ESPE Abstracts

Background: The Low dose dexamethasone suppression test (LDDST) is an important investigation for suspected Cushing’s Syndrome (CS). The traditional definition of normal suppression of serum cortisol to ≤50 nmol/l during the LDDST (0.5 mg 6 hrly × 48 h) comes from a time when biochemical autoanalysers did not routinely detect very low values. Previous studies reported 5.1–8.3% of patients with Cushing’s Disease (CD) suppressed to <50 nmol/l at 48 h during LDDST. Many clinicians experienced in the assessment of suspected CS consider that ‘normal’ individuals should suppress to ≤20 nmol/l during a LDDST and that LDDST values of 20–50 nmol/l represent a range of uncertainty. Current sensitivity and specificity is reported as 90% and 100% for a cut off of ≤50 nmol/l.

Methods: We reviewed a retrospective cohort of paediatric patients referred to our centre with suspected CS between 1982 and 2018.

Results: Of 70 suspected CS patients, 49 had Cushing’s Disease (CD), 7 had Primary Pigmented Nodular Adrenocortical Disease (PPNAD) and 14 ‘control’ subjects, in whom the diagnosis of CS was excluded following detailed biochemical evaluation and prolonged clinical/auxological follow-up. The serum cortisol remained >50 nmol/l in 42/49 (86%) CD patients (29 males, median age 13.22 years) during LDDST. In contrast, cortisol during LDDST was >20 nmol/l in 48/49 (98%) CD patients. One patient with cortisol ≤20 nmol/l during LDDST had a high clinical suspicion of CD and investigations including bilateral simultaneous inferior petrosal sinus sampling confirmed this. The sensitivity and specificity of a LDDST cut off value of ≤20 nmol/l is 97.96% (95% CI 89.15%–99.95%) and 100% (76.84%–100%). None of the 7 PPNAD patients (four male, median age 12.2 years) had cortisol levels of ≤50 nmol/l during LDDST. Cortisol levels in all 14 controls (three males, median age 12.7 years) suppressed to ≤20 nmol/l during LDDST.

Conclusion: Whilst the numbers are small, changing the LDDST cut off from ≤50 nmol/l to ≤20 nmol/l improves the sensitivity of the test from 85.71% to 97.96% in our paediatric CD patients. This does not adversely affect the specificity which remains 100%. We therefore suggest using serum cortisol of ≤20 nmol/l as a new diagnostic cut off value.