Adverse Effects of Antipsychotics and Mood Stabilizers
Adverse Effects of Antipsychotics and Mood Stabilizers
Although psychotropic medications have revolutionized the treatment of many psychiatric disorders, the benefits sometimes come at a price. Unwanted adverse effects may limit the use of these agents and may, in some instances, result in serious morbidity and mortality.
In recent years, the adverse effects of second-generation antipsychotics on body weight have become a significant concern for clinicians and their patients; however, other metabolic effects of these drugs, as well as metabolic effects of first-generation antipsychotics and mood stabilizers, also require our attention and awareness. In this article we focus on the phenomenology, evaluation, and management of endocrine and metabolic adverse effects of these commonly used psychotropic agents.
Lithium and thyroid function
At therapeutic serum levels, the main action of lithium is to inhibit hormone release from the thyroid gland, which may result in hypothyroidism. Patients with underlying Hashimoto thyroiditis or radiation-induced thyroid injury may be particularly susceptible to this inhibitory effect.1 In adults, the prevalence of lithium-induced hypothyroid-ism generally increases with longer duration of therapy--about 20% of women and 5% of men are affected with hypothyroidism after 10 years of continuous treatment.2 The results of a pediatric study also suggest that patients taking both lithium and valproate may be at particularly high risk of having hypothyroidism develop, since the weak inhibitory action of valproate on the thyroid may synergize with the effects of lithium.3
Patients should have baseline thyroid function tests (at least for serum thyroid-stimulating hormone [TSH]) performed before beginning lithium treatment; since lithium-induced hypothyroidism may occur at any time during lithium therapy.1-3 Tests should be repeated 1 to 2 months, 6 months, and 12 months after starting lithium, and yearly thereafter. If hypothyroidism occurs during lithium treatment, replacement thyroid hormone therapy with levothyroxine should be started, and the dosage increased as needed to normalize serum TSH. Since lithium-induced hypothyroidism may occasionally be transient, it may be useful to try to slightly decrease the daily thyroxine dose after several years of treatment; if the serum TSH remains normal on the lower dose, further dose reductions may then be attempted.1
Lithium and parathyroid function
Mild hypercalcemia may develop in patients receiving lithium, generally in the range of 10.5 to 11.5 mg/dL, with normal or mildly elevated serum levels of parathyroid hormone, usually along with decreased urinary calcium excretion. This constellation of abnormalities appears to be due to a lithium-induced loss of sensitivity of the parathyroid calcium-sensing receptor; the parathyroid gland perceives a higher than normal serum calcium level as normal, and parathyroid hormone secretion is regulated around an elevated set point.4 About 5% of patients receiving long-term lithium therapy have been found to have hypercalcemia.5,6 The hypercalcemia is usually not progressive, and kidney stones are uncommon (since the patients are usually hypocalciuric).
Although the hypercalcemia is usually reversible if lithium is discontinued, in some instances it has been persistent, and parathyroid adenomas or multigland parathyroid hyperplasia have been found in some patients with lithium-induced hypercalcemia.7 Because hypercalcemia may occur at any time during lithium therapy,4-7 serum calcium should be measured 1, 6, and 12 months after starting lithium, and yearly thereafter. If serum calcium levels exceed 11.5 mg/dL or if clinical signs and symptoms of hyperparathyroidism (eg, fatigue, nausea, constipation, worsening osteoporosis) develop, lithium may need to be discontinued.
Antipsychotics and hyperprolactinemia
Prolactin secretion is primarily regulated through the suppressive action of dopamine, which is secreted from hypothalamic neurons and delivered to the anterior pituitary through the hypothalamic-pituitary portal blood vessels. Medications that block the pituitary dopamine D2 receptor will interfere with this tonic inhibitory effect of dopamine and thereby increase serum prolactin levels. Most typical (first-generation) and some atypical (second-generation) antipsychotics raise serum prolactin acutely; after the initial rise, serum prolactin frequently declines gradually (months to years), and in some cases may return to normal, despite continued administration of the drug.8
The atypical antipsychotic agents are variable in their effects on prolactin, since they vary in their affinity for the pituitary dopamine D2 receptor. One of the atypical agents, aripiprazole, acts on the dopamine receptor as a partial agonist (which lowers serum prolactin levels). Women tend to have higher prolactin responses to antipsychotics than men because estrogen stimulates prolactin synthesis and responsiveness.9 The approximate relative potency of antipsychotics in causing hyperprolactinemia is (in order of decreasing potency) risperidone, haloperidol, olanzapine, ziprasidone, quetiapine, clozapine, and aripiprazole.
Prolactin acts on the hypothalamus to suppress the secretion of gonadotropin-releasing hormone, which results in decreased secretion of luteinizing hormone and follicle-stimulating hormone and, ultimately, hypogonadism. In women, the hypogonadism is manifest most commonly as secondary amenorrhea and infertility; decreased libido and osteoporosis can occur in both women and men as a result of decreased gonadal production of sex steroids. Breast enlargement and breast pain can be seen in both men and women; galactorrhea is more common in women, who have estrogen-primed breasts. Although hyperprolactinemia-induced hypogonadism might be expected to slow pu- bertal development, normal pubertal progression over a 12-month period was observed in the only study to date that specifically addressed this issue.10 Erectile dysfunction may also occur, perhaps as a direct effect of prolactin on the nervous system11 as well as an effect of hypogonadism.
A recent study based on adverse-event reports submitted to the FDA has suggested that antipsychotic-induced hyperprolactinemia may be associated with an increased incidence of benign, presumably prolactin-secreting pituitary adenomas.12 There were 77 cases of pituitary tumors reported in patients taking antipsychotic agents; 54 of these involved risperidone, the drug with the greatest propensity to cause hyperprolactinemia. Although a caus-al relationship cannot be excluded, it seems more likely that symptomatic hyperprolactinemia developed in patients receiving antipsychotics, particularly risperidone, who then underwent MRI scanning of the pituitary as part of a diagnostic evaluation. Since incidental pituitary tumors (usually nonfunctioning) are found on MRI scans in about 10% of the general population,13 the tumors found in the patients treated with antipsychotics may bear no relationship to the drug therapy.
During drug therapy, the physician should inquire about menstruation, nipple discharge, breast enlargement, sexual functioning, and (if appropriate) pubertal development. If these are normal, there is no need to measure serum prolactin. However, if problems in these areas are uncovered that are temporally related to antipsychotic therapy, serum prolactin level should be measured.
If the level of serum prolactin is elevated but less than 200 ng/mL in a symptomatic patient, other causes of hyperprolactinemia (such as pregnancy, hypothyroidism, and renal failure) should be excluded by measuring serum human chorionic gonadotropin, TSH, and creatinine levels. If these tests are negative, one might try to decrease the dosage of the antipsychotic medication or change treatment to an agent less likely to elevate serum prolactin. If it is not possible to lower the dosage or switch to a different medication, it may be helpful to add a small dose of aripiprazole, which will usually suppress serum prolactin by virtue of its partial dopamine-agonist properties.14 Dopaminergic drugs, such as bromocriptine or cabergoline, can also be given but may occasionally worsen the underlying psychosis.15
If the serum prolactin level is greater than 200 ng/mL or remains elevated despite a change to a prolactin-sparing agent, an MRI scan of the sella turcica should be performed to exclude a pituitary or parasellar tumor. If the appearance on the MRI scan is normal, one might add aripiprazole or bromocriptine (as above), prescribe sex steroids (eg, oral contraceptives for women of menstrual age, testosterone for men), or give drugs to prevent osteoporosis (eg, bisphosphonates).9
Valproate and polycystic ovary syndrome
Polycystic ovary syndrome (PCOS) is defined as chronic anovulation and hyperandrogenism, with or without actual polycystic ovaries. Although not all PCOS patients are obese, insulin resistance is the norm, even in patients of normal weight.16 Recent reports have suggested that women of menstrual age who received valproate were at increased risk for the development of PCOS, characterized by oligomenorrhea, increased serum androgens with resulting hirsutism and acne, and, in some cases, a polycystic ovarian morphology seen on pelvic sonography. Joffe and colleagues17 recently reported that PCOS developed in about 10% of women with bipolar disorder who were started on valproate therapy (compared with 1.5% of women treated with lithium and other antiepileptic mood stabilizers), most within the first 6 to 12 months of treatment.
In vitro studies have demonstrated a direct action of valproate on ovarian tissue to increase the expression of androgen biosynthetic enzymes, resulting in increased ovarian production of androgens.18 Both valproate and drug-induced weight gain appear to raise serum insulin levels; the resulting hyperinsulinemia further stimulates ovarian androgen production.17,19
Patients should be counseled about diet and exercise at the start of valproate therapy in an effort to avoid weight gain since obesity will worsen insulin resistance and hyperandrogenemia.16 At each visit, patients taking valproate should be asked about menstrual irregularity, hirsutism, and acne. If these are present, the patient should be referred to a gynecologist or endocrinologist for further evaluation and treatment. Therapy for PCOS may include an oral contraceptive or progestin, insulin sensitizers such as metformin, antiandrogens, and local/topical treatments for hirsutism and acne. Since other mood stabilizers such as lithium and lamotrigine do not appear to cause PCOS, a switch to either one of these medications could also be helpful.17
Antipsychotics and metabolic adverse events
Recently, the focus on the tolerability of antipsychotics has shifted from extrapyramidal syndrome (EPS) toward adverse effects on body weight and related metabolic consequences that seem to be more common with atypical than with typical antipsychotics.20 The concerns about weight gain are fueled by ample evidence that increased body weight and obesity, particularly increased visceral fat agglomeration, are associated with greatly increased risks for diabetes, dyslipidemia, and hypertension (the metabolic syndrome; see Table 1) and are related to increased coronary heart disease morbidity and mortality.21,22 In addition, antipsychotic-induced weight gain has been associated with a lower quality of life23 and impaired medication adherence.24
Data suggest that there is an increased prevalence of obesity, diabetes, and dyslipidemia in patients with schizophrenia and bipolar disorder compared with the general population. Increased relative risk estimates range from 2 to 3 times the risk for obesity25,26 and diabetes27,28 to 4 to 5 times the risk for dyslipidemia.22 Likewise, prevalence rates for the metabolic syndrome and for calculated 10-year coronary heart disease events have been shown to be about twice as high in patients with schizophrenia29-31 and in patients with mixed diagnoses receiving atypical antipsychotics.32 Rates in patients with bipolar disorder have also been higher than in the general population, although they seem to be less elevated than in patients with schizophrenia.33