American Association for Clinical Chemistry
Better health through laboratory medicine
December 2008 Clinical Laboratory News: Endocrine Testing


December 2008: Volume 34, Number 12

Endocrine Testing
Improving Patient Care with Tandem Mass Spectrometry
By Ravinder Jit Singh, PhD


Series Article Image for Dec 2008 CLN 

The current focus on patient safety has increased emphasis on the need for accurate and precise lab test results. To deliver high quality results for steroids in a timely manner, labs use a wide array of technologies, including competitive radio-immunoassay (RIA), chemiluminescence immunoassay (CLIA), and fluorimetric immunoassay (FIA). Although these immunoassays are well suited for small–molecular-weight compounds, sandwich immunoassays, such as ELISAs, are the preferred methods for proteins and small peptide hormones. Most of these assay formats have been adapted to commercial platforms, which are highly automated and allow labs to process many patient specimens per hour.

But FDA-cleared immunoassays are not available for all endocrine analytes that labs need to measure, for example catecholamines and metanephrines. For these analytes, large labs commonly use gas (GC) or high-pressure liquid chromatographic (HPLC) techniques coupled with various detection methods, such as flame ionization, ultraviolet light, electrochemical or mass spectrometry. In the past decade, HPLC coupled with tandem mass spectrometry (LC-MS/MS) has revolutionized measurement of endocrine analytes. In addition, LC-MS/MS allows labs to conveniently measure new biomarkers for which no commercial assays exist.

The most significant advance in MS technology that has facilitated routine analyses in clinical labs is the invention of an electropsray source by Nobel laureate John B. Fenn, PhD. This technology facilitates ionization of the analytes present in liquid droplets and sprays the molecules directly into the mass spectrometer from the HPLC. With this advancement, labs can achieve greater throughput of patient samples compared to GC-MS. Now considered the gold standard for measuring steroid hormones, LC-MS/MS analysis offers improved precision and accuracy (1).

Over the last few years, we have implemented this technology at the Mayo Clinic for routine analysis of steroids and now perform more than a million endocrine tests per year. In addition to being a reliable method for tests related to endocrine disorders, some labs and accrediting agencies now use LC-MS/MS as a reference method. This article examines diagnostic applications of the technology for several endocrine disorders.

Cushing’s Syndrome

Cortisol, a glucocorticoid hormone, plays an essential role in adaptation to stress, regulation of metabolism, and inflammatory responses. Hypercortisolism is associated with the rare condition known as Cushing’s syndrome and has been linked to hypertension, diabetes, and obesity. It is frequently difficult to distinguish mild or moderate hypercortisolism, so called pseudo-Cushing’s syndrome, from full blown Cushing’s.

Endogenous hypercortisolism can result from either an ACTH-secreting pituitary tumor or a cortisol-secreting tumor in the adrenal gland. A small molecular weight molecule, cortisol’s structure is very similar to other endogenous and exogenous steroids, resulting in artificially high values in immunoassays.

Analysis of urinary-free cortisol (UFC) is most commonly used for the diagnosis of Cushing’s syndrome; however, physicians also order plasma cortisol, plasma-free cortisol, midnight plasma cortisol, and midnight salivary cortisol tests to confirm the diagnosis. Measurements of UFC and its metabolite, cortisone, are useful in evaluating apparent mineral-corticoid excess, congenital adrenal hyperplasia, and adrenal insufficiency.

Historically, labs have used RIA for analysis of UFC and plasma cortisol. Today, labs have replaced RIA with automated non-radioactive CLIA. Even with an extraction step to eliminate polar compounds, these assays are prone to interferences from other endogenous steroid metabolites and exogenous synthetic glucocorticoids and can produce false-positive results (Figure 1).

Figure 1
Correlation Between Cortisol Assay Methods

Urinary-free cortisol data from 24-hour urine specimen of different patients. IA = immunoassay

Comparisons between immunoassay and chromatographic methods have been reported for UFC measurements, and researchers have concluded that UFC immunoassays are not highly precise or accurate. Therefore, more specific methods have been developed using LC-UV, LC-MS and GC-MS. These chromatographic methods have reduced interference for cortisol quantification and allow quantification of cortisone, an endogenous metabolite of UFC.

The LC-UV method for cortisol and cortisone however, require a lengthy analysis time to obtain resolution between cortisol and cortisone. The analysis must also ensure that the commonly used synthetic corticosteroids and the more hydrophilic cortisol metabolites do not interfere with the cortisol and cortisone peaks (2).

Despite its superiority to immunoassays, LC-UV analysis is still prone to interferences, most notably from carbamazepine and its hydroxy metabolites. To resolve carbamazepine interference, some labs have turned to LC-MS or GC-MS analysis. Although chromatographic methods with a single MS detector provide specific quantitation of cortisol, these methods have not been widely implemented due to low throughput and higher instrument cost. However, triple-quad, tandem mass detectors have provided exponentially better throughput for endocrine testing, and the cost per sample is now comparable to commercial immunoassays.

Sex Steroids

In vivo formation of cortisol, aldosterone, and sex steroids involves cleavage of cholesterol, followed by minor oxidations and reductions at various carbon sites, which results in various steroid intermediates. These steroid intermediates are very similar in structure and chemical properties, presenting an analytical challenge for various methods used in clinical labs.

Inherited defects in steroid biosynthesis, in particular 21-hydroxylase deficiency, results in a disease known as congenital adrenal hyperplasia (CAH). This enzyme deficiency increases levels of an intermediate steroid, 17α-hydroxyprogesterone (17-OHP), a precursor for androgens (testosterone) and a cause of virilization of females.

Steroid intermediates and metabolites have been reported to cross-react with the immunoassay reagents, and in particular, labs have reported calibration issues for 17-OHP immunoassays. Chromatographic separation and detection using tandem mass spec, however, eliminates cross reactants and allows for specific determination of sex steroids.

To confirm the diagnosis of CAH, it is critical to determine multiple steroid intermediates in a single blood sample collected from the patient. Analysis of individual analytes by RIA or CIA can be very time consuming and expensive, but with LC-MS/MS, labs can measure multiple steroids in small-volume serum samples, which is particularly advantageous for pediatric samples. (Figure 2) (3).

Figure 2
Multiple Steroid Analysis by LC-MS/MS

A 100-µL serum sample was processed using solid-phase extraction and injected into an LC-MS/MS. The steroid peaks are cortisone (1), cortisol (2), 21-dexycortisol (3), cortiscosterone (4), 11-deoxycortisol (5), androstenedione (6), dexycortiscosterone (7), 17-hydroxy progesterone (8), progesterone (9), and pregnenolone (10).

Estrogen immunoassays also present analytical issues for labs, and recently the quality of the epidemiologic data collected from estrogen immunoassays has been questioned. Using these assays, researchers have reported variable 17 β-estradiol serum levels in postmenopausal women, and median normal values by these methodologies differ by approximately 6-fold (4). Therefore, some labs have developed tandem MS based assays to increase sensitivity and specificity of these assays. Because these analyses require expensive instruments and highly trained personnel, only large reference labs currently offer this type of analysis.

Similar issues exist for testosterone testing, especially for results from women and children (5, 6). Current assays lack accuracy and precision, and they have large interlaboratory variation in these patient populations. Large reference labs have adopted the LC-MS/MS methodology to address some of these concerns. Although the LC-MS/MS assays provide testosterone results with good precision and accuracy, improvements need to be made to increase the sensitivity and precision at the lowest detection limits for children and women.

Labs typically develop in-house LC-MS/MS assays for steroid hormones. These assays have labor-intense, manual steps that can produce interlaboratory variations. Some reference labs also claim their procedures are proprietary. To maintain assay quality, CAP and the New York State Department of Health offer proficiency programs for sex steroids. In a recent survey, the New York State Department of Health reported the CVs to be >20% among the labs performing testosterone analysis by MS/MS methods (7). They concluded that although LC-MS/MS instruments are readily available, improvements in comparability are needed, and a high-level reference-method protocol should be developed through formal interlaboratory collaboration. In a recent position statement, endocrine experts also cited problems in proficiency testing for testosterone (8).


Pheochromocytoma is a rare but potentially fatal tumor arising from chromaffin cells. It can produce episodic secondary hypertension, along with headaches, sweating, and palpitations. Screening for pheochromocytoma is typically part of an evaluation for secondary causes of hypertension, unexplained fainting spells, incidental adrenal masses, or less commonly, for patients with a family history of pheochromocytoma. Patients with pheochromocytoma can present with adrenal incidentaloma, hypertensive paroxysms, sustained apparent polygenic hypertension, hypertension in pregnancy, and hypertensive crisis induced by anesthesia. Although pheochromocytoma is lethal, it can usually be cured with surgery.

Biochemical testing for pheochromocytoma typically has included measurements of metanephrines and catecholamines. But analysis of plasma concentrations of free metanephrines is challenging, since less than 5% exist in an unconjugated state. Researchers recently reported that measurement of fractionated, plasma-free metanephrines by HPLC with electrochemical detection (HPLC-EC) had 100% sensitivity and 89% specificity for detecting pheochromocytoma (9).

Other methods for analysis of metanephrine and normetanephrine include colorimetric assays, immunoassays, HPLC, and GC-MS. Drug interferences and the lack of an internal standard limit the utility of the colorimetric assay. Although new immunoassays for metanephrines have been shown to be free of drug interference, they still lack an internal standard to monitor recovery through the extraction process. Recent modifications in HPLC methods have resolved known drug interferences for metanephrines, but analytical run times have been increased. To overcome drug interferences, researchers have also developed an isotope-dilution GC-MS method, which is specific but requires a time-consuming derivatization step and takes longer to run. Alternatively, some labs have successfully implemented LC-MS/MS methods that use stable deuterium-labeled isotopes of metanephrines and normetanephrine and have good throughput.

The LC-MS/MS assay has several advantages. It is sensitive and offers automated on-line extraction and high-throughput processing of samples. In addition, the method can measure the dopamine metabolite, methoxytyramine, an added utility for detection of dopamine-producing paragangliomas. The capability to detect all three O-methylated metabolites in as little as 50 µL of plasma or urine also makes the assay suitable for diagnosis of other neuroendocrine tumors—particularly neuroblastomas (10).

Vitamin D

Experts now project a vitamin D deficiency epidemic in North America. Labs most often measure serum 25-hydroxy vitamin D (25-OH-D), an accepted marker for vitamin D nutritional status. With this increased interest, more than 5 million 25-OH-D tests are expected to be performed in the U.S. this year.

The methods currently used for 25-OH-D include low-throughput assays, such as HPLC-UV and RIA, or high-throughput automated CLIA and LC-MS/MS assays. Although various methods are available for measuring circulating concentrations of 25-OH-D, none of these assays are standardized against a common calibrator. Surveys from CAP and the UK-based Vitamin D External Quality Assessment Scheme provide independent approaches to monitor the performance of laboratories that use various methods for testing of 25-OH-D. Based on this data, the CV for the same method has been reported to be >20% among labs. Recent CAP data also indicate that labs performing immunoassays report results ranging from 41 to 96 µg/L for a survey sample with a value of 75 µg/L determined by LC-MS/MS (Figure 3). There could be many reasons for these variations, including drifts in calibrator reagents. Regardless of the reason, there is a clear and urgent need for harmonization and standardization (11).

Figure  3
Accuracy of Vitamin D Assays

Shown here are the 2007 CAP survey data for labs participating in 25-OH-D proficiency testing. The range of results for survey materials BGS-01 through BGS-04 is shown for labs using either RIA (solid lines, n = 16) or automated CLIA (broken lines, n = 18). A single lab used LC-MS/MS (closed circles). (12)

Reprinted from Clinical Chemistry with permission of AACC, Washington, DC.

Calcium homeostasis is also frequently monitored in vitamin D deficient patients. Fortunately, the performance of calcium tests by most manufacturers is very good and has a CV of <1%. In order to deliver high quality vitamin D results, however, labs need to have vitamin D tests with similar precision. NIST is currently developing quality control materials (human serum, SRM 972) to deal with this problem. The control material will contain 25-OH-D2, 25-OH-D3, and the metabolite 3-epi-25-OH-D at four different concentrations, as characterized by LC-MS/MS. Preparation of this SRM is especially important for immunoassays for which the cross-reactivity with 25-OH-D2 and 25-OH-D3 is not well defined.

Endocrine Testing: Future Challenges

Immunoassays allow for the sensitive detection of a wide range of endocrine hormones. They are very precise, come in a wide range of formats, and most importantly, are highly automated. This familiar assay format has been used to establish reference ranges for a variety of endocrine analytes, and the problems and limitations of the reagents used in these assay formats are also widely understood in clinical labs. As labs have gained experience with MS analysis, investigators have reported discrepancies between results for this method and those of immunoassays, especially at the low end of the concentration range for steroid hormones, creating a heightened awareness of the limitations of immunoassays.

But LC-MS/MS technology is relatively new to clinical labs. Its problems and challenges are not as well understood for routine clinical analysis, and only a small number of labs have adopted this highly specific technology. Recently, CDC, in partnership with NIST and the Endocrine Society, initiated a project to standardize assays for endocrine hormones. The goal of this effort is not only to help clinicians and laboratorians better understand MS analysis, but also to work with the industry to standardize and improve endocrine immunoassays. These efforts are clearly worthwhile and will undoubtedly help improve the quality of these lab tests.


  1. Kinter M. Toward broader inclusion of liquid chromatography-mass spectrometry in the clinical laboratory. Clin Chem 2004;50:1500–1502.
  2. Taylor RL, Machacek D, Singh RJ. Validation of a high-throughput liquid chromatography-tandem mass spectrometry method for urinary cortisol and cortisone. Clin Chem 2002:48;1511–1519.
  3. Guo TD, Taylor RL, Singh RJ, et al. Simultaneous determination of 12 steroids by isotope dilution liquid chromatography-photo spray ionization tandem mass spectrometry. Clin Chim Acta 2006;372:76–82.
  4. Nelson RE, Grebe SK, O'Kane DJ, et al. Liquid chromatography-tandem mass spectrometry assay for simultaneous measurement of estradiol and estrone in human plasma. Clin Chem 2004;50:373–384.
  5. Albrecht L, Styne D. Laboratory testing of gonadal steroids in children. Pediatr Endocrinol Rev 2007;5 Suppl 1:599–607.
  6. Kane J, Middle J, Cawood M. Measurement of serum testosterone in women; what should we do? Ann Clin Biochem 2007;44:5-15.
  7. Cao Z, Soldin S, Rej R. Poor interlaboratory agreement of testosterone measurements using HPLC-tandem mass spectrometry. Clin Chem 2008;54 Suppl S:A113.
  8. Rosner W, Auchus RJ, Azziz R, Sluss PM, Raff H. Position statement: Utility, limitations, and pitfalls in measuring testosterone: An endocrine society position statement. J Clin Endocrinol Metab 2007;92:405–413.
  9. Taylor RL, Singh RJ. Validation of liquid chromatography-tandem mass spectrometry method for analysis of urinary conjugated metanephrine and normetanephrine for screening of pheochromocytoma. Clin Chem 2002;48:533–539.
  10. Singh RJ. Eisenhofer G. High-throughput, automated, and accurate biochemical screening for pheochromocytoma: are we there yet? Clinical Chemistry 2007;53:1565–7.
  11. Carter GD, Carter R, Jones J, Berry J. How accurate are assays for 25-hydroxyvitamin D? Data from the international vitamin D external quality assessment scheme. Clin Chem 2004;50:2195–2197.
  12. Singh RJ. Are clinical laboratories prepared for accurate testing of 25-Hydroxy vitamin D? Clin Chem 2008;54:221–223.

Ravinder Jit Singh, PhD is co-director of the endocrine laboratory at Mayo Clinic, Rochester, Minn.