Ann Clin Biochem 1990; 27: 378-381
Case Reports
Delayed diagnosis of Addison's disease J R Paterson, W D Neithercut and R J Spooner From the Department of Pathological Biochemistry, Western Infirmary and Gartnavel General Hospital. Glasgow. UK
Addison's disease may be difficult to diagnose because of the nonspecificity of the most frequently occurring symptoms. Hyponatraemia, hyperkalaemia, and uraemia are commonly detected in sera from patients with Addison's disease. We report two patients in whom the diagnosis was delayed, despite typical biochemical abnormalities in their sera during initial presentation. A computer search of 4 weeks of biochemical data (9862 requests for electrolytes) indicated that these typical biochemical abnormalities occurred commonly in other patients, increasing the difficulty of the clinical biochemist in detecting Addison's disease. Awareness of this problem, in combination with adequate clinical information, should enable the clinical biochemist to assist in the earlier diagnosis of Addison's disease.
SUMMARY.
Additional key phrases: hyponatraemia; hyperkalaemia: uraemia: serum cortisol Addison's disease is an uncommon condition, the diagnosis of which can be difficult to make and indeed may be missed.!'! The most frequently occurring symptoms are non-specific, including poor appetite, dyspepsia, nausea and vomiting.' The clinical signs of weight loss and hypotension may be overlooked although the presence of hyperpigmentation should alert the examining clinician to the possibility of Addison's disease." Biochemical analysis of serum from patients with Addison's disease during their initial presentation often reveals electrolyte abnormalities. In a Danish series of 108 Addisonian patients, hyponatraemia and hyperkalaemia were reported at presentation in 88 % and 64 %, respectively.' In a smaller Scottish series, hyponatraemia, hyperkalaemia and uraemia were reported in 75 %,40 % and 70 %, respectively." Our aim in this study was to highlight the potential of clinical biochemistry in the detection of Addison's disease. We also report two patients in whom the diagnosis was delayed although typical biochemical abnormalities were present in their initial serum samples. CASE 1 A 20 year old male was admitted to the acute surgical ward. He had a 3-day history of nausea and right iliac fossa pain. Examination revealed right iliac fossa tenderness and guarding. A diagcorrespondence; Dr J R Paterson, Institute of Biochemistry, Royal Infirmary, Glasgow G4 OSF, UK.
nosis of acute appendicitis was considered and discarded. He was observed overnight and discharged home the following day because his abdominal pain had settled. The only treatment given was I L of 0·9 % saline and I· 5 L of 5 % dextrose by intravenous infusion. On the day of his discharge he experienced recurrence of his symptoms of nausea and abdominal pain. He attended his general practitioner who noticed that he was hyperpigmented. The general practitioner contacted the biochemistry department and requested that the cortisol concentration be measured in a serum sample collected when the patient had been in the surgical ward. The cortisol concentration in this sample was 35 nmol{L (reference interval 160475 nmol{L). The electrolyte concentrations in this sample are included in Table I. He was admitted to a medical unit for investigation. A short tetracosactrin (Synacthen) test gave the following results: 8 am cortisol 37 nmol{L, 8·30am cortisol 36nmol{L. Adrenal calcification was not found on plain abdominal X-ray. A 1{100 titre of thyroid microsomal antibodies was detected in his serum although antiadrenal antibodies were not found. Thyroxine and thyroid stimulating. hormone concentrations were within our reference intervals at 93nmol{L and 0'6mU{L, respectively. A mild normochromic normocytic anaemia with a haemoglobin concentration of 12·0 g{dL was also present. Initially on admission he was treated with an intravenous infusion of 0·9 % saline. Following the short Synacthen test he was
378
Downloaded from acb.sagepub.com by guest on June 5, 2016
Delayed diagnosis of Addison's disease prescribed 20 mg hydrocortisone in the morning and IOmg in the evening along with 0·1 mg of fludrocortisone once daily. He was discharged home on this treatment. CASEl A 70-year-old man was admitted to a geriatric ward with ~2-week history of nausea, poor appetite and dyspepsia. His past medical history included osteoarthritis of the right knee, and childhood tuberculosis which had resulted in fusion of the left hip and recurrent right shoulder abscesses. He had recently been prescribed coproxamol after an earlier course of non-steroidal anti-inflammatory drugs for joint pain. His blood pressure on admission was 130/90 mm Hg. An upper gastrointestinal endoscopy was arranged which revealed a distal gastritis and duodenitis. He was prescribed l50mg ranitidine twice daily. Other investigations arranged included biochemical analysis of his serum which showed hyponatraemia and renal impairment (Table 1). A full blood count revealed a normocytic anaemia, a chest radiograph revealed changes attributed to old tuberculosis in his right lung field, and an abdominal ultrasound examination was normal. Over the next 2 weeks he was given physiotherapy to improve his poor mobility. Hyponatraemia persisted during this period (Request 2, Patient 2, Table 1). He became pyrexial, drowsy, disorientated and hypotensive 3 weeks following admission. His renal function deteriorated and hyponatraemia and hyperkalaemia persisted (Request 3, Patient 2, Table I). Blood cultures grew Staphylococcus aureus and treatment with intravenous cefotaxime I g t.i.d., and flucloxacillin 500 mg q.d.s. was commenced. He deteriorated clinically over the next 5 days. Intravenous rifampicin, I g t.d.s., was commenced because of the possibility of reactivated tuberculosis. He suffered a cardiac arrest and could not be resuscitated. A post mortem examination revealed bilateral tuberculous destruction of the adrenal glands. Retrospective analysis of two stored serum samples revealed cortisol concentrations of 89 and 84 nmol/L, Anti-adrenal antibodies were not present in the serum samples. METHODS AND RESULTS A computer search of 4 weeks of biochemical results held in the department's database was used to identify patients whose serum clinical
379
chemistry contained abnormalities similar to those found in Addison's disease. Initially patients with hyponatraemia were identified by the computer search. These patient's computer records were then searched again to identify those who also had serum urea concentrations between 6·0 and l4'Ommol/L, those who also had serum potassium concentrations greater than 4·5 mmol/L, and those who met all three biochemical criteria. The original request forms for patients who met all three biochemical criteria in a second 4 week period were located and the clinical information that had been available with the original request was reviewed. To our knowledge only one new patient with Addison's disease presented during the period of the searches. Table 2 contains the results of the computer search for the electrolyte pattern of Addison's disease. Hyponatraemia, a serum sodium of less than 135 mmol/L, was found in 7·9 %, a serum potassium concentration greater than 4·5 mmol/L in association with hyponatraemia in 3·0 %, and a serum urea concentration between 6·0 and 14·0 mmol/L in association with hyponatraemia in 2·8 %. All three biochemical criteria were found in 87 (0'9 %) patients. Clinical information was available on all but four of the original request forms. The diagnosis of Addison's disease was not mentioned on any form. Twentyone per cent of requests identified postoperative patients, 18 % of patients had cardiovascular disease, 15 % had malignant disease, 6·5 % had renal disease, 4 % of patients were diabetic and 2 % were receiving total parenteral nutrition. A wide range of other diagnoses, from porphyria to meningitis, were given on a further 24 % of forms. DISCUSSION Autoimmune adrenalitis and adrenal destruction by tuberculosis are the two most common causes of Addison's disease.':" In both our patients there was a delay in making the diagnosis. Patients with Addison's disease commonly have abnormalities of their serum biochemistry. An awareness of these ought to lead to an earlier diagnosis. The characteristic abnormalities are mild hyponatraemia, mild hyperkalaemia and mild uraemia, with occasional reports of metabolic acidosis, hypercalcaemia and hypoglycaemia. 3.5.6 In both of our patients mild hyponatraemia was present on admission (Table I). Hyperkalaemia was not present in serum samples from either patient on admission although the admission
Downloaded from acb.sagepub.com by guest on June 5, 2016
Downloaded from acb.sagepub.com by guest on June 5, 2016
I.
2.
134 132 128 130 133 131 133
129 127
4·7 4·5 6·1 6·3 5·5 5·7 5·5
4·6 4·8
Urea (mmol/L)
13·0 13-6 28·2 29·4 27·0 25·1 23·7
10·3 8·3
(2'5-7'5)
328 245 267
172 210
95
778
Na < 135 rnmol/L 279
Na < 135 mmol/L Urea > 6 and < 14 mmol/L
299
Na < 135 mmol/L K > 4·5 mmol/L
Old TB dyspeptic (on NSAIDS) Arthritis was on (NSAIDS) Septicaemia 2Y to UTI ? Septicaemia TB Septicaemia and pre-renal failure Septic shock. ? Source of infection. BM's erratic? cause
? Appendicitis ? Appendicitis
Information available on request form
87
Na < 135 mmol/L, K > 4·5 mmol/L Urea > 6 and < 14 mmol/L
Creatinine (Jlmol/L) (60-110)
Results of a computer search for the electrolyte pattern of Addison's disease in 9862 electrolyte requests
No. of patients
TABLE
1 2 3 4 5 6 7
Patient 2
I 2
Patient 1
(135-144)
Potassium (mmol/L) (3'5-5·1)
;:s
Serum concentrations
Sodium (mmol/L)
~ ;;:
'".,
;:s
I:>
~ I:>
;:s l:l..
l:l
~
'"., "
~ §:
I:>
ib
Serum concentrations and request form information
Request No.
TABLE
w
00
o
Delayed diagnosis of Addison's disease
potassium concentrations of both patients were greater than 4·5 mmol/L. Hyponatraemia and hyperkalaemia were reported in 88 % and 64 %, respectively, in one series of 108 Addisonian patients.' However, the occurrence of this electrolyte pattern is not uncommon in clinical biochemistry. The computer search detected 299 individuals in a 4-week period who had serum sodium and 'potassium concentrations consistent with Addison's disease. Examination of the clinical information available on the request forms showed that most patients with an Addisonian electrolyte pattern were known to have other serious illnesses which could have caused the electrolyte pattern or could also have triggered a crisis in a patient with Addison's disease. Measurements of the serum cortisol concentrations of these patients would perhaps help identify undiagnosed patients with Addison's disease. From the information available here it is not possible to assess whether or not the numbers of undiagnosed patients with Addison's disease that would be detected would justify the cost and time involved in carryingout extra cortisol assays. We have, however, commenced a prospective study which will try to answer this question. In Case 2 rifampicin was given to the patient. This drug has been reported to precipitate adrenal crises in patients with adrenal insufficiency by induction of cortisol metabolizing enzymes.' However, it was unlikely to have been of clinical significance in this case because of the short time interval between starting rifampicin and death (2 weeks for maximal enzyme induction). As the biochemical abnormalities of Addison's disease are often not severe and the clinical information available to the clinical chemist is usually restricted to key diagnostic phrases, this often downgrades the value of biochemical data and
381
interpretation. The difficulty that may therefore occur in making the diagnosis was illustrated by the delay in diagnosis that occurred in the two patients reported here, in whom typical biochemical abnormalities were observed at initial presentation. Although combination of the clinical and biochemical data should improve the probability of the diagnosis being made quickly, the common occurrence of similar electrolyte disturbances in other patients and the protean nature of the signs and symptoms continue to require vigilance on the part of clinician and clinical chemist.
Acknowledgements We thank Dr Connell and Dr Williams for their kind permission to report these two patients. We also thank Dr G O'Donovan and Dr St J O'Reilly for their help. REFERENCES I Mason AS, Meade TW, Lee JAR, Morris IN.
2 3 4 5
6
7
Epidemiological and clinical picture of Addison's disease. Lancet 1968; ii: 744-7 Small M, MacCuish AC, Thomson JA. MissedAddisonian crisis in surgical wards. Postgrad Med J 1987; 63: 367-9 Nerup J. Addison's disease--clinical studies. A report of 108 cases. Acta Endocrinol1974; 76: 127-41 IrvineWJ, BarnesEW. Adrenocortical insufficiency. Clin Endocrinol Metab 1972; 1: 549-94 Burke CWo Adrenocortical diseases. In: Oxford Textbood of Medicine, 2nd Edition. Weatherall OJ, Ledingham JGG, Warrell OA, eds. Oxford: Oxford University Press 1987; 10.69-10.83 Bandy PK. Disorders of the adrenal cortex. In: Williams Textbook of Endocrinology, 7th Edition. Wilson JD, Foster OW, eds. Philadelphia: WB Saunders Company 1985; 853-6 Eiansary ER, Earis JE. Rifampicin and adrenal crisis. Dr Med J 1983; 286: 1861-2
Accepted for publication 13 February 1990
Downloaded from acb.sagepub.com by guest on June 5, 2016
Case Reports
Delayed diagnosis of Addison's disease J R Paterson, W D Neithercut and R J Spooner From the Department of Pathological Biochemistry, Western Infirmary and Gartnavel General Hospital. Glasgow. UK
Addison's disease may be difficult to diagnose because of the nonspecificity of the most frequently occurring symptoms. Hyponatraemia, hyperkalaemia, and uraemia are commonly detected in sera from patients with Addison's disease. We report two patients in whom the diagnosis was delayed, despite typical biochemical abnormalities in their sera during initial presentation. A computer search of 4 weeks of biochemical data (9862 requests for electrolytes) indicated that these typical biochemical abnormalities occurred commonly in other patients, increasing the difficulty of the clinical biochemist in detecting Addison's disease. Awareness of this problem, in combination with adequate clinical information, should enable the clinical biochemist to assist in the earlier diagnosis of Addison's disease.
SUMMARY.
Additional key phrases: hyponatraemia; hyperkalaemia: uraemia: serum cortisol Addison's disease is an uncommon condition, the diagnosis of which can be difficult to make and indeed may be missed.!'! The most frequently occurring symptoms are non-specific, including poor appetite, dyspepsia, nausea and vomiting.' The clinical signs of weight loss and hypotension may be overlooked although the presence of hyperpigmentation should alert the examining clinician to the possibility of Addison's disease." Biochemical analysis of serum from patients with Addison's disease during their initial presentation often reveals electrolyte abnormalities. In a Danish series of 108 Addisonian patients, hyponatraemia and hyperkalaemia were reported at presentation in 88 % and 64 %, respectively.' In a smaller Scottish series, hyponatraemia, hyperkalaemia and uraemia were reported in 75 %,40 % and 70 %, respectively." Our aim in this study was to highlight the potential of clinical biochemistry in the detection of Addison's disease. We also report two patients in whom the diagnosis was delayed although typical biochemical abnormalities were present in their initial serum samples. CASE 1 A 20 year old male was admitted to the acute surgical ward. He had a 3-day history of nausea and right iliac fossa pain. Examination revealed right iliac fossa tenderness and guarding. A diagcorrespondence; Dr J R Paterson, Institute of Biochemistry, Royal Infirmary, Glasgow G4 OSF, UK.
nosis of acute appendicitis was considered and discarded. He was observed overnight and discharged home the following day because his abdominal pain had settled. The only treatment given was I L of 0·9 % saline and I· 5 L of 5 % dextrose by intravenous infusion. On the day of his discharge he experienced recurrence of his symptoms of nausea and abdominal pain. He attended his general practitioner who noticed that he was hyperpigmented. The general practitioner contacted the biochemistry department and requested that the cortisol concentration be measured in a serum sample collected when the patient had been in the surgical ward. The cortisol concentration in this sample was 35 nmol{L (reference interval 160475 nmol{L). The electrolyte concentrations in this sample are included in Table I. He was admitted to a medical unit for investigation. A short tetracosactrin (Synacthen) test gave the following results: 8 am cortisol 37 nmol{L, 8·30am cortisol 36nmol{L. Adrenal calcification was not found on plain abdominal X-ray. A 1{100 titre of thyroid microsomal antibodies was detected in his serum although antiadrenal antibodies were not found. Thyroxine and thyroid stimulating. hormone concentrations were within our reference intervals at 93nmol{L and 0'6mU{L, respectively. A mild normochromic normocytic anaemia with a haemoglobin concentration of 12·0 g{dL was also present. Initially on admission he was treated with an intravenous infusion of 0·9 % saline. Following the short Synacthen test he was
378
Downloaded from acb.sagepub.com by guest on June 5, 2016
Delayed diagnosis of Addison's disease prescribed 20 mg hydrocortisone in the morning and IOmg in the evening along with 0·1 mg of fludrocortisone once daily. He was discharged home on this treatment. CASEl A 70-year-old man was admitted to a geriatric ward with ~2-week history of nausea, poor appetite and dyspepsia. His past medical history included osteoarthritis of the right knee, and childhood tuberculosis which had resulted in fusion of the left hip and recurrent right shoulder abscesses. He had recently been prescribed coproxamol after an earlier course of non-steroidal anti-inflammatory drugs for joint pain. His blood pressure on admission was 130/90 mm Hg. An upper gastrointestinal endoscopy was arranged which revealed a distal gastritis and duodenitis. He was prescribed l50mg ranitidine twice daily. Other investigations arranged included biochemical analysis of his serum which showed hyponatraemia and renal impairment (Table 1). A full blood count revealed a normocytic anaemia, a chest radiograph revealed changes attributed to old tuberculosis in his right lung field, and an abdominal ultrasound examination was normal. Over the next 2 weeks he was given physiotherapy to improve his poor mobility. Hyponatraemia persisted during this period (Request 2, Patient 2, Table 1). He became pyrexial, drowsy, disorientated and hypotensive 3 weeks following admission. His renal function deteriorated and hyponatraemia and hyperkalaemia persisted (Request 3, Patient 2, Table I). Blood cultures grew Staphylococcus aureus and treatment with intravenous cefotaxime I g t.i.d., and flucloxacillin 500 mg q.d.s. was commenced. He deteriorated clinically over the next 5 days. Intravenous rifampicin, I g t.d.s., was commenced because of the possibility of reactivated tuberculosis. He suffered a cardiac arrest and could not be resuscitated. A post mortem examination revealed bilateral tuberculous destruction of the adrenal glands. Retrospective analysis of two stored serum samples revealed cortisol concentrations of 89 and 84 nmol/L, Anti-adrenal antibodies were not present in the serum samples. METHODS AND RESULTS A computer search of 4 weeks of biochemical results held in the department's database was used to identify patients whose serum clinical
379
chemistry contained abnormalities similar to those found in Addison's disease. Initially patients with hyponatraemia were identified by the computer search. These patient's computer records were then searched again to identify those who also had serum urea concentrations between 6·0 and l4'Ommol/L, those who also had serum potassium concentrations greater than 4·5 mmol/L, and those who met all three biochemical criteria. The original request forms for patients who met all three biochemical criteria in a second 4 week period were located and the clinical information that had been available with the original request was reviewed. To our knowledge only one new patient with Addison's disease presented during the period of the searches. Table 2 contains the results of the computer search for the electrolyte pattern of Addison's disease. Hyponatraemia, a serum sodium of less than 135 mmol/L, was found in 7·9 %, a serum potassium concentration greater than 4·5 mmol/L in association with hyponatraemia in 3·0 %, and a serum urea concentration between 6·0 and 14·0 mmol/L in association with hyponatraemia in 2·8 %. All three biochemical criteria were found in 87 (0'9 %) patients. Clinical information was available on all but four of the original request forms. The diagnosis of Addison's disease was not mentioned on any form. Twentyone per cent of requests identified postoperative patients, 18 % of patients had cardiovascular disease, 15 % had malignant disease, 6·5 % had renal disease, 4 % of patients were diabetic and 2 % were receiving total parenteral nutrition. A wide range of other diagnoses, from porphyria to meningitis, were given on a further 24 % of forms. DISCUSSION Autoimmune adrenalitis and adrenal destruction by tuberculosis are the two most common causes of Addison's disease.':" In both our patients there was a delay in making the diagnosis. Patients with Addison's disease commonly have abnormalities of their serum biochemistry. An awareness of these ought to lead to an earlier diagnosis. The characteristic abnormalities are mild hyponatraemia, mild hyperkalaemia and mild uraemia, with occasional reports of metabolic acidosis, hypercalcaemia and hypoglycaemia. 3.5.6 In both of our patients mild hyponatraemia was present on admission (Table I). Hyperkalaemia was not present in serum samples from either patient on admission although the admission
Downloaded from acb.sagepub.com by guest on June 5, 2016
Downloaded from acb.sagepub.com by guest on June 5, 2016
I.
2.
134 132 128 130 133 131 133
129 127
4·7 4·5 6·1 6·3 5·5 5·7 5·5
4·6 4·8
Urea (mmol/L)
13·0 13-6 28·2 29·4 27·0 25·1 23·7
10·3 8·3
(2'5-7'5)
328 245 267
172 210
95
778
Na < 135 rnmol/L 279
Na < 135 mmol/L Urea > 6 and < 14 mmol/L
299
Na < 135 mmol/L K > 4·5 mmol/L
Old TB dyspeptic (on NSAIDS) Arthritis was on (NSAIDS) Septicaemia 2Y to UTI ? Septicaemia TB Septicaemia and pre-renal failure Septic shock. ? Source of infection. BM's erratic? cause
? Appendicitis ? Appendicitis
Information available on request form
87
Na < 135 mmol/L, K > 4·5 mmol/L Urea > 6 and < 14 mmol/L
Creatinine (Jlmol/L) (60-110)
Results of a computer search for the electrolyte pattern of Addison's disease in 9862 electrolyte requests
No. of patients
TABLE
1 2 3 4 5 6 7
Patient 2
I 2
Patient 1
(135-144)
Potassium (mmol/L) (3'5-5·1)
;:s
Serum concentrations
Sodium (mmol/L)
~ ;;:
'".,
;:s
I:>
~ I:>
;:s l:l..
l:l
~
'"., "
~ §:
I:>
ib
Serum concentrations and request form information
Request No.
TABLE
w
00
o
Delayed diagnosis of Addison's disease
potassium concentrations of both patients were greater than 4·5 mmol/L. Hyponatraemia and hyperkalaemia were reported in 88 % and 64 %, respectively, in one series of 108 Addisonian patients.' However, the occurrence of this electrolyte pattern is not uncommon in clinical biochemistry. The computer search detected 299 individuals in a 4-week period who had serum sodium and 'potassium concentrations consistent with Addison's disease. Examination of the clinical information available on the request forms showed that most patients with an Addisonian electrolyte pattern were known to have other serious illnesses which could have caused the electrolyte pattern or could also have triggered a crisis in a patient with Addison's disease. Measurements of the serum cortisol concentrations of these patients would perhaps help identify undiagnosed patients with Addison's disease. From the information available here it is not possible to assess whether or not the numbers of undiagnosed patients with Addison's disease that would be detected would justify the cost and time involved in carryingout extra cortisol assays. We have, however, commenced a prospective study which will try to answer this question. In Case 2 rifampicin was given to the patient. This drug has been reported to precipitate adrenal crises in patients with adrenal insufficiency by induction of cortisol metabolizing enzymes.' However, it was unlikely to have been of clinical significance in this case because of the short time interval between starting rifampicin and death (2 weeks for maximal enzyme induction). As the biochemical abnormalities of Addison's disease are often not severe and the clinical information available to the clinical chemist is usually restricted to key diagnostic phrases, this often downgrades the value of biochemical data and
381
interpretation. The difficulty that may therefore occur in making the diagnosis was illustrated by the delay in diagnosis that occurred in the two patients reported here, in whom typical biochemical abnormalities were observed at initial presentation. Although combination of the clinical and biochemical data should improve the probability of the diagnosis being made quickly, the common occurrence of similar electrolyte disturbances in other patients and the protean nature of the signs and symptoms continue to require vigilance on the part of clinician and clinical chemist.
Acknowledgements We thank Dr Connell and Dr Williams for their kind permission to report these two patients. We also thank Dr G O'Donovan and Dr St J O'Reilly for their help. REFERENCES I Mason AS, Meade TW, Lee JAR, Morris IN.
2 3 4 5
6
7
Epidemiological and clinical picture of Addison's disease. Lancet 1968; ii: 744-7 Small M, MacCuish AC, Thomson JA. MissedAddisonian crisis in surgical wards. Postgrad Med J 1987; 63: 367-9 Nerup J. Addison's disease--clinical studies. A report of 108 cases. Acta Endocrinol1974; 76: 127-41 IrvineWJ, BarnesEW. Adrenocortical insufficiency. Clin Endocrinol Metab 1972; 1: 549-94 Burke CWo Adrenocortical diseases. In: Oxford Textbood of Medicine, 2nd Edition. Weatherall OJ, Ledingham JGG, Warrell OA, eds. Oxford: Oxford University Press 1987; 10.69-10.83 Bandy PK. Disorders of the adrenal cortex. In: Williams Textbook of Endocrinology, 7th Edition. Wilson JD, Foster OW, eds. Philadelphia: WB Saunders Company 1985; 853-6 Eiansary ER, Earis JE. Rifampicin and adrenal crisis. Dr Med J 1983; 286: 1861-2
Accepted for publication 13 February 1990
Downloaded from acb.sagepub.com by guest on June 5, 2016
