Hyponatremia and SIADH
Hyponatremia, with or without syndrome of inappropriate antidiuretic hormone secretion (SIADH), is a relatively rare complication of treatment with virtually all antidepressants, some anticonvulsants (notably, oxcarbazepine and carbamazepine), and many antipsychotics. In the absence of a known noniatrogenic cause for hyponatremia, psychotropic medications may need to be withheld pending determination of the etiology. Some authors advocate changing to an alternative psychotropic agent with a different pharmacological profile and closely monitoring serum sodium levels. Fluid restriction is usually the first-line treatment for mild psychotropically induced hyponatremia. Adjunctive demeclocycline or lithium sometimes may also help to reduce the risk of a recurrence of hyponatremia during continued therapy with a psychotropic drug suspected of contributing to SIADH or hyponatremia.
Hypotonic hyponatremia (often defined as a serum sodium level <130 or 135 mmol/L, depending on the study) has been reported in association with numerous psychotropic agents, including bupropion, carbamazepine, divalproex, lithium, lamotrigine, MAOIs, oxcarbazepine, SSRIs, SNRIs (including duloxetine and venlafaxine), mirtazapine, tertiary amine TCAs, other norepinephrine reuptake inhibitors such as reboxetine, and most FGAs and SGAs. Hyponatremia that has been observed to occur during treatment with an SSRI (citalopram) has been shown to recur after substitution with an SNRI (duloxetine). The mechanism(s) by which antidepressants can cause increased release of hypothalamic antidiuretic hormone (ADH), causing water retention and serum hypo-osmolarity, are not well understood. SIADH may be three times more likely among patients taking SSRIs than among patients taking other antidepressants (Movig et al. 2002). A review of relative risks for SIADH among antidepressants suggests that mirtazapine and tricyclics may be less likely than other antidepressants to cause SIADH (De Picker et al. 2014). Case reports also exist of antipsychotic-associated hyponatremia leading to NMS (Spigset and Hedenmalm 1995).
Known patient-specific risk factors for psychotropic-induced hyponatremia include the following (Spigset and Hedenmalm 1995):
Increasing age (especially >65 years)
Smoking (because nicotine increases antidiuretic hormone [ADH] secretion)
Mechanisms by which psychotropic drugs may cause SIADH are not well understood, although one proposal is that SSRIs may cause release of ADH or renal responsiveness to ADH. However, as noted earlier, nonserotonergic drugs such as bupropion also can cause SIADH, although in our experience bupropion often may be a reasonable and relatively low-risk alternative to SSRIs or SNRIs. SIADH has been reported to occur anywhere from 2 days to many months after the initiation of a causal psychotropic agent; typically it occurs during the first few weeks of treatment with an antidepressant but may have a more variable time of onset during therapy with other psychotropic agents. Many nonpsychotropic agents also may cause hyponatremia or SIADH; these drugs include, among others, diuretics (e.g., amiloride and thiazide diuretics), antineoplastic drugs (e.g., vinca alkaloids), and the fibrate clofibrate.
The formal diagnosis of SIADH entails the presence of hypotonic hyponatremia plus lower serum osmolality (usually ~240–275 mOsm/kg) than urine osmolality (usually >100–200 mOsm/kg), as well as increased urine sodium (>30 mEq/L). Patients are typically euvolemic and have normal hepatic, renal, cardiac, thyroid, and adrenal function. (Importantly, hyponatremia with high serum osmolality points to other medical etiologies than SIADH, such as hyperglycemia or hyperosmolar hyperglycemic nonketotic coma.) Hyponatremia may manifest without symptoms or may involve nonspecific features such as weakness, lethargy, headache, and weight gain. Neuropsychiatric signs (e.g., confusion, seizures) may occur when serum sodium levels fall below ~120 mmol/L. The differential diagnosis of noniatrogenic hyponatremia is vast, as summarized in Table 10–1.
Laboratory assessment of hyponatremia includes measurement of urine electrolytes and osmolality. A diagnosis of SIADH is supported by concentrated urine (i.e., elevation of both urine Na+ [>20 mmol/L] and urine osmolality [>100 mmol/L]), whereas dilute urine (i.e., urine Na+ <20 mmol/L and urine osmolality <100 mmol/L) more likely suggests psychogenic polydipsia.
In mild euvolemic hyponatremia caused by psychotropic agents, elimination of the suspected causal agent in addition to fluid restriction to about 1 L/day is usually the preferred first-line intervention and often leads to resolution of the abnormality. This intervention may require at least temporarily withholding most classes of major psychotropic drugs or, if necessary, choosing an alternative pharmacotherapy that involves a novel mechanism from that of the suspected causal agent. Most authorities caution against rapid correction of low serum sodium levels (e.g., via aggressive infusion of hypertonic saline) because of the risk for inducing central pontine myelinolysis. Severe acute hyponatremia should be gradually corrected in a medical setting via intravenous infusion of sodium chloride solution, usually no faster than an initial rate of 1–2 mmol/L per hour.
The tetracycline antibiotic demeclocycline is sometimes used in the treatment of hyponatremia because of its ability to inhibit the renal effects of ADH. For instances in which continued therapy with a suspected causal psychotropic agent is deemed clinically necessary, adjunctive demeclocycline initially dosed at 900–1,200 mg/day may be advisable (Spigset and Hedenmalm 1995). Polyuria typically occurs within 1–2 weeks, and the dosage may then be gradually reduced, usually to 300–900 mg/day, in order to maintain normal serum sodium levels. Notably, lithium also inhibits the renal effects of ADH and is sometimes recommended as an alternative strategy in the management of hyponatremia.
In medical patients with SIADH who are taking an antidepressant, psychiatrists are often asked to render an opinion on the likelihood that the antidepressant is a causal factor, along with recommendations about continuing versus stopping an existing antidepressant and advice about the choice and timing of starting an alternative antidepressant. Generally speaking, one is obliged to stop existing antidepressants unless another known etiology exists. In the ideal, until serum sodium normalizes, one would not initiate another agent that could potentially also cause hyponatremia. However, in our experience, if ongoing antidepressant pharmacotherapy is indicated, it is reasonable to favor the initiation of either mirtazapine or bupropion, or sometimes a tricyclic or possibly a psychostimulant, when the risk–benefit ratio demands uninterrupted antidepressant therapy.
Acute intermittent porphyria
Congestive heart failure
Subarachnoid or subdural hemorrhage
Note. COPD=chronic obstructive pulmonary disease.
aHypokalemia and renal insufficiency typically present.
Metabolic Acidosis and Alkalosis
A very limited number of psychotropic medications are associated, although rarely, with metabolic acidosis or alkalosis. Most notably, renal tubular acidosis can be induced by topiramate and lithium; toxicity is associated with TCA overdose; and hypokalemic hyperglycemic nonketotic acidosis can occur during treatment with SGAs. Mild acidosis may involve no presenting signs, but more severe acid-based derangements can involve an array of physical symptoms. Diagnosis is made by laboratory assessment, including arterial blood gas sampling. Treatment involves discontinuation of a presumptive causal agent, followed by supportive medical management.
Disorders of acid-base chemistry may manifest with metabolic acidosis or alkalosis, and may be either acute or chronic. When resulting from psychotropic medications, mild acidosis may be asymptomatic; or when more severe, it may manifest with variable symptoms that can include nausea, vomiting, malaise, headache, chest or abdominal pain, muscle or bone pain, and hyperpnea (long, deep breaths).
The anion gap represents the concentration of unmeasured ions in serum, as determined by calculating the sum of routinely measured cations (i.e., [Na+]+[K+]) minus routinely measured anions (i.e., [Cl–]+ [HCO3–]). A normal anion gap is usually 8–12 mEq/L. Causes of an increased anion gap are summarized in the mnemonic MUDPILES: Methanol, Uremia, Diabetic ketoacidosis, Propylene glycol, Isoniazid, Lactic acidosis, Ethylene glycol, Salicylates.
A normal anion gap acidosis may be caused by carbonic anhydrase inhibitors (e.g., topiramate), spironolactone, hyperparathyroidism, ammonium chloride, Addison’s disease, diarrhea, hyperchloremia, organic solvents, diarrhea, or excess saline.
Hyperchloremic metabolic acidosis, without an anion gap, has been reported with the use of topiramate, due to renal bicarbonate loss resulting from carbonic anhydrase inhibition by topiramate. Acidosis may manifest clinically with dyspnea and confusion or other acute mental status changes, alongside laboratory values reflecting a rising serum chloride level with diminishing serum bicarbonate. This rare phenomenon has been reported to occur at any point during topiramate treatment and warrants consideration and evaluation of a basic metabolic laboratory panel in any patient receiving topiramate who develops acute mental status changes. Existing case reports suggest that the process generally ceases upon discontinuation of topiramate.
Other acid-base disturbances that can develop with specific psychotropic drugs include renal tubular acidosis caused by lithium and respiratory alkalosis resulting from TCA overdoses. Hypokalemic ketoacidosis may occur in association with sudden dramatic rises in blood glucose levels (e.g., as seen in hyperosmolar hyperglycemic nonketotic coma), a rare but potential risk associated with most if not all SGAs.
Lactic acidosis also can be a complication of treatment with the oral hypoglycemic agent metformin, usually when given at higher dosages. Patients taking metformin for possible reversal of psychotropic-induced weight gain (see Chapter 15, “Metabolic Dysregulation and Weight Gain,” and Table 15–6) should be cautioned to pay attention to the development of sore muscles, which could indicate the presence of lactic acidosis.