Use of 60 min cortisol levels decreases sensitivity of short synacthen tests
JT George, Antony Jacob, I Barlow & Mohamed Malik
Introduction: Short synacthen tests (SST) are widely performed as a screening test in patients with suspected hypoadrenalism. However, the timing of sampling and interpretation of cortisol levels remain a matter of controversy. Though known to be inferior to 30 min cortisol levels, some centres continue to analyse 60 min values (along with 30 min levels) to avoid resource-intensive tests like insulin stress tests and glucagon stimulation tests.
Objective: To determine the sensitivity and reliability of 60 min cortisol in patients with borderline (400550 nmol/l) 30 min cortisol.
Methods: Four hundred and eighty two consecutive SSTs processed in our laboratory, with 30 and 60 min samples (as per standard practice) were analysed retrospectively. Patients with a 30 min cortisol below 550 nmol/l and 60 min cortisol above 550 nmol/l underwent glucagon stimulation tests.
Results: Thirty-nine (8%) had cortisol levels below 550 nmol/l at 30 min. While 27 (5.6%) had similar results at 60 min, 12 (2.5%) patients had a 60 min cortisol value greater than 550 nmol/l and glucagon stimulation tests were carried out on 10 of these patients. Two patients could not be followed up. Only 2 (20 and 0.4% of total sample) of these patients achieved peak cortisol levels >550 nmol/l, while the remaining 8 (80 and 2.1% of total sample) failed to achieve this. Reliance on 60 min cortisol levels could therefore have resulted in missing the diagnosis in 8 patients with impaired hypothalamicpituitaryadrenal axis equating to a 23% decrease in the specificity of SST.
Conclusion: Using 60 min cortisol levels in conjunction with 30 min values can result in a decrease in sensitivity of SST. Previous studies have shown greater than 90% sensitivity for SST using 30 min values and we recommend avoiding 60 min testing in screening SSTs.