Figure 8.1. Location of some of the genes on chromosomes that have been implicated in bipolar disorder.
Ol which, in summary, states that depression arises as a consequence of biogenic amine deficit, while mania is due to an excess of these amines in central synapses. In mania, evidence in support of this hypothesis comes from the limited studies that have been undertaken on patients before and after effective treatment. An alternative approach has been to study drugs such as lithium that have been used to treat the condition.
Most studies of the changes in the urine concentration of the main central metabolite of noradrenaline, 3-methoxy-4-hydroxyphenylglycol (MHPG), have shown abnormalities in manic patients. However, there is a discrepancy in the literature regarding the duration and extent of the change. Urinary noradrenaline concentrations have, however, been found to be increased during the active phase of the illness and to return to normal following effective treatment; the increase is said to reflect a rise in the concentration of MHPG in the cerebrospinal fluid (CSF). The concentration of the main dopamine metabolite, homovanillic acid (HVA), is also reported to rise in mania, but whether such changes are causally related to the core symptoms of the illness is debatable as it would be anticipated that an increase in the sympathetic drive would be necessarily associated with the illness.
While there have been a number of studies of changes in the sympathetic system in mania, few studies have attempted to assess changes in the serotonergic system. Hypomania has been reported to occur in depressed patients being treated with 5-hydroxytryptophan, the precursor amino acid of 5-HT, in combination with the peripheral decarboxylase inhibitor carbidopa. Mania has also been reported to occur in depressed patients following treatment with tryptophan in combination with clomipramine, a 5-HT uptake inhibitor. Nevertheless, there are no reports of a 5-HT agonist exacerbating the symptoms of mania in patients who are hypomanic! This suggests that a serotonergic stimulus may trigger a manic episode but alone is not a sufficient cause. Regarding the changes in serotonergic function in mania, only one study to date has investigated [3H]5-HT transport into the platelets of patients before and after effective treatment. Unlike depression, where the [3H]5-HT uptake is reduced, in mania the uptake is enhanced before treatment and normalized on recovery.
Regarding the dopaminergic system, there is experimental evidence to show that dopaminomimetic agents such as amphetamine, piribedil, bromocriptine and L-dopa can initiate mania in predisposed patients during remission. Indeed, the behavioural excitation and hypomania following D-amphetamine withdrawal has been proposed as a model of mania. Other evidence implicating a change in the dopaminergic system has been derived from the efficacy of neuroleptics (dopamine antagonists), which effectively attenuate the symptoms of the illness.
Unlike the biogenic amines, the cholinergic system has received relatively little attention as a possible factor in mania. Experimental evidence shows that cholinomimetic drugs and anticholinesterases have antimanic properties, although their effects appear to be short-lived. Furthermore, their effects appear to be associated with a reduction in the affective core symptoms and locomotor components of the illness, but not in the grandiose thinking and expansiveness.
Which, if any, of these different types of neurotransmitters is causally involved in the illness is still a matter for conjecture.
In SUMMARY, it would appear that the noradrenergic, serotonergic and dopaminergic systems are all overactive in acute mania while the inhibitory GABAergic system is underactive. This view would concur with the actions of drugs used to treat mania which primarily enhance inhibitory neurotransmission (which also explains their use as antiepileptics) and/or decrease excitatory neurotransmission. The changes thought to occur in mania with respect to these neurotransmitter pathways are illustrated in Figures 8.2, 8.3, 8.4 and 8.5.
Pharmacological treatment of mania
Of the various types of psychotropic drugs which have been used to treat mania, lithium salts are universally acclaimed to be the most important and effective treatment of mania and manic-depression.
It can be argued that the introduction of lithium salts into the practice of psychiatry in 1949 heralded the beginning of psychopharmacology, as it predated the discovery of chlorpromazine, imipramine, monoamine oxidase inhibitors and resperine. Lithium came into clinical use serendipitously, the Australian psychiatrist Cade having by chance given it to a small group of manic patients and found that it had beneficial effects,
whereas it appeared to lack activity when given to schizophrenics and depressives. However, lithium salts did not come into regular use in most industrialized countries until the early 1970s, partly because of the toxicity of the drug and partly because of the lack of commercial interest in a drug that could be dug out of the soil!
Lithium salts, generally in the form of the carbonate or bicarbonate, are rapidly absorbed from the gastrointestinal tract and reach a peak plasma concentration after 2-4 hours. Extreme fluctuations in blood lithium levels, which are associated with side effects such as nausea, diarrhoea and abdominal cramp, are reduced by using sustained release preparations. Lithium is not protein bound and therefore is widely distributed throughout the body water, which accounts for the adverse effects it has on most organ systems should it reach toxic levels. To avoid toxicity, and ensure optimal
efficacy, it is essential to monitor the plasma levels at regular intervals to ensure that they lie between 0.6 and 1.2 mEq/litre; there is evidence that lower levels (0.4-0.6 mEq/litre) may be sufficient when lithium salts are used to prevent relapse in the case of patients with unipolar depression.
As lithium is an alkaline earth metal which readily exchanges with sodium and potassium, it is actively transported across cell membranes. The penetration of kidney cells is particularly rapid, while that of bone, liver and brain tissue is much slower. The plasma: CSF ratio in man has been calculated to be between 2:1 and 3:1, which is similar to that found for the plasma: red blood cell (RBC) ratio. This suggests that the plasma: RBC ratio might be a useful index of the brain concentration and may be predictive of the onset of side effects, as these appear to correlate well with the intracellular concentration of the drug.
Most of the lithium is eliminated in the urine, the first phase of the elimination being 6-8 hours after administration, followed by a slower phase which may last for 2 weeks. Sodium-depleting diuretics such as frusemide, ethacrynic acid and the thiazides increase lithium retention and therefore toxicity, while osmotic diuretics as exemplified by mannitol and urea enhance lithium excretion. The principal side effects of lithium are summarized in Table 8.1.
Table 8.1. Main side effects of lithium
Neuromuscular changes General muscle weakness Ataxia Tremor
Fasciculation and twitching Choreoathetoid movements Hyperactive tendon reflexes*
Central nervous system Slurred speech* Blurring of vision Dizziness Vertigo
Cardiovascular system Hypotension Pulse irregularities ECG changes Circulatory collapse
General fatigue* and lethargy* Dehydration
*Side effects usually associated with the toxic effects of lithium.
The mode of action of lithium is still the subject of debate! Because of its similarity to sodium it was initially believed that it acted by competing with sodium in the brain and other tissues. However, it is now known that lithium interacts equally well with potassium, calcium and magnesium ions, all of which are widely distributed and essential for the functioning of most biological processes. It seems likely that lithium displaces sodium and potassium from their intracellular compartments and thereby substitutes for them; calcium, magnesium and phosphate concentrations are also altered. These effects of lithium on the electrolyte balance were once considered to be related to an action of the drug on sodium/potassium dependent adenosine triphosphatase (Na+K+-ATPase), an enzyme primarily involved in the repolarization of excitable membranes. Lithium appears to compete with a common binding site on the carrier, the site having a greater affinity for the lithium than the sodium ion. This could account for the ability of the drug to slow the speed of repolarization of nervous tissue. Other effects on brain function may be associated with an increase in the permeability of the blood-brain barrier resulting from an interaction of lithium with membrane phospholipids. The increased concentration of amino acids in the CSF may be a reflection of this.
It has long been apparent that the uptake, storage, release and metabolism (i.e. the turnover) of biogenic amines can be affected by both mono- and divalent cations. The effect on dopamine receptor sensitivity may be a particularly important action of lithium. It has been speculated that dopamine receptor hypersensitivity is closely associated with the onset of mania. Both acutely and chronically administered lithium can reduce the supersensitivity of both pre- and postsynaptic dopamine receptors, an effect which may help to explain its mood-stabilizing action but not its somewhat controversial ability to initiate tardive dyskinesia. Regarding the effects of lithium on noradrenergic function, it has long been known that it increases the reuptake of noradrenaline into neurons, increases the turnover of this amine in the brain without markedly affecting its turnover in the periphery, decreases its release and enhances its metabolism. The net effect of lithium is therefore to lead to a reduction in noradrenergic function which presumably is reflected in its antimanic properties.
At the postsynaptic level, lithium has been shown to reduce the function of beta adrenoceptors, presumably by affecting the coupling between the receptor and the secondary messenger system. This effect only becomes apparent following chronic treatment, which may help to explain the delay of several days, or even weeks, before an optimal beneficial effect is observed. All antidepressants are known to reduce the functional activity of postsynaptic beta receptors, which may explain why lithium has both an antimanic and an antidepressant effect in patients with manic-depression.
Relationship between the mode of action of lithium and its side effects
Many of the adverse effects of lithium can be ascribed to the action of lithium on adenylate cyclase, the key enzyme that links many hormones and neurotransmitters with their intracellular actions. Thus antidiuretic hormone and thyroid-stimulating-hormone-sensitive adenylate cyclases are inhibited by therapeutic concentrations of the drug, which frequently leads to enhanced diuresis, hypothyroidism and even goitre. Aldosterone synthesis is increased following chronic lithium treatment and is probably a secondary consequence of the enhanced diuresis caused by the inhibition of antidiuretic-hormone-sensitive adenylate cyclase in the kidney. There is also evidence that chronic lithium treatment causes an increase in serum parathyroid hormone levels and, with this, a rise in calcium and magnesium concentrations. A decrease in plasma phosphate and in bone mineralization can also be attributed to the effects of the drug on parathyroid activity. Whether these changes are of any clinical consequence is unclear.
Prolactin secretion, at least in experimental animals, is increased following chronic lithium treatment, probably as a consequence of the enhanced sensitivity of postsynaptic 5-HT receptors and the decreased sensitivity of dopamine receptors. In patients on therapeutic doses of the drug, however, the plasma prolactin levels would not appear to be markedly altered. There is little evidence that circulating gonadotrophin concentrations are affected by therapeutic doses of lithium.
One major side effect of lithium that causes great concern to patients is weight gain; this has been estimated to occur in up to 60% of patients according to some investigators. In addition to increased food intake, lithium also has an effect on the intermediary metabolism of carbohydrates. During the acute phase of lithium administration, insulin release is decreased leading to a raised plasma glucose; the insulin concentration then rises and increased fat synthesis could then occur. This appears to be due to an inhibition of several enzymes at the beginning of the glycolytic pathway, which could lead to enhanced lipid synthesis.
Recently research has focused on the action of lithium on serotonergic function. Lithium has been shown to facilitate the uptake and synthesis of 5-HT, to enhance its release and to increase the transport of tryptophan into the nerve terminal, an effect which probably contributes to the increased 5-HT synthesis. The net effect of these changes is to produce postsynaptic receptor events, which might explain why lithium, in combination with tryptophan and a monoamine oxidase inhibitor or a 5-HT uptake inhibitor, is often effective in therapy-resistant depression.
Drugs which enhance the activity of the central cholinergic system have been shown to have antimanic effects. Experimental studies have shown that lithium increases acetylcholine synthesis in the cortex, which is probably associated with an increase in the high affinity transport of choline into the neuron; the release of this transmitter is also increased. Whether these effects on the cholinergic system are relevant to its therapeutic action in manic patients remains to be proven.
So far attention has concentrated on the effects of lithium on excitatory transmitters. There is evidence that the drug can also facilitate inhibitory transmission, an effect that has been attributed to a desensitization of the presynaptic gamma-aminobutyric acid (GABA) receptors, which results in an increase in the release of this inhibitory transmitter. The increased conversion of glutamate to GABA may also contribute to this process. Thus it would appear that lithium has a varied and complex action on central neurotransmission, the net result being a diminution in the activity of excitatory transmitters and an increase in GABAergic function.
When receptors are directly linked to ion channels, fast excitatory or inhibitory postsynaptic potentials occur. However, it is well established that slow potential changes also occur and that such changes are due to the receptor being linked to the ion channel indirectly via a secondary messenger system. For example, the stimulation of beta adrenoceptors by noradrenaline results in the activation of adenylate cyclase. The antimanic and antidepressant effects of lithium are linked to a reduction in the functional activity of postsynaptic beta adrenoceptor-linked cyclase, combined with a reduction in the activity of the presynaptic noradrenergic neuron. The adverse effects of the drug on renal and thyroid function are due to the inhibition of the hormone-linked cyclases in these organs. Undoubtedly, transmitter receptor changes (e.g. serotonergic, noradrenergic, dopaminergic and GABAergic) play a major role in the therapeutic effects of lithium. Such changes may be related to the ability of the drug to re-synchronize disrupted circadian rhythms, which appear to be an essential feature of the affective disorders.
The long-term toxic effects of lithium, such as nephrogenic diabetes insipidus, which has been calculated to occur in up to 5% of patients, and the rare possibility of lithium combined with neuroleptics being neurotoxic, has stimulated the research for other drug treatments. However, apart from the neuroleptics, these drugs have not been studied as extensively in the treatment of acute mania, but are worthy of consideration because of their reduced side effects.
Most psychotic and non-compliant patients are difficult to treat with lithium alone and need to be treated with neuroleptics. Haloperidol has been widely used alone to control the more florid symptoms of mania, but doubts have arisen concerning its toxic interactions with lithium. Such considerations are based on a report that such a combination caused neurotoxicity in a small group of manic patients, but it should be emphasized that a variety of other neuroleptics have also been rarely found to cause these effects. The symptoms of neurotoxicity include ataxia, confusion, hyperactive reflexes, chorea, slurred speech and even coma. It seems likely that some of these patients suffered from the malignant neuroleptic syndrome rather than enhanced lithium toxicity, but problems such as dehydration and over-sedation may have enhanced the drug interaction. More recently, atypical antipsychotics such as olanzapine and risperidone have been shown to be effective in the treatment of acute mania. These drugs have advantages over haloperidol, and the first-generation neuroleptics, due to the improved side-effect profiles and better patient compliance.
Tardive dyskinesia can occur in manic patients on neuroleptics alone, the frequency may be greater than in schizophrenics who are more likely to be on continuous medication. One possible explanation for this lies in the fact that neuroleptics are often administered to manic patients for short periods only, sufficient to abort the active episode, and then abruptly stopped. Thus high doses of neuroleptics are separated by drug-free periods, leading to a situation most likely to precipitate tardive dyskinesia. The recent increase in prescribing high potency neuroleptics such as haloperidol instead of low potency drugs such as chlorpromazine or thioridazine has undoubtedly increased the frequency of tardive dyskinesia. Clearly, use of the atypical antipsychotics with the very low frequency of EPS makes them the treatments of choice.
Valproic acid (dipropylacetic acid) is a single branched chain carboxylic acid that is structurally unlike any of the other drugs used in the treatment of bipolar disorder or epilepsy. The amide derivative, valproamide, is available in Europe as a more potent form of valproate. Valproate was first developed in France as an antiepileptic agent in 1963. As an antiepileptic agent, it was shown to be active against a variety of epilepsies without causing marked sedation.
The mechanism of action of valproate is complex and still the subject of uncertainty. The drug appears to act by enhancing GABAergic function. Thus it increases GABA release, inhibits catabolism and increases the density of GABA-B receptors in the brain. There is also evidence that it increases the sensitivity of GABA receptors to the action of the inhibitory transmitter. Other actions that may contribute to its therapeutic effects include a decrease in dopamine turnover, a decrease in the activity of the NMDA-glutamate receptors and also a decrease in the concentration of somatostatin in the CSF. Unlike carbamazepine, valproate does not bind to peripheral benzodiazepine receptors (see p. 230).
Numerous open studies, and seven controlled studies, have shown that valproate is effective in the treatment of acute mania. It has also been claimed to have an antidepressant action. Recent studies have shown that valproate is effective in the long-term treatment of bipolar disorder.
Valproate is generally well tolerated and less likely to cause cognitive impairment than other antiepileptic drugs such as carbamazepine. It does frequently cause gastrointestinal upset and a benign elevation of liver transaminases however. Because valproate is highly plasma protein bound, and is partially metabolized by the cytochrome P450 system, it can interact with many other drugs. For example, aspirin can enhance the efficacy and toxicity of valproate by displacing it from the plasma proteins while microsomal enzyme-inducing drugs such as carbamazepine can decrease its plasma and tissue concentrations. The general properties of valproate are further discussed in Chapter 12.
This is a tricyclic compound somewhat similar to imipramine that is an anticonvulsant widely used in the treatment of temporal lobe epilepsy. Following its widespread use as an antiepileptic, it soon became evident that it had psychotropic effects. These included an improvement in mood, reduced aggressiveness and improved cognitive function. Kindling refers to the development of seizures after repeated delivery of a series of subthreshold stimuli to any region of the brain. This phenomenon can most readily be induced in limbic structures and, whereas conventional anticonvulsants such as phenytoin and phenobarbitone have little effect in attenuating kindled seizures, carbamazepine and the benzodiazepine anticonvulsants prevent such seizure development. It is now well established that carbamazepine is relatively selective in attenuating seizure activity in the hippocampus and amygdala, which suggests that it acts preferentially at limbic sites in the brain.
The mechanism of action of carbamazepine is complex, and is complicated by the fact that it has a long half-life metabolite, carbamazepine epoxide, which also has pronounced psychotropic properties.
The anticonvulsant properties of the drug would appear to be due to its ability to inhibit fast sodium channels, which may be unrelated to its psychotropic effects. Like lithium, it has been shown to decrease the release of noradrenaline and reduce noradrenaline-induced adenylate cyclase activity; unlike lithium, it seems to have little effect on tryptophan or 5-HT levels in patients at therapeutically relevant concentrations. It also reduces dopamine turnover in manic patients and increases acetylcholine synthesis in the cortex, an effect also seen with lithium. The effect of carbamazepine on GABAergic function appears to be related to its interaction with GABA-B type receptors, which may be relevant to its usefulness in the treatment of trigeminal neuralgia. There is no evidence that it changes GABA levels in the CSF of patients. Furthermore, while it would appear that the drug has no effect on central benzodiazepine receptors, there is evidence that it has a high affinity for the peripheral type of benzodiazepine receptor. These receptors are found in the mammalian brain but differ from the central receptors in that they are not linked to GABA receptors and therefore do not affect chloride ion flux. The main function of the peripheral type of benzodiazepine receptor would seem to be to control calcium channels. This may help to explain some of the psychotropic effects of carbamazepine, particularly as calcium channel antagonists such as verapamil have antimanic effects.
Changes in the activity of adenosine receptors have been implicated in the stimulant effects of drugs like caffeine. Carbamazepine exhibits a partial agonist effect on adenosine receptors, and experimental evidence suggests that the reduced reuptake and release of noradrenaline caused by the drug are due to its interaction with these receptors. The precise relevance of these findings to its anticonvulsant and psychotropic effects is presently unclear.
Of the various peptides (e.g. the opioids, vasopressin, substance P and somatostatin) thought to be involved in the actions of carbamazepine, there is evidence that the reduction in the CSF concentration of somatostatin might be important in explaining its effects on cognition and also on the hypothalamo-pituitary-adrenal axis; somatostatin is a major inhibitory modulator of this axis and hypercortisolism frequently occurs in patients following carbamazepine administration.
There is still controversy regarding the general usefulness of carbama-zepine as an alternative to lithium. It is apparent that the nature of the illness alters throughout the lifetime of the patient, so that pharmacological interventions may differ according to the stage of the illness. Preliminary clinical studies suggest that lithium may be particularly beneficial during the early and intermediate stages of the illness, whereas carbamazepine and related anticonvulsants may be more useful, either alone or in combination with lithium, at later stages, particularly, when the patient shows rapid, continuous cycling between mania and depression.
Other drugs that are reported to have beneficial effects but which have not undergone such extensive evaluation as the neuroleptics or carbamazepine include the calcium channel antagonists such as verapamil. A small open study has suggested that the alpha2 adrenoceptor agonist clonidine may have some activity. More substantial studies have been conducted on the benzodiazepines lorazepam and clonazepam, and the anticonvulsant sodium valproate. All these drugs facilitate GABAergic function in some way, the first two by acting as agonists at benzodiazepine receptor sites and the latter by desensitizing the GABA autoreceptor and thereby enhancing the release of this inhibitory transmitter. Lastly, electroconvulsive shock treatment (ECT) has been claimed to be effective in attenuating the symptoms of an acute manic attack, but there is evidence that patients treated with ECT should not receive lithium concomitantly to reduce the possibility of neurotoxic side effects.
In addition to these drugs, many of the newer antiepileptic drugs such as lamotrigine have found a place in the therapeutic management of mania. These are extensively covered in Chapter 12.
The pharmacological management of bipolar disorder involves treatment of both the acute and the longer-term maintenance phase of the illness. Long-term maintenance is necessary to reduce or prevent the recurrence of the symptoms, and to minimize the risk of suicide.
For many years, lithium salts have been used for maintenance treatment. However, naturalistic studies have reported a relatively high failure rate in patients on lithium and therefore other therapeutic approaches have been considered.
With regard to the use of lithium in maintenance therapy, the studies which were published in the 1970s clearly demonstrated the efficacy of lithium in preventing relapse into mania or depression in patients with bipolar disorder. However, subsequent longer-term naturalistic studies raised doubts over the validity of these findings. In particular, these studies have shown that up to 50% of patients respond poorly to lithium. Some of the reasons for the re-evaluation of the early reports on the efficacy of lithium as a maintenance treatment are due to the methodological limitations of the placebo-controlled studies which include a lack of diagnostic criteria and a limited consideration of those patients withdrawing from the clinical trial prematurely.
In contrast to the large number of studies that have investigated lithium as a maintenance treatment for bipolar disorder, relatively few studies have been made of divalproex sodium, despite its widespread use in the acute treatment of mania. There is evidence from one placebo-controlled study in which lithium was compared with divalproex sodium that the latter drug was better tolerated but that the prevention of relapse did not differ between the drugs. It would therefore appear that a switch to divalproex sodium may be particularly useful in bipolar patients who are experiencing cognitive deficits, loss of creativity and functional impairments consequent on lithium use.
Again there are relatively few studies that have investigated the use of carbamazepine in maintenance therapy. The results of the studies published suggest that carbamazepine is not as effective as lithium or divalproex. In the controlled studies of carbamazepine, the majority of patients required adjunctive treatment to prevent a breakthrough for the manic or depressive symptoms.
Despite the widespread use of neuroleptics in maintenance treatment of bipolar disorder, there have not been any systematic studies of their suitability for this role. Through clinical experience it has been widely accepted that neuroleptics are useful adjunctive treatments to lithium and related drugs. Treatment refractory patients frequently respond to atypical antipsychotics such as clozapine or risperidone. Such adverse effects as EPS, cognitive dysfunction and weight gain frequently limit the long-term use of classical neuroleptics. For this reason, the atypical neuroleptics such as olanzapine and risperidone should now be considered as alternatives for maintenance treatment.
• Treatment of choice - mood stabilizer with or without an antidepressant (e.g. lithium, valproate, carbamazepine, lamotrigine). Antidepressants include an SSRI, venlafaxine, mirtazepine as possibilities but few controlled trials to substantiate choice.
• Switching - alternative mood stabilizer plus alternative second-generation antidepressant.
• Augmentation of the response - combine two mood stabilizers; add thyroid hormone to mood stabilizer.
• Other options - ECT; possibly calcium channel blockers such as verapamil or nimodipine.
In CONCLUSION, lithium is universally accepted as a mood-stabilizing drug and an effective antimanic agent whose value is limited by its poor therapeutic index (i.e. its therapeutic to toxicity ratio). Neuroleptics are effective in attenuating the symptoms of acute mania but they too have serious adverse side effects. High potency typical neuroleptics appear to increase the likelihood of tardive dyskinesia. Of the less well-established treatments, carbamazepine would appear to have a role, particularly in the more advanced stages of the illness when lithium is less effective.
Was this article helpful?