Mechanism of action

The monoamine hypothesis proposes that, in depression, there is deficiency of the neurotransmitters noradrenaline/norepinephrine and serotonin in the brain which can be restored by antidepressants. Drugs that alleviate depression also enhance monoamine availability and release (Fig. 20.2), increasing activity at postsynaptic receptors. It is relevant that (older) antihypertensive agents, e.g. reserpine, which reduced the availability of noradrenaline/norepinephrine, caused depression.

Fig. 20.2 Mechanism of action of antidepressant drugs at the synapse.

SSRIs act, as their name indicates, predominantly by preventing serotonin reuptake by blocking the cell-surface serotonin transporter; with little effect on noradrenaline/norepinephrine reuptake. Tricyclic antidepressants and reboxetine inhibit noradrenaline/norepinephrine reuptake, but tricyclic effects on serotonin reuptake vary widely; desipramine and protriptyline have no effect, whereas clomipramine is about five times more potent at blocking serotonin than noradrenaline/norepinephrine reuptake. SNRIs are capable of inhibiting reuptake of both transmitters. However, for venlafaxine a dose of at least 150 mg/day is required for noradrenaline/norepinephrine uptake blockade to be exerted. Mirtazapine also achieves an increase in noradrenergic and serotonergic neurotransmission, but through antagonism of presynaptic α₂-autoreceptors (receptors that mediate negative feedback for transmitter release, i.e. an autoinhibitory feedback system). Other novel antidepressants include trazodone, which blocks several types of serotonin receptor (including the 5HT2A and 5HT2C receptors) as well as α-adrenoceptors and histaminergic receptors and acts as a partial agonist at the 5HT1A receptor, and agomelatine which acts both as an agonist of melatonin receptors and a blocker of the serotonin 5HT2C receptor, the combined effects of these actions leading to a rise in frontal cortex dopamine availability.

MAOIs increase the availability of noradrenaline/norepinephrine and serotonin by preventing their destruction by the monoamine oxidase type A enzyme in the presynaptic terminal (see Ch. 21, Table 21.3). The older MAOIs, phenelzine, tranylcypromine and isocarboxazid, bind irreversibly to monamine oxidase by forming strong (covalent) bonds. The enzyme is thus rendered permanently ineffective such that amine metabolising activity can be restored only by production of fresh enzyme, which takes weeks. These MAOIs are thus called ‘hit and run’ drugs as their effects greatly outlast their detectable presence in the blood.

But how do changes in monoamine transmitter levels produce an eventual elevation of mood? Raised neurotransmitter concentrations produce immediate alterations in postsynaptic receptor activation, leading to changes in second-messenger (intracellular) systems and to gradual modifications in cellular protein expression. Antidepressants increase a cyclic AMP response element binding (CREB) protein, which in turn is involved in regulating the transcription of genes that influence survival of other proteins, including brain-derived neurotrophic factor (BDNF), which exerts effects on neuronal growth.

Although the monoamine hypothesis of depression is conceptually straightforward, it is in reality an oversimplification of a complicated picture. Other systems that are implicated in the aetiology of depression (and which provide potential targets for drug therapy) include the hypothalamic–pituitary–thyroid axis and the hypothalamic–pituitary–adrenal (HPA) axis. The finding that 50% of depressed patients have raised plasma cortisol concentrations constitutes evidence that depression may be associated with increased HPA drive.

Drugs with similar modes of action to antidepressants find other uses in medicine. Bupropion (amfebutamone) inhibits reuptake of both dopamine and noradrenaline/norepinephrine. It was originally developed and used as an antidepressant but is now more frequently used to assist smoking cessation (see p. 321). It is also prescribed in attention deficit hyperactivity disorder (see p. 345). Sibutramine, licensed as an anorectic agent, is a serotonin and noradrenaline/norepinephrine reuptake inhibitor (SNRI).

Pharmacokinetics

The antidepressants listed in Table 20.1 are generally well absorbed after oral administration. Steady-state plasma concentrations of TCAs show great individual variation but correlate with therapeutic effect. Where there is a failure of response, measurement of plasma concentration can be useful as the failure may be attributable to low plasma levels due to ultra-rapid metabolism (though it is often not available). Antidepressants in general are metabolised principally by hepatic cytochrome P450 enzymes. Of the many isoenzymes identified, the most important in antidepressant metabolism are CYP P450 2D6 (Table 20.2A) and CYP 3A4 (Table 20.2B). Other important P450 enzymes are CYP 1A2 (inhibited by the SSRI fluvoxamine, induced by cigarette smoking; substrates include caffeine and the atypical antipsychotics clozapine and olanzapine) and the CYP 2 C group (inhibition by fluvoxamine and fluoxetine, involved in breakdown of escitalopram and moclobemide). Sometimes several CYP enzymes are capable of mediating the same metabolic step. For example, at least six isoenzymes, including CYP 2D6, 3A4 and 2 C9, can mediate the desmethylation of the SSRI sertraline to its major metabolite.

Table 20.2A Psychotropic (and selected other) drugs known to be CYP 2D6 substrates, inhibitors and inducers

A substrate is a substance that is acted upon and changed by an enzyme. Where two substrates of the same enzyme are prescribed together, they will compete and, if present in sufficient quantities, the metabolism of one or other, or both, drugs may also be inhibited, resulting in increased plasma concentration and possibly in enhanced therapeutic or adverse effects. An enzyme inducer accelerates the metabolism of co-prescribed drugs that are substrates of the same enzyme, reducing their effects. An enzyme inhibitor retards metabolism of co-prescribed drugs, increasing their effects.

a CYP 2D6 is involved only in the breakdown of venlafaxine to its active metabolite and therefore implications of 2D6 interactions for efficacy are of limited significance.

Table 20.2B Psychotropic (and selected other) drugs known to be CYP 3A4 substrates, inhibitors and inducers

A substrate is a substance that is acted upon and changed by an enzyme. Where two substrates of the same enzyme are prescribed together, they will compete and, if present in sufficient quantities, the metabolism of one or other, or both, drugs may also be inhibited, resulting in increased plasma concentration and possibly in enhanced therapeutic or adverse effects. An enzyme inducer accelerates the metabolism of co-prescribed drugs that are substrates of the same enzyme, reducing their effects. An enzyme inhibitor retards metabolism of co-prescribed drugs, increasing their effects.

a mCPP (meta-chlorophenylpiperazine), the active metabolite of trazodone, is a CYP 2D6 substrate; observe for unwanted effects when trazodone is co-administered with the 2D6 inhibitors fluoxetine or paroxetine.

Several of these drugs produce active metabolites that prolong their action (e.g. fluoxetine is metabolised to norfluoxetine, t_½ 200 h). The metabolic products of certain TCAs are antidepressants in their own right, e.g. nortriptyline (from amitriptyline), desipramine (from lofepramine and imipramine) and imipramine (from clomipramine). Half-lives of TCAs lie generally in a range from 15 h (imipramine) to 100 h (protriptyline), and those for SSRIs from 15 h (fluvoxamine) to 72 h (fluoxetine).

Around 7% of the Caucasian population have very limited CYP 2D6 enzyme activity. Such ‘poor metabolisers’ may find standard doses of TCAs intolerable, and it is often worth prescribing them at a very low dose. If the drug is then tolerated, plasma concentration assay may confirm the suspicion that the patient is a poor metaboliser. There is also a genetic polymorphism influencing CYP 2 C19 activity which has a clinically important effect on metabolism of escitalopram.

Therapeutic efficacy

Provided antidepressant drugs are prescribed at an adequate dose and taken regularly, 60–70% of patients with depression should respond within 3–4 weeks. Conventional meta-analyses have shown little evidence that any particular drug or class of antidepressant is more efficacious than others, but there are some possible exceptions to this general statement:

Selection

An antidepressant should be selected to match individual patients’ requirements, such as the need or otherwise for a sedative effect, or the avoidance of antimuscarinic effects (especially in the elderly). In the absence of special factors, the choice rests on the weight of evidence of efficacy, the tolerability, the safety in overdose and the likelihood of an effective dose being reached. SSRIs and other newer drugs (e.g. mirtazapine and venlafaxine) are highlighted as best meeting these needs.

Mode of use

Antidepressants usually require 3–4 weeks for the full therapeutic effect to be achieved. When a minimal response is seen, an antidepressant can usefully be extended to 6 weeks to see whether further benefit is achieved. By contrast, patients may experience unwanted effects, especially ‘jitteriness’ or ‘activation’ symptoms such as increased anxiety and irritability, soon after starting treatment (they should be warned about this possibility), but such symptoms usually diminish with time. Some drugs have the advantage that they can be started at a dose which would be considered adequate for the therapeutic effect (e.g. most SSRIs) but in contrast many others, including all tricylic antidepressants, need to be started at a low and generally tolerable starting dose to the therapeutic dose. For example, imipramine should be started at 25–50 mg/day, with gradual increments to a recognised ‘minimum therapeutic’ dose, around 125 mg/day (140 mg/day for lofepramine). Low starting doses are particularly important for elderly patients. Only when the drug has reached the minimum therapeutic dose and been taken for at least 4 weeks can response or non-response be adequately established. However, some patients do achieve response or remission at subtherapeutic doses, for reasons of drug kinetics and limited capacity to metabolise, the self-limiting nature of depression, or by a placebo effect (reinforced by the experience of side-effects suggesting that the drug must be having some action).

For SSRIs dose titration is often unnecessary as the minimum therapeutic dose can usually be tolerated as a starting dose. Divided doses are not usually required, and administration is by a single morning or evening dose. Evidence suggests that patients commencing treatment on SSRIs are more likely to reach an effective dose than those starting on TCAs. Of the novel compounds, trazodone usually requires titration to a minimum therapeutic dose of at least 200 mg/day. Response to reboxetine, venlafaxine and mirtazapine may occur at the starting dose, but some dose titration is commonly required. Venlafaxine is licensed for treatment-resistant depression by gradual titration from 75 to 375 mg/day. There is some need for dose titration when using MAOIs. Unlike other drug classes, reduction to a lower maintenance dose is recommended after a response is achieved if unwanted effects are problematical.

Changing and stopping antidepressants

When an antidepressant fails through lack of efficacy despite an adequate trial or due to unacceptable adverse effects, a change to a drug of a different class is generally advisable. For a patient who has not responded to an SSRI it is logical to try a novel compound such as venlafaxine, mirtazapine or reboxetine, or sometimes a tricylic antidepressant. Any of these options may offer a greater increase in synaptic noradrenaline/norepinephrine than the ineffective SSRI.

Evidence also suggests that patients failing on one SSRI may respond to a different drug within the class, an approach that is particularly useful where other antidepressant classes have been unsuccessful previously, are contraindicated, or have characteristics that the patient or doctor feels are undesirable. Awareness of biological differences between drugs within a class may also be helpful when patients cannot tolerate other drug classes. For instance, among SSRIs, paroxetine has the most affinity for the serotonin transporter and fluoxetine the least, while among TCAs, clomipramine has more important serotonergic enhancing effects than the others.

When changing between antidepressant doses, a conservative approach would be to reduce the first antidepressant progressively over 2 or more weeks before starting the new drug. The gradual reduction is particularly important with paroxetine and venlafaxine which are known to cause ‘discontinuation syndromes’ if stopped abruptly, and less important with fluoxetine due to its long half-life active metabolite which offers ‘built-in’ protection against withdrawal problems. A more proactive approach would involve ‘cross-tapering’ the second antidepressant – i.e. starting it while the first antidepressant is being reduced and gradually titrating the dose up. However, an important exception concerns changes to or from MAOIs, which must be handled with great caution due to the dangers of interactions between antidepressants (see below). Therefore MAOIs cannot safely be introduced within 2 weeks of stopping most antidepressants (3 weeks for imipramine and clomipramine; combination of the latter with tranylcypromine is particularly dangerous), and not until 5 weeks after stopping fluoxetine, due to its long half-life active metabolite. Similarly, other antidepressants should not be introduced until 2–3 weeks have elapsed from discontinuation of MAOI (as these are irreversible inhibitors; see p. 313). No washout period is required when using the reversible MAOI, moclobemide.

When a patient achieves remission, the antidepressant should be continued for at least 9 months at the dose that returned mood to normal. Premature dose reduction or withdrawal is associated with increased risk of relapse. In cases where three or more depressive episodes have occurred, evidence suggests that long-term continuation of an antidepressant offers protection, as further relapse is almost inevitable in the next 3 years.

When ceasing use of an antidepressant, the dose should be reduced gradually to avoid discontinuation syndromes (symptoms include anxiety, agitation, nausea and mood swings). Discontinuation of SSRIs and venlafaxine are associated additionally with dizziness, electric shock-like sensations and paraesthesia. Short t_½ drugs that do not produce active metabolites (e.g. paroxetine, venlafaxine) and TCAs are most likely to cause such problems.

Augmentation

The development of many new antidepressants in recent years has reduced the need to use augmentation strategies. Nevertheless augmentation, i.e. the addition of a second drug to an existing antidepressant, can be used when two or more standard antidepressants have successively failed to alleviate depressive symptoms despite treatment at an adequate dose for an adequate time. Some of the augmentations discussed may even be used earlier than this if there is an indication or justification for the augmenting drug specific to the individual patient.

One strategy which has come to prominence is to augment (or combine) an SSRI or SNRI antidepressant with the novel antidepressant mirtazapine. The initial justification for this combination stems from mirtazapine’s unorthodox mechanism of action – the idea being that the presynaptic adjustments effected by mirtazapine could act additively or even synergistically with the monoamine reuptake inhibition of SSRIs and SNRIs. A second justification is more practical – mirtazapine is known to improve the quality of sleep and serotonin reuptake inhibitors may initially disrupt this, thus mirtazapine can be added both to boost the antidepressant effect and to address an unresolved problem with sleep. An evidence base does exist both for mirtazapine–venlafaxine and mirtazapine–fluoxetine co-prescription in depression, with both combinations reported as providing significantly higher remission rates than fluoxetine alone. Ease of initiation of these combinations, along with the evidence of enhanced effectiveness, means that these are currently the most commonly used augmentation strategies for treatment of depression in psychiatry inpatients in the UK.

Another important augmentation strategy employs the mood stabiliser lithium carbonate. Controlled trials suggest that up to 50% of patients who have not responded to standard antidepressants can respond after lithium augmentation but the evidence is stronger for augmenting tricyclics than for augmenting SSRIs. Addition of lithium requires careful titration of the plasma concentration up to the therapeutic range, with periodic checks thereafter and monitoring for toxicity (see p. 331).

More recently, augmentation of SSRIs with atypical antipsychotics has been effective in clinical trials. Trial evidence is strongest using olanzapine, and also exists for quetiapine, risperidone and aripiprazole. Antipsychotics also have important potential for side-effects which must be taken into account before their introduction (see p. 322).

Tri-iodothyronine (T3) also aids antidepressant action, and most evidence points to added benefit with TCAs. When co-prescribing TCAs with thyroid hormone derivatives, be aware that the combination of lofepramine with the levo isomer of thyroxine is contraindicated. The amino acid L-tryptophan and the β-adrenoceptor blocker pindolol may also be used to augment. Tryptophan increases 5-hydroxytryptamine (5HT) production, and pindolol may act by blocking negative feedback of 5HT on to 5HT_1A-autoreceptors.

Other indications for antidepressants

Antidepressants may benefit most forms of anxiety disorder, including panic disorder, generalised anxiety disorder, post-traumatic stress disorder, obsessive–compulsive disorder and social phobia (see p. 331). SSRIs are recognised as first-line drug treatment for all five of these anxiety disorders.

Fluoxetine is effective in milder cases of the eating disorder bulimia nervosa, in higher doses (60 mg/day) than are required for depression. This effect is independent of that on depression (which may coexist), and may therefore involve a different mode of action. Antidepressants appear to be ineffective in anorexia nervosa. Antidepressants also have a role in pain control. The SNRI drug duloxetine is licensed for diabetic neuropathy while considerable trial evidence exists for use of the original tricyclic antidepressants imipramine and amitriptyline for the control of chronic pain.

Adverse effects

As most antidepressants have similar therapeutic efficacy, the decision regarding which drug to select often rests on adverse effect profiles and potential to cause toxicity.

Interactions

Antidepressant use offers considerable scope for adverse interaction with other drugs and it is prudent always to check specific sources for unwanted outcomes whenever a new drug is added or removed to a prescription list that includes an antidepressant.

St John’s wort

The herbal remedy St John’s wort (Hypericum perforatum) has found favour in some patients with mild to moderate depression. The active ingredients in the hypericum extract have yet to be identified, and their mode of action is unclear. Several of the known mechanisms of action of existing antidepressants are postulated, including inhibition of monoamine reuptake and the MAO enzyme, as well as a stimulation of GABA receptors. Much of the original research into the efficacy of St John’s wort was performed in Germany, where its use is well established. Several direct comparisons with tricyclic antidepressants have shown equivalent rates of response, but the interpretation of these studies is complicated by the fact that many failed to use standardised ratings for depressive symptoms, patients tended to receive TCAs below the minimum therapeutic dose, and sometimes received St John’s wort in doses above the maximum recommended in commercially available preparations. Use of St John’s wort is further complicated by the lack of standardisation of the ingredients. A large multi-centre trial found only limited evidence of benefit for St John’s wort over placebo in significant major depression.²

Despite these reservations, there is certainly a small proportion of patients who, when presented with all the available facts, express a strong desire to take only St John’s wort, perhaps from a preference for herbally derived compounds over conventional medicine. For patients with mild depression, it seems reasonable on existing evidence to accede to this preference rather than impair the therapeutic alliance and risk prescribing a conventional antidepressant that will not be taken.

Electroconvulsive therapy

ECT involves the passage of a small electrical charge across the brain by electrodes applied to the frontotemporal aspects of the scalp with the aim of inducing a tonic–clonic seizure. Reference to it is made here principally to indicate its place in therapy.

ECT requires the patient to be under a general anaesthetic, so carrying the small risks associated with general anaesthesia in minor surgical operations. It may cause memory problems, although these are generally transient. For these reasons, as well as the relative ease of use of antidepressant drugs, ECT is usually reserved for psychiatric illness where pharmacological treatments have been unsuccessful or where the potential for rapid improvement characteristic of ECT treatment is important. This may arise where patients are in acute danger from their mental state, for instance the severely depressed patient who has stopped eating or drinking, or those with florid psychotic symptoms. Modern-day ECT is a safe and effective alternative to antidepressant treatment and remains a first-line option in clinical circumstances where a rapid response is desired, when it can be life saving.

Classification

Originally tested as an antihistamine, chlorpromazine serendipitously emerged as an effective treatment for psychotic illness in the 1950s. Chlorpromazine-like drugs were originally termed ‘neuroleptics’ or ‘major tranquillisers’, but the preferred usage now is ‘antipsychotics’. Classification is by chemical structure, e.g. phenothiazines, butyrophenones. Within the large phenothiazine group, compounds are divided into three types on the basis of the side-chain, as this tends to predict adverse effect profiles (Table 20.3). The continuing search for greater efficacy and better tolerability led researchers and clinicians to reinvestigate clozapine, a drug that was originally licensed in the 1960s but subsequently withdrawn because of toxic haematological effects. Clozapine appeared to offer greater effectiveness in treatment-resistant schizophrenia, to have efficacy against ‘negative’ in addition to ‘positive’ psychiatric symptoms (see Table 20.4), and to be less likely to cause extrapyramidal motor symptoms. It regained its licence in the early 1990s with strict requirements on dose titration and haematological monitoring. The renewed interest in clozapine and its unusual efficacy and tolerability stimulated researchers to examine other ‘atypical’ antipsychotic drugs.

Table 20.3 Antipsychotic drugs

a No recognised classification system exists for atypical antipsychotics. Tentative terms based on receptor binding profiles have been applied to certain drug groupings, e.g. ‘broad-spectrum atypicals’ for clozapine, olanzapine and quetiapine, whereas risperidone and ziprasidone have been described as ‘high-affinity serotonin–dopamine antagonists’.

b Amisulpride and sulpiride are structurally related.

c Sertindole is available only on a named-patient basis when at least one previous antipsychotic has failed owing to lack of efficacy or adverse effects.

Table 20.4 Symptoms of schizophrenia

Thus the most important distinction in modern-day classification of antipsychotic drugs is between the classical (typical) agents, such as chlorpromazine, haloperidol and zuclopenthixol, and the so-called atypical antipsychotics, which include clozapine and now risperidone, olanzapine, quetiapine, amisulpride, aripiprazole and others. These latter are ‘atypical’ in their mode of action, their lack of extrapyramidal motor symptoms and adverse effect profiles. Categorisation of atypical agents by their chemical structure is of limited value clinically as they are very heterogeneous. Similarly classification through a shared affinity for a particular receptor system has not been possible; as discussed below, the atypical antipsychotics are very heterogeneous in their receptor-binding profiles.

Indications

Antipsychotic drugs are used for the prophylaxis and acute treatment of psychotic illnesses including schizophrenia and psychoses associated with depression and mania. They also have an important role as an alternative or adjunct to benzodiazepines in the management of the acutely disturbed patient, for both tranquillisation and sedation. There is some evidence that atypical antipsychotics may be used for treatment-resistant anxiety disorders after other treatments (e.g. SSRIs, SNRIs, pregabalin) have failed to produce an adequate response. Certain antipsychotics have an antidepressant effect that is distinct from their ability to alleviate the psychosis associated with depression, and as such can be used to augment the effect of antidepressants. Antipsychotics have also proved useful in the tic disorder Tourette’s syndrome, and for recurrent self-harming behaviour.

Mechanism of action

The common action of conventional antipsychotics is to decrease brain dopamine function by blocking the dopamine D₂ receptors (Fig. 20.3). However, the atypical drugs act on numerous receptors and modulate several interacting transmitter systems. All atypicals (except amisulpride) exhibit greater affinity for 5HT₂ receptors than D₂ receptors, unlike the classical agents. Atypical drugs that do antagonise dopamine D₂ receptors appear to have some selectivity of effect for those in the mesolimbic system (producing antipsychotic effect) rather than the nigrostriatal system (associated with unwanted motor effects). Clozapine and risperidone exert substantial antagonism of α₂-adrenoceptors, a property that may explain their benefits against negative symptoms. Blockade of muscarinic acetylcholine receptors as with chlorpromazine and clozapine reduces the occurrence of extrapyramidal effects. Aripiprazole is a unique drug because it is a partial dopamine D₂-receptor agonist that acts conversely as an antagonist in regions where dopamine is overactive, such as the limbic system. It increases dopamine function where this is low (such as in the frontal cortex) and has little motor effect.

Fig. 20.3 Sagittal brain section illustrating dopaminergic pathways. (1) Mesolimbic pathway (thought to be overactive in psychotic illness according to the dopamine hypothesis of schizophrenia). (2) Nigrostriatal pathway (involved in motor control, underactive in Parkinson’s disease and associated with extrapyramidal motor symptoms). (3) Tuberoinfundibular pathway (moderates prolactin release from the hypothalamus). VTA, ventrotegmental area.

Pharmacokinetics

Antipsychotics are well absorbed after oral administration and distribute widely. They are metabolised mainly by hepatic cytochrome P450 isoenzymes, e.g. CYP 2D6 (risperidone, perphenazine; see Table 20.2A), CYP 3A4 (sertindole; see Table 20.2B), CYP 1A2 (olanzapine, clozapine). Metabolism of some compounds is complex, e.g. zuclopenthixol and haloperidol are metabolised by both CYP 2D6 and CYP 3A4, which means that co-prescription of inhibitors of either enzyme can cause raised plasma concentrations of these antipsychotics. Amisulpride is an exception to the general rule as it is mainly eliminated unchanged by the kidneys with little hepatic metabolism. Elimination t_½ values range from quetiapine 7 h, clozapine 12 h, haloperidol 18 h to olazapine 33 h. Depot preparations usefully release drug over 2–4 weeks after intramuscular injection (see below).

Efficacy

Symptoms in schizophrenia are defined as positive and negative (Table 20.4). Although a classical antipsychotic drug should provide adequate treatment of positive symptoms including hallucinations and delusions, at least 60% of patients may have unresolved negative symptoms such as apathy, flattening of affect and alogia. Evidence suggests that clozapine may have a significant advantage against negative symptoms. This drug has a further advantage over all other antipsychotics, whether classical or atypical, in that it is the most effective agent for ‘resistant’ schizophrenia, i.e. where other antipsychotics prescribed at adequate doses fail to produce improvement or are not tolerated.

Schizophrenia often runs a chronic relapsing and remitting course. Less than a quarter of patients avoid further episodes, with the most common reason for relapse being the stopping of medication against medical advice.

Mode of use

Atypical antipsychotics are all licensed for use in schizophrenia and are now recommended as the first-line treatment for newly diagnosed cases. Some also have licences for psychosis in mania and for control of agitated or disturbed behaviour in the context of a psychotic illness.

The longer a psychosis is left untreated, the less favourable is the outcome, and drug treatment should be instigated as soon as an adequate period of assessment has allowed a provisional diagnosis to be established. Patients who are ‘neuroleptic naïve’, i.e. have never previously taken any antipsychotic agent, should start at the lowest available dose (see Table 20.5, below). For most atypical agents a period of dose titration by protocol from the starting dose up to a stated lowest therapeutic dose is usual, e.g. risperidone 4 mg/day, quetiapine 300 mg/day. Dose increases are indicated when there is no response until the desired effect on psychotic symptoms or calming of disturbed behaviour is achieved, the urgency of the situation determining the interval between increments until the maximum licensed dose is achieved. Conservative dose titration is advisable for the elderly and patients with learning disabilities (who may require antipsychotics for psychosis or severe behavioural disturbance).

Prescription of conventional antipsychotics follows similar rules to those for atypical drugs, starting at low doses in neuroleptic-naïve patients, e.g. haloperidol 0.5 mg/day in case the patient is particularly susceptible to adverse, especially extrapyramidal, motor side-effects. There is a wide range of effective doses for many classical agents (e.g. chlorpromazine, once the most prescribed antipsychotic but now rarely used, has a dose range from 25 to 1000 mg/day). As the potency (therapeutic efficacy in relation to dose) of antipsychotic agents varies markedly between compounds, it is useful to think of the effective antipsychotic dose of classical agents in terms of ‘chlorpromazine equivalents’ (see Table 20.5). For example, haloperidol has a relatively high antipsychotic potency, such that 2–3 mg is equivalent to chlorpromazine 100 mg, whereas 200 mg sulpiride (low potency) is required for equivalent antipsychotic effect. Prescribing in excess of this requires specialist involvement. With co-prescribed antipsychotics, their total maximum antipsychotic dose should not exceed 1000 mg chlorpromazine equivalents per day, except under specialist supervision.

Clozapine may be initiated only under specialist supervision and usually after at least one other antipsychotic has failed through lack of efficacy or unacceptable adverse effects. Additionally, monitoring of leucocyte count is mandatory (danger of agranulocytosis) and blood pressure checking is required (for hypotensive effect). Patients are most vulnerable to agranulocytosis on initiation of therapy, with 75% of cases occurring in the first 18 weeks. The dose titration schedule must be followed strictly, starting with clozapine 12.5 mg at night and working up over a period of 4 weeks to a target therapeutic dose of 450 mg/day.

Adverse effects (Table 20.5)

Active psychotic illnesses often cause patients to have poor insight into their condition. Adverse drug effects can be the final straw in compromising already fragile compliance, leading to relapse. When atypical antipsychotics first came to prominence in the mid-1990s much was made of their lower propensity to cause several of the most troublesome side-effects of classical antipsychotics, especially extrapyramidal motor effects. However, while these problems are encountered less frequently, atypical drugs have a range of troublesome metabolic side-effects which had not been reported in the previous era of classic antipsychotics. Thus, to understand the current position relating to the pros and cons of atypical antipsychotics, it is necessary first to describe the side-effect profile of classical antipsychotic drugs.

Comparison of conventional and atypical antipsychotics

Atypical antipsychotics may have advantages in two areas:

In some countries finance may be the overriding factor in favour of retaining classical agents rather than atypicals as first choice in schizophrenia. The basis for any such decision must extend beyond crude drug costs and take account of the capacity of atypicals to lessen extrapyramidal symptoms, improve compliance, and thus prevent relapse of psychotic illness and protect patients from the lasting damage of periods of untreated psychosis. Additionally, greater efficacy in relation to negative symptoms affords schizophrenic patients the opportunity to reintegrate into the community and make positive contributions to society, when the alternative is long-term residence in hospital. Recognising drugs as therapeutic entities as well as units of cost is an important element in deciding between classical and atypical drugs, and indeed about decision-making in the purchase of all drugs by institutions or countries.

Lithium

Lithium salts were known anecdotally to have beneficial psychotropic effects as long ago as the middle of the 19th century, but scientific evidence of their efficacy was not obtained until 1949, when lithium carbonate was tried in manic patients; it was found to be effective in the acute state and, later, to prevent recurrent attacks.³

Carbamazepine

Carbamazepine is licensed as an alternative to lithium for prophylaxis of bipolar affective disorder, although clinical trial evidence is actually stronger to support its use in the treatment of acute mania. While lithium monotherapy is more effective than carbamazepine for prophylaxis of classical bipolar affective disorder, carbamazepine appears to be more effective than lithium for rapidly cycling bipolar disorders, i.e. with recurrent swift transitions from mania to depression. It is also effective in combination with lithium. (See also Epilepsy, p. 349.)

Valproate

Three forms of valproate are available: valproic acid, sodium valproate and semisodium valproate. The latter two are metabolised to valproic acid which exerts the pharmacological effect. It is the semisodium valproate preparation (Depakote) which is licensed in the UK for acute treatment of mania, underpinned by a reasonably strong evidence base. Valproate has become the drug of first choice for prophylaxis of bipolar affective disorder in the USA, despite the lack of robust supporting clinical trial evidence. Treatment with valproic acid is easy to initiate (especially compared to lithium) with no requirement for plasma concentration monitoring needed, although baseline and periodic checks of full blood count and liver function are recommended following reports of occasional blood dyscrasias or hepatic failure. It is generally well tolerated but weight gain can be a problem, so body mass index should be recorded and evaluated at intervals. Other potential side-effects include gastric irritation, lethargy, confusion, tremor and hair loss.

Treatment with carbamazepine or valproic acid appears not to be associated with the ‘rebound effect’ of relapse into manic symptoms that may accompany early withdrawal of lithium therapy.

Other drugs

Evidence is emerging for the efficacy of lamotrigine in prophylaxis of bipolar affective disorder, especially when depressive episodes predominate. Another drug class which is developing increasing importance in bipolar affective disorder is the atypical antipsychotics. Antipsychotic agents (both conventional and atypical) have long been used for control of acute manic symptoms, including both the grandiose delusions and thought disorder associated with manic psychosis and the acute behavioural disturbance which can occur in an extremely agitated patient. However, atypical antipsychotics such as olanzapine, quetiapine and aripiprazole are now known to be useful alternatives to conventional mood stabilisers in long-term prevention of manic relapse. Quetiapine appears to be particularly useful in preventing depressive relapse.

Other drugs that have been used in augmentation of existing agents include the anticonvulsants oxcarbazepine and gabapentin, the benzodiazepine clonazepam, and the calcium channel blocking agents verapamil and nimodipine.

Panic disorder (PD)

The main feature is recurrent, unexpected panic attacks. These are discrete periods of intense fear accompanied by characteristic physical symptoms such as skipping or pounding heart, sweating, hot flushes or chills, trembling/shaking, breathing difficulties, chest pain, nausea, diarrhoea and other gastrointestinal symptoms, dizziness or light-headedness. The first panic attack often occurs without warning but may subsequently become associated with specific situations, e.g. in a crowded shop, driving. Anticipatory anxiety and avoidance behaviour develop in response to this chain of events. The condition must be distinguished from alcohol withdrawal, caffeinism, hyperthyroidism and (rarely) phaeochromocytoma.

Patients experiencing panic attacks often do not know what is happening to them, and because the symptoms are similar to those of cardiovascular, respiratory or neurological conditions, often present to non-psychiatric services, e.g. casualty departments, family doctors, medical specialists, where they may either be extensively investigated or given reassurance that there is nothing wrong. A carefully taken history reduces the likelihood of this occurrence.

Treatment

The choice lies between a fast-acting benzodiazepine, such as lorazepam or alprazolam, and one with delayed efficacy but fewer problems of withdrawal, such as a tricyclic antidepressant (TCA), e.g. clomipramine or imipramine, or an SSRI, e.g. paroxetine. The different time course of these two classes of agent in panic disorder is depicted in Figure 20.4 (see also Table 20.6).

Fig. 20.4 Schematic representation of the time course of panic treatments.

Benzodiazepines rapidly reduce panic frequency and severity and continue to be effective for months; significant tolerance to the therapeutic action is uncommon. On withdrawal of the benzodiazepine, even when it is gradual, increased symptoms of anxiety and panic attacks may occur, reaching a maximum when the final dose is stopped. Indeed, some patients find they are unable to withdraw and remain long-term on a benzodiazepine.

Antidepressants (both SSRIs and TCAs) have a slower onset of action and may cause an initial increase in both anxiety and panic frequency, such that a patient may discontinue medication, even after a single dose. This provoking reaction usually lasts for no more than 2–3 weeks after which panic frequency and severity improve quickly, but patients need help to stay on treatment in the first weeks. The doctor needs to give a clear explanation of the likely course of events and the antidepressant should be started at half the usual initial dose to reduce the likelihood of exacerbation. The dose of antidepressant required to treat panic disorder is generally as high as, or higher than, that for depression and maximal benefit may not emerge for 8–12 weeks. Patients should therefore receive as large a dose as can be tolerated for this length of time.

Social anxiety disorder

The essential feature of social phobia is a marked and persistent fear of performance situations when patients feel they will be the centre of attention and will do something humiliating or embarrassing. The situations that provoke this fear can be quite specific, for example public speaking, or be of a much more generalised nature involving fear of most social interactions, for example initiating or maintaining conversations, participating in small groups, dating, speaking to anyone in authority. Exposure to the feared situation almost invariably provokes anxiety with similar symptoms to those experienced by patients with panic attacks, but some seem to be particularly prominent and difficult, i.e. blushing, tremor, sweating and a feeling of ‘drying up’ when speaking.

Post-traumatic stress disorder (PTSD)

Symptoms characteristically follow exposure to an extreme traumatic stressor event. These include persistent re-experiencing of the traumatic event, persistent avoidance of stimuli associated with the trauma and numbing of general responsiveness, and persistent symptoms of increased arousal. In taking a history the association with the event is usually obvious. PTSD is differentiated from acute stress disorder (below) by its persistence – the symptoms of the latter resolve within about 4 weeks. Depression quite commonly coexists with PTSD and should be enquired for in the history.

Acute stress disorder/adjustment disorder

Acute stress disorder is anxiety in response to a recent extreme stress. Although in some respects it is a normal and understandable reaction to an event, the problems associated with it are not only the severe distress the anxiety causes but also the risk that it may evolve into a more persistent state.

Generalised anxiety disorder (GAD)

The essential feature of this condition is chronic anxiety and worry. To the non-sufferer the focus of the worry often seems to be trivial, e.g. getting the housework done or being late for appointments, but to the patient it is insurmountable. The anxiety is often associated with other symptoms, which include restlessness, difficulty in concentrating, irritability, muscle tension and sleep disturbance. The course of the disorder is typically chronic with exacerbations at times of stress, and is often associated with depression. Its chronic nature with worsening at times of stress helps to distinguish GAD from anxiety in the form of episodic panic attacks with associated anticipatory anxiety (panic disorder). Hyperthyroidism and caffeinism should also be excluded.

Simple phobia

A specific phobia is a fear of a circumscribed object or situation, for instance fear of spiders, of flying, of heights. The diagnosis is not usually in doubt. A course of treatment by a trained therapist, involving graded exposure to the feared stimulus is the treatment of choice and can be very effective. By its nature such therapy generates severe anxiety and a benzodiazepine (or prior treatment with an SSRI, e.g. paroxetine) may be necessary to allow patients to engage in therapy.

Obsessive–compulsive disorder (OCD)

Obsessive–compulsive disorder has two main components:

OCD on its own often starts in adolescence and has a chronic and pervasive course unless treated. OCD starting later on in life is often associated with affective or anxiety disorders. Symptoms often abate briefly if the individual is taken to a new environment.

General comments about treating anxiety disorders (Table 20.7)

Normal sleep

Humans spend about a third of their lives asleep, but why we sleep is not yet fully understood. Sleep is a state of inactivity accompanied by loss of awareness and a markedly reduced responsiveness to environmental stimuli. The time of falling asleep is determined by three factors, which in normal sleepers occur at bedtime. These are: (1) circadian rhythm, i.e. the body’s natural clock in the hypothalamus triggers the rest/sleep part of the sleep–wake cycle; (2) ‘tiredness’, i.e. time since last sleep, usually about 16 h, which is in part mediated by the build-up of sleep-promoting chemicals in the brain; and (3) lowered mental and physical arousal. If one of these processes is disrupted then sleep initiation is difficult, and it is these three factors that are addressed by standard teaching of good sleep habits (see below).

Types of sleep disorder

Several types of sleep disorder are recognised, and their differentiation is important; a simplified summary is given below but reference to DSM-IV, ICD-10 and ICSD⁴ will clarify the exact diagnostic criteria:

• Insomnia: not enough sleep or sleep of poor quality; problems of falling asleep (initial insomnia) or staying asleep (maintenance insomnia), or waking too early.

• Hypersomnia: excessive daytime sleepiness.

• Parasomnia: unusual happenings in the night:

nightmares

night terrors

sleep walking

REM behaviour disorder.

• Other, e.g.:

sleep scheduling disorders (circadian rhythm disorder)

restless legs syndrome

periodic leg movements of sleep.

Insomnia

Insomnia is the complaint of poor sleep, with difficulty either in initiating sleep or maintaining sleep throughout the night, and causing significant distress and impairment in social, occupational or other important areas of functioning. It is by far the most common sleep disorder, with 7–10% of adults fulfilling these diagnostic criteria in surveys. It can occur exclusively in the course of a physical disorder such as pain, a mental disorder such as depression, or a sleep disorder such as restless legs syndrome, or it can be a primary sleep disorder. One survey showed similar deficits in quality of life in insomniacs as in patients with long-term disorders such as diabetes.

Insomnia may or may not be accompanied by daytime fatigue, but is not usually accompanied by subjective sleepiness during the day. If sleep propensity in the daytime is measured by objective means (time to EEG sleep), these patients are in fact less sleepy (take longer to fall asleep) than normal subjects.

A summary of precipitating factors for insomnia is shown in Box 20.1.

Timely treatment of short-term sleep disturbance is valuable, as it may prevent progression to a chronic condition, which is much harder to alleviate. Approach is:

• Treat any precipitating cause (see above).

• Educate and reassure.

• Establish good sleep habits.

Keep regular bedtimes and rising times.

Avoid daytime napping.

Daytime (but not evening) exercise and exposure to daylight.

Avoid stimulants, alcohol and cigarettes in the evening.

Establish bedtime routine: ‘wind-down’ – a milk drink may be helpful.

Avoid dwelling on problems in bed – set aside a time early in the evening for planning.

Bed should be comfortable.

• Consider hypnotic medication.

In the treatment of chronic insomnia the most important factor is anxiety about sleep, arising from conditioning behaviours that predispose to heightened arousal and tension at bedtime. Thus the bedroom is associated with not sleeping, and automatic negative thoughts about the sleeping process occur in the evening. Cognitive behavioural therapy for insomnia (CBTi), if available, is an effective treatment for insomnia delivered either individually or in small group format and has been found to be as effective as prescription medications for short-term treatment of chronic insomnia. Moreover, there are indications that the beneficial effects of CBT may last well beyond the termination of active treatment. Cognitive behavioural therapists help the patient to change behaviour and thoughts about sleep, particularly concentrating on learned, sleep-incompatible behaviours and automatic negative thoughts at bedtime.

Drugs for insomnia

Sleep–wake function involves a complex balance between arousing and sleep-inducing physiological systems. Current research suggests that arousal and wakefulness are promoted by parallel neurotransmitter systems whose cell bodies are located in brainstem or midbrain centres, with projections to the thalamus and forebrain. These activating neurotransmitters are noradrenaline/norepinephrine, serotonin, acetylcholine, dopamine and histamine. In addition the newly discovered orexin system with cell bodies in the hypothalamus promotes wakefulness through regulating arousal ‘pathways’ (and inhibiting sedative ones). For all these arousal neurotransmitters, sleep can be promoted by blocking their post-synaptic actions, leading to reduced arousal. For example many over-the-counter sleep-promoting agents contain antihistamines, which block the histamine H₁-receptor and so decrease arousal. The relatively low efficacy of these compounds may be explained by the fact that they target only one of the parallel arousal systems. The same is true for any drug which blocks one of the other arousal systems; they produce a degree of sedation but are not generally effective hypnotics.

The promotion of sleep is regulated by a number of other neurotransmitters; primary among these is γ-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the brain. The majority of brain cells are inhibited by GABA so increasing its function reduces arousal and produces sleep, and eventually anaesthesia. There are many subsets of GABA neurones distributed throughout the brain but a particular cluster in the hypothalamus (ventrolateral preoptic nucleus) can be considered to be the sleep ‘switch’.⁵ These neurones switch off brain arousal systems at the level of the cell bodies and therefore promote sleep. GABA receptors in the cortex can also promote sedation and sleep by inhibiting the target neurones of the arousal system. Most drugs used in insomnia act by increasing the effects of GABA at the GABA_A receptor.

The GABA_A–benzodiazepine receptor complex

γ-Aminobutyric acid (GABA) is the most important inhibitory transmitter in the central nervous system, comprising up to 40% of all synapses. GABAergic neurones are distributed widely in the CNS and GABA controls the state of neuronal excitability in all brain areas. The balance between excitatory inputs (mostly glutamatergic) and inhibitory GABAergic activity determines neuronal activity. If the balance swings in favour of GABA, then sedation, amnesia, muscle relaxation and ataxia appear and nervousness and anxiety are reduced. The mildest reduction of GABAergic activity (or increase in glutamate) elicits arousal, anxiety, restlessness, insomnia and exaggerated reactivity.

When GABA binds with the GABA_Areceptor, the permeability of the central pore of the receptor complex opens, so allowing more chloride ions into the neurone and decreasing excitability. Classical benzodiazepines (BZDs) in clinical use bind to another receptor on the complex (the benzodiazepine receptor) and enhance the effectiveness of GABA, producing a larger inhibitory effect (Fig. 20.5). These drugs are agonists at the receptor and an antagonist, flumazenil, prevents agonists from binding at the receptor site; it is used clinically to reverse benzodiazepine actions.

Fig. 20.5 Schematic representation of the binding sites on the GABA_A–benzodiazepine receptor complex. Note that drugs binding to the benzodiazepine (BZD) receptor site do not open the chloride channel directly, but rather augment the capacity of GABA to do so. Conversely agents such as the barbiturates both potentiate GABA and at higher concentrations have a direct effect on the chloride channel. This explains why barbiturates and alcohols are more toxic in overdose than the benzodiazepines.

Hypersomnia

Parasomnias

Other sleep disorders

Restless legs syndrome (RLS) is a disorder characterised by disagreeable leg sensations usually prior to sleep onset, and an almost irresistible urge to move the legs. The sensation is described as ‘crawling’, ‘aching’, ‘tingling’ and is partially or completely relieved with leg motion returning after movement ceases. Most, if not all, patients with this complaint also have periodic limb movements of sleep (PLMS), which may occur independently of RLS. These periodic limb movements consist of highly stereotyped movements, usually of the legs, that occur repeatedly (typically every 20–40 s) during the night. They may wake the patient, in which case there may be a complaint of daytime sleepiness or occasionally insomnia, but often only awaken the sleeping partner, who is usually kicked. RLS may respond to formulations of levodopa or dopamine agonists.

Sleep scheduling disorders. Circadian rhythm disorders are often confused with insomnia and both can be present in the same patient. With such sleep scheduling disorders, sleep occurs at the ‘wrong’ time, i.e. at a time that does not fit with work, social or family commitments. A typical pattern may be a difficulty in initiating sleep for a few nights due to stress, whereupon once asleep the subject continues sleeping well into the morning to ‘catch up’ the lost sleep. Thereafter the ‘time since last sleep’ cue for sleep initiation is delayed, and the sleep period gradually becomes more delayed until the subject is sleeping in the day instead of at night. A behavioural programme with strategic light exposure is appropriate, with pharmacological treatment as an adjunct, e.g. melatonin, to help reset the sleep–wake schedule.