An update on the biochemical diagnosis of congenital ACTH insufficiency

Congenital ACTH insufficiency due to disorders of the hypothalamo-pituitary axis is life threatening. However, the optimal method for establishing the diagnosis remains controversial. The standard Synacthen test (SST) is safe and easy to perform in infancy, but its performance has not been well evaluated in infants and children with ACTH insufficiency.

We aimed to determine the value of the SST by comparing the 30-min SST cortisol concentration with physiological (spontaneous) cortisol secretion (2-hourly measurements for 24 h) in 28 patients (13 male; age range 0-5.1 years) with one or a combination of optic nerve (25) midline forebrain (17) and other pituitary hormone (17) abnormalities. Spontaneous cortisol secretion (mean < 145 nmol/l) was arbitrarily used as the 'gold-standard' in the diagnosis of ACTH insufficiency as it offered a reflection of endogenous cortisol mileu. A normal SST was defined as a 30-min serum cortisol > 540 nmol/l.

The SST basal, 30-min and 60-min serum cortisol concentrations were significantly (P < 0.001) lower in patients with ACTH insufficiency (mean spontaneous cortisol < 145 nmol/l). 3/13 patients without and 12/15 patients with ACTH insufficiency failed the SST (specificity 76.9%; sensitivity 80%). Two out of three patients with ACTH insufficiency who passed the test were symptomatic with fatigue, poor growth and hypoglycaemia that resolved with hydrocortisone treatment. An 08:00 h serum cortisol cut-off of 175 nmol/l provided a sensitivity of 93.3% and a specificity of 92.3% as compared with spontaneous cortisol secretion. The 30-min increment and the 60-min SST cortisol concentrations although highly specific (100%) had reduced sensitivity (40% and 60%, respectively).

A normal SST does not rule out ACTH insufficiency and measurement of physiological cortisol secretion may be required to confirm the diagnosis. A combination of 08:00 h serum cortisol > 175 nmol/l and a 30-min SST serum cortisol > 540 nmol/l to exclude ACTH insufficiency (sensitivity 69%, specificity 100%) would ensure that no child with ACTH insufficiency would be missed and hence would prevent serious consequences, although it may lead to over-diagnosis in some cases.