Endocrine Abstracts

Most tests of the hypothalamo-pituitary-adrenal axis (HPA) use total cortisol measurements, although only free cortisol is biologically active. Total cortisol is 90% bound to cortisol binding globulin (CBG). CBG varies between individuals, is elevated by oestrogens and reduced by acute illness or surgery. Low CBG levels may result in individuals apparently failing such tests, and receiving inappropriate steroid replacement. We hypothesised that correcting total cortisol for CBG would improve the diagnostic accuracy of total cortisol measurements.

Methods: A) 46 healthy volunteers (including 11 women on oestrogen) underwent a 250microgram Short Synacthen Test. Total cortisol, CBG and free cortisol (steady state gel filtration method) were measured. From analysis of the raw data we derived the modified free cortisol index (mFCI) (total cortisol²/CBG) B) Free cortisol was measured in post-operative samples from 5 patients with low post-operative total cortisols (range 399-541nmol/L) following major elective surgery.

Results: A) Post-synacthen, free cortisol ranged from 66-153nmol/L and mFCI from 5000-19700nmol²/mg/L. The correlation between total cortisol and free cortisol was 0.5. The mFCI demonstrated greatly improved correlation with free cortisol (r=0.9, P<0.0001). In women on oestrogen CBG was 73% higher compared to women not on oestrogen. The post-synacthen total cortisol was 25% higher (P<0.05). However there was no difference in post-synacthen free cortisol (P=0.7) or mFCI (P=0.7) between the two groups. B) All surgical patients made uneventful recoveries with no evidence of HPA insufficiency. CBG fell by 40% immediately post-operatively. Despite the low total cortisol levels all patients had normal free cortisol (79-105nmol/L) and mFCI values (7800-12200nmol²/mg/L).

Conclusion: The mFCI markedly improves the correlation between total cortisol and free cortisol. The mFCI allows interpretation of the SST in women taking oestrogen and improves the interpretation of total cortisol measurements in patients with altered CBG levels. We recommend the mFCI be calculated when interpreting dynamic HPA tests.