Postoperative serum cortisol concentration and adrenal insufficiency in neonates undergoing open-heart surgery

We sought to determine whether immediate postoperative serum cortisol concentration predicts adrenal insufficiency in neonates after cardiac surgery with cardiopulmonary bypass. We hypothesized that cortisol <10 µg/dL would be associated with increased catecholamine requirements and fluid resuscitation and would predict hemodynamic responsiveness to exogenous steroids.

Retrospective study of 41 neonates was carried out for the levels of cortisol in the immediate postoperative period; of whom, 15 received steroids due to high levels of inotropic support. Laboratory and clinical outcomes were collected.

Median cortisol was 12 µg/dL (interquartile range: 5.2-27.4). Levels of cortisol <10 µg/dL was not associated with any clinical variable indicative of increased illness severity. Peak lactate (9.1 vs 11.8 mmol/L, P = .04) and maximum arteriovenous saturation difference ([Sao 2 - Svo 2] 28% vs 32%, P = .05) were both lower among patients with levels of cortisol <10 µg/dL. Six (40%) patients had a significant hemodynamic improvement within 24 hours after receiving steroids (responders), although there was no statistical difference between levels of cortisol in responders versus nonresponders. Level of cortisol was positively correlated with maximum lactate (P < .001), maximum Sao 2 - Svo 2 (P < .001), maximum inotrope score (P = .014), initial 24-hour fluid intake (P = .012), and time to negative fluid balance (P = .008) and was negatively correlated with initial 24-hour urine output (P < .001).

Low cortisol obtained in the immediate postoperative period is not associated with worse postoperative outcomes or predictive of steroid responsiveness. In contrast, elevated levels of cortisol are positively correlated with severity of illness. The use of an absolute cortisol threshold to identify adrenal insufficiency and/or guide steroid therapy in neonates after cardiac surgery is unjustified.