FIRST PUBLISHED IN ISSUE 111 (2014)

 

Hormones have been measured by immunoassay for more than four decades, since Berson and Yalow discovered that antibodies could be used as diagnostic agents. They were awarded a Nobel Prize and, altruistically, did not patent their findings.

However, the story of hormone assays begins earlier, with a variety of in vivo biological tests. Subsequently, colourimetric assays were discovered and used to good effect, particularly for cortisol. Since 1970, immunoassays have become increasingly routine and, after a period of dominance within the clinical laboratory, it is only now after a further 40 years that a new technology, tandem mass spectrometry (TMS), is starting to emerge as the method of choice.

EARLY TESTS USING BIOLOGICAL EFFECTS OF HORMONES – BIOASSAYS
The early bioassays included the chick cockscomb test. Androgen-like material was applied to the cockscomb of a chick, and the increase in size was used as a measure of testosterone. One can only imagine the errors that might have been introduced while trying to hold the animal still in front of a lamp, whilst the shadow of the comb was outlined on a screen to measure its size.

Another qualitative test used the Xenopus toad, which ovulated in response to human chorionic gonadotrophin injections, and was used as an early pregnancy test. Probably more precise was the ovarian weight augmentation assay in immature rodents, used to measure follicle-stimulating hormone (FSH).

DEVELOPMENT OF MODERN IMMUNOASSAYS
Early immunoassays using antibodies from immunised animals with 125I-labelled hormones in competitive assays were very welcome. These assays were time-consuming. In the 1970s, luteinising hormone and FSH assays took 5 days, from pipetting the samples on a Monday through to calculation and reporting of results on a Friday, but by the late 1980s a batch of 30 samples could be measured in a day.

In those days, the assays were all built in-house, and our older colleagues will remember the ‘hot labs’ where they iodinated their ligands, and the local university animal house where they venesected animals for their polyvalent antibodies. I remember being regaled with tales of a colleague being flung skyward at the university farm by an angry ram who was in no mood to donate his blood to science.

The antibodies were not always very precise in their detection of specific molecules, hence preparatory separation of serum samples was required, using techniques such as paper and column chromatography or organic solvent extraction to increase the specificity of the antigen available to be bound. These methods are very time-consuming and require considerable technical skill. This issue of specificity was addressed by the next milestones.

BENEFICIAL REFINEMENTS
First we saw the development of monoclonal antibody production from hybridomas (associated with another Nobel Prize). Secondly there was the development of non-isotopic labels, such as enzymes, luminescence, delayed fluorescence, polarisation fluorescence etc. Thirdly we benefited from the merging of biological with mechanical sciences as robotic instruments were developed. These instruments can pipette faster and more precisely than humans and, equally importantly, could ensure that incubations were precisely timed, so that assay numbers could be increased from 30 per day to several hundred per day without any assay drift across batches.

The increase in productivity comes with a cost. Direct assays for steroid molecules are still prone to interference from other molecules. Monoclonal antibody kits from different commercial providers are proprietary agents, and so there are variations between methods that persist even with international reference preparations. Moreover, monoclonal antibodies may not necessarily be the best tool for measuring peptide hormones that exist with many glycoforms and oligomeric forms.

Immunoassays have developed outside the clinical and research laboratory environment and are now used by clinical staff as near patient tests, by field toxicologists for environmental poisons, and by the lay public as pregnancy and HIV tests.

The existence of interference by molecules similar to the ones under investigation has already been mentioned but, over the years, every time an assay for a novel analyte has been produced, it has been followed by reports of antibody interference. These are usually only noted when assay results clearly diverge from the clinical picture, but our group has shown that more subtle interference occurs in at least 1:200 patient samples. The interference is quite promiscuous and not limited to a single analyte.

TANDEM MASS SPECTROMETRY – THE FUTURE?
The scaling down of mass spectrometers from the size of a double bedroom to a desk top instrument has permitted the introduction of this technology to the clinical laboratory. TMS using quadrupoles is only useful for the measurement of small molecules such as steroids, drugs and intermediary metabolites, and more sophisticated instruments are necessary for peptide hormone measurement. It is remarkable that, despite it being less than 10 years since the first reports of the use of TMS in endocrinology, Journal of Clinical Endocrinology & Metabolism has stated that it will only accept papers that use TMS methods for steroid analysis in future.

TMS is presently a sophisticated technique that needs skilled staff, but it will undoubtedly become a more friendly technique in the future. After all, mass spectrometry is used in airport security for the detection of volatile explosives, so it is only time before this comes to routine clinical laboratories. Inevitably, as TMS is more widely used, newer types of analytical problem will be found. We are already aware of interference by substances that co-elute in the preparatory columns and by epimers and structural isomers. So watch this space…

Endocrinology has developed in tandem with immunoassays, and both sciences have progressed in leaps and bounds since the seminal work of Berson and Yalow in 1959. The next game changer will be the analysis of hormones in real time by the patient.

JULIAN H BARTH
Consultant Chemical Pathologist, University of Leeds (correct at the time of first publication)