Hypomagnesaemia due to use of proton pump inhibitors--a review.

Abstract

Magnesium homeostasis is essential for many intracellular processes and depends on the balance of intestinal absorption and renal excretion. Hypomagnesaemia may arise from various disorders. We review the literature on hypomagnesaemia due to the use of proton pump inhibitors, as illustrated by a case of a 76-year-old woman with muscle cramps and lethargy caused by hypomagnesaemia and hypocalcaemia with a low parathyroid hormone level while using esomeprazole, a proton pump inhibitor (PPI). After oral magnesium repletion both abnormalities resolved. Fractional magnesium excretion was low, excluding excessive renal loss. A causal relation with PPI use was supported by the recurrence of hypomagnesaemia after rechallenge. In the past decade our understanding of transcellular magnesium transport was enhanced by the discovery of several gene mutations i.e. transient receptor potential melastin (TR PM) 6 and 7. In this light we discuss the possible aetiology of proton pump inhibitor related hypomagnesaemia.

Full text

Page 1

169
may 2009, Vol. 67, No. 5
© 2009 Van Zuiden Communications B.V. All rights reserved.
reVieW
Hypomagnesaemia due to use of proton pump
inhibitors – a review
M.T. Kuipers1, H.D. Thang2, A.B. Arntzenius1*
Departments of 1Internal Medicine and 2Nephrology, Spaarne Hospital, Hoofddorp, the Netherlands,
*corresponding author: tel.: +31 (0)23-890 71 44, fax: +31 (0)23-890 89 89,
e-mail: ABArntzenius@spaarneziekenhuis.nl
a b st r aC t
magnesium homeostasis is essential for many intracellular
processes and depends on the balance of intestinal
absorption and renal excretion. Hypomagnesaemia may
arise from various disorders.
We review the literature on hypomagnesaemia due to the
use of proton pump inhibitors, as illustrated by a case of
a 76-year-old woman with muscle cramps and lethargy
caused by hypomagnesaemia and hypocalcaemia with a
low parathyroid hormone level while using esomeprazole,
a proton pump inhibitor (PPi). after oral magnesium
repletion both abnormalities resolved. fractional
magnesium excretion was low, excluding excessive renal
loss. a causal relation with PPi use was supported by the
recurrence of hypomagnesaemia after rechallenge. in the
past decade our understanding of transcellular magnesium
transport was enhanced by the discovery of several gene
mutations i.e. transient receptor potential melastin (trPm)
6 and 7. in this light we discuss the possible aetiology of
proton pump inhibitor related hypomagnesaemia.
K e y Wor ds
Hypocalcaemia, hypomagnesaemia, proton pump
inhibitors, TRPM6, TRPM7
i n t rodUC t ion
Magnesium is implicated in many biochemical and
physiological processes in our body as an essential cation.
Although homeostasis of magnesium is tightly regulated
by a dynamic interplay of intestinal absorption, exchange
with bone and renal excretion, hypermagnesaemia
and hypomagnesaemia occur. Hypomagnesaemia
may arise from various disorders and can cause tetany,
convulsions, cardiac arrhythmias, hypoparathyroidism
and hypocalcaemia. Recently three case reports described
six adults with hypomagnesaemia and hypocalcaemia
which resolved after withdrawal of the proton pump
inhibitor.1-3 In this case we present another patient with
severe hypomagnesaemia due to extra renal magnesium
loss associated with the use of esomeprazole.
C a se r e P or t
A 76-year-old woman presented at our emergency
department with lethargy and muscle cramps in the
abdomen and extremities. Her medical history included
an appendectomy, diverticulosis, gastro-oesophageal
reflux symptoms, sliding diaphragmatic hernia, benign
paroxysmal positional vertigo and ischaemic heart disease.
There was no history of other gastrointestinal operations,
chronic alcoholism or use of laxatives. She did not complain
of nausea, vomiting or diarrhoea. For one year she had been
taking esomeprazole (40 mg/day) because of the indication
for aspirin and the history of reflux symptoms. Her other
medications were aspirin, diltiazem, oxazepam, psyllium
fibres and simvastatin. Physical examination showed no
abnormalities, Trousseau’s and Chovstek’s signs were
negative. On admission the laboratory findings showed
hypocalcaemia (1.26 mmol/l), hypomagnesaemia (0.18
mmol/l), hypokalaemia (3.2 mmol/l) and a low parathyroid
hormone (0.9 pmol/l). Her serum vitamin D3 level was
normal. There was neither alkalosis nor acidosis. We
started intravenous calcium and magnesium suppletion.
In three days the calcium, potassium, magnesium and
parathyroid hormone (PTH) level normalised and her
symptoms slowly resolved (figure 1). We interpreted these
results as severe hypomagnesaemia with secondary
hypoparathyroidism. After 11 days she was discharged

Page 2

170
may 2009, Vol. 67, No. 5
Kuipers, et al. Hypomagnesaemia due to use of proton pump inhibitors.
with oral magnesium supplements (magnesium oxide 500
mg three times daily). Two months later the magnesium
was discontinued. Within four weeks a dramatic drop
in the serum magnesium and calcium followed. The
fractional magnesium excretion was 0.69% (table 1).
This suggests a total body magnesium deficiency due to
extra renal magnesium wasting. About two years later
we challenged the notion that proton pump inhibitors
can cause hypomagnesaemia by impairing the intestinal
magnesium uptake. We discontinued the magnesium
supplements and the esomeprazole. The serum magnesium
and calcium level did not change in four weeks. Because of
increasing symptoms of gastro-oesophageal reflux we let
her resume the esomeprazole. Within four weeks the serum
magnesium level dropped to 0.4 mmol/l. The pattern we
observed in our patient fits nicely with the theory of proton
pump inhibitor induced intestinal magnesium loss that
was suggested in three recent publications on similar case
histories.1-3
disC Ussion
We describe an adult patient presenting with severe
hypomagnesaemia and hypocalcaemia while using
a proton pump inhibitor. Discontinuation of her
treatment with magnesium supplements resulted in
hypomagnesaemia and hypocalcaemia. We also
documented that during withdrawal of magnesium
supplements and esomeprazole the electrolyte balance
did not change. There are several disorders which may
cause hypomagnesaemia. In recent years several genes
were shown to be involved in renal magnesium and to
some extent intestinal magnesium transport. FXYD2
gene mutation causes autosomal dominant renal wasting
with hypocalciuria due to mutations in the γ subunit of
Na,K-ATPase at the basolateral membrane of the distal
convoluted tubule (DCT). SLC12A3 gene mutation causes
Gitelman’s syndrome due to a problem in the Na,K-2Cl-
co-transporter. Mutations in genes encoding claudin 16
and claudin 19, structural proteins found as part of the
tight junction between cells of the thick ascending limb of
Henle’s loop, cause familial hypomagnesaemia syndrome
with hypercalciuria and nephrocalcinosis. Groenestege
et al. showed that EGF plays an important role in renal
magnesium regulation.4 Pro-EGF is a membrane protein
expressed at high levels in the luminal and low levels in
the basolateral membranes of the DCT. P1070L mutation
in the EGF gene prevents EGF secretion, especially in the
basolateral space. This reduces TRPM6 activity, the key
figure 1. Course of laboratory parameters over time
0.0
Magnesium oxide
Esomeprazole
02610 14 110
0.2
0.4
0.6
1.0
0.8
1.4
1.8
2.2
2.6
1.2
1.6
2.0
2.4
2.8
3.0
Serum magnesium
Serum calcium
Serum parathyroid hormone
0
2
4
6
10
8
12
14
48 12 1618 2022 100102 104106108
Serum magnesium and calcium level (mmol/l)
Serum parathyroid hormone (pmol/l)
Week

Page 3

171
may 2009, Vol. 67, No. 5
Kuipers, et al. Hypomagnesaemia due to use of proton pump inhibitors.
channel that allows magnesium reabsorption. Therefore
EGF antagonists i.e. gefitinib, cetuimab, erlotinib and
panitumumab induce hypomagnesaemia.
In this case there was no SLC12A3 gene mutation.
Renal conservation of magnesium and calcium during
esomeprazole use was normal (table 1), which was also
described earlier by Epstein et al. and Shabajee et al.2,3 This
suggests that the problem did not arise in the kidney.
Gastrointestinal causes are diarrhoea, short bowel
syndrome, fistula, a magnesium free diet and very rarely a
malabsorption syndrome. The average daily diet contains
approximately 360 mg (=15 mmol) magnesium.5 About 24
to 75% of this ingested magnesium is absorbed, depending
on dietary contents and body stores. In a normal secreting
stomach the high H+ concentration is competing with
the metal ions for the ligand binding sites in food and
so liberating the metals.6 Gastric acid suppression
therapy with proton pump inhibitors causes low gastric
hydrochloric acid secretion. Theoretically we can assume
that the absorption of magnesium falls in hypochlorhydric
subjects but evidence for magnesium malabsorption in
proton pump inhibitor users is lacking. Limited studies
show conflicting conclusions.6 The intestinal magnesium
absorption occurs in two ways: in a passive paracellular
manner and in an active saturable transcellular manner.
About 90% of the absorption happens passively and linear
with the dietary magnesium content.7 Free magnesium
follows its concentration gradient and flows down the
paracellular pathway. The fact that treatment with
oral magnesium supplements is effective suggests a
normal passive paracellular pathway and a problem in
the active transcellular route. The active transcellular
pathway has become clearer since the identification of
TRPM6 and TRPM7. These channels conduct divalent
cations (magnesium and calcium) into the cell following
the transmembrane electrochemical gradient.7 TRPM6
is expressed along the entire gastro intestinal tract,
in the kidney (predominantly in the DCT), in testis
and lung tissue, TRPM7 is omnipresent in tissues. A
mutation in TRPM6 was found responsible for primary
hypomagnesaemia with secondary hypocalcaemia (HSH).
This rare autosomal recessive disorder was first described
by Paunier and colleagues in 1968.8 Patients present in
early infancy with diarrhoea, tetany and/or convulsions.
Premature stop mutations, exon deletions, frame shifts
and inserted splice sites are the identified mutations
and lead to loss of function of the TRPM6 channel.
This disease, together with the expression pattern of
TRPM6, showed us that TRPM6 is responsible for the
absorption of magnesium renally and in the intestines.9
There are several hormonal and other factors that have an
influence on TRPM6/7 expressions and their magnesium
transport. Li et al. showed that the TRPM6/7 channel
activity is potentiated by external protons.10 In another
in vitro study they demonstrated that glutamates in the
pore-forming region of TRPM6/7 are responsible for
the divalent selectivity and pH sensitivity. Mutations in
these negatively charged amino acids produced dramatic
changes in the channel function with a decreased calcium
and magnesium permeation. Proton pump inhibitor
users lack intestinal protons, i.e. they have a more alkalic
intraluminal pH. To our knowledge clinical manifestations
have not been described in heterozygotic mutations in
TRPM6/7. Could it be that variants in TRPM6/7 may
predispose to the side effect of proton pump inhibitors
as described in this article? The effect of proton pump
table 1. Electrolyte balance and other biochemical data at admittance, during and after magnesium supplements
and esomeprazole
normal valueday 13 months*
6 months2 years and
1 month**
0.74
2.09
4.1
n.d.
112
n.d.
7.0
2 years and
2 months***
0.4
1.85
3.7
n.d.
99
n.d.
4.6
Serum magnesium
Serum calcium
Serum potassium
Serum albumin
Serum creatinine
Serum vitamin D3
Serum parathyroid
hormone
Urinary magnesium
Urinary calcium
Urinary creatinine
Urinary volume
0.65-1.05 mmol/l
2.10-2.60 mmol/l
3.5-5.0 mmol/l
35-55 g/l
50-100 umol/l
30-130 nmol/l
1.0-7.5 pmol/l
0.18
1.26
3.3
40.9
91
39
0.9
0.39
1.99
4.1
44.3
108
n.d.
3
0.67
2.18
4.6
n.d.
n.d.
58
3.9
3.0-5.0 mmol/24 h
1.25-10 mmol/24 h
10-12 mmol/24 h
ml/24 h
n.d.
n.d.
n.d.
n.d.
0.11
0.08
6.4
1600
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d.
n.d. = not determined. *three weeks after stopping magnesium supplements with continuing esomeprazole. **Plasma magnesium still remains
normal four weeks after stopping magnesium supplements and esomeprazole. ***Plasma magnesium drops dramatically four weeks after
resuming esomeprazole without magnesium supplements.

Page 4

172
may 2009, Vol. 67, No. 5
Kuipers, et al. Hypomagnesaemia due to use of proton pump inhibitors.
inhibitors on TRPM6/7 has not yet been determined, so
this field still remains an active area of research.
While awaiting additional studies, we suggest that
magnesium loss in proton-pump inhibitor users
should be considered in the differential diagnosis of
hypomagnesaemia of any age. This newly recognised side
effect may not even be extremely rare once prescribers
become aware of its existence.
r e f e r e nC es
1. Cundy T, Dissanayake A. Severe hypomagnesaemia in long-term users of
proton-pump inhibitors. Clin Endocrinol (Oxf). 2008;69(2):338-41.
2. Epstein M, McGrath S, Law F. Proton-pump inhibitors and hypomagnesemic
hypoparathyroidism. N Engl J Med. 2006;26:355(17):1834-6.
3. Shabajee N, Lamb EJ, Sturgess I, Sumathipala RW. Omeprazole and
refractory hypomagnesaemia. BMJ. 2008;10;337:a425.
4. Groenestege WM, Thebault S, van der Wijst J, et al. Impaired basolateral
sorting of pro-EGF causes isolated recessive renal hypomagnesemia.
J Clin Invest. 2007;117(8):2260-7.
5. Todd Alexander R, Hoederop JG, Bindels RJ. Molecular Determinants
of Magnesium Homeostasis: insight from human disease. J Am Soc
Nephrol. 2008;19:1451-8.
6. Evenepoel P. Alteration in digestion and absorption of nutrients
during profound acid suppresion. Best Pract Res Clin Gastroenterol.
2001;15(3):539-51.
7. Quamme G. Recent developments in intestinal magnesium absorption.
Curr Opin Gastroenterol. 2008;24:230-5.
8. Paunier L, Radde IC, Kooh SW, Conen PE, Fraser D. Primary
hypomagnesemia with secondary hypocalcemia in an infant. Pediatrics.
1968;41:385-402.
9. Schlingmann KP, Waldegger S, Konrad M, Chubanov V, Gudermann T.
TRPM6 and 7 Gatekeepers of human magnesium metabolism. Biochim
Biophys Acta. 2007;1772(8):813-21.
10. Li M, Jiang J, Yue L. Functional characterization of homo-and heteromeric
channel kinases TRPM6 and TRPM7. J Gen Physiol. 2006;127(5):525-37.
11. Li M, Du J, Jiang J, et al. Molercular determinants of Mg and Ca
permeability and pH sensitivity in TRPM6 and TRPM7. J Biol Chem.
2007;282(35):25817-30.