Relative Adrenal Insufficiency in Decompensated Cirrhotic Children: Does It Affect Outcome?

INTRODUCTION: Relative adrenal insufficiency (RAI) is associated with poor outcome in adult cirrhotics. So far, pediatric studies are not available on the same. We aimed to prospectively study the presence and outcome of RAI in children with decompensated cirrhosis over 180 days. METHODS: Hemodynamically stable children with decompensated cirrhosis were sampled for serum basal cortisol and peak cortisol (after 30 minutes of 1-mu g intravenous Synacthen) at day 1 and day 21. RAI was diagnosed as peak cortisol <500 nmol/L. Serum cytokines (interleukin-6 and tumor necrosis factor-alpha) and lipid profile were correlated with RAI. Cohort was followed up for outcomes over 180 days for complications and survival. With the identified risk factors, prognostic models were derived and compared with pediatric end-stage liver disease (PELD) and Child-Turcotte-Pugh scores. RESULTS: Prevalence of RAI was 54% at baseline and 61% at day 21 in the enrolled patients (n = 63, aged 128 +/- 48 months, male 78%). No significant differences in cytokines and serum lipid levels were seen between RAI and normal adrenal function groups. Patients with RAI at baseline (D1-RAI) developed higher complications at follow-up as compared to the normal adrenal function group (53% vs 24%, P = 0.02). The PELD score (odds ratio 1.08, confidence interval 1.05-1.12, P < 0.01) and D1-RAI (odds ratio 3.19, confidence interval 1.32-7.73, P = 0.01) were independent predictors of follow-up complications. The PELD-delta cortisol model (area under the receiver operating curve 0.84, P < 0.001, 92% sensitivity; 60% specificity) predicted morbidity better than isolated PELD or Child-Turcotte-Pugh scores. DISCUSSION: RAI is a risk factor for development of complications in pediatric cirrhosis over short-term follow-up. The PELD-delta cortisol score is a promising prognostic model for predicting follow-up complications.