HNF1B variant without hyperglycaemia as a cause of isolated profound hypomagnesaemia
- 1 Department of Endocrinology, Skane University Hospital, Lund, Sweden
- 2 Department of Clinical Sciences in Lund, Lund University, Lund, Sweden
- Correspondence to Dr Shobitha Vollmer; Shobitha.Vollmer@skane.se
Abstract
A young man presented unconscious with severe hyponatraemia, hypokalaemia, hypomagnesaemia and metabolic alkalosis. After 4 months of treatment in hospital, the hypomagnesaemia persisted. The patient had no signs of diabetes mellitus, and radiology showed no abnormalities of the kidneys, pancreas or genitourinary tract. A parenteral magnesium load demonstrated renal wasting with increased fractional urinary excretion of magnesium. Genetic tests for Gitelman as well as Bartter syndromes were negative. However, a wider genetic panel revealed that the patient was heterozygous for a deletion on chromosome band 17q12, encompassing the whole HNF1B gene.
This case highlights the importance of considering pathogenic HNF1B variants in isolated profound hypomagnesaemia caused by renal wasting. Pathogenic HNF1B variants may partly mimic hypomagnesaemia found in Gitelman and Bartter syndromes and may be present without other features linked to HNF1B variants, including diabetes mellitus.
Background
It is thought that up to 5% of all diabetes mellitus is monogenic, of which HNF1B-maturity-onset diabetes of the young (MODY), or MODY type 5, accounts for 2%–5% of all MODY cases.1 HNF1B is a transcription factor, which is expressed in the embryonic development of the pancreas, kidneys, liver and genital tract. Accordingly, HNF1B variants may present with developmental abnormalities in any of these organs but is most often diagnosed in patients with hyperglycaemia.2 Mean age at diagnosis is 24 years and approximately 50%–60% of HNF1B-associated cases are caused by de novo variants.1 It has previously been reported that 40% of cases with HNF1B-MODY involve extrapancreatic abnormalities, with renal disease being the most common.3 Among those with renal disease, renal cysts appear to be the most frequently occurring followed by collecting system defects.2 In one study, hypomagnesaemia was found in 75% of cases.4
A pathogenic HNF1B variant impairs sodium–potassium ATPase function in the distal convoluted tubule, thereby, reducing magnesium reabsorption.5 Thus, HNF1B-MODY should be considered as a differential diagnosis in patients with hypomagnesaemia secondary to renal magnesium wasting.
Case presentation
A young man in his early 20s was found unconscious in the bathroom, suspected of having seizures, and brought into the emergency department where he was quickly intubated and transferred to the intensive care unit. The patient was previously fit and well, and was studying at university. He had no family history of diabetes mellitus or renal disease. However, his mother was treated for hypothyroidism and his maternal grandmother had undergone surgery for a goitre.
At the intensive care unit, he was found to be hypertensive with a blood pressure of 230/165 mm Hg and a venous blood gas showed metabolic alkalosis with severe hyponatraemia, hypokalaemia, hypomagnesaemia and hypochloraemia (see table 1). He had been to a party the previous night but plasma ethanol was below levels of detection and urinary toxicology screening was negative.
The patient’s laboratory findings in plasma and urine at presentation compared with at follow-up
| Parameter | At presentation | At follow-up (18–24 months) |
Reference range |
| Height, cm | 175 | 175 | |
| Weight, kg | 58 | 77 | |
| Plasma indices | |||
| Na+, mmol/L | 118 | 137 | 137–145 |
| K+, mmol/L | 2.7 | 3.8 | 3.5–4.4 |
| Mg2+, mmol/L | 0.39 | 0.58 | 0.70–0.95 |
| Ca2+, mmol/L | 1.03 | 1.33 | 1.15–1.33 |
| PO4 3-, mmol/L | 0.76 | 1.0 | 0.70–1.60 |
| HCO3 -, mmol/L | 32 | – | 22–27 |
| Cl−, mmol/L | 73 | – | 98–110 |
| Creatinine, µmol/L | 52 | 75 | 60–105 |
| Urate, µmol/L | – | 462 | 230–480 |
| Lactate, mmol/L | 1.6 | – | 0.5–2.2 |
| Alanine transaminase, µkat/L | 0.69 | 0.75 | 0.15–1.1 |
| Aspartate aminotransferase, µkat/L | 1.0 | 0.45 | 0.25–0.75 |
| Alkaline phosphatase, µkat/L | 1.4 | 1.6 | 0.70–1.9 |
| Gamma-glutamyl transferase, µkat/L | 0.50 | 0.44 | 0.15–1.3 |
| Conjugated bilirubin, µmol/L | 26 | 8 | 5–25 |
| Pancreatic amylase, µkat/L | 0.18 | <0.33 | 0.15–1.1 |
| C-reactive peptide, mg/L | 52 | <0.60 | <3.0 |
| Fasting glucose, mmol/L | – | 5.7 | 4.2–6.0 |
| Stimulated C-peptide, nmol/L | – | 2.1 | 0.37–1.5 |
| HbA1c, mmol/moL | – | 35 | 27–42 |
| Thyroid-stimulating hormone, mIU/L | <0.01 | 1.2 | 0.40–3.7 |
| Free T4, pmol/L | 36 | 16 | 12–23 |
| Urinary indices | |||
| Ca2+, mmol/24 hours | 2.1 | – | 2.5–7.5 |
| Mg2+, mmol/24 hours | 4.0 | 5.8 | 2.5–7.5 |
| Fractional excretion of Mg2+, % | – | 9 | <4 |
Investigations
The patient developed hypoxic brain injury with cortical ischaemic lesions seen on MRI shortly after admission. Electroencephalogram showed no epileptiform activity and lumbar puncture showed no abnormalities. Repeat MRIs showed that the ischaemic lesions eventually regressed.
The patient’s liver function tests, plasma urate and plasma amylase were normal. HbA1c was normal, as well as fasting plasma glucose and stimulated plasma C-peptide (see table 1).
Ultrasound of the kidneys and urinary tract was normal. Serum aldosterone and plasma renin were normal. The patient had no microalbuminuria and urine osmolality was 646 mOsm/kg after fasting. Plasma osmolality remained around 290 mOsm/kg.
Thyroid function tests, taken prior to receiving iodine-containing contrast for CT, showed a suppressed plasma thyroid-stimulating hormone (TSH) and elevated plasma thyroid hormones. Thyroid peroxidase antibodies and TSH receptor antibodies were negative.
CT of the neck and thorax was normal, with the exception of a 2.5 cm cyst in the left thyroid lobe. The cyst was inactive on a thyroid scintigraphy but the rest of the thyroid showed normal uptake. Ultrasound confirmed a cystic nodule, showed normal blood flow in the rest of the thyroid gland, and fine-needle aspirations of the cystic nodule were classified as Bethesda 1.6
CT of the abdomen did not show any abnormalities in the pancreas, kidneys and genitourinary tract.
Despite hypomagnesaemia, total urinary magnesium excretion was 5.8 mmol/24 hours, and urinary calcium excretion was low (see table 1). To confirm renal magnesium wasting, an intravenous magnesium challenge was performed by infusing 10 mmol of magnesium sulfate over 4 hours during which plasma magnesium rose from 0.55 mmol/L to 0.88 mmol/L.7 The following morning, plasma magnesium was 0.58 mmol/L. Fractional magnesium excretion was calculated to be 9% (reference <4%).8
A single-gene test for Gitelman syndrome and a 10-gene panel sequence analysis for Bartter syndromes were negative. Wider next-generation sequencing thereafter revealed a 1.26 Mb heterozygous deletion on chromosome band 17q12 encompassing HNF1B, PIGW and ZNHIT3.
Treatment
In the intensive care unit, infusions of 0.9% sodium chloride combined with potassium chloride were given until plasma sodium reached 139 mmol/L 24 hours later. The patient was initially also given magnesium sulfate infusions.
The unexpected laboratory finding of hyperthyroidism was considered to motivate block-and-replace treatment with methimazole and levothyroxine, which normalised thyroid function. Treatment was given for 8 months and after discontinuation, thyroid function has remained normal.
After 7 days in the intensive care unit, the patient was extubated and transferred to the acute medical unit. Later, he was transferred to neurology and rehabilitation wards. Treatment with oral potassium chloride and magnesium hydroxide tablets normalised plasma potassium and sodium levels, but plasma magnesium would not increase above 0.55 mmol/L. The addition of a mineralocorticoid receptor antagonist did not affect plasma magnesium levels.
Following 4 months in hospital, the patient was discharged from the rehabilitation unit with potassium and magnesium tablets. Due to the suspicion of seizures at presentation, the patient was also discharged with oral sodium valproate.
Outcome and follow-up
The patient had to take a break from university for nearly 2 years but has since managed to return to his studies. Following rehabilitation at the brain injury unit, the patient briefly moved in with his parents but is now living independently on his own again. Genetic testing of the patient’s parents and siblings has been offered but not completed.
The patient has regular follow-up at the endocrine clinic. He is currently taking 30 mmol of magnesium daily, has a magnesium level of 0.58 mmol/L and does not have any gastrointestinal side-effects. The patient has had no further seizures and his antiepileptic treatment will be discontinued in the near future. The patient will be followed up annually with electrolytes, blood glucose, HbA1c and thyroid function tests.
Discussion
The serious nature in which the patient presented could probably be explained by several factors, predominantly the multiple severe electrolyte abnormalities contributing to reduced consciousness and possible seizures. Thus, hypomagnesaemia is commonly seen with hyponatraemia, hypokalaemia and hypocalcaemia.9 Also, magnesium is required to maintain normal levels of intracellular potassium and extracellular sodium via sodium–potassium ATPase, sodium–potassium–chloride cotransporter and potassium channels.10 Furthermore, hyponatraemia can lead to reduced consciousness and seizures. The subsequent fluid loss, secondary to sodium loss, activates the renin–angiotensin–aldosterone system, which perpetuates potassium loss.8 Finally, hypomagnesaemia inhibits parathyroid hormone, which may lead to hypocalcaemia.8
This patient’s hypomagnesaemia is caused by the heterozygous deletion encompassing HNF1B, PIGW and ZNHIT3. The latter two genes are associated with autosomal recessive disorders. HNF1B variants can have variable phenotype and penetrance, which explains why the patient does not have genitourinary malformations or diabetes mellitus.11 Although genetic testing of first-degree relatives has not yet been completed, this variant is frequently discovered de novo.2 11 HNF1B is a transcription factor that binds to the promoter of FXYD2 (FXYD domain containing ion transport regulator 2), which encodes a subunit of sodium–potassium ATPase. Magnesium reabsorption is achieved via transient receptor potential melastatin type 6 magnesium channels in the distal convoluted tubule and is dependent on the voltage gradient set by sodium–potassium ATPase. Thus, impaired sodium–potassium ATPase by a pathogenic HNF1B variant causes renal magnesium and sodium wasting.11–14
A European case study described three patients where hypomagnesaemia was the first clinical manifestation of a pathogenic variant in HNF1B with high fractional urinary excretion of magnesium as well as hypocalciuria.13 Another report described a boy who presented with hypokalaemic hypochloraemic alkalosis, hypomagnesaemia and hypocalciuria. Initially, he was diagnosed with Gitelman syndrome but was later found to have an HNF1B variant.15 However, in most cases, renal disease, diabetes mellitus or genitourinary malformations are the first presenting symptom or sign.1–3 16 17 Whole-gene HNF1B deletion, which occurs in approximately 50% of patients, has also been described to be associated with cognitive impairment or autistic spectrum disorder, but this has not been shown in our patient.17–19
In our patient, hyperthyroid laboratory values were seen at presentation but, after initial block-and-replace treatment, they had normalised at follow-up without further treatment. Hyperthyroidism has, to our knowledge, not previously been described in patients with HNF1B variants and its aetiology remains unclear. It might be speculated whether hyperthyroidism triggered the severe electrolyte derangement or if the electrolyte abnormalities caused thyroid hormone release from the thyroid gland but it may also be an incidental finding unrelated to HNF1B.
Genetic counselling will be offered to the patient and his family. Considering that HNF1B variants are inherited in an autosomal dominant manner, any offspring of the patient are at 50% risk of inheriting the variant and of being affected.
In summary, our case highlights the importance of considering HNF1B variants in patients with severe hypomagnesaemia with hypocalciuria, which may partly mimic Gitelman syndrome. Another important aspect is that this patient only displayed hypomagnesaemia without any of the other features of HNF1B-MODY, including diabetes mellitus and renal cysts.
Patient’s perspective
It hit me hard out of seemingly nowhere. I was at a party and there I drank more than usual. Before that, I had eaten pizza. Afterwards it felt like my head had hit a rock. My first thought was that it was a hangover since I’ve never experienced something equal to that pain before, and knowing I had drunk an amount of alcohol I wasn’t used to. Honestly, I’m not a heavy drinker at all. In the desperate state of pain, I had tried different ways to ease it. I called our national health support to no avail, probably because the person I spoke with didn’t think anything seemed out of the ordinary. Because of my limited experience with hangovers, I asked other people how they dealt with it. Their tip was pain-relieving pills (the kind which you don’t need a prescription for in my country). Also, I tried drinking water since none of their help helped me. While my memory of the chronological chain of events may be a bit off, I know I started to vomit and drink water in cycles, since the pain lingered on. In hindsight, that wasn’t very wise.
Moving on, today I wonder if I have always had a worse energy level than other people. I look back on events in my life in the year prior to my accident with signs of me yawning and feeling very tired. It’s hard to know something is wrong as long as you get by though.
Learning points
-
Hypomagnesaemia may be the first and only presenting feature of a pathogenic HNF1B variant.
-
Cases of hypomagnesaemia, where renal wasting is suspected, should be screened for HNF1B variants with a wide genetic panel.
-
HNF1B-related disorders can partly mimic Gitelman syndrome with hypomagnesaemia and hypocalciuria.
-
It is important to identify if hypomagnesaemia is caused by an HNF1B variant as it can prompt screening for other features such as diabetes mellitus or genitourinary malformations.
-
If a person is found to have a pathogenic HNF1B variant, screening should be offered to first-degree relatives and genetic counselling to the patient and their family.
Ethics statements
Patient consent for publication
Footnotes
-
Contributors SV and PK drafted the case report. SV, PK and ML revised the case report and approved the final version.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Case reports provide a valuable learning resource for the scientific community and can indicate areas of interest for future research. They should not be used in isolation to guide treatment choices or public health policy.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2023. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer