Severe hypomagnesaemia due to lansoprazole

Abstract

A 71-year-old woman, who had been taking lansoprazole for 18 months for dyspepsia, presented with vomiting, thought to be due to gallstones, and was found to have severe hypomagnesaemia. She was treated with intravenous and then oral magnesium, and discharged, but was soon readmitted with symptoms due to hypomagnesaemia, and again treated with magnesium supplementation. No other recognised cause for hypomagnesaemia was found. Because of recent reports of hypomagnesaemia due to other proton pump inhibitors, lansoprazole was changed to ranitidine. Her symptoms resolved and the serum magnesium returned to normal. Oral magnesium supplementation was stopped with no return of symptoms or hypomagnesaemia. Such an association must be borne in mind with suggestive symptoms in patients on long term proton pump inhibitors; their cessation or change to H₂ receptor antagonists is likely to correct the situation rapidly.

Background

Proton pump inhibitors are very commonly used in clinical practice and patients may remain on these agents for months or years. There are recent case reports of omeprazole induced hypomagnesaemia.^1–3 Serum magnesium is not always measured by multichannel analysers, and the symptoms may be vague. It is possible that this important side effect is missed. This effect is thought to be due to their common action, and we present such a case involving a patient taking lansoprazole.

Case presentation

A 71-year-old woman was admitted as an emergency with protracted vomiting for 2 days. She had a history of hypertension, which was being treated with lisinopril, atorvastatin and bendroflumethiazide, and had been taking lansoprazole for 18 months for dyspepsia. Her recent symptoms were thought to be due to gallstones, found on abdominal ultrasound. Blood tests (9 November 2008) revealed hypocalcaemia and hypomagnesaemia (table 1), which was thought to be secondary to the vomiting and was treated with intravenous and then oral supplementation. Another possible factor was bendroflumethazide, which was stopped. She was discharged with a plan for routine cholecystectomy.

Table 1.

Serum magnesium and calcium concentrations

Date:	9/11/08	11/11/08	3/12/08	29/12/08	9/1/09	16/1/09	9/7/09
Mg++ (mmol/l) NR: 0.7–1.0	0.23	1.32 post iv Mg	<0.08	<0.08	0.28	0.83	0.85
Corrected calcium (mmol/l) NR: 2.1–2.6	2.0	1.9	2.1	2.1	2.4		2.5

03/12/08: Na⁺ 141 mmol/l; K⁺ 3.9 mmol; urea 3.8 mmol/l; creatinine 97 μmol/l.

Parathyroid hormone: 2.7 pmol/l (NR 1.3–6.8).

09/01/09: Urine magnesium <0.2 mmol/l.

Urine Mg/creatinine ratio <0.03 mmol/mmol.

Urine creatinine 5.78 mmol/l.

NR, normal range.

She was readmitted with symptoms of malaise, leg cramps and tingling of the face. Her doctor had remeasured her serum magnesium, while she was on oral magnesium supplementation, and had found it to be unrecordable (3 December 2008). Intravenous supplementation was given and oral supplementation increased, but the magnesium values remained very low (29 December 2008). No other recognised cause for hypomagnesaemia was found. Apart from hypertension she had been well in the past, and she had had no relevant bowel surgery or diarrhoea. There was no relevant family history. She does not drink alcohol and denied regular use of laxative.

Differential diagnosis

• Gastrointestinal: vomiting (resolved rapidly), malabsorption (excluded).

• Renal loss: renal tubular dysfunction, Bartter and Gitelman syndrome (excluded).

• Drugs: amphotericin, cisplatin, pentamidine, cyclosporine, gentamicin (no exposure), thiazide or loop diuretics (stopped with no benefit).

Treatment

Because of recent reports of hypomagnesaemia due to other proton pump inhibitors, the patient’s lansoprazole treatment was changed to ranitidine.

Outcome and follow-up

She was again discharged from hospital; her symptoms resolved and the serum magnesium returned to normal (16 January 2009). After 2 weeks her oral magnesium supplementation was stopped and she has remained well, with a normal serum magnesium value 6 months later (9 July 2009).

Discussion

This case demonstrates the problem of severe hypomagnesaemia in a patient treated for some time (18 months) with lansoprazole. Cessation of the drug resulted in a rapid resolution of the problem, which did not recur with the H₂ blocker ranitidine. Association of hypomagnesaemia with use of the proton pump inhibitors omeprazole and esomeprazole,^1,3,5,6 and very recently lansoprazole,⁴ has been reported. The mechanism seems unclear—the possibility of decreased magnesium absorption from the gut has been proposed.^1,2 In our case there was no evidence of increased magnesium urinary excretion, and hypocalcaemia with inappropriately low parathyroid hormone occurred as expected due to the hypomagnesaemia.

Proton pump inhibitors are widely used, and the symptoms of magnesium deficiency may easily be missed. There might be a case for monitoring magnesium values, especially with long term use, as has already been suggested.⁷ Cessation of proton pump inhibitors or change to H₂ receptor antagonists is likely to correct this potentially serious situation rapidly.

Learning points

• Lansoprazole may cause hypomagnesaemia, as is known to be the case with other proton pump inhibitors.

• This potentially serious condition can easily be missed, unless the clinician is aware of the association.

• Cessation of lansoprazole together with substitution with an H₂ blocker such as ranitidine rapidly corrects the condition.

• There may be a case for measuring serum magnesium in patients treated for long periods with proton pump inhibitors.