Endocrine Abstracts

Background: Beckman Access Cortisol immunoassays, utilising monoclonal antibodies, produce more accurate serum cortisol concentrations but approximately 20% lower compared with older assays necessitating reduction in thresholds to maintain diagnostic accuracy of SST. Cortisol thresholds were lowered from 550 nmol/l to 460 nmol/l at 30- or 60-minutes at the Great Western Hospital in 2016, and to 410 and 460 nmol/l at 30- and 60-minutes respectively in 2022. We reassessed the effectiveness of newly established SST (250 mg ACTH) thresholds and 8-10 am cortisol which may negate the need for dynamic tests.

Methods: A retrospective analysis included 101 patients who underwent SST from June to September 2022.

Results: Of the steroid-naïve patients (n = 74), 92% passed the SST and 8% failed who later were diagnosed with AI. One test result was indeterminate. Of the steroid-exposed patients (n = 27), 34.6% passed and 65.4% failed the SST and among them, 4 patients had suboptimal cortisol response (between 350 – 459). Threshold concordance between 30- and 60-minutes was 99%. 4 patients (5.1%) with false negative results had further testing with Insulin Tolerance test (ITT) and received a final diagnosis of secondary AI. Average morning cortisol in steroid-naïve patients, who passed the SST and did not end up with a diagnosis of AI, was 283 nmol/l (106-565), and in those who failed SST, 42.8 nmol/l (11-77).

Conclusion: The high concordance between 30- and 60-minute thresholds suggests that 30-minute sampling alone may be sufficient to maintain diagnostic accuracy. The revised SST thresholds effectively eliminated false positives, but false negatives still occurred. Therefore, if there is high clinical suspicion of secondary AI, ITT should be considered. In steroid-naïve patients morning cortisol <100 nmol/l measured by Beckman Access warrants starting management for AI, cortisol >210 nmol/l likely indicates low pretest probability of AI.