Evaluating the low-dose ACTH stimulation test in neonates: ideal times for cortisol measurement.

Context Low-dose adrenocorticotropic hormone stimulation testing (LDST) can be used to diagnose central adrenal insufficiency. However, uncertainty remains over optimal times to draw serum cortisol levels. Objective To determine optimal times to draw serum cortisol levels for the LDST in neonates. Design A retrospective chart review of LDSTs performed on neonates from January 1, 2009 to September 30, 2017. Setting Children's Hospital of Eastern Ontario (CHEO), a tertiary-care outborn pediatric center. Patients Forty-nine patients were included: 23 (46.9%) born at term, 12 (24.5%) born very preterm to late preterm, and 14 (28.6%) born extremely preterm. Intervention Cortisol levels were drawn at baseline and 15, 30, and 60 minutes following administration of Cortrosyn 1 mcg/kg (maximum dose 1 mcg). Main outcome measure Timing of peak cortisol level and marginal value of drawing a second and third cortisol sample at 15, 30, or 60 minutes was determined. Results Cortisol peaked at 15-, 30-, and 60-minute sampling times for 4%, 27%, and 69% of patients, respectively. The probability that a failed LDST changes to a pass by adding a 15- or 30-minute sample to the superior 60 minute sample is 5.6% (1% to 25.8%) and 11% (3.1% to 32.6%), respectively, for a cortisol pass threshold of 18.1mcg/dL (500 nmol/L). Conclusions In contrast to studies of older children, we found that the majority of neonatal LDST cortisol peaks occurred at the 60-minute sampling time with the addition of a 30-minute sample providing substantial benefit. It is questionable if a 15-minute sample provides any benefit, making a case to revise LDST protocols to sample cortisol later for neonates.