NDT Advance Access originally published online on August 24, 2004
Nephrology Dialysis Transplantation 2004 19(11):2914-2917; doi:10.1093/ndt/gfh358
Nephrol Dial Transplant Vol. 19 No. 11 © ERA-EDTA 2004; all rights reserved
Teaching Point
(Section Editor: K. Kühn)
Supported by an educational grant from 
Severe hypokalaemia: is one reason enough?
Alexander Woywodt,
Alena Herrmann,
Hermann Haller and
Marion Haubitz
Division of Nephrology, Department of Medicine, Hannover Medical School, Hannover, Germany
Correspondence and offprint requests to: Alexander Woywodt, MD, Division of Nephrology, Department of Medicine, Hannover Medical School, Carl-Neuberg-Strasse 1, 30625 Hannover, Germany. Email: Woywodt.Alexander{at}MH-Hannover.de
Keywords: aetiology; hypertension; liquorice; severe hypokalaemia
 |
Introduction
|
|---|
Hypokalaemia is a common electrolyte disorder with a straightforward
differential diagnosis. The usual suspects are diarrhoea, vomiting
or intake of diuretics. In some patients, a spot urinary chloride
may help to tell the tale. Occasionally, a comprehensive workup
is required [
1] to elucidate disorders such as Conn's, Bartter's
or Gitelman's syndromes. Sometimes, however, the cause of hypokalaemia
may be all too obvious and one of the usual suspects later turns
out to be an innocent bystander rather than the primary culprit.
We present such a case to illustrate a common but dangerous
error in medicine, namely our intention to frame all available
information into an all too obvious diagnosis.
|
Case
|
|---|
A 50-year-old woman was referred to our casualty department
with muscle pain and weakness. She had experienced pain and
weakness in her calf muscles, shoulders and hands for some 3
weeks without any precipitating or aggravating factors. The
symptoms were constant during the day and did not depend on
exercise. The patient declined malaise, fever, weight loss and
dyspnoea. The previous medical history was notable for chronic
headache and hypertension, which was controlled with telmisartan
and torasemide. Both had been started 9 months prior to admission.
There was a history of gynaecological complaints, which were
still unresolved. The patient worked as an accountant and the
family history was unrevealing. Earlier on the day of referral,
she had been seen by a general practitioner, who had admitted
her with a tentative diagnosis of connective tissue disease.
On examination, the patient was afebrile and did not appear
acutely ill. Blood pressure was 150/80 mmHg. The deltoid and
gastrocnemial muscles were tender to palpation and a malar rash
was seen. The remainder of the clinical examination was unremarkable.
The admitting physician concurred that polymyositis was indeed
a tempting diagnosis, owing to the age and gender of the patient
and the preponderance of polymyositis to affect proximal muscles.
Laboratory studies revealed profound hypokalaemia of 1.8 mol/l,
pH 7.6, creatine kinase 13 051 U/l and myoglobin 5633 U/l. Urine
chloride was 72 mmol/l (elevated). The initial diagnosis was
refuted and the hypokalaemia was believed to be due to the intake
of torasemide. Both intravenous and oral potassium were administered,
yet the serum potassium remained as low as 1.9 mmol/l and the
patient was transferred to the intensive care unit. On this
occasion, another attempt was made to uncover a second cause
of hypokalaemia. The patient declined vomiting, diarrhoea or
intake of any medication other than the above mentioned. Intake
of laxatives was specifically asked for, again to no avail.
When asked for liquorice intake, however, the patient confessed.
In particular, she preferred strong liquorice candies, which
she obtained in considerable quantities from a local pharmacy.
This had been going on for quite some time, perhaps 1.5 years.
We were unable to retrieve any previous measurements of electrolytes.
Plasma renin was <0.2 ng/ml/h, plasma aldosterone was <6
pg/ml and urinary aldosterone was 3.5 µg/24 h. The patient
made an uneventful recovery.
|
Discussion
|
|---|
One of the most dangerous errors in medicine is to try and fit
all the available information into a tentative diagnosis. First,
we tried to fit all symptoms into a diagnosis of polymyositis,
until the laboratory results ridiculed such an assumption and
demonstrated that our patient had red cheeks rather than a malar
rash. Unfortunately, we quickly repeated this type of error
in the differential diagnosis of hypokalaemia. Here, we accused
one of the usual suspects and assumed that diuretic intake was
the only cause of hypokalaemia. This form of error derives form
a particular style of clinical reasoning, namely pattern recognition
[
2]. Other types of reasoning include algorithmic, exhaustive
(gathering every possible piece of data) and hypothetico-deductive
(generating and rejecting) hypotheses as more and more data
are collected. The latter is widely held up as the ideal, while
the exhaustive approach is often used for rare diseases or unusual
presentations. The pattern-recognition style is commonly used
by primary care physicians and in casualty, where instant recognition
of a disease is paramount. This strategy has both merits and
dangers. In particular, physicians are keen to look for evidence
to support the initial diagnosis and ignore any data that might
refute it. In other words, physicians try to fit ‘a square
peg in a round hole’ [
3]. In our case, one pattern was
present but missed: hypertension and hypokalaemia.
Recognition of this pattern would have prompted a search for Glycyrrhiza glabra (Figure 1). Unfortunately, we missed the point and believed that diuretics were the sole culprit. Elevated urinary chloride was still in keeping with this assumption (as long as diuretic action is still present), while the results of further laboratory studies were not: use of diuretics provides a stimulus for the renin–angiotensin– aldosterone system, while our patient had low plasma renin and aldosterone, suggesting the presence of an aldosterone-like substance. In addition, the patient reported here had slightly elevated blood pressure, while hypokalaemia due to diuretic intake is commonly associated with low or low-normal blood pressure. Finally, it has been noted that hypokalaemia due to loop diuretics is usually mild and transient [4], unless another effect is at work. Therefore, in the presence of a normal diet, severe hypokalaemia due to a loop diuretic should always prompt a search for additional factors.
View larger version (154K):
[in this window]
[in a new window]
|
Fig. 1. Literally the root of the matter: radix Glycyrrhizae glabrae (root of Glycyrrhiza glabra, the European liquorice).
|
|
Dangerous effects due to intake of liquorice were, in fact,
noted as early as in the 1960s, when the pathophysiology of
liquorice-induced disorders was not even nearly understood.
After liquorice had been noted as a cause of hypertension it
was first believed to exert a direct mineralocorticoid effect
[
5]. Accordingly, liquorice treatment for Addison's was recommended.
Patients with peptic ulcers were also treated with liquorice
derivatives. Today, liquorice ingestion differs markedly from
one region to another. Extraordinary amounts of liquorice are
consumed in Iceland, succeeded by the Netherlands and Scandinavia.
In Germany, the habit is considerably more prevalent in coastal
provinces. In the UK, ‘Pontefract cakes’ are notorious
[
6] and pay testimony to the centuries-old use of liquorice
(liquorice was introduced into Pontefract, Yorkshire, by monks
in 1562; sugar was later added by a local chemist and the novel
product named Pontefract cakes). Severe hypokalaemia, rhabdomyolysis
and tetraparesis due to these cakes have been reported [
7].
The causative ingredient of liquorice is glycyrrhicinic acid, a pro-drug. Interestingly, glycyrrhicinic acid is also found in a broad variety of herbal preparations, pastis and surprisingly Belgian beer. Its metabolite, glycyrrhetenic acid, acts as an inhibitor of 11ß-hydroxysteroid dehydrogenase (11ß-HSD), a key enzyme of steroid metabolism. The function of this enzyme is to convert cortisol (which has glucocorticoid and mineralocorticoid activity) into cortisone (glucocorticoid but no mineralocorticoid activity). To fully appreciate the importance of this enzyme, we must step back and understand that mineralocorticoid action of aldosterone is needed in a minority of cells, such as the distal tubule, colon and salivary glands. But how can these tissues sense aldosterone, given that cortisol with its inherent mineralocorticoid action is abundantly present? The answer lies in the action of 11ß-HSD. The enzyme is present where aldosterone action is needed and degrades cortisol so that aldosterone alone binds the mineralocorticoid receptor (Figure 2). Like many other endocrine disorders, these events were delineated with a little help from a rare genetic syndrome, apparent mineralocorticoid excess. The hallmark of this disease is absence of 11ß-HSD and binding of cortisol to the mineralocorticoid receptor [8]. These observations ultimately led to the finding that glycyrrhicinic acid acts as an inhibitor of 11ß-HSD. Thus, cortisol can escape inactivation and bind to the mineralocorticoid receptor with equal affinity, leading to an apparent excess of aldosterone. Recent research has elucidated two isoforms of 11ß-HSD: 11ß-HSD1 acts as an oxo-reductase that generates cortisol, while 11ß-HSD2 is crucial for inactivation of cortisol as described above. It has been speculated that 11ß-HSD1 may be involved in truncal obesity and metabolic syndrome, with the exciting prospect of therapeutic manipulation. Hypertension is the most common disorder caused by liquorice, but life-threatening electrolyte disturbances have been reported also [9]. It is conceivable that our patient first developed liquorice-induced hypertension, then received a loop diuretic and finally sustained severe life-threatening hypokalaemia. Finally, it has been suggested that furosemide also inhibits 11ß-HSD, but this assumption has also been refuted [10].
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 2. Steroid metabolism and the syndrome of apparent mineralocorticoid excess. Under normal circumstances, mineralocorticoid receptors in aldosterone-sensitive tissues (kidney, colon, salivary glands) are protected from the action of cortisol by 11ß-HSD, which converts cortisol to its keto-derivative. Glycyrrhetenic acid inhibits the enzyme and the mineralocorticoid receptor is ‘swamped’ by cortisol.
|
|
|
Teaching points
|
|---|
- In cases of severe hypokalaemia, consider more than one underlying reason.
- Pattern recognition is an important but potentially dangerous approach during the diagnostic process. Beware of the patient with a label, be prepared to question your tentative diagnosis and try to construct other patterns from the available information.
- Liquorice ingestion is an important cause of hypokalaemia and hypertension. All patients with hypokalaemia and/or hypertension must be evaluated for liquorice ingestion. Uncommon sources of glycyrrhicinic acid, such as herbal preparations, must not be forgotten.
- Hypokalaemia is a very common finding in patients who receive a loop diuretic. However, the effect often disappears over time and severe life-threatening hypokalaemia should always prompt a search for other factors.
Conflict of interest statement. None declared.
|
References
|
|---|
- Reimann D, Gross P. Chronic, diagnosis-resistant hypokalaemia. Nephrol Dial Transplant 1999; 14: 2957–2961[Free Full Text]
- Leape LL. Error in medicine. JAMA 1994; 272: 1851–1857[Abstract/Free Full Text]
- Saint S, Saha S, Tierney LM,Jr. Clinical problem-solving. a square peg in a round hole. N Engl J Med 1998; 338: 379–383[Free Full Text]
- Brater DC. Benefits and risks of torasemide in congestive heart failure and essential hypertension. Drug Saf 1996; 14: 104–120[Web of Science][Medline]
- Frey FJ, Ferrari P. Pastis and hypertension – what is the molecular basis? Nephrol Dial Transplant 2000; 15: 1512–1514[Free Full Text]
- Dellow EL, Unwin RJ, Honour JW. Pontefract cakes can be bad for you: refractory hypertension and liquorice excess. Nephrol Dial Transplant 1999; 14: 218–220[Abstract/Free Full Text]
- Hussain RM. The sweet cake that reaches parts other cakes can’t! Postgrad Med J 2003; 79: 115–116[Abstract/Free Full Text]
- Stewart PM, Corrie JE, Shackleton CH, Edwards CR. Syndrome of apparent mineralocorticoid excess. A defect in the cortisol–cortisone shuttle. J Clin Invest 1988; 82: 340–349[Web of Science][Medline]
- Barrella M, Lauria G, Quatrale R, Paolino E. Hypokalaemic rhabdomyolysis associated with liquorice ingestion: report of an atypical case. Ital J Neurol Sci 1997; 18: 217–220[CrossRef][Web of Science][Medline]
- Palermo M, Armanini D, Shackleton CH et al. Furosemide and 11beta-hydroxysteroid dehydrogenase activity, in man. Exp Clin Endocrinol Diabetes 2002; 110: 272–276[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
Top
Introduction
Case