In plasma, cortisol circulates in three fractions: 80% circulates bound to cortisol binding globulin (CBG), 10-15% is bound to albumin, whereas the remaining 5-10% circulates as free and biologically active cortisol. However, at present, all clinically applicable assays measure total cortisol, and consequently fail to discriminate between bound (inactive) and free (active) cortisol. The failure poses a problem as CBG is markedly affected by estrogens and inflammation. Estrogens potently stimulate CBG; during pregnancy and treatment with oral contraception (OC), women demonstrate elevated total cortisol concentrations, making it difficult to diagnose cortisol disorders. Indeed, to correctly diagnose cortisol deficiency in women receiving OC, stopping medication for 6 weeks prior to cortisol measurements, is required. Conversely, in severe illness, the stress imposed upon the patient causes CBG to decline (mediated via increased interleukin-6 levels; IL-6). This results in low total plasma cortisol levels, which may be interpreted as cortisol insufficiency. However, as IL-6 at the same time stimulates cortisol production, levels of free cortisol are in fact elevated. In this situation, the present cortisol assays may yield erroneous low cortisol readings, making the diagnosis of true adrenal insufficiency in severe illness very difficult to confirm. Thus, many patients admitted to intensive care units (ICU) may receive unnecessary GC therapy, potentially causing adverse effects.
At present, serum total cortisol concentrations are mostly analysed by automated immunoassays in clinical laboratories. A recent national-wide comparison study of assays used at the hospitals (organized by Danish Endocrine Society and The Danish EQA-organizer DEKS) showed great variation in assay specificity and accuracy, primarily caused by lack of discrimination between cortisol and related steroid hormones and medication such as prednisolone. In fact, an internal investigation among Danish clinical chemistry laboratories demonstrated that most cortisol immunoassays showed direct bias in levels of cortisol in samples from women taking OC, when compared to the reference method (liquid chromatography-tandem mass spectrometry; LC-MS). Thus, the introduction of new and more specific cortisol assays directed against free cortisol in serum may circumvent the current methodological drawbacks and improve our abilities to diagnose cortisol disorders considerably.
Research questions
Given the above considerations, we find it highly relevant to develop a clinically applicable, fast track method that is able to isolate free cortisol in serum and hereafter to analyze its concentration by high-throughput LC-MS based assay. Thus, we seek to develop a highly specific and accurate assay for free cortisol that can be rapidly implemented in daily endocrine practice.
Research plan
1) To develop a clinically applicable, fast track high-throughput LC-MS based assay for free cortisol. We have three approaches for assessing free cortisol: a) Our first attempt will be to raise monoclonal antibodies against free cortisol. Coupling strategy of the cortisol to enhance the immunogenicity prior to the immunization will be evaluated. An optimal screening and contra-screening setup for selection of hybridoma cells will be established to enable production and isolation of specific antibodies with high affinity towards free cortisol without interference with bound cortisol. The selection and development of antibodies will be performed to ensure functional affinity, specificity and robustness when immobilized on the beads. b) In parallel, as a second attempt, we will raise monoclonal antibodies targeting cortisol binding proteins (i.e. CBG and albumin). These reagents will be used for an initial immunodepletion step for removal of bound cortisol from plasma. Subsequently, we will setup and measure total cortisol, free cortisol as well as CBG by state-of-the-art LC-MS. Concomitant measurements of total cortisol, free cortisol and CBG allow us to determine whether abnormal total cortisol concentrations are caused primarily by defects in cortisol, CBG or both. Also, in this way, we are able to decide whether it is necessary to remove albumin-bound cortisol too. c) Our third approach will be to use commercial Rapid Equilibrium Dialysis (RED) device plates to separate bound from unbound cortisol prior to LC-MS analysis.
2) Making the method clinically applicable. This will be accomplished by the establishment of reference values in healthy subjects and then compare the obtained reference values to free cortisol concentrations in patients diagnosed with endogenous cortisol deficiency and excess, respectively. According to standard procedures when establishing reference values for morning cortisol concentrations, we will collect donor blood from 120 healthy males and 120 healthy females from the Blood Bank at OUH.
Perspectives
The development of a clinically applicable analysis for serum free cortisol will be a breakthrough as no such method is available. Thus, such as analysis harbors many options. First, it may significantly improve our ability to diagnose cortisol-related disorders, and the introduction of an assay for free cortisol can as such have huge impact on daily clinical practice in many different medical specialties. Second, the assay may be suitable for commercialization and thereby return royalties to our institution. Finally, the assay may attract many new scientific collaborations and put attention to our institution.