Pseudo-Gitelman Syndrome Presenting with Hypokalemic Metabolic Alkalosis and Hypocalciuria

Both Bartter syndrome (BS) and Gitelman syndrome (GS) are characterized by hypokalemic metabolic alkalosis resulting from the renal loss of potassium and sodium chloride. However, their inherited molecular etiologies differ along the thick ascending limb of Henle’s loop and distal convoluted tubule. Accordingly, their clinical phenotypes differ. The most significant difference lies in hypercalciuria in BS, and hypocalciuria and hypomagnesemia in GS1).

Pseudo-BS is a well-known differential diagnosis from BS in cases of normotensive hypokalemic metabolic alkalosis. It is commonly associated with loop diuretic or laxative abuse and surreptitious vomiting2). Pseudo-BS and pseudo-GS are often collectively referred to as pseudo-BS/GS3, 4). However, we believe that pseudo-GS should be considered a distinct entity because it is characterized by thiazide diuretic-induced hypocalciuria. In contrast to pseudo-BS, the scope and etiology of pseudo-GS remain unclear.

Here, we report the cases of two young adult female patients who presented with clinical features similar to those of GS: hypokalemic metabolic alkalosis with renal salt wasting, hypocalciuria, and hypomagnesemia. There were no indications of pseudo-BS, including loop or thiazide diuretic abuse. Initial sequencing studies for SLC12A3, CLCKNB, and KCNJ10 revealed no mutations, and subsequently performed whole-exome sequencing revealed no pathogenic variants. The metabolic alkalosis was saline-responsive in one case as is typical, and steroid therapy was necessary in the other to relieve chronic tubulointerstitial nephritis, which was diagnosed with kidney biopsy.

Case 1: A 20-year-old woman visited our emergency room (ER) in August 2012 for severe dizziness. She had two episodes of fainting in the past 3 months. Her blood pressure was 90/60 mmHg, and her hemoglobin level was 10.4 g/dL. Laboratory tests revealed a serum sodium concentration of 136 mmol/L, potassium level of 2.5 mmol/L, chloride level of 79 mmol/L, and total CO₂ of 50.3 mmol/L. Urinalysis revealed a specific gravity of 1.015, pH of 8.5, albumin +, glucose -, 1-4 red blood cells (RBCs)/high-power field (HPF), and 1-4 white blood cells (WBCs)/HPF. Urine electrolyte levels were as follows: sodium, 89 mmol/L; potassium, 95 mmol/L; chloride, 52 mmol/L; and creatinine, 88 mg/dL. Arterial blood gas analysis revealed a pH of 7.62, PaCO₂ of 52 mmHg, PaO₂ of 193 mmHg, and HCO₃- of 52 mmol/L. Serum calcium level was 9.3 mg/dL, phosphorus level was 2.6 mg/dL, magnesium level was 2.3 mg/dL, blood urea nitrogen (BUN) was 15.3 mg/dL, and creatinine level was 0.85 mg/dL. The urine calcium-to-creatinine ratio was 0.08 mg/mg, and fractional excretions of sodium, potassium, and chloride were calculated as 0.6%, 36.7%, and 0.6%, respectively. Mutation analysis for SLC12A3 revealed no abnormalities.

The patient neglected outpatient care and was readmitted 10 times over a decade for paraparesis and/or fainting. Hypomagnesemia was often observed. There was no evidence of diuretic or laxative abuse or surreptitious vomiting. Dental examination revealed no tooth erosion caused by gastric acid. Further sequencing studies for CLCKNB and KCNJ10 revealed no mutations. During each hospital stay, saline was infused with intravenous potassium chloride and oral magnesium oxide to correct the hypokalemic metabolic alkalosis and hypomagnesemia, respectively. Table 1 summarizes the treatment response during her most recent admission in January 2023. Her metabolic alkalosis improved with saline infusion, and azotemia and hyperuricemia were ameliorated. However, hypocalciuria and hypomagnesemia were not saline-responsive. Whole-exome sequencing revealed no pathogenic variants.

Table 1
Laboratory data during saline infusion and KCl supplementation in Case 1

• Click for larger image

Case 2: A 28-year-old woman was referred from a local hospital for the evaluation of hypokalemic metabolic alkalosis. She was previously healthy throughout 2016 with no abnormal findings during her check-up. However, in 2018, she visited the ER twice due to collapse while traveling in the United States and was diagnosed with hypokalemia and had a serum potassium level of ~2 mmol/L. Upon returning to South Korea, she visited a university hospital in Seoul and was suspected of having GS. Oral potassium and magnesium tablets were prescribed. However, she was admitted to a local hospital in 2020 because of fainting and limb paralysis.

The patient’s family history was unremarkable. She did not take any self-medication, including diuretics or laxatives. No gastrointestinal problems, including vomiting or diarrhea, were observed. At the first admission to our hospital in March 2021, her blood pressure and body mass index were 90/60 mmHg and 20.3 kg/m², respectively.

The patient’s hemoglobin level was 10.2 g/dL. Other laboratory findings were serum sodium level of 137 mmol/L, potassium level of 2.5 mmol/L, chloride level of 86 mmol/L, and total CO₂ of 42.9 mmol/L. Urine electrolyte levels were as follows: sodium, 159 mmol/L; potassium, 49 mmol/L; chloride, 21 mmol/L; and creatinine, 72 mg/dL. Arterial blood gas analysis revealed a pH of 7.57, PaCO₂ of 47 mmHg, PaO₂ of 89 mmHg, and HCO₃- of 43 mmol/L. Urinalysis revealed a specific gravity of 1.015, pH of 9.0, albumin +, glucose -, 3-4 RBCs/HPF, and 0-2 WBCs/HPF.

Serum calcium level was 9.7 mg/dL, phosphorus level was 1.7 mg/dL, uric acid level was 11.3 mg/dL, magnesium level was 1.5 mg/dL, BUN was 18.9 mg/dL, and creatinine level was 1.56 mg/dL. Fractional excretions of sodium, potassium, and chloride were 2.5%, 42.5%, and 0.5%, respectively. The urine calcium-to-creatinine ratio was 0.01 mg/mg, urine protein-to-creatinine ratio was 301 mg/g, and urine 2-microglobulin was 2.27 mg/L (reference <0.19 mg/dL).

A kidney biopsy was performed to evaluate azotemia. The glomerular pathology was unremarkable; however, moderate interstitial inflammation, moderate tubular atrophy, and mild interstitial fibrosis were observed, which were consistent with chronic tubulointerstitial nephritis (Fig. 1). Mutation analysis for SLC12A3, CLCKNB, and KCNJ10 revealed no abnormalities.

Fig. 1
The kidney biopsy findings in Case 2. The high power image reveals that there were no remarkable morphological changes in the glomeruli (A). The image with low power reveals that mild interstitial fibrosis was associated with moderate degrees of tubular atrophy and interstitial inflammation (B).

• Click for larger image

Initially, the patient’s metabolic alkalosis appeared to be temporarily responsive to saline infusion. However, the condition persisted until December 2021, when she was readmitted because of progressive azotemia. Intravenous methylprednisolone (500 mg) was administered daily for 3 days, and then switched to oral prednisolone (50 mg) once daily. However, the patient could not tolerate high-dose prednisolone and required reduced steroid therapy. Figure 2 shows the laboratory parameter trends between November 2021 and May 2023, suggesting that the aggravated metabolic alkalosis and azotemia were improved by increasing prednisolone dosage. To reduce the prednisolone dosage, we sequentially administered cyclosporine and mycophenolate mofetil. However, both these agents were ineffective and intolerable. Whole-exome sequencing of her genomic DNA revealed no pathogenic variants.

Fig. 2
The course of hypokalemic metabolic alkalosis and azotemia in Case 2. Changes in serum potassium, total CO₂, and creatinine levels in response to glucocorticoid therapy are shown.

• Click for larger image

In the differential diagnosis of hypokalemic metabolic alkalosis, whether or not hypertension is present should be first identified. If hypertension is present, mineralocorticoid excess, such as primary aldosteronism, must be considered after excluding the possibility of diuretic use in hypertensive patients.

Our cases had low-normal blood pressure, severe hypokalemia, and remarkable metabolic alkalosis, as evidenced by pH >7.55 and HCO₃- >40 mmol/L. The spot urine potassium-to-creatinine ratio clearly indicated renal potassium loss5). Urine chloride is also useful in the differential diagnosis of metabolic alkalosis because a chloride deficit is crucial for the maintenance of metabolic alkalosis6). Our patients had urine chloride levels of >20 mmol/L, excluding the possibility of self-induced vomiting and chloride diarrhea2). Importantly, they had additional diagnostic features of GS: fractional excretion of chloride >0.5%, hypocalciuria defined as a urine calcium-to-creatinine ratio <0.2, and hypomagnesemia <1.70 mg/dL1).

The established criteria for the diagnosis of GS include the identification of biallelic inactivating mutations in SLC12A3, which encodes the Na⁺-Cl^--cotransporter (NCC) expressed in the distal convoluted tubule1); however, we found no mutations in SLC12A3 in our cases. In addition, whole-exome sequencing excluded the possibility of newly identified pathogenic variants in various genes (e.g., CLCKNB, KCNJ10, FXYD2, and HNF1B) that may indirectly reduce NCC activity7). Diuretic or laxative screening was unavailable in our practice. Nevertheless, we believe that our cases were not related to drug abuse or surreptitious vomiting because of our patient-physician relationships for more than 5-10 years. Therefore, our findings are compatible with a diagnosis of pseudo-GS. Interestingly, Mori et al. reported that approximately 50% of 70 clinically diagnosed patients with GS were mutation-negative based on gene panel sequencing8).

Our two patients shared other clinical features. Both patients were young women with mild anemia, and the dizziness they experienced was so severe that fainting or collapse occurred. Thus, we believe that the chief complaints were due to severe metabolic alkalosis rather than hypokalemia. Remarkable bicarbonaturia was evidenced by a urine pH of >8.5 in most circumstances, and urine chloride level was slightly lower than urine sodium level. Hypomagnesemia was caused by renal magnesium wasting and hyperuricemia was ameliorated by fluid repletion (Table 1). Hyperuricemia has been proposed as a characteristic feature of pseudo-BS9), supporting the idea that volume contraction underlies the pathophysiology of metabolic alkalosis induced by pseudo-BS and pseudo-GS. Interestingly, an association between tophaceous gout and GS has rarely been reported in young men10).

In contrast to case 1, case 2 appears to have a poor long-term prognosis. Renal tubulointerstitial injury was evident, as predicted by persistent azotemia and tubular proteinuria. Autoimmune pathogenesis, including anti-Ro and anti-La antibodies, was investigated, but no systemic causes were found. In patients with Sjögren's syndrome, the presence of circulating autoantibodies against NCC may cause GS11).

In summary, we report the cases of two young female patients with pseudo-GS who presented with hypokalemic metabolic alkalosis, hypocalciuria, and hypomagnesemia; however, no pathogenic variants were detected on whole-exome sequencing. Volume contraction appears to have caused metabolic alkalosis, but the cause of the chloride deficit remains unidentified. An uncertain immunopathogenesis that produces tubulointerstitial nephritis may also present as a Gitelman-like syndrome.