting barbiturates were used more commonly (after correcting for the number of prescriptions) in a recent Australian study [39]. Over-the-counter medications (such as paracetamol) can also be lethal in overdose, yet there is no suggestion that these should be withdrawn from the market or not used. By the same token, other effective psychotropic agents used in the treatment of bipolar depression, such as lithium, are potentially lethal in overdose yet there is no suggestion that we should not use them because they can be potentially harmful if taken in overdose.
The important issue is the appropriate management of potentially suicidal depressed patients [37]. The ‘safety in overdose’ argument with the SSRIs would suggest that good clinical care doesn't come into it, that the patient can just be given tablets that won't kill them if they overdose on them. The essence of good management of the suicidal patient is comprehensive clinical care including: careful and comprehensive initial and ongoing clinical assessment; risk assessment; treatment in an appropriate setting; and the use of appropriate pharmacological, psychological and social interventions. Simply prescribing a medication that won't kill the patient if they take a large dose is not the answer. Rather than representing good care, this approach may encourage less comprehensive and vigilant treatment of a potentially suicidal patient and pharmacological sloppiness, presenting another situation where treatment falls wide of the standard of reasonable skill and care required of medical practitioners in Australia [3].
Second, younger persons (currently the most at-risk group for suicide) who may have a non-melancholic depression may have responded better to non-pharmacological treatments than the injudicious and inappropriate use of an antidepressant in the first place.
Summary
It is premature to write off the tricyclics or relegate them to second-line treatment for depression. We argue that they have a place as a first-line treatment for severe (melancholic) depression where they seem to have a therapeutic advantage over the newer agents. If prescribed properly, along with the appropriate counselling, they are well tolerated and have few drug interactions to worry about. The risk of suicide for depressed patients will always be there; this needs to be dealt with by good clinical care and taking the appropriate precautions with a potentially lethal medication rather than selecting a drug which is safe in overdose.
While we do not advocate the wholesale usage of the TCAs (the newer agents have an important place in the pharmacotherapy of depression), we consider that they have an important role in the treatment of severe depression. The TCAs do have a different side-effect profile to the SSRIs. This different side-effect profile may be a considerable advantage to some patients, whereas to others it may be an impediment to taking the TCA. In considering tolerability issues, it is important to keep in mind potential drug interactions related to the P450 system, and the possibility of a serotonin syndrome arising. Finally, while the new agents are safer when taken in overdose, managing suicidal patients involves high quality clinical care, not just prescribing a safe agent.