Misleading normal TSH and persistently elevated creatine kinase: clues to the diagnosis of chronic Sheehan’s syndrome

Abstract

A 53-year-old woman was referred for medical evaluation of therapy-resistant dyslipidaemia accompanied by elevated creatine kinase levels. Because cessation or alteration of her medication did not improve laboratory abnormalities, hypothyroidism was considered, despite the fact that thyroid stimulating hormone levels were within the reference interval. On further evaluation, she was found to have panhypopituitarism and empty sella turcica as shown by MRI. These findings were unexpected since there was no clinical suspicion during detailed evaluation. When supplementary questions were asked, she brought up a history of severe postpartum haemorrhage 30 years ago, for which she underwent a hysterectomy. Based on these findings, the patient was diagnosed with Sheehan’s syndrome. This syndrome is a rare but potentially life-threatening complication of postpartum haemorrhage, characterised by varying degrees of hypopituitarism that are most commonly presented many years after delivery. The patient recovered after adequate hormone replacement therapy.

Background

Postpartum haemorrhage is a common obstetric emergency in both high and low per capita income countries. One of the rare but potentially life-threatening complications of this condition is Sheehan’s syndrome, that is, postpartum hypopituitarism. During the pregnancy, the pituitary gland is considerably enlarged, due to the estrogen-stimulated lactotroph hyperplasia and prone to infarction from hypovolaemic shock secondary to excessive blood loss during or after child birth. Damage to the pituitary can range from mild to severe, and may affect the secretion of one, several or all of its hormones. A common clinical presentation is the combination of failure to lactate postdelivery and amenorrhoea or oligomenorrhoea, but the spectrum is broad, ranging from non-specific complaints (eg, anorexia, weakness, fatigue and anaemia) to severe manifestations of hypopituitarism (eg, hypotension, hyponatraemia and hypothyroidism). Many patients remain undiagnosed and untreated due to diagnostic delay that can be attributed to the lack of awareness of this rare condition. Since symptoms may present at any time, from immediately postpartum to even many years after delivery, taking a detailed and complete medical history is indispensable.

Case presentation

A 53-year-old woman was referred to the outpatient department of our hospital by her general practitioner due to persistent dyslipidaemia under statin treatment and an elevated creatine kinase (CK). Five years earlier, she was referred with the same problem. The internist at that first visit recorded a history of a caesarian section, a hysterectomy because of persistent bleeding, and familial hypercholesterolaemia, for which she was treated with atorvastatin. No information on lactation was recorded in her file at that time. Physical examination showed no abnormalities; she was euthyroid and demonstrated no signs of hypocortisolism (eg, a normal blood pressure and electrolytes, and she did not describe weight loss or recurrent infections). There were no indications for secondary dyslipidaemia, that is, she was normoglycaemic, and her liver function was normal. Thyroid function was considered normal based on her serum thyroid stimulating hormone (TSH) level (1.9 mIU/L) that fell within the reference interval (0.35–5.00 mIU/L). In the Netherlands, free thyroxine (fT4) is only measured as a reflex test, when TSH levels exceed the reference interval; measurement of her fT4 levels had thus not been requested before by the general practitioner. Due to lack of clinical suspicion, the internist at the first visit did not explicitly request fT4 either. Renal function was mildly impaired (estimated glomerular filtration rate 49 mL/min/1.73 m²) without albuminuria. Despite slightly elevated CK levels (308 IU/L, ULN <225 IU/L), she had no muscle complaints and tolerated statin therapy quite well, with a good response of the plasma lipid profile. Therefore, the patient was advised to continue her treatment and was referred back to the general practitioner for guidance on the management of CK and renal function. In the Netherlands, it is common practice for patients to be monitored by their general practitioner, based on good regional appointments and accessible contact with a specialist when needed.

A gradual increase in plasma CK levels over the course of 5 years (up to 667 IU/L, ULN <170 IU/L), which did not decrease after discontinuation of atorvastatin was the reason for the current referral.

Although the patient had no obvious complaints initially, she recently noticed a mild exhaustion of her thighs after routine physical activities (ie, walking or biking). She was not on medication other than atorvastatin (40 mg/day) and did not use any toxic substances (nicotine, alcohol and drugs).

On physical examination, no abnormalities in heart, lungs, abdomen or muscle strength in arms and legs were found. Her blood pressure was 107/65 mm Hg with a regular and equal pulse rate of 81/min, and her body weight was 56 kg with a height of 1.67 m (body mass index 20.1 kg/m²).

Investigations

Comprehensive laboratory testing was performed (table 1). The internist additionally requested CK isoenzyme levels (MM: skeletal muscle, MB: heart muscle and BB: brain) and repeated TSH together with fT4. Laboratory tests confirmed previous findings such as mild normocytic anaemia, dyslipidaemia, moderate renal function impairment, normal liver function and normoglycaemia. CK isoenzyme analysis showed that mainly CK-MM was responsible for the elevated CK levels. In addition normal TSH levels along with low fT4 were detected. Since there was no medical history nor clinical signs indicating abnormal thyroid function, the reliability of these results was questioned, and additional tests were requested.

Table 1.

Extensive laboratory testing at the moment of referral

Parameter	Unit	Reference interval	Result
Total blood count
Haemoglobin	mmol/L	7.5–10.0	7.4
MCV	fL	80–100	87
Erythrocytes	×1012/L	4.0–5. 0	4.1
Leucocytes	×109/L	4.5–11.0	6.2
Platelets	×109/L	150–350	157
Blood chemistry
Glucose	mmol/L	4.0–7.8	5
Haemoglobin A1c	mmol/mol	26–42	32
Sodium	mmol/L	135–145	144
Chloride	mmol/L	97–108	110
Potassium	mmol/L	3.5–5.1	4.2
Urea	mmol/L	3.5–7.2	6.6
Creatinine	µmol/L	53–97	117
eGFR	mL/min/1.72 m2	>90	42
ALAT	IU/L	5–34	18
ASAT	IU/L	5–31	50
ALP	IU/L	<125	40
GGT	IU/L	5–38	13
LDH	IU/L	100–250	228
Bilirubin (total)	µmol/L	2–17	7
CK	IU/L	130–170	721 (>98% CK-MM, <1% CK-MB, <1% CK-BB)
Total protein	g/L	64–83	73
Albumin	g/L	34–48	35
Lipid profile
Total cholesterol	mmol/L	3.0–6.5	7.8
Triglyceride	mmol/L	0.4–2.7	7.6
LDL cholesterol	mmol/L	<2.5	3.3
HDL cholesterol	mmol/L	>1.2	0.7
Lipoprotein (a)	mg/L	<300	120
Apolipoprotein A1	g/L	1.05 – 2.05	1.13
Apolipoprotein B	g/L	0.55 – 1.30	1.47
Apolipoprotein E			E3/E3
Hormones
TSH	mIU/L	0.35 – 5.00	1.9
fT4	pmol/L	9 - 19	<6
fT3	pmol/L	2.9 – 4.9	2.4

ALAT, alanine aminotransferase; ALP, alkaline phosphotase; ASAT, aspartate aminotransferase; CK, creatine kinase; eGFR, estimated glomerular filtration rate; fT3, free triiodothyronine; fT4, free thyroxine; GGT, gamma-glutamyl transferase; HDL, high-density lipoprotein; LDH, lactate dehydrogenase; LDL, low-density lipoprotein; MCV, mean corpuscular volume; TSH, thyroid-stimulating hormone.

The same sample was analysed with different assays and with symmetric equilibration dialysis (table 2), confirming both low fT4 and fT3, thus excluding assay interference. The fact that inappropriately normal or low-normal TSH levels contradict low free hormone levels, pointed towards secondary hypothyroidism. Therefore, additional hormone assays were performed (table 3), revealing panhypopituitarism. MRI of the pituitary gland was performed, showing a partial empty sella turcica (figure 1).

Table 2.

Tests performed in the same sample using different immunoassays in order to exclude analytical interference

	Parameter	Different immunoassays
	Abbott Architect i2000SR	Roche Cobas E602	Orthos Vitros ECiQ	Equilibrium Dialysis (symmetric)	ELISA
TSH	1.9 (0.35–5.0 mIU/L)	2.34 (0.35–5.0 mIU/L)	1.7 (0.4–4.3 mIU/L)
fT4	<6 (9–19 pmol/L)	1.53 (9–19 pmol/L)	1.1 (11–25 pmol/L)	1.82 (13–25 pmol/L) Free dialysis percentage 0.011 (0.014–0.025)
fT3	2.37 (2.6–5.7 pmol/L)		1.3 (3.4–6.3 pmol/L)
tT4			12 (25–128 nmol/L)	17 (55–155 nmol/L)
tT3			0.96 (1.4–2.5 nmol/L)
HAMA					<5, non-detectable (<25 ng/mL)

fT3, free triiodothyronine; fT4, free thyroxine; HAMA, human antimouse antibody; TSH, thyroid-stimulating hormone; tT3, total triiodothyronine; tT4, total thyroxine.

Table 3.

Additional hormonal testing for assessment of pituitary function

Parameter	Unit	Reference interval	Result
Cortisol	nmol/L	100–540	<0.02
LH	IU/L	5–60	2
FSH	IU/L	25–130	8
Oestradiol	nmol/L	0.04–0.10	<0.04
IGF-1	ng/mL	74–242	10 (−5 SD)
Prolactin	U/L	<0.6	0.1

FSH, follicle-stimulating hormone; IGF-1, insulin-like growth factor; IGF-1, insulin-like growth factor; LH, luteinising hormone; SD, standard deviation.

Figure 1.

MRI of the pituitary gland (arrows), showing a partial empty sella turcica with a minimal amount of pituitary tissue at the bottom of the sella turcica.

Differential diagnosis

Elevated CK levels are associated with a number of myopathies, including those based on inflammation, infection or endocrine, metabolic and toxic disorders.

Given the fact that our patient had only minor symptoms without pain or other alarming signs, the most relevant differentials of elevated CK and low fT4 levels with normal TSH in this case were:

• Drug-induced myopathy as a result of statin therapy.

• Analytical interference.

• Endocrine myopathy due to secondary hypothyroidism with pituitary hypofunction.

Drug-induced myopathy as a result of statin treatment was ruled out by persistently elevated CK levels, despite discontinuation of medication for 3 months. Assay interference was also eliminated by conducting extensive testing with different assays and a human antimouse antibody test. Additional testing of different pituitary axes led to the final diagnosis of panhypopituitarism with secondary hypothyroidism.

Treatment

Replacement therapy for individual pituitary and downstream hormonal deficiencies was commenced as soon as the diagnosis of Sheehan’s syndrome was made. (Inter)national clinical practice guidelines recommend administration of 15–20 mg hydrocortisone per day, divided into two or three doses to mimic physiological daily production rates.^{1 2} Correspondingly, our patient received hydrocortisone 10-5-5 mg per day, followed by the addition of thyroid hormone replacement therapy after 1 week, and titrated up to 100 µg/day after 6 weeks. Finally, growth hormone (GH) therapy 0.4 mg/day was added, and the patient was instructed how to use hydrocortisone injections in emergency situations. Her lipid-lowering medication was continued.

Outcome and follow-up

Following the initiation of hormonal replacement therapy, our patient was closely monitored through clinical evaluation and laboratory tests, monthly for the first 3 months, followed by every 3 months intervals in the first year and continued twice a year. The plasma lipid profile, elevated CK levels and thyroid function were normalised within the first 2 months of therapy (table 4). The patient reported increased level of energy and muscular strength.

Table 4.

Laboratory tests 2 months after commencement of substitution therapy

Parameter	Unit	Reference interval	Result
Total blood count
Haemoglobin	mmol/L	7.5–10.0	9.0
MCV	fL	80–100	91
Erythrocyte	×1012/L	4.0–5. 0	4.9
Blood chemistry
Glucose	mmol/L	4.0–7.8	4.0
Sodium	mmol/L	135–145	144
Potassium	mmol/L	3.5–5.1	3.8
Creatinine	µmol/L	53–97	96
eGFR	mL/min/1.72 m2	>90	56
CK	IU/L	130–170	115
Lipid profile
Total cholesterol	mmol/L	3.0–6.5	4.6
Triglyceride	mmol/L	0.4–2.7	1.5
LDL cholesterol	mmol/L	<2.5	2.6
HDL cholesterol	mmol/L	>1.2	1.4
Hormones
TSH	mIU/L	0.35–5.00	0.8
fT4	pmol/L	9–19	18.3
fT3	pmol/L	2.9–4.9	3.1
IGF-1	ng/mL	74–242	221 (+1.9 SD)

CK, creatine kinase; eGFR, estimated glomerular filtration rate; fT3, free triiodothyronine; fT4, free thyroxine; HDL, high-density lipoprotein; IGF-1, insulin-like growth factor; LDL, low density lipoprotein; MCV, mean corpuscular volume; SD, standard deviation; TSH, thyroid-stimulating hormone.

Bone densitometry showed osteopenia (T-scores −1.9 in left and right femoral necks, −1.5 in the lumbar spine) for which combination therapy with bisphosphonate 70 mg and vitamin D 20 µg per day was initiated and the patient was advised to increase her calcium intake.

Discussion

The clinical manifestations of Sheehan’s syndrome vary widely depending on the extent of tissue destruction. Pituitary gland enlargement during pregnancy, small size of sella turcica, vasospasm and thrombosis are frequently described as possible etiopathogenetic mechanisms.^3–5 Abrupt onset of severe hypopituitarism immediately after delivery is far less common than presentation after many years. This progressive course may partly be explained by the leakage of antigens from the damaged pituitary gland and the concomitant formation of antibodies. Most patients have vague complaints for a long period of time before showing clinical symptoms suggestive of partial or panhypopituitarism.^{5 6} Failure of postpartum lactation and/or menstruation are the most common presenting findings of patients with Sheehan’s syndrome. Delayed diagnosis, even up to several decades, in developed countries may also be the result of a lower incidence because of improved obstetrical care and as a consequence unawareness among physicians regarding the syndrome.⁵

This pattern is clearly visible in the history of our patient. It took nearly three decades to diagnose Sheehan’s syndrome, caused by a complicated caesarean birth with a severe postpartum haemorrhage resulted in 5 L of blood transfusion. Diagnosis was delayed because irregularities in menstruation could not be noticed in her case as a result of her hysterectomy. She had not attempted to breast feed her baby and hence failure of lactation was not discovered either. Apart from being more exhausted after a significant effort, such as a bicycle tour, she hardly had any complaints at presentation.

The last two decades she was treated with lipid-lowering medication for familial combined hyperlipidaemia. Routine controls revealed therapy-resistant dyslipidaemia and elevated plasma CK levels. Cessation or alteration of her medication did not improve dyslipidaemia, nor did it normalise CK levels. Since TSH levels were normal (within the reference interval), and clinical signs were absent, hypothyroidism was initially considered unlikely. In the Netherlands, it is advised to use only TSH as a screening tool for thyroid function. fT4 is measured as a reflex test in case of an abnormal TSH result or in case of clinical suspicion.

To evaluate the patient’s persistent dyslipidaemia and elevated CK levels, she was referred to our outpatient department twice. At first instance, an expectative policy was decided on, since there was no clinical suspicion for an underlying endocrine myopathy. At her second presentation, 5 years later, plasma CK levels were gradually increasing and dyslipidaemia was still persisting. During this visit additional laboratory tests, that is, CK-isoenzyme analysis, total lipid panel and fT4 were performed. The results pointed at secondary hypothyroidism for which further endocrine investigation was performed, thus revealing adrenal insufficiency, hypogonadism and GH-deficiency. MRI of the pituitary gland showed a partial empty sella, confirming Sheehan’s syndrome.

Nearly one-third of patients with hypothyroidism has persistently elevated serum concentrations of CK.⁷ Although this situation was considered in our patient, the absence of symptoms delayed diagnosis. The screening policy in the Netherlands to measure only TSH levels and not fT4 poses a pitfall, since not all care providers are aware that TSH secretion in secondary hypothyroidism may be normal or even elevated in some cases.^{6 8 9}

Pathophysiology of persistently elevated CK levels by overt hypothyroidism is suggested to be a result of diminished glycolysis and oxidative phosphorylation and thus insufficient ATP levels to confer protection against cell membrane alteration.¹⁰ Besides, abnormal accumulation of glycogen in muscle cells is correlated with the severity of hypothyroidism.¹⁰ The persistency of dyslipidaemia in our patient can also be traced back to hypothyroidism, which is associated with unfavourable effects on lipid profiles. Hypothyroid patients have decreased activities of low-density lipoprotein (LDL) receptors, lipoprotein lipase, hepatic lipase and 3-hydroxy-3-methylglutaryl-coenzyme A reductase. As a result, levels of total cholesterol, LDL and intermediate-density lipoprotein (IDL), triglyceride-rich lipoproteins (very low-density lipoprotein, VLDL), high-density lipoprotein (HDL)and lipoprotein(a) (Lp(a)) are increased.^11–14

Severe hypotension due to postpartum haemorrhage can result in varying degrees of renal function impairment by Sheehan’s syndrome.^{3 5 6} In our patient, a moderate stable renal function impairment and a mild anaemia were observed. Haematological abnormalities such as anaemia, thrombocytopaenia, pancytopenia have also been described in a variable proportion Sheehan’s syndrome patients, as a result of cortisol and thyroid hormone deficiencies with secondarily decreased erythropoietin levels.^{3 5 6} Osteopenia, as found in our patient, may result from multiple hormone deficiencies (GH and gonadotrophin hormones).

Treatment of our patient consisted of replacement of deficient hormones, appropriate for her stage of life, that is, thyroid hormone and glucocorticoids with the addition of medication to prevent osteoporosis. Her lipid-lowering medication was continued and follow-up visits are scheduled every half year.

Learning points.

• Only thyroid-stimulating hormone (TSH) as an initial test for thyroid function or as a start of reflex testing cannot rule out secondary hypothyroidism. TSH levels can be normal and within the reference interval, but still relatively low compared with free thyroxine (fT4).

• fT4 measurement should always explicitly be requested together with TSH, when hypothyroidism is suspected.

• Clinicians should be aware that hypothyroidism can lead to therapy-resistant dyslipidaemia and persistently elevated plasma creatine kinase levels.

• Accurate and detailed anamnesis remains one of the most important tools towards diagnosis of Sheehan’s syndrome.

Ethics statements

Patient consent for publication

Obtained.