Steroid hormones are among the most frequently measured analytes in clinical and research laboratories. Testosterone, estradiol, cortisol, progesterone and their precursors underpin investigations in endocrinology, fertility, paediatrics and oncology. For most of the past few decades, they have been measured by immunoassay: fast, automated, and available on high-throughput analysers in almost every laboratory.

The difficulty is that steroids are small molecules that share the same four-ring backbone. Many differ by little more than a hydroxyl group or a double bond. An antibody designed to recognise one steroid can also bind its relatives, its metabolites, or a drug with a similar structure. At high concentrations, the error is often small. At low concentrations, it can dominate the result.

Where immunoassay struggles

Low concentrations. Testosterone in women and children, and estradiol in men and postmenopausal women, sit at concentrations where many immunoassays lose accuracy. These are the groups where a small absolute error matters most.

Cross-reactivity with drugs. Fulvestrant, a breast cancer treatment, is structurally similar to estradiol. Several manufacturers have issued field safety notices warning that it can cause falsely elevated results in their estradiol immunoassays, which could lead clinicians to wrongly assess a patient’s menopausal status and alter treatment. The TGA, EMA and US FDA have all updated the product labelling to warn of the interference. Manufacturers recommended LC-MS as an alternative method, since it is not expected to cross-react with the drug.

Newborn screening. Screening for congenital adrenal hyperplasia (CAH) relies on measuring 17-hydroxyprogesterone in dried blood spots, usually by immunoassay, an approach known for high false positive rates. Many programmes now use LC-MS/MS steroid profiling as a second-tier test on the same blood spot, which has substantially reduced false positives where it has been adopted. Closer to home, New Zealand has implemented second-tier LC-MS/MS testing for CAH.

What LC-MS/MS does differently

LC-MS/MS identifies a steroid by three independent properties: its retention time on the column, its precursor mass, and the mass of its fragments. Together these give far greater specificity than antibody recognition alone.

It also allows isotope dilution. A stable isotope-labelled version of each steroid is added to every sample at the start, correcting for losses during extraction and for matrix effects in the source.

And it measures panels. A single run can quantify several steroids at once, such as 17-hydroxyprogesterone, androstenedione and cortisol. The ratios between them often say more than any single value, which is why steroid profiling has become so useful in newborn screening and in investigating adrenal disorders.

Standardisation has followed. The CDC Hormone Standardization Program, which uses LC-MS/MS reference methods, certifies testosterone and estradiol measurements against defined accuracy and precision criteria.

A shift with some nuance

The move towards mass spectrometry has been strongly encouraged, but not mandated. In 2007, the Endocrine Society recommended mass spectrometry-based methods as the reference for measuring testosterone. In 2013, the Journal of Clinical Endocrinology & Metabolism went further, announcing it would require mass spectrometry for sex steroid measurements in published research. That requirement was suspended before it took effect, pending review.

The Endocrine Society task force that followed settled on requiring assays to be demonstrably valid and reliable, without mandating any particular method. Discussion at the time noted that mass spectrometry was not yet in general use, and that rigorously validated immunoassays can still give accurate results.

That nuance matters, because LC-MS/MS is not automatically better. It brings its own challenges:

  • Isobaric steroids. Several steroids share the same mass and fragment in similar ways, such as 11-deoxycortisol, 21-deoxycortisol and corticosterone. They have to be separated chromatographically, or the specificity advantage is lost.
  • Agreement between laboratories. Results from different LC-MS/MS methods do not always agree. A European comparison across nine centres found good agreement for some steroids, such as testosterone, but poor agreement for others, including cortisol and 17-hydroxyprogesterone. Traceable calibrators and participation in external quality assessment help, but each method’s own performance matters as much as its calibration.
  • Throughput and expertise. Immunoassay analysers are fully automated and run by general laboratory staff. LC-MS/MS methods usually involve more sample preparation and need staff with specialist experience.

Where this could be heading

For the low-concentration applications and look-alike molecules where immunoassays are weakest, LC-MS/MS is increasingly the method of choice. As sample preparation becomes more automated and more laboratories gain experience with the technique, steroid panels by mass spectrometry may become more widely available beyond specialist and reference laboratories. Immunoassays are likely to remain part of routine testing for some time, with the choice of method depending on the analyte, the concentration and the clinical or research question.