Developing a urinary steroid profiling service within the NHS: a personal journey recorded by Dr Norman Taylor

Nothing prepared me for 46 years (so far) of analysing steroids, a period which is more than half of the time since the first steroid was crystallised from human urine in 1929. 

My early years
My birth year of 1947 was just months before the start of the NHS. I remember my parents' passionate support of its principles and with gratitude the local GPs, who would consent to visit for every childhood illness. Aged nine, I spent a week in hospital with appendicitis, returning home with a proud scar and a determination to become a doctor. Soon after, my best friend developed a sharp pain and after checking him and asking questions, I pronounced that he had appendicitis, which he did, but my scar was neater than his! I didn’t make it to medical school, so turned to applied biology. After an HND and 2 years of voluntary work in Africa, I returned to find jobs suddenly scarce but landed one at St Thomas’ Hospital as a technician tasked with analysing a single steroid. One year on, I transferred to the newly opened Clinical Research Centre where Cedric Shackleton had just returned from a post-doc and had started a lab for multicomponent steroid analysis. He coined the term ‘urine steroid profiling’. I have been practising it ever since.

The early years of steroid testing
Steroids first came to the fore in the modern era during the study of soluble physiologically-active agents that could be extracted from the adrenals, ovaries and testes. The thirties and forties saw crystallisation of individual steroids, testing of hormonal activity and structural determination. Steroid synthesis in quantities sufficient to be useful therapeutically began in the early fifties, revolutionising several areas of medicine, including moderation of the immune system, assisting fertility and in contraception. The first inborn error of steroid metabolism was fully characterised at this time. This needed methods for the analysis of natural steroids in body fluids. Although steroid hormones in blood were the major focus, urinary metabolites proved to be very informative in providing a composite picture of steroid metabolism. This remains true to this day, with all significant new findings on steroid metabolism still originating with observations in urine. The continuing challenge is to separate and detect individual steroids from within a ‘soup’ containing an uncountable number of related compounds. 

During a time of rapid development of new techniques, Cedric Shackleton went through chromatography on paper, thin layer plates, and finally gas chromatography-mass spectrometry (GC-MS) within a mere 4 years (1966-70). He was the first in the UK to use capillary column gas chromatography for steroid analysis: it provided enormously improved resolution. By the time I joined him in 1972, these techniques were all in place and have not changed in principle since. Over the next few years, along with my co-worker, John Honour (who went on to run the other Supraregional Assay Service for this within England), disorder after disorder came under our scrutiny. We soon realised that each had its own ‘fingerprint’ and were able to identify all the major components, very much helped by the resources of the Medical Research Council Steroid Reference Collection, which kept hundreds of crystalline pure steroids and provided small amounts for identification purposes. We found many previously undescribed steroids. In 1980, we jointly published an atlas of steroid profiles to aid steroid disorder identification by others. 

An evolving service
A service to provide steroid profiling on patients was never systematically planned, but simply grew from a creative interaction with clinicians and clinical biochemists. The organisation and ethos of the NHS has been critical: we continue to learn together in a free, non-hierarchical exchange. Few, if any, countries in the world can match the UK for this. Among the great benefits of GC-MS is that it can be used non-selectively: whatever is there in greatest amounts will be seen. All the data are electronically stored and can always be interrogated, so steroids that were not at first considered significant can be looked for if new information emerges. The combination of ‘fingerprint’ and searchability has also meant that, with careful record-keeping, we can use previous findings on a given disorder to aid identification of new cases. A highly developed database helps in finding parallels, such as locating other patients who are being treated with the same drug. It also enables us to identify previous encounters with a patient and so provide continuity: trends in single patients can be monitored and earlier findings made known to a clinician who may not be aware of them. 

Using GC allied with tandem mass spectrometry (GC-MS/MS) increases sensitivity and specificity and has an incredible and still largely unrealised potential to delve further into steroid metabolism. A doctoral project by Sofia Christakoudi in our laboratory revealed around 450 different steroid metabolites in newborns with 21-hydroxylase deficiency, of which only 15 or so had been previously described in the literature1-6.

Serum steroid analysis does, of course, remain the mainstay of clinical investigation, for which immunoassay reigns supreme. 

Some of its limitations of specificity are well known, such as when quantifying testosterone in women, whilst others are less familiar. For instance, steroids at pathologically high levels may unexpectedly interfere, an example being when serum from newborns with severe 21-hydroxylase deficiency shows a ‘normal’ serum cortisol concentration when in fact the true value is very low. 

From urine steroid profiling to serum steroid panelling
Liquid chromatography-tandem mass spectrometry (LC-MS/MS), which is far more specific than immunoassay, is now coming into use, but is not yet available on automatic analysers. It has mostly been used to directly replace single steroid assays, but it is perfectly possible to quantify many steroids simultaneously. We now have a method, developed by David Taylor and Lea Ghataore7, that targets all the major intermediates in the steroid hormone synthesis pathways. This will be launched as a service soon. Since the technique requires prior setting up for the particular steroid targets, it should be called serum steroid panelling and not profiling. Because many steroids have the same general structure and so have identical molecular weights, differentiating the critical ones is a challenge. Comparing findings in urine and serum on single patients has already revealed errors in results obtained by LC-MS/MS by other services8.  The learning continues… 

For further information on Viapath’s Steroid Analysis, please contact:
Dr Norman Taylor
Email: norman [dot] taylor1 [at] nhs [dot] net

References 

1. Christakoudi S, Cowan D A, Taylor N F. Steroids excreted in urine by neonates with 21-hydroxylase deficiency: Characterization, using GC–MS and GC–MS/MS, of the D-ring and side chain structure of pregnanes and pregnenes. Steroids (2010) 75 34-52 Epub 2009 Sep 30
2. Christakoudi S, Cowan D A, Taylor N F. A new marker for early diagnosis of 21-hydroxylase deficiency: 3,16,17-trihydroxy-5-pregnane-7,20-dione. J Steroid Biochem Mol Biol (2010) 121 574-581 Epub 2010 Mar 17
3. Christakoudi S, Cowan D A, Taylor N F. Steroids excreted in urine by neonates with 21-hydroxylase deficiency: 2.  Characterization, using GC-MS and GC-MS/MS, of pregnanes and pregnenes with an oxo group on the A- or B- ring. Steroids (2012)  77 382-393 Epub 2011 Dec 21
4. Christakoudi S, Cowan D A, Taylor N F.  Steroids excreted in urine by neonates with 21-hydroxylase deficiency. 3. Characterization, using GC-MS and GC-MS/MS, of androstanes and androstenes. Steroids. (2012)  77 1487-1501 Epub 2012 Sep 3
5. Christakoudi S, Cowan D A, Taylor N F.  Steroids excreted in urine by neonates with 21-hydroxylase deficiency. 4. Characterization, using GC-MS and GC-MS/MS, of 11oxo-pregnanes and 11oxo-pregnenes. Steroids. (2013) 78 468-475 Epub 2013 Feb 26
6. Christakoudi S, Cowan D A, Christakoudi G, Taylor N F. 21-Hydroxylase deficiency in the neonate – Trends in steroid anabolism and catabolism during the first weeks of life J Steroid Biochem Mol Biol (2013) 138 334-347 Epub 2013 Jul 31
7. Taylor DR, Ghataore L, Couchman L, Vincent RP, Whitelaw B, Lewis D, Diaz-Cano S, Galata G, Schulte K-M, Aylwin S, Taylor NF   A 13-steroid serum panel based on LC-MS/MS: use in detection of adrenocortical carcinoma. Clinical Chemistry (2017) 63 1836-1846 Epub 2017 Sept 13
8. Boughton C, Taylor D, Taylor N, Whitelaw BC  Mineralocorticoid hypertension and hypokalaemia induced by posaconazole.  Endocr Diabetes Metab (2018)  DOI 10.1530/EDM-17-0157 Epub 2018