CLINICAL PHARMACOLOGY OF LITHIUM
Until recently, lithium carbonate was the universally preferred treatment for bipolar disorder, especially in the manic phase. With the approval of valproate, aripiprazole, olanzapine, quetiapine, risperidone, and ziprasidone for this indication, a smaller percent-age of bipolar patients now receive lithium. This trend is rein-forced by the slow onset of action of lithium, which has often been supplemented with concurrent use of antipsychotic drugs or potent benzodiazepines in severely manic patients. The overall success rate for achieving remission from the manic phase of bipo-lar disorder can be as high as 80% but lower among patients who require hospitalization. A similar situation applies to maintenance treatment, which is about 60% effective overall but less in severely ill patients. These considerations have led to increased use of com-bined treatment in severe cases. After mania is controlled, the antipsychotic drug may be stopped and benzodiazepines and lithium continued as maintenance therapy.
The depressive phase of manic-depressive disorder often requires concurrent use of an antidepressant drug . Tricyclic antidepressant agents have been linked to precipitation of mania, with more rapid cycling of mood swings, although most patients do not show this effect. Selective serotonin reuptake inhibitors are less likely to induce mania but may have limited efficacy. Bupropion has shown some promise but—like tricyclic antidepressants—may induce mania at higher doses. As shown in recent controlled trials, the anticonvulsant lamotrigine is effective for many patients with bipolar depression. For some patients, however, one of the older monoamine oxidase inhibitors may be the antidepressant of choice. Quetiapine and the combination of olanzapine and fluoxetine has been approved for use in bipolar depression.
Unlike antipsychotic or antidepressant drugs, which exert sev-eral actions on the central or autonomic nervous system, lithium ion at therapeutic concentrations is devoid of autonomic blocking effects and of activating or sedating effects, although it can pro-duce nausea and tremor. Most important is that the prophylactic use of lithium can prevent both mania and depression. Many experts believe that the aggressive marketing of newer drugs has inappropriately produced a shift to drugs that are less effective than lithium for substantial numbers of patients.
Recurrent endogenous depression with a cyclic pattern is con-trolled by either lithium or imipramine, both of which are supe-rior to placebo.Schizoaffective disorder, another condition with an affectivecomponent characterized by a mixture of schizophrenic symptoms and depression or excitement, is treated with antipsychotic drugs alone or combined with lithium. Various antidepressants are added if depression is present.
Lithium alone is rarely successful in treating schizophrenia, but adding it to an antipsychotic may salvage an otherwise treat-ment-resistant patient. Carbamazepine may work equally well when added to an antipsychotic drug.
An interesting application of lithium that is relatively well sup-ported by controlled studies is as an adjunct to tricyclic antide-pressants and SSRIs in patients with unipolar depression who do not respond fully to monotherapy with the antidepressant. For this application, concentrations of lithium at the lower end of the recommended range for manic-depressive illness appear to be adequate.
Clinicians rely on measurements of serum lithium concentrations for assessing both the dosage required for treatment of acute mania and for prophylactic maintenance. These measurements are customarily taken 10–12 hours after the last dose, so all data in the literature pertaining to these concentrations reflect this interval.
An initial determination of serum lithium concentration should be obtained about 5 days after the start of treatment, at which time steady-state conditions should have been attained. If the clinical response suggests a change in dosage, simple arithmetic (new dose equals present dose times desired blood level divided by present blood level) should produce the desired level. The serum concentration attained with the adjusted dos-age can be checked after another 5 days. Once the desired concentration has been achieved, levels can be measured at increasing intervals unless the schedule is influenced by inter-current illness or the introduction of a new drug into the treat-ment program.
The decision to use lithium as prophylactic treatment depends on many factors: the frequency and severity of previous episodes, a cre-scendo pattern of appearance, and the degree to which the patient is willing to follow a program of indefinite maintenance therapy. If the present attack was the patient’s first or if the patient is unreliable, one might prefer to terminate treatment after the episode has subsided. Patients who have one or more episodes of illness per year are candi-dates for maintenance treatment. Although some patients can be maintained with serum levels as low as 0.6 mEq/L, the best results have been obtained with higher levels, such as 0.9 mEq/L.
Renal clearance of lithium is reduced about 25% by diuretics (eg, thiazides), and doses may need to be reduced by a similar amount. A similar reduction in lithium clearance has been noted with several of the newer nonsteroidal anti-inflammatory drugs that block synthesis of prostaglandins. This interaction has not been reported for either aspirin or acetaminophen. All neuroleptics tested to date, with the possible exception of clo-zapine and the newer atypical antipsychotics, may produce more severe extrapyramidal syndromes when combined with lithium.
Many adverse effects associated with lithium treatment occur at varying times after treatment is started. Some are harmless, but it is important to be alert to adverse effects that may signify impend-ing serious toxic reactions.
Tremor is one of the most common adverse effects of lithiumtreatment, and it occurs with therapeutic doses. Propranolol and atenolol, which have been reported to be effective in essential tremor, also alleviate lithium-induced tremor. Other reported neurologic abnormalities include choreoathetosis, motor hyper-activity, ataxia, dysarthria, and aphasia. Psychiatric disturbances at toxic concentrations are generally marked by mental confusion and withdrawal. Appearance of any new neurologic or psy-chiatric symptoms or signs is a clear indication for temporarily stopping treatment with lithium and for close monitoring of serum levels.
Lithium probably decreases thyroid function in most patients exposed to the drug, but the effect is reversible or nonprogressive. Few patients develop frank thyroid enlargement, and fewer still show symptoms of hypothyroidism. Although initial thyroid testing followed by regular monitoring of thyroid function has been pro-posed, such procedures are not cost-effective. Obtaining a serum TSH concentration every 6–12 months, however, is prudent.
Polydipsia and polyuria are common but reversible concomitants of lithium treatment, occurring at therapeutic serum concentra-tions. The principal physiologic lesion involved is loss of respon-siveness to antidiuretic hormone (nephrogenic diabetes insipidus). Lithium-induced diabetes insipidus is resistant to vasopressin but responds to amiloride.
Extensive literature has accumulated concerning other forms of renal dysfunction during long-term lithium therapy, including chronic interstitial nephritis and minimal-change glomerulopathy with nephrotic syndrome. Some instances of decreased glomerular filtration rate have been encountered but no instances of marked azotemia or renal failure.
Patients receiving lithium should avoid dehydration and the asso-ciated increased concentration of lithium in urine. Periodic tests of renal concentrating ability should be performed to detect changes.
Edema is a common adverse effect of lithium treatment and may be related to some effect of lithium on sodium retention. Although weight gain may be expected in patients who become edematous, water retention does not account for the weight gain observed in up to 30% of patients taking lithium.
The bradycardia-tachycardia (“sick sinus”) syndrome is a definite contraindication to the use of lithium because the ion further depresses the sinus node. T-wave flattening is often observed on the electrocardiogram but is of questionable significance.
Renal clearance of lithium increases during pregnancy and reverts to lower levels immediately after delivery. A patient whose serum lithium concentration is in a good therapeutic range during preg-nancy may develop toxic levels after delivery. Special care in monitoring lithium levels is needed at these times. Lithium is transferred to nursing infants through breast milk, in which it has a concentration about one third to one half that of serum. Lithium toxicity in newborns is manifested by lethargy, cyanosis, poor suck and Moro reflexes, and perhaps hepatomegaly.
The issue of lithium-induced dysmorphogenesis is not settled. An earlier report suggested an increase in cardiac anomalies— especially Ebstein’s anomaly—in lithium babies, and it is listed as such in Table 59–1 in this book. However, more recent data suggest that lithium carries a relatively low risk of teratogenic effects. Further research is needed in this important area.
Transient acneiform eruptions have been noted early in lithium treatment. Some of them subside with temporary discontinuance of treatment and do not recur with its resumption. Folliculitis is less dramatic and probably occurs more frequently. Leukocytosis is always present during lithium treatment, probably reflecting a direct effect on leukopoiesis rather than mobilization from the marginal pool. This adverse effect has now become a therapeutic effect in patients with low leukocyte counts.
Therapeutic overdoses of lithium are more common than those due to deliberate or accidental ingestion of the drug. Therapeutic over-doses are usually due to accumulation of lithium resulting from some change in the patient’s status, such as diminished serum sodium, use of diuretics, or fluctuating renal function. Since the tissues will have already equilibrated with the blood, the plasma concentrations of lithium may not be excessively high in proportion to the degree of toxicity; any value over 2 mEq/L must be considered as indicating likely toxicity. Because lithium is a small ion, it is dialyzed readily. Both peritoneal dialysis and hemodialysis are effective, although the latter is preferred.
Valproic acid (valproate), as an antiepileptic, has been demonstrated to have antimanic effects andis now being widely used for this indication in the USA. (Gabapentin is not effective, leaving the mechanism of action of valproate unclear.) Overall, valproic acid shows efficacy equivalent to that of lithium during the early weeks of treatment. It is significant that valproic acid has been effective in some patients who have failed to respond to lithium. Moreover, its side-effect profile is such that one can rapidly increase the dosage over a few days to produce blood levels in the apparent therapeutic range, with nausea being the only limiting factor in some patients. The starting dosage is 750 mg/d, increasing rapidly to the 1500–2000 mg range with a recommended maximum dosage of 60 mg/kg/d.
Combinations of valproic acid with other psychotropic medica-tions likely to be used in the management of either phase of bipolar illness are generally well tolerated. Valproic acid is an appropriate first-line treatment for mania, although it is not clear that it will be as effective as lithium as a maintenance treatment in all subsets of patients. Many clinicians advocate combining valproic acid and lith-ium in patients who do not fully respond to either agent alone.
Carbamazepine has been considered to be a reasonable alternative to lithium when the latter is less than optimally efficacious. The mode of action of carbamazepine is unclear, and oxcarbazepine is not effective. Carbamazepine may be used to treat acute mania and also for prophylactic therapy. Adverse effects are generally no greater and sometimes less than those associated with lithium. Carbamazepine may be used alone or, in refractory patients, in combination with lithium or, rarely, valproate.
The use of carbamazepine as a mood stabilizer is similar to its use as an anticonvulsant . Dosage usually begins with 200 mg twice daily, with increases as needed. Maintenance dosage is similar to that used for treating epilepsy, ie, 800–1200 mg/d. Plasma concentrations between 3 and 14 mg/L are considered desirable, although no therapeutic range has been established. Blood dyscrasias have figured prominently in the adverse effects of carbamazepine when it is used as an anticonvulsant, but they have not been a major problem with its use as a mood stabilizer. Overdoses of carbamazepine are a major emergency and should generally be managed like over-doses of tricyclic antidepressants .
Lamotrigine has been reported to be useful in preventing the depression that often follows the manic phase of bipolar disorder. A number of novel agents are under investigation for bipolar depression, including riluzole, a neuroprotective agent that is approved for use in amyotrophic lateral sclerosis; ketamine, a noncompetitive NMDA antagonist previously discussed as a drug believed to model schizophrenia but thought to act by producing relative enhancement of AMPA receptor activity; and AMPA receptor potentiators.