Introduction: The Short Synacthen Test (SST) is the most popular diagnostic investigation of adrenal insufficiency (AI) amongst adult and paediatric endocrinologists. AI can present insidiously and symptoms may be non-specific. The number of medical indications for glucocorticoids is growing and SST usage is correspondingly increasing. There is evidence that an early morning plasma cortisol (EMC) of below ~100150 nmol/l is highly predictive of failing the SST and the corollary is seen with an EMC above ~336400 nmol/l. Using an EMC to screen patients for AI has been advocated to reduce the number of invasive, time-consuming and resource-intensive SSTs, although there is a paucity of studies in children. More sensitive and specific modern cortisol assays make deriving local diagnostic thresholds important. We analysed our SST data since the introduction of a new cortisol assay to derive our own screening thresholds for SST and examined the relationship between the basal, incremental and peak plasma cortisol following synacthen stimulation.
Methods: All SST performed at Sheffield Childrens Hospital, UK, between September 2014 and 2017 were retrospectively analysed. Cortisol quantification was performed on the Abbott Architect i1000 chemiluminescent immunoassay (CVs <5%). A pass for both the high and low-dose SST is currently 450 nmol/l. Basal cortisol was used as a surrogate for EMC and correlation coefficients with increment and peak examined. Subgroup analysis was performed using sex and an age-approximate for pubertal status (09 and 1016 years old). Positive and negative predictive values using a basal plasma cortisol of <160 nmol/l and >340 nmol/l respectively were calculated.
Results: Overall 393 SSTs were included (209 M, 184 F, 175 prepubertal, 218 post-pubertal). The correlation coefficient for basal and peak cortisol was 0.63, (0.63 female, 0.62 male; 0.65 09 years, 0.66 1016 years of age). There was no relationship in any of the groups between basal and incremental cortisol (overall data correlation coefficient 0.061). Of the cohort 28% had basal cortisols <160 nmol/l of which 54% failed the SST, PPV=0.54. Correspondingly 13% had basal cortisols >339 nmol/l, none of whom failed the SST, NPV=1.
Conclusions: There is a reasonably strong relationship between basal and peak cortisol on the SST, but no relationship exists between basal and incremental cortisol. Subgroup analysis did not significantly strengthen the correlations. On the Abbott Architect plasma cortisol assay an EMC of >339 nmol/l appears to safely predict passing the SST and <160 nmol/l yields a high PPV for failing the SST.
27 - 29 Sep 2018
European Society for Paediatric Endocrinology