Hypercalcemia and “Primary” Hyperparathyroidism During Lithium Therapy

A 66-year-old Caucasian man with schizoaffective disorder was admitted to the psychiatric unit for a relapse of psychotic mania. The patient’s syndromal affective and psychotic symptoms had been in remission for the previous 11 years, during which time the patient had been treated with lithium, 1200 mg/day, and olanzapine, 15 mg/day. He took no additional medications or supplements and was otherwise healthy. Four months before admission, routine primary care screening revealed an elevated serum calcium level, at 11.0 mg/dL (reference range, 8.6–10.2 mg/dL). Records showed elevated calcium levels dating back 9 years. Three months before admission, the patient’s serum lithium level was 1.0 mmol/L (reference range, 0.6–1.2 mmol/L) and his intact parathyroid hormone (iPTH) level was 92.2 pg/mL (reference range, 12–88 pg/mL). His primary care provider conferred with an endocrinologist, who noted that lithium therapy is a well-documented cause of hypercalcemia and suggested discontinuing lithium.

Lithium was discontinued and divalproex initiated at 1000 mg/day, 17 days before admission. At admission, the emergency department evaluation noted pressured speech, disorganized thought process, religious preoccupations, paranoia, five sleepless nights, elevated energy, labile mood, and uncharacteristically impulsive behaviors. Laboratory studies showed a serum calcium level of 10.7 mg/dL. Results of physical examination, comprehensive metabolic screening, CBC, and toxicology screening were otherwise normal, as was the patient’s thyroid-stimulating hormone level. An increase in the dosage of divalproex to 2000 mg/day, along with trials of olanzapine (20 mg/day), haloperidol (15 mg/day), and risperidone (4.5 mg/day), were ineffective in controlling the patient’s mania.

The endocrinology department was consulted once again to determine the risks and advisability of reinstituting lithium. The consultant recommended a further hypercalcemia workup. On hospital day 10, the patient’s calcium level normalized to 9.8 mg/dL, while his iPTH level was an inappropriately high-normal 76.7 pg/mL. His urinary calcium level was elevated at 362.5 mg/24 hours (reference range, 100–300 mg). His 25-hydroxy vitamin D level was noted to be low at 21.0 ng/mL (reference range, 30–100 ng/mL). A technetium-99m sestamibi scan revealed increased uptake in one parathyroid gland, suggestive of adenoma. Bone mineral density testing by DXA (dual-energy X-ray absorptiometry) scan showed mild osteopenia in the left hip only (T-score=1.4 SD). ECG results were unremarkable. Based on hypercalciuria and enhanced localized parathyroid uptake, the consultant diagnosed primary hyperparathyroidism. Because it was considered asymptomatic, surgical intervention was not deemed necessary. The consultant concluded that it was a reasonable risk to reinitiate lithium with follow-up monitoring in the endocrinology clinic.

Lithium was restarted on hospital day 10 and titrated to the previously effective dosage. The patient’s manic psychosis remitted, and he was discharged on a regimen of lithium, 1200 mg/day, and risperidone, 3 mg/day. The patient remained in psychiatric remission at the 5-week and 15-week follow-up appointments. Twelve weeks after discharge, his serum calcium and iPTH levels were again elevated at 10.7 mg/dL and 86.3 pg/mL, respectively, and his serum lithium level was 0.98 mmol/L.

We conducted a PubMed all-fields search combining “lithium” and “hypercalcemia” and/or “hyperparathyroidism” and assessed the resulting 151 English-language articles, the majority in nonpsychiatric journals, for relevance to lithium-associated calcium disorders. This growing literature of case reports, case series, small prospective studies, and controlled cross-sectional studies establishes that lithium therapy increases the prevalence of hypercalcemia and hyperparathyroidism. A recent meta-analysis of 60 studies (1) found an absolute risk for hypercalcemia and hyperparathyroidism of 10% in a total of 730 lithium-treated patients compared with 730 unexposed subjects. A recent case-control cross-sectional study of 112 patients with bipolar disorder (2) found a prevalence of 8.6% for hyperparathyroidism and 24.1% for hypercalcemia in 58 lithium-treated patients compared with 54 unexposed patients. These represent much higher rates of hypercalcemia/hyperparathyroidism than the general population rate, which ranges from 0.1% to 0.7% (3). Because of the lack of population-based data, the true prevalence remains undetermined. Although several major psychiatric consensus guidelines reference the phenomenon, to date only one recommends routine calcium screening in lithium-treated patients (4).

Lithium is hypothesized to alter calcium homeostasis by several mechanisms. Lithium acts directly on renal tubules and bowel, stimulating calcium reabsorption, likely contributing to hypercalcemia independently of parathyroid hormone renal effects (5–7). Lithium has direct stimulatory effects on parathyroid hormone release (8). The literature widely discusses the phenomenon of lithium increasing the “set point” at which the parathyroid glands slow the release of parathyroid hormone in response to increasing serum calcium levels, an effect thought to be mediated by lithium’s actions on parathyroid cell calcium-sensing mechanisms (2, 6–9). Calcium and iPTH levels may increase in the days or weeks following lithium initiation or may do so later in the course of treatment (2, 6, 7, 9–11). Lithium is thought to effect parathyroid glandular pathology through chronic stimulation or by “unmasking” underlying subclinical pathology (6, 12). Duration of lithium therapy is correlated with increasing parathyroid mass; in one study (13), patients who had been treated with lithium for more than 3 years exhibited threefold increases in parathyroid mass compared with patients who had less than 6 months of lithium exposure. Once parathyroid hyperplastic or adenomatous changes occur during lithium therapy, the clinical presentation may closely resemble primary hyperparathyroidism and the disturbance may become independent of the presence or absence of lithium (6, 7, 10, 11). This phenomenon is referred to in the literature as “lithium-induced (or associated) ‘primary’ hyperparathyroidism” (7, 12). The case presented here demonstrates both acute and chronic lithium effects on calcium metabolism. Acute effects are seen in the close temporal reversible, reinducible association between lithium intake and levels of calcium and iPTH, and the chronic effects manifest in the glandular pathology evidenced on sestamibi scanning. The hypercalciuria in this case is not typical in lithium-induced hyperparathyroidism, although it is not unusual in the literature, particularly once demonstrable parathyroid pathology occurs (11).

Parathyroid hormone, secreted by the parathyroid glands, tightly regulates calcium homeostasis. Ionized calcium level (although we commonly measure serum calcium) is the main modulator of parathyroid hormone release (7, 10). Parathyroid gland pathology may cause inappropriately high parathyroid hormone secretion, triggering elevations in serum calcium and iPTH levels, the condition termed primary hyperparathyroidism. Symptomatic hypercalcemia/primary hyperparathyroidism classically presents with “stones, bones, groans, and moans,” manifesting as nephrolithiasis, renal dysfunction, osteoporosis, pancreatitis, peptic ulcer disease, behavioral and mood changes, and delirium (14). Additional, less specific symptoms may include weakness, fatigue, bone pain, and paranoia (6, 9, 12, 15). It is noted that vitamin D insufficiency, as seen in this case, may have an independent stimulatory effect on parathyroid hormone release in older patients (9).

With the widespread availability of automated metabolic assays since the 1970s, hypercalcemia and primary hyperparathyroidism are identified more sensitively; up to 80% of cases have no symptomatic manifestations (16, 17)—dubbed “asymptomatic primary hyperparathyroidism”—which has stimulated discussion regarding appropriate intervention guidelines (16–18). Current consensus guidelines recommend surgical referral when the serum calcium level exceeds 1 mg/dL above normal range, when creatinine clearance is below 60 mL/minute, when there is a bone density T-score below −2.5 SD at any site, when the patient is under age 50, or when surveillance is not possible (18). Patients who do not meet these criteria are considered appropriate for surveillance. A recent large-scale epidemiologic study of patients with primary hyperparathyroidism found that 51% developed surgical criteria and 28% received surgery over a 13-year study period (3). Demographic analysis has demonstrated a female-to-male ratio of 2.5:1 in primary hyperparathyroidism, with a threefold higher prevalence in patients over age 80 compared with patients ages 20–29 (3). Demographic trends in lithium-associated cases are similar to those in cases of primary hyperparathyroidism, with higher prevalences in women and the elderly (9, 11). While these guidelines and demographic factors may provide a framework for management of lithium-associated asymptomatic cases, understanding differences between lithium-associated and true primary hyperparathyroidism, along with consideration of the risks of discontinuing lithium, will additionally influence management considerations.

Lithium-associated hypercalcemia/hyperparathyroidism often presents with typical differences from primary hyperparathyroidism, including serum calcium levels ranging from slightly to very elevated, iPTH levels of high-normal to elevated, normal phosphate levels, hypermagnesemia, and hypocalciuria. Primary hyperparathyroidism, in contrast, typically presents with greater elevations of calcium and iPTH levels, low serum phosphate levels, normal magnesium levels, and hypercalciuria (6, 7, 9, 14, 19). The earlier literature commonly asserts that lithium-related calcium disorders are frequently inconsequential and result in lower morbidity than does primary hyperparathyroidism (7, 10, 19, 20). The evidence to support these claims tends to be fragmentary and questionable, however. The hypocalciuria typical of lithium-associated cases likely leads to the lower rates of nephrolithiasis observed in the literature (5, 7, 9, 19). Recent evidence supports the thesis that lithium may have protective effects on bone, possibly counteracting osteolytic effects it may indirectly cause through parathyroid hormone stimulation (21). Only one controlled study (20) has directly compared patients treated with lithium and patients with primary hyperparathyroidism. Ionized calcium and iPTH levels were lower in 26 lithium-treated patients compared with primary hyperparathyroidism patients. The lithium group was not a defined hypercalcemic group, although 54% had elevated ionized calcium or iPTH levels, and none had symptomatic hypercalcemic manifestations. The incidence and prevalence of symptomatic versus asymptomatic lithium-associated cases have not been clearly determined. One surgical outcome study that retrospectively surveyed 348 lithium patients who had been referred for parathyroidectomy found 15 with hypercalcemic symptoms and mean serum calcium levels of 11.7 mg/dL (12). While these studies suggest that lithium may be associated with relatively benign manifestations of hypercalcemia/hyperparathyroidism, particularly regarding the risks of nephrolithiasis and osteoporosis, the evidence is not derived from controlled studies defining study groups of lithium patients with demonstrated calcium abnormalities. In contrast, numerous case reports and series describe classical symptomatic hypercalcemic states in lithium-treated patients (7, 11, 12, 15). Patients with co-occurring nephrogenic diabetes insipidus, other renal diseases, osteoporotic disorders, and other conditions increasing their susceptibility to metabolic instability or delirium would likely have an additive risk of significant morbidity with lithium-associated hypercalcemic conditions. Lithium-associated hypercalcemia may also lead to additive risk for cardiac conduction disturbances (9).

The absence of formal management guidelines necessitates individualized risk assessment. Management options include surgical referral, lithium discontinuation, monitored lithium continuation, and calcimimetic therapy. The mainstay of treatment for symptomatic primary hyperparathyroidism is either localized or subtotal parathyroidectomy (11, 15). This remains the standard of care whether or not the pathology is lithium associated (14). Surgical interventions have been curative even with postsurgical lithium use, although there are postsurgical recurrences with and without lithium use (11, 15). The condition is reported as often, although not always, reversible with lithium discontinuation (6, 7, 10). Discontinuation poses obvious psychiatric risks, which must be carefully considered; controlled studies demonstrate high rates of manic relapse with lithium discontinuation (22). Monitored lithium continuation in asymptomatic cases may be a prudent option for many patients, modeled on commonly practiced asymptomatic primary hyperparathyroidism surveillance. Monitored surveillance should proceed cautiously and with an appreciation for nuance. Subtle symptoms of hypercalcemia/hyperparathyroidism may mimic underlying psychiatric disorders, with disturbances of mood, energy, and cognition in patients who are otherwise classified as asymptomatic (17). Several surgical outcome studies showed postsurgical improvements in scores on the Comprehensive Psychological Rating Scale and the SF-36 Health Survey in “asymptomatic” primary hyperparathyroidism patients (17). Cinacalcet, a calcimimetic agent and standard treatment for secondary hyperparathyroidism from renal failure, gained U.S. Food and Drug Administration approval in 2011 for treatment of primary hyperparathyroidism in patients unsuitable for parathyroidectomy. It has been reported to have reversed lithium-associated hypercalcemia/hyperparathyroidism in a total of five cases to date. In consultation with an endocrinologist, cinacalcet represents an important additional treatment option in symptomatic patients for whom lithium discontinuation poses substantial psychiatric risks, for whom surgical intervention has failed, or for whom surgery is contraindicated (9, 11, 14). This off-label use was well tolerated in the reported cases, although high rates of gastrointestinal side effects were cited in larger trials for primary hyperparathyroidism (23), and high cost may be a barrier to use (14).

The clinical course in the case presented here raises the question of whether our patient might have benefited from a more considered evaluation of the risks and benefits of lithium discontinuation. The hypercalcemic condition persisted unnoticed and without toxic consequences for at least 9 years before lithium discontinuation. Manic relapse then occurred within days after lithium discontinuation and proved unresponsive to alternative therapies. The patient’s history of clearly diagnosed manic episodes effectively controlled with lithium, the absence of symptomatic hypercalcemia/hyperparathyroidism manifestations, and the lack of risk-amplifying medical comorbidities suggest that this patient might have been an appropriate candidate for monitored surveillance of his asymptomatic hypercalcemia/hyperparathyroidism.

Hypercalcemia/hyperparathyroidism is a well documented but underappreciated complication of lithium therapy. Many patients may tolerate chronic asymptomatic hypercalcemia/hyperparathyroidism without adverse outcomes. However, given the potential unfavorable consequences of hypercalcemic states, it would be prudent for psychiatrists to increase their awareness of these risks and to practice surveillance. Although our patient had not experienced significant morbidity related to hypercalcemia/hyperparathyroidism, this does not guarantee that complications will not develop with continued aging and lithium exposure. The prevalence and potential consequences of this phenomenon support adding calcium monitoring to existing lithium monitoring standards, along with consultation with endocrine specialists when indicated. We recommend at minimum instituting the 2009 International Society for Bipolar Disorders guidelines for baseline calcium testing with follow-up testing at 6 months and annually thereafter (4). This testing may be easily integrated into existing practices for monitoring renal and thyroid parameters. More frequent monitoring may be necessary and appropriate for patients with additive risk factors, particularly the elderly, women, and patients with medical comorbidities for whom lithium-induced effects may amplify underlying risks.

The authors acknowledge Marcia Valenstein, M.D., for general and editorial guidance during preparation of the manuscript.