Impact of collection method and sample handling on measured levels of circulating ACTH

Introduction

Cortisol, which is produced and released by the adrenal gland under the regulation of the adrenocorticotropic hormone (ACTH), plays a crucial role in the survival mechanisms involved in critical illness. A dissociation between high plasma cortisol and low ACTH in patients has been observed in the more chronic phase of critical illness induced by severe sepsis and multiple trauma [1]. In order to assess the potential confounding impact of errors in the measurement of ACTH, we evaluated the stability of ACTH during sample processing.

Methods

Two tests were performed. For the first test, two blood samples per patient were taken from 10 randomly selected patients in the surgical ICU. One sample was collected warm (that is, samples were collected at room temperature (RT), left for 60 to 90 minutes at RT followed by 24 hours at 4°C, centrifuged (10 minutes, 1,000 rpm), and serum was stored at -80°C), the other sample was collected cold (that is, recommended collection procedure for ACTH detection: samples were collected on ice, centrifuged (10 minutes, 1,000 rpm), and plasma was stored at -80°C). For a second test, again, two blood samples were taken per patient (n = 10) in the surgical ICU. One sample was kept frozen until ACTH testing (original), whereas the other sample was defrosted several times at RT before analysis (defrosted). ACTH levels were measured using an ACTH RIA kit (BRAHMS) and the different sampling methods and sample treatments were compared using Bland-Altman statistics.

Results

No clinically relevant differences in ACTH levels were observed between cold (28.2 ± 22.3 pg/ml) and warm (24.3 ± 20.6 pg/ml) sampling (bias 3.8 ± 7.3 pg/ml) or between original (22.1 ± 10.8 pg/ml) and defrosted (22.3 ± 10.5 pg/ml) samples (bias 0.2 ± 1.9 pg/ml). Moreover, the differences in ACTH levels measured between the two collection methods or sample treatments were independent of the amount of ACTH present in the samples.

Conclusions

We showed that warm collection or frequent defrosting of samples does not induce a clinically significant error in ACTH levels measured, implying that nonideally collected or treated samples can still be used for rough ACTH analysis.