Skip to main content
BMJ Case Reports logoLink to BMJ Case Reports
. 2020 Sep 7;13(9):e235450. doi: 10.1136/bcr-2020-235450

False elevation of serum cortisol in chemiluminescence immunoassay by Siemens Advia Centaur XP system in 21-hydroxylase deficiency: an ‘endocrine laboma’

Neeti Agrawal 1, Partha Pratim Chakraborty 1,, Anirban Sinha 1, Animesh Maiti 1
PMCID: PMC7477984  PMID: 32900728

Abstract

Liquid chromatography–mass spectrometry, the gold standard method for cortisol measurement, is expensive and not widely available in the developing countries. Chemiluminescent immunoassay, commonly used for cortisol measurement is prone to clinically meaningful inter-assay variability in some analysers. This occurs due to non-specific nature of anticortisol antibodies used in different platforms, having cross reactivity with structurally similar cortisol precursors like 17α-hydroxyprogesterone (17OHP), 11-deoxycortisol and 21-deoxycortisol. In patients with 21-hydroxylase deficiency, where 17OHP and 21-deoxycortisol are significantly elevated, older generation machines like Siemens Advia Centaur XP provide spuriously high cortisol concentration compared with values measured by Roche Cobas e 411 or Siemens Immulite 1000. Diagnosis of potentially life-threatening salt-wasting 21-hydroxylase deficiency may be missed and treatment may be delayed due to such interference. Two children with classic 21-hydroxylase deficiency are being reported here, in whom high cortisol values were observed in Siemens Advia Centaur XP system.

Keywords: adrenal disorders, paediatric intensive care, congenital disorders, neonatal and paediatric intensive care

Background

Defect in 21-hydroxylase, a crucial enzyme involved in adrenal steroidogenesis, is the most common cause (90%–95% of all cases) of congenital adrenal hyperplasia (CAH) with overall incidence ranging from 1 in 10 000 to 1 in 20 000 live births.1 Depending on residual enzymatic activity, patients with 21-hydroxylase deficiency present with varying degrees of glucocorticoid deficiency, androgen excess with/without mineralocorticoid deficiency. The classic salt-wasting form is the most severe variety of 21-hydroxylase deficiency, wherein newborns usually present with failure to thrive, dehydration, hyponatraemia, hyperkalaemia, cortisol deficiency and ambiguous genitalia (in girls) or peripheral precocious puberty (in boys) due to complete loss of enzymatic activity. The prototypical hormonal profile in this form is low cortisol, high 17α-hydroxyprogesterone (17OHP), high adrenocorticotrophic hormone (ACTH), high plasma renin activity (PRA) and high androgens. Elevated 21-deoxycortisol, though not measured routinely, is characteristic of 21-hydroxylase deficiency. In resource-restricted settings, where definitive diagnosis by CYP21A2 mutational analysis is often difficult, diagnosis of this potentially life-threatening condition is achieved on the bases of clinical history, relevant biochemical and hormonal investigations. Most, if not all, of the laboratories in the developing countries perform these tests by chemiluminescent immunoassay (CLIA). Discrepancy in measured cortisol concentration by CLIA in different available analysers is problematic owing to variable cross-reactivity of cortisol with compounds like 17OHP and 21-deoxycortisol, which circulate in high concentrations in these patients. Cortisol, measured by older generation analysers, may be elevated, and at times mislead the treating physician about making a definitive diagnosis and appropriate treatment of 21-hydroxylase deficiency.

Case presentation

Case 1

A 5-month-old child with ambiguous genitalia was referred for evaluation of discordant biochemical parameters (table 1). Born out of non-consanguineous union, the child had been admitted with salt-losing crisis on postnatal day 45. With a provisional diagnosis of virilising CAH, the child had been put on hydrocortisone and oral salt. There was no history of maternal virilisation or androgenic drug intake by the mother during pregnancy.

Table 1.

Summary of investigations

Parameters Case 1 Case 2 Age and sex specific reference range
During initial evaluation
Karyotype 46, XX 46, XX
Sodium 125 142 139–146 mmol/L
Potassium 7.6 4.7 4.1–5.3 mmol/L
08:00 serum cortisol
(Siemens Advia Centaur XP)
24.04 and 28.8 24.26 4.3–22.4 µg/dL
08:00 plasma adrenocorticotrophic hormone (iced sample) >1250 498 <46 pg/mL
17α-hydroxyprogesterone 231.9 316.5 <2 ng/mL
Testosterone 213 167 <20 ng/dL
Dehydroepiandrosterone sulfate 264.4 293 At 5 months: <74 µg/dL
At 4 years:
≤22 µg/dL
Androstenedione 10.28 0.7–3.6 ng/mL
Serum aldosterone 5.8 2.52–39.2 ng/dL
Plasma renin activity >24 >24 1.9–5.2 ng/mL/hour
Evaluation post-referral
08:00 serum cortisol
Roche Cobas e 411 (II Generation)
1.64 3.5 3–21 µg/dL
08:00 serum cortisol
(Siemens Immulite 1000)
<1 1.9 5–25 µg/dL

Clinical examination revealed the following:

Blood pressure: 68/42 mm Hg (<50th percentile for age and height). Prader stage 3 genitalia in form of clitoromegaly (clitoral length: 20 mm), posterior labial fusion and single perineal opening. The labio-scrotal folds were hyperpigmented and gonads were non-palpable. Clitoral index and anogenital ratio were 48 mm2 and 0.68, respectively.

Case 2

A 4-year-old child, reared as a girl, referred to us with ambiguous genitalia and premature pubarche with elevated serum cortisol. The genital ambiguity had been noticed at birth and the clitorophallus had been progressively increasing in length since birth. There was no history of prior hospitalisation or adrenal crisis.

Relevant clinical findings were:

Blood pressure: 102/58 mm Hg (systolic: 50th–90th percentile, diastolic: 50th–90th percentile), height: 112.6 cm (>97th percentile, +3 standard deviation score (SDS), height age: 6 years), weight: 18.6 kg (75th–97th percentile, weight age: 5.5 years). She had Prader stage 1 genitalia in form of clitoromegaly (clitoral length: 25 mm), no posterior labial fusion, separate urethral and vaginal opening. Clitoral index was 42 mm2; anogenital ratio was 0.45; Tanner stage was B1P2.

Investigations

Relevant investigations have been summarised in table 1.

Differential diagnosis

Isolated mineralocorticoid deficiency or resistance was not considered in any of these children because of genital ambiguity. The possible causes of high cortisol in a 46, XX child with ambiguous genitalia are 21-hydroxylase deficiency with prior hydrocortisone use, 21-hydroxylase deficiency and coexistent high cortisol-binding globulin (CBG) and primary glucocorticoid resistance syndrome (PGRS). Prior hydrocortisone use was meticulously ruled out in case 1 from history and cross-checking medical records. Case 2 was treatment naive. Absence of hypertension and elevated PRA pointed against PGRS in both these children. Low cortisol values in Roche Cobas e 411 and Siemens Immulite 1000 also rejected elevated CBG as a cause of high cortisol. Though we could not perform mutational analysis of CYP21A2, CAH due to 21-hydroxylase deficiency (classic salt-wasting variety in case 1 and classic simple virilising variety in case 2) was established after ruling out all other possibilities and using the principle of ‘Occam’s razor’ in medicine.

Treatment

Both these children were initiated with hydrocortisone (25 mg/m2/day), that gradually tapered to 10 mg/m2/day after normalisation of androgens. Both of them also received fludrocortisone. The dose is being adjusted to keep androgens (androstenedione and testosterone) in age and sex specific mid-normal range and 17OHP values between 5 ng/mL and 10 ng/mL. Parents of both children decided to raise them as women. Genital reconstruction surgery has been planned for a later date in case 1.

Outcome and follow-up

No further episodes of adrenal crisis recurred in case 1. General well-being, appetite increased and the child gained weight. Electrolytes and deranged hormonal parameters also normalised. Height velocity in case 2 decreased. Androgens and PRA were also normalised with hydrocortisone and fludrocortisone.

Discussion

Although liquid chromatography–mass spectrometry (LC–MS) is the gold standard for cortisol measurement, its limited availability and high cost are major barriers in resource-restricted settings.2 3 Serum cortisol is commonly measured by CLIA using different analysers like Siemens Advia Centaur XP, Roche Cobas e 411, Abbott Architect Plus and Siemens Immulite 1000. The antibodies used in these assays are non-specific and cross-react with structurally similar compounds like 17OHP, 21-deoxycortisol and 11-deoxycortisol. This interference may not be significant normally but in conditions where these precursors are grossly elevated, such cross-reactivity leads to clinically meaningful positive bias in cortisol measurements.4–6 The degree of cross-reactivity varies significantly in different analysers resulting in discordant cortisol values when same sample is measured in different platforms using CLIA. Cross-reactivity of cortisol with 17OHP and/or 21-deoxycortisol in Siemens Advia Centaur XP and Roche Cobas e 411 systems correlates with 17OHP and 21-deoxycortisol concentrations in the tested sample. Siemens Advia Centaur XP is more prone to overestimate serum cortisol compared with Roche Cobas e 411 in 21-hydroxylase deficiency, a condition with elevated circulatory 17OHP and 21-deoxycortisol concentrations.7

Unlike in girls, CAH often remains undiagnosed in boys during neonatal periods due to lack of genital ambiguity. Absence of low cortisol, when measured in Siemens Advia Centaur XP, may also deceive the clinician and result in therapeutic misadventure, even if 21-hydroxylase deficiency is suspected in those boys presenting with isosexual precocity. Moreover, it also leads to unnecessary investigations and futile attempts to search for alternative diagnosis of precocious puberty.

Another important issue is cross-reactivity of aldosterone with high levels of adrenal steroid precursors like 17OHP and 21-deoxycortisol in immunoassays, radioimmunoassay in particular.8 9 Children with isolated defects of the mineralocorticoid pathway, present with failure to thrive, hyponatraemia, hyperkalaemia, elevated PRA and low (in hypoaldosteronism) or elevated aldosterone (in pseudohypoaldosteronism (PHA)). Adrenal androgens are typically normal, while cortisol and ACTH are normal or high depending on the underlying stress at the time of evaluation of these conditions.

Boys with classic salt-wasting CAH admitted with low sodium, high potassium, metabolic acidosis, thus, were misdiagnosed as PHA because of normal/high cortisol and high aldosterone due to assay interference even in CLIA.10 Both of our patients have elevated cortisol when measured by Advia Centaur XP, but, low in Cobas e 411 and Immulite 1000 systems. Aldosterone in case 2 was measured by CLIA in Roche Cobas 6000 system and was low normal consistent with her diagnosis of 21-hydroxylase deficiency.

Interestingly, Siemens Immulite 1000 system uses most specific anticortisol antibodies and no cross-reactivity has been demonstrated either with 17OHP (at a concentration of 400 µg/dL) or 21-deoxycortisol (at a concentration of 500 µg/dL). Our findings are also concordant with this observation.

So, how do clinicians working in the resource-restricted settings and not having access to mutational analysis or LC–MS facilities confidently diagnose 21-hydroxylase deficiency, a disease frequently encountered in neonatal practice? Clinical suspicion remains the mainstay, supplemented by biochemical and hormonal parameters, measured by newer generation CLIA machines. Interestingly, Siemens Advia Centaur XP may also suggest the diagnosis if urinary cortisol is measured instead of serum cortisol. Urine in patients with 21-hydroxylase deficiency contains high 17-hydroxypregnanolone, pregnanetriol, pregnanetriolone, which are the urinary metabolites of 17OHP and 21-deoxycortisol. As per the users’ manual of Siemens Advia Centaur XP, no cross-reactivity has been detected between cortisol and 17-hydroxypregnanolone or pregnanetriol. So, despite falsely normal or high serum cortisol values in this platform, urinary cortisol is invariably low in classic 21-hydroxylase deficiency even if measured in Siemens Advia Centaur XP.

Physicians should be aware of the method of hormonal analysis, the analysers used and disease-specific analytical issues while performing hormonal assays to avoid misdiagnosis, a condition satirically termed as ‘endocrine laboma’. This knowledge is required to prevent unnecessary concerns, needless investigations and inappropriate treatment.

Learning points.

  • Chemiluminescent immunoassay-based measurement of serum cortisol in patients with 21-hydroxylase deficiency is prone to positive bias in different analysers used in routine practice. The interference is maximum in Siemens Advia Centaur XP, least in Roche Cobas e 411 and nil in Siemens Immulite 1000.

  • Salt-wasting variety of 21-hydroxylase deficiency and isolated mineralocorticoid defect share many clinical (dehydration, failure to thrive), biochemical (hyponatraemia, hyperkalaemia) and hormonal (elevated plasma renin activity) parameters. Salt-losing congenital adrenal hyperplasia may be misdiagnosed as pseudohypoaldosteronism due to spurious elevation of cortisol and aldosterone in older generation assays.

  • Such a misdiagnosis is not infrequent in men with salt-wasting crisis during neonatal periods. Ambiguous genitalia in 46, XX individuals points against isolated mineralocorticoid defect.

  • If measured cortisol value is discordant with the clinical diagnosis of 21-hydroxylase deficiency, cortisol needs to be reanalysed in different platforms with least interference. Elevated 17α-hydroxyprogesterone, adrenocorticotrophic hormone and adrenal androgens are other supportive clues to underlying 21-hydroxylase defect.

  • Clinicians, having access to Siemens Advia Centaur XP only, should measure 24-hour urinary cortisol, instead of serum cortisol, in an adequately collected sample to confirm cortisol deficiency in classic 21-hydroxylase deficiency.

Footnotes

Contributors: NA, PPC, AS and AM were involved in diagnosis and management of both the patients. NA and PPC did the literature search and wrote the manuscript.

Funding: The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

Competing interests: None declared.

Patient consent for publication: Parental/guardian consent obtained.

Provenance and peer review: Not commissioned; externally peer reviewed.

References

  • 1.Speiser PW, Azziz R, Baskin LS, et al. Congenital adrenal hyperplasia due to steroid 21-hydroxylase deficiency: an endocrine Society clinical practice guideline. J Clin Endocrinol Metab 2010;95:4133–60. 10.1210/jc.2009-2631 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Stanczyk FZ, Clarke NJ. Advantages and challenges of mass spectrometry assays for steroid hormones. J Steroid Biochem Mol Biol 2010;121:491–5. 10.1016/j.jsbmb.2010.05.001 [DOI] [PubMed] [Google Scholar]
  • 3.Keevil BG. Lc-Ms/Ms analysis of steroids in the clinical laboratory. Clin Biochem 2016;49:989–97. 10.1016/j.clinbiochem.2016.04.009 [DOI] [PubMed] [Google Scholar]
  • 4.Haddad RA, Giacherio D, Barkan AL. Interpretation of common endocrine laboratory tests: technical pitfalls, their mechanisms and practical considerations. Clin Diabetes Endocrinol 2019;5:12. 10.1186/s40842-019-0086-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Klee GG. Interferences in hormone immunoassays. Clin Lab Med 2004;24:1–18. 10.1016/j.cll.2004.01.003 [DOI] [PubMed] [Google Scholar]
  • 6.Krasowski MD, Drees D, Morris CS, et al. Cross-Reactivity of steroid hormone immunoassays: clinical significance and two-dimensional molecular similarity prediction. BMC Clin Pathol 2014;14:33. 10.1186/1472-6890-14-33 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hawley JM, Owen LJ, Lockhart SJ, et al. Serum cortisol: an up-to-date assessment of routine assay performance. Clin Chem 2016;62:1220–9. 10.1373/clinchem.2016.255034 [DOI] [PubMed] [Google Scholar]
  • 8.Balcells C, Gili T, Pérez J, et al. Pseudohypoaldosteronism without nephropathy masking salt-wasting congenital adrenal hyperplasia genetically confirmed. BMJ Case Rep 2013;2013:bcr2012008281. 10.1136/bcr-2012-008281 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Tuhan HU, Catli G, Anik A, et al. Cross-reactivity of adrenal steroids with aldosterone may prevent the accurate diagnosis of congenital adrenal hyperplasia. J Pediatr Endocrinol Metab 2015;28:701–4. 10.1515/jpem-2014-0170 [DOI] [PubMed] [Google Scholar]
  • 10.Boddu SK, Madhavan S. High aldosterone and cortisol levels in salt wasting congenital adrenal hyperplasia: a clinical conundrum. J Pediatr Endocrinol Metab 2017;30:1327–31. 10.1515/jpem-2017-0166 [DOI] [PubMed] [Google Scholar]

Articles from BMJ Case Reports are provided here courtesy of BMJ Publishing Group

RESOURCES