Endocrine Abstracts

Introduction: Measurements of midnight salivary cortisol, 24-hour urinary free cortisol (UFC) and 1mg overnight Dexamethasone (DXM) suppression test (short DST) represent classical first line screening tests to diagnose Cushing syndrome. In women using combined Estrogen–Progestogen contraception, cortisol response during short DST is often decreased, leading to false positive (FP = morning plasma cortisol post-DST > 50 nmol/l).

Purpose: The main purpose of our work is to compare overnight (1 mg at midnight) and 48 h (0.5 mg/6 h for 48 h) low dose dexamethasone suppression tests (long DST) in volunteers using oral contraceptives. If the long DST shows less false positive than the short DST, it could be an alternative to stopping contraception in patients with a suspicion of Cushing syndrome.

Method: Thirty healthy volunteers using combined Estrogen–Progestogen contraception participated to the study. Plasma cortisol, ACTH and CBG were measured at baseline (first visit), after short DST (second visit) and after long DST (third visit). Measurements were performed by a 2nd generation Cobas E immunoassay. Results to DST were divided into three categories, the two last corresponding to false positive:

- morning plasma cortisol ≤ 50 nmol/l (normal)

- morning plasma cortisol 51 - 138 nmol/l (doubtful)

- morning plasma cortisol > 138 nmol/l (abnormal)

Results: Plasma cortisol levels decreased to a median of 69 nmol/l (percentiles 10–90: 31.8–162.6 nmol/l) after short DST and to a median of 36.5 nmol/l (percentiles 10–90: 18.7–68.7 nmol/l) after long DST. The results showed 63% and 27% of false positive (FP) after respectively short and long DST (significant difference: P=0.008 [χ²]). When we focused on categories of response, no value was higher than 138 nmol/l after the long DST compared to 11% after the short DST, 27% of the values were in the doubtful zone after the long DST compared to 52% after the short DST. In this population, the higher specificity of the long DST was confirmed by a repeated measure ANOVA (P=0.001). Finally, when we divided the population into tertiles on the base of their initial plasma cortisol level we observed:

- in the highest tertile (> 900 nmol/l): the short test had 89% FP vs 66% FP for the long test;

- in the average tertile (between 700 and 900 nmol/l): the short test had 66% FP versus 11% FP for the long test;

- in the lowest tertile (< 700 nmol/l), the short test showed still 33% FP vs 0% for the long test.

After multivariate analysis, we showed that the 2 major factors determining a participant’s risk of FP were: the basal plasma cortisol level followed by the type of Estrogen–Progestogen contraception. In conclusion, our results suggest that, if a DST is required, the long-DST may be a better option than the short one in a population using an oral contraception and that the cessation of oral contraception is probably not mandatory to interpret the results. These conclusions must be confirmed in a population suspected of hypercortisolism.