Finding the cause of severe hypokalemia: A 4-step approach

INITIAL EVALUATION IN THE EMERGENCY DEPARTMENT

His temperature was 36.5°C (97.7°F), heart rate 83 beats per minute, blood pressure 174/92 mm Hg, respiratory rate 14 breaths per minute, and oxygen saturation 98% while breathing room air. He was alert and oriented. His oral mucous membranes were moist. Cardiopulmonary and abdominal examinations were unremarkable.

Neurologic examination revealed no signs of aphasia, and cranial nerve assessment showed no pathological findings. However, he had proximal and symmetrical muscle weakness, predominantly affecting the lower extremities (Medical Research Council Strength Score 3 on a scale of 5), associated with hyporeflexia, without sensory deficits. There were no clinical signs of trauma from his fall.

In initial laboratory tests, his complete blood count showed no remarkable alterations, and his renal function was preserved, but he had severe hypokalemia, with a serum potassium level of 2.2 mmol/L (reference range 3.5–5.5) and metabolic alkalosis, with a pH of 7.52 (7.33–7.43) and a serum bicarbonate level of 31 mmol/L (21–25), while his sodium, chloride, calcium, and magnesium levels were normal (Table 1). An electrocardiogram showed sinus rhythm, without any potential signs of hypokalemia such as conduction block, ST-segment depression, or flattened T waves in the precordial leads.

Hypokalemia was confirmed with a second measurement before potassium replacement therapy was started.

DIFFERENTIAL DIAGNOSIS: WHERE IS THE POTASSIUM GOING?

Hypokalemia is common, affecting up to 15% of patients hospitalized with acute illness (16.8% in 1 study), and leads to increased mortality.¹

Transient hypokalemia is generally related to increased potassium uptake into cells, whereas persistent hypokalemia is most often due to excessive potassium loss.² A thorough evaluation of potassium intake, distribution, and excretion—based on a complete clinical history, physical examination, and urinary and serum electrolyte measurements—can identify the underlying cause of hypokalemia in most cases.³

Most cases of hypokalemia due to increased losses have obvious extrarenal causes such as vomiting, diarrhea, or excessive sweating.² However, if the hypokalemia is severe and the patient has no evident gastrointestinal losses, the initial approach should focus on the kidneys.

DETERMINATION OF RENAL POTASSIUM LOSSES

1. Which test is most appropriate for evaluating renal potassium losses in a patient such as ours, with severe hypokalemia and no evident gastrointestinal losses?

• 24-hour urine potassium excretion

• Spot urine potassium concentration

• Transtubular potassium gradient

• Spot urine potassium-to-creatinine ratio

Twenty-four-hour urine potassium excretion has been the gold standard test for evaluating renal potassium losses. In patients with hypokalemia, values lower than 15 mmol/day suggest extrarenal losses or cellular redistribution, whereas higher values indicate renal potassium wasting.² Li et al³ calculated that a cutoff value of 36 mmol/day had the best sensitivity and specificity.

Spot urine potassium concentration has been proposed as an alternative to 24-hour excretion because 24-hour urine collection is not feasible in urgent care settings.³ A spot urine potassium concentration below 15 mmol/L suggests extrarenal losses, while a concentration over 40 mmol/L indicates renal losses. However, its interpretation depends on urine osmolality: if the osmolality is high (ie, the urine is maximally concentrated), the urine potassium level may appear elevated despite low 24-hour excretion, whereas in dilute urine, the potassium level may be falsely low.⁴

The transtubular potassium gradient has been proposed as a way to account for urine osmolality. It is calculated as follows⁵:

In hypokalemia, a transtubular potassium gradient greater than 3 reflects an inadequate renal response and hence renal loss. However, when we use the transtubular potassium gradient we are assuming that osmotic reabsorption beyond the cortical collecting duct is negligible. Therefore, it overestimates potassium losses in cases of significant sodium and urea reabsorption beyond this point, which modifies urine osmolality. Consequently, for the transtubular potassium gradient to be useful, the urine sodium concentration must be at least 25 mmol/L and the urine osmolality must be at least as high as the plasma osmolality, that is, the urine must be concentrated.⁴

The urine potassium-to-creatinine ratio is less influenced by variations in urine osmolality, since creatinine is excreted at a relatively constant rate in healthy individuals. A value of 13 millimoles of potassium per gram of creatinine (mmol/g) or higher indicates renal loss.⁴ In a comparison of the different methods mentioned above, the urine potassium-to-creatinine ratio performed nearly as well as the reference test, 24-hour urinary potassium excretion.³ Therefore, it is recommended as a quick, simple, and accurate assessment of renal potassium losses.⁶

Although used less often in routine practice, a fractional potassium excretion value greater than 9% has been shown to be the most sensitive and specific method for estimating renal losses in hypokalemia.³ If the serum concentrations of potassium and creatinine are measured at the same time as urinary potassium and creatinine, fractional potassium excretion can be calculated as follows:

Our patient’s urine potassium-to-creatinine ratio before potassium supplementation was started was 78 mmol/g and his fractional potassium excretion was 20.05%, confirming that he was losing too much potassium through his kidneys (Table 2).

WHY ARE THE KIDNEYS WASTING POTASSIUM?

2. What is the most useful next step in the etiologic evaluation of hypokalemia due to renal losses?

• Assess blood pressure and volume status

• Calculate the transtubular potassium gradient

• Assess the acid-base balance and measure serum bicarbonate

• Request urine sodium and chloride levels

Assessing blood pressure and volume status is essential in identifying the underlying cause of hypokalemia, and based on these, other tests can be done.⁷

Kaliuresis is regulated by the interplay between mineralocorticoid activity and sodium delivery to the distal nephron. In hypokalemia due to renal losses, causes related to increased distal sodium delivery are generally associated with volume depletion and normal or low blood pressure, whereas those related to increased mineralocorticoid activity are typically associated with the opposite: volume expansion and hypertension.

If the blood pressure is low or normal, ie, in cases of increased distal sodium delivery, the next step is to assess acid-base status. If the patient has metabolic acidosis, suspect renal tubular acidosis types I or II.⁴ If, on the other hand, they have metabolic alkalosis, then we should measure urine electrolytes. Elevated urine sodium and chloride levels suggest diuretic use, Gitelman or Bartter syndrome, or hypomagnesemia.^8,9 Conversely, low urine chloride levels (< 20 mmol/L) may suggest vomiting, nasogastric suction, or other causes of gastrointestinal chloride loss associated with volume depletion.¹⁰

Hypomagnesemia deserves special mention, as it is commonly associated with severe or treatment-resistant hypokalemia. Low magnesium enhances renal potassium losses by inhibiting the sodium-potassium adenosine triphosphatase pump and increasing the activity of renal outer medullary potassium channels. Similar to what is observed in Gitelman syndrome, which features a genetic coding defect of the sodium-chloride cotransporter, hypomagnesemia decreases the abundance of sodium-chloride cotransporter in the distal tubule, thereby reducing sodium and chloride reabsorption and increasing their urinary excretion. This increase in distal sodium supply intensifies potassium secretion in the collecting duct, aggravating hypokalemia.⁸

If the blood pressure is high and increased mineralocorticoid activity is suspected, the transtubular potassium gradient has been proposed as a diagnostic tool; however, numerous limitations prevent its recommendation.⁶ Its values tend to be higher in hyperaldosteronism than in other causes of renal potassium loss,³ but it does not reliably distinguish between high and low mineralocorticoid states.¹¹ Moreover, sometimes potassium secretion remains appropriate despite excess mineralocorticoids.¹²

ALDOSTERONE IS THE MAIN MINERALOCORTICOID

Our patient had hypokalemia due to renal losses and he had hypertension. Therefore, we suspected he had increased mineralocorticoid activity.

Mineralocorticoids (eg, aldosterone) stimulate sodium reabsorption through the epithelial sodium channel in the principal cells of the distal and collecting ducts, generating a negative electrochemical gradient in the tubular lumen that favors secretion of potassium and hydrogen. This process promotes urine acidification and bicarbonate retention in the plasma. In addition, hypokalemia stimulates ammonium production, which contributes to perpetuating alkalosis.

Metabolic alkalosis secondary to excess mineralocorticoids typically presents with elevated urine chloride levels (≥ 20 mmol/L). It can be differentiated from other causes of hypokalemic metabolic alkalosis on the basis of urine chloride concentration and the presence or absence of hypertension (Table 3).^4,7,10,13 The net result of excess mineralocorticoids is a sustained increase in sodium and water reabsorption, intravascular volume expansion, elevated blood pressure, hypokalemia, and metabolic alkalosis.^7,13,14

Because aldosterone is the main mineralocorticoid, hyperaldosteronism is the prototypical entity of increased mineralocorticoid activity. Hyperaldosteronism is present in up to 10% of patients with hypertension and is the most common cause of resistant hypertension, accounting for up to 20% of cases. Resistant hypertension is defined as blood pressure that remains above target despite the use of 3 antihypertensive agents, including a diuretic.^15,16

Not all patients with hyperaldosteronism have hypokalemia: the prevalence has been reported at 9% to 37%, with a correlation between hypokalemia, aldosterone levels, and severity of hypertension.¹⁶ In patients with resistant hypertension and hypokalemia, the prevalence of hyperaldosteronism can be as high as 25% to 50%.¹⁷ Although coexisting metabolic alkalosis is frequently encountered, its prevalence in patients with primary hyperaldosteronism has not been clearly established.^13,18

CASE CONTINUED: HIS POTASSIUM STAYS LOW

In the emergency department, an intravenous infusion of potassium chloride 40 mmol/L was started at 10 mmol/hour, along with oral supplements. However, 12 hours later, even though our patient had received at least 120 mmol of potassium chloride, his serum potassium level was still low—in fact, it had declined to 2.0 mmol/L.

Because his serum potassium level had not responded to adequate initial replacement, he was admitted to the intensive care unit for central line replacement and monitoring. Based on clinical suspicion, in our patient with hypokalemia due to renal losses, resistant hypertension, and metabolic alkalosis, a targeted hormone study was requested.

BIOCHEMICAL DIAGNOSIS

3. Which test would you request next, based on your diagnostic suspicion?

• Isolated plasma aldosterone

• Plasma renin and aldosterone

• Saline infusion test

Our next step would be to measure plasma renin and aldosterone concurrently.

Hyperaldosteronism is the excessive and nonsuppressible production of aldosterone. It can be primary, ie, due to renin-independent adrenal hypersecretion of aldosterone,¹⁶ or secondary, ie, due to excessive renin synthesis.¹⁹ Therefore, renin is suppressed in primary cases and abnormally elevated in secondary cases.¹⁴ On the other hand, when the clinical picture of hyperaldosteronism presents with suppressed renin and low aldosterone levels, it is called pseudohyperaldosteronism.²⁰

Hypokalemia due to increased mineralocorticoid activity can be caused by increased secretion of renin, aldosterone, a nonaldosterone mineralocorticoid, or a nonmineralocorticoid molecule mimicking its effects.³ The diagnosis of primary hyperaldosteronism is based on a complex sequence of screening, confirmatory, and localization tests^16,17; we propose a simplified, 4-step strategy for studying hypokalemia due to increased mineralocorticoid activity (Figure 1).^4,7,13,17

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Diagnostic algorithm for hypokalemia due to increased mineralocorticoid activity.

Based on information from references 4, 7, 13, and 17.

The plasma aldosterone level varies greatly in individuals and as a test for primary hyperaldosteronism it has a high false-negative rate, arguing against its use in isolation.²¹ Screening relies on simultaneous measurement of plasma renin concentration or activity and aldosterone, with calculation of the aldosterone-renin ratio. Cutoff values vary depending on the assay used in each laboratory. The characteristic biochemical profile is defined by low or suppressed renin with inappropriately elevated aldosterone, resulting in an increased aldosterone-renin ratio.

Renin is considered suppressed if its activity is less than 1.0 ng/mL/hour or its concentration is less than 8.2 mU/L.¹⁷

DEFINITIVE DIAGNOSIS

4. What is the most likely definitive diagnosis in our patient?

• Pseudohyperaldosteronism by enzymatic inhibition

• Pseudohyperaldosteronism due to apparent excess of mineralocorticoids

• Pseudohyperaldosteronism due to congenital adrenal hyperplasia

• Pseudohyperaldosteronism due to Liddle syndrome

Pseudohyperaldosteronism may arise from endogenous or exogenous causes, either genetic or acquired.

Congenital adrenal hyperplasia causes hypertensive hypokalemia as the result of congenital deficiency of 17-alpha-hydroxylase, which leads to hypoandrogenism with sexual infantilism, and deficiency of 11-beta-hydroxylase, which is associated with hyperandrogenism and virilization. The impaired synthesis of glucocorticoids leads to increased adrenocorticotropic hormone levels, which results in accumulation of mineralocorticoid precursors.¹⁴

Liddle syndrome and apparent mineralocorticoid excess should be considered in the absence of excess aldosterone or other nonaldosterone precursors.¹¹

Liddle syndrome is an autosomal dominant disorder characterized by independent activation of the epithelial sodium channel in the collecting tubule, leading to excessive sodium and water reabsorption and increased potassium elimination. It is rare, accounting for only 1% to 2% of cases of hypertension in patients younger than 40 without other secondary causes.²³

Apparent mineralocorticoid excess is caused by mutations in the gene for 11-beta-hydroxysteroid dehydrogenase 2, the enzyme that inactivates cortisol. In these cases, the resulting excess of cortisol activates the mineralocorticoid receptor, triggering its effects.²⁴

Our patient was 78 years old, and the congenital conditions described here are rare. Therefore, apparent mineralocorticoid excess, congenital adrenal hyperplasia, and Liddle syndrome were considered unlikely. However, it is essential to consider that acquired conditions can mimic genetic defects that cause apparent mineralocorticoid excess and congenital adrenal hyperplasia. Endogenous hypercortisolism can saturate 11-beta-hydroxysteroid dehydrogenase type 2, and licorice consumption can inhibit this enzyme. Additionally, certain drugs, such as azole antifungals,²⁵ can inhibit steroidogenesis enzymes (Table 5).^14,27

Pseudohyperaldosteronism due to enzymatic inhibition was suspected because the patient was receiving abiraterone. Androgen deprivation associated with abiraterone and prednisone is the therapy of choice in metastatic castration-sensitive prostate cancer. Abiraterone reduces androgen synthesis by inhibiting 17-alpha-hydroxylase and C17,20-lyase.²⁶

However, this effect also impairs glucocorticoid synthesis, producing compensatory adrenocorticotropic hormone stimulation and, consequently, accumulation of mineralocorticoid precursors upstream of the 17-alpha-hydroxylase blockade. In this case, deoxycorticosterone is the primary molecule responsible for increased mineralocorticoid activity (Figure 2).^26–30

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Pseudohyperaldosteronism due to abiraterone.

Based on information from references 26–30.

Prednisone not only improves survival but also attenuates abiraterone-induced adrenocorticotropic hormone stimulation. Nevertheless, up to 17% of patients treated with abiraterone and prednisone develop hypokalemia,²⁷ and 37% develop new-onset hypertension or worsening hypertension.²⁸ Although our patient was receiving prednisone 5 mg/day, this dose may be insufficient to fully suppress adrenocorticotropic hormone and thereby prevent abiraterone-induced pseudohyperaldosteronism.^29,30

CONCLUSION OF THE CASE

Further blood tests in our patient revealed elevated deoxycorticosterone levels, low cortisol levels, and nonsuppressed adrenocorticotropic hormone levels (Table 4). This confirmed the diagnosis of pseudohyperaldosteronism due to enzymatic inhibition by abiraterone. It is plausible that the recent initiation of hydrochlorothiazide unmasked a previously latent state of mineralocorticoid excess.

The patient was treated with potassium supplements until his serum potassium levels were normalized. He was started on eplerenone, which is a mineralocorticoid receptor antagonist, and the prednisone dose was increased to 10 mg/day. The patient recovered satisfactorily and returned home. In subsequent checkups, he was asymptomatic, with optimal blood pressure control and no new episodes of hypokalemia.

This case illustrates the approach to a group of conditions associated with hypokalemia that usually pose a diagnostic challenge. A 4-step algorithm is proposed for evaluating hypokalemia due to increased mineralocorticoid activity (Figure 1). Mineralocorticoid excess states should be considered in patients with renal leak hypokalemia associated with new-onset, resistant, or recently worsened hypertension, recurrent episodes of hypokalemia, or hypokalemia disproportionate to diuretic use.

TAKE-HOME POINTS

• The urinary potassium-creatinine ratio or fractional potassium excretion are recommended as the methods of choice for assessing renal potassium loss in the acute setting.

• Although the transtubular potassium gradient is widely used in practice, it does not provide reliable data in the diagnostic approach to hypokalemia due to renal losses, nor reliable data in determining mineralocorticoid activity.

• When mineralocorticoid excess is suspected because the patient has hypertension and hypokalemia due to renal losses, measuring renin and aldosterone can be diagnostic.

• Pseudohyperaldosteronism includes both genetic and acquired disorders characterized by excess nonaldosterone mineralocorticoid activity, which should be considered in the evaluation of hypertensive hypokalemia.

• Clinicians should remain vigilant about the impact of newly introduced medications on the internal milieu, including electrolyte balance and hormonal regulation.

DISCLOSURES

The authors report no relevant financial relationships which, in the context of their contributions, could be perceived as a potential conflict of interest.