Often considered a magic bullet for analytical measurements, liquid chromatography tandem mass spectrometry (or simply LC-MS/MS) has revolutionised clinical chemistry departments up and down the country. Since its earliest application for monitoring immunosuppressant drug concentrations in transplant patients, LC-MS/MS has expanded to provide services for many clinical disciplines, with large, multi-analyte profiles now available.
Here, we provide a brief introduction to the key stages in LC-MS/MS, before discussing the state of play in the UK for selected endocrine assays, including those for oestradiol, 5-hydroxyindoleacetic acid (5HIAA), metanephrines, cortisol and dexamethasone. Finally, we look at what the future holds.
'Since its earliest application for monitoring immunosuppressant drug concentrations in transplant patients, LC-MS/MS has expanded to provide services for many clinical disciplines, with large, multi-analyte profiles now available.'
WHAT EXACTLY IS LC-MS/MS?
Broadly, the process of generating LC-MS/MS results can be thought of as three distinct processes, all of which must be given the appropriate attention to produce meaningful, accurate results.
The first of these is sample preparation. The choice here is dependent on several factors, including the concentration of the target analyte (nmol/l versus pmol/l), sample matrix (serum, urine, saliva), turnaround time, the chemical properties of the analyte and, last but not least, the cost of the consumables. In general, the more challenging the analyte is to measure, the greater the sample volume and the more intensive the sample clean-up required.
Next, we turn our attention to liquid chromatography. This helps separate the analyte of interest from other compounds, based on the analyte’s physicochemical properties. It is paramount to ensure that the analyte of interest is separated from potential interferents and is not compromised by matrix effects, which can be the Achilles’ heel of many methods. Whereas matrix effects can be difficult to predict, attention to detail can go some way to ensure your target analyte is not interfered with by another compound.
It is practically impossible to distinguish isobaric compounds (i.e. those of the same molecular weight) from one another using only mass spectrometry. This is especially true of steroid hormones, where the masses are commonly the same as neighbouring precursors or metabolites. In these situations, chromatography really comes into its own, and is an essential component of developing specific methods.
The final stage, of course, involves the mass spectrometer itself. Although it is simply a detector, it is a very sensitive one. Being reliant on mass-to-charge, it holds a distinct advantage over other approaches (e.g. assays reliant on antibody–antigen reactions). Furthermore, with tandem mas spectrometry, you have the ability to fragment the analyte of interest to look for products of your target analyte. This adds another layer of assurance that what you are detecting is what you believe it to be.
When these processes are combined, you can develop robust and accurate assays that are clinically useful. There is a huge interest in LC-MS/MS within the UK, and clinical laboratories are increasingly active in assay development. In case you may have missed it, the following are some of the major applications that have become available.
'There is a huge interest in LC-MS/MS within the UK, and clinical laboratories are increasingly active in assay development.'
Sensitive oestradiol measurements may be required to help investigate precocious puberty, gynaecomastia, post-menopausal females and patients taking aromatase inhibitors. This measurement can prove particularly problematic for immunoassays as, with functional sensitivities of approximately 100pmol/l, they simply don’t measure low enough to be informative. Taken in conjunction with their inherent poor specificity, when you do have a result, you can’t be sure it’s meaningful (e.g interference within the immunoassay with fulvestrant1). With LC-MS/MS assays that measure down to 3pmol/l now available in the UK, the diagnosis and monitoring of the aforementioned patient groups are now achievable.
Mass spectrometry assays have also been successfully used to help advance the investigation of neuroendocrine tumours. As concentrations are an order of magnitude greater in urine than in serum, traditionally only 24-h urine collection could be used to quantify 5HIAA. However, following the development of more sensitive LC-MS/MS instrumentation, several assays for serum and plasma 5HIAA are now available in the UK. These offer more convenience for patients, are less susceptible to dietary influences, and are not prone to under- or over-collection of urine.2
Screening for phaeochromocytoma has also benefited from moving from urine to plasma in recent years. This has largely been facilitated by the development of LC-MS/MS assays. Some centres also offer 3-methoxytyramine as standard to help identify dopamine-secreting tumours, as well as head and neck paragangliomas.3
The questionable performance of some serum cortisol immunoassay measurements is well reported.4 Both the poor specificity of antibodies and the inability of some assays to liberate cortisol from its binding globulin have established a requirement for accurate measurements. This is especially true for patients on existing glucocorticoid regimens or those taking metyrapone to treat Cushing’s syndrome.
'With the provision of mass spectrometry, a new debate is emerging with regard to how urgently cortisol results should be provided.'
These problems have successfully been circumnavigated using LC-MS/MS, and several services are now available within the UK. With the provision of mass spectrometry, a new debate is emerging with regard to how urgently cortisol results should be provided. Is there a need for an accurate result in <24h, or is it imperative to provide a result that may or may not be close to the true value in under an hour (or a little over)? (Please note: if you feel particularly strongly about this, I would welcome your informed opinion!)
In a similar way, although cortisol immunoassays tend not to suffer from dexamethasone interference, some patients may not adhere to the 1mg overnight dexamethasone protocol and others may have CYP3A4 mutations that mean they will quickly metabolise the drug. As serum LC-MS/MS dexamethasone assays are now routinely available, it is possible to identify false-positive results before going on to arrange imaging.5
As we move into the next decade, there are several exciting developments in endocrinology that promise to be further elucidated using mass spectrometry. Considerable research is being undertaken to look at 11-oxoandrogens in polycystic ovary syndrome, congenital adrenal hyperplasia and precocious puberty. With serum and saliva assays available,6,7 there promises to be further input from UK-based laboratories in this emerging field. Indeed, with salivary testing still in its infancy, growing evidence suggests that the use of salivary cortisone may more accurately correlate with serum free cortisol concentrations. This is finding new applications in the diagnosis and monitoring of adrenal disorders.8
The promise of complementing mass spectrometry with data science has been realised by the Arlt group in Birmingham, as they have shown that LC-MS/MS urine steroid profiling can be used alongside machine learning to enhance the detection of adrenocortical carcinomas.9 This revolutionary approach could be applied to other conditions that rely on LC-MS/MS profiling rather than absolute quantification of a target analyte.
Finally, the use of LC-MS/MS to quantify protein markers has been under-utilised in UK laboratories. As methods for insulin10 and thyroglobulin11 have been published, it is likely that this deficiency will be addressed sooner rather than later.
If you are interested in knowing more about the assays discussed above, or mass spectrometry in general, please feel free to contact me at email@example.com.
JAMES HAWLEY, JO ADAWAY AND BRIAN KEEVIL
Wythenshawe Hospital, Manchester University NHS Foundation Trust
- Owen LJ et al. 2019 British Journal of Cancer 120 404–406.
- Adaway JE et al. 2016 Annals of Clinical Biochemistry 53 554–560.
- Adaway JE et al. 2015 Annals of Clinical Biochemistry 52 361–369.
- Hawley JM et al. 2016 Clinical Chemistry 62 1220–1229.
- Hawley JM et al. 2018 Annals of Clinical Biochemistry 55 665–672.
- Hawley JM et al. 2020 Clinical Chemistry & Laboratory Medicine 58 741–752.
- Schiffer L et al. 2019 Annals of Clinical Biochemistry 56 64–71.
- Blair J et al. 2017 Current Opinion in Endocrinology, Diabetes & Obesity 24 161–168.
- Bancos I et al. 2020 Lancet Diabetes & Endocrinology 8 773–781.
- Taylor SW et al. 2016 Clinica Chimica Acta 455 202–208.
- Hoofnagle AN et al. 2008 Clinical Chemistry 54 1796–1804.