Adrenal insufficiency in decompensated cirrhotic patients without infection: prevalence, predictors and impact on mortality

Abstract

Background: Relative adrenal insufficiency (RAI) is common in compensated and decompensated chronic liver disease in the presence of sepsis. This study was performed to find out the prevalence of RAI in decompensated cirrhotic patients presenting with hepatic encephalopathy and variceal bleeding without any evidence of infection.

Methods: The study prospectively included 75 cirrhotic patients with signs of decompensation. The short Synacthen test (SST) was performed on all patients after ruling out infection. Patients with positive blood, urine, sputum, ascitic and pleural fluid cultures or evidence of infection on chest X-ray and those with elevated procalcitonin levels (>0.05 ng/ml) were excluded. RAI in critical illness was defined by a delta cortisol level (difference between basal and post-stimulation cortisol) of ≤9 μg/dl after SST.

Results: The mean age of the study population was 54 ± 11 years. Upper gastrointestinal bleed and hepatic encephalopathy were seen in 56.6% and 41.5%, respectively, and both were seen in 1.9%. Of the 75 patients, 55 (73%) were in Child-Turcotte-Pugh (CTP) class C and the mean model for end-stage liver disease (MELD) score was 21 ± 7. Forty-five patients (60%) met our criteria for RAI. Those with RAI had lower serum albumin (2.4 ± 0.5 g/dl vs 2.7 ± 0.5 g/dl, p = 0.03) and higher MELD scores (22 ± 7 vs 19 ± 6, p = 0.03). Prevalence of RAI in CTP class C was 65% (36 out of 55 patients) compared to 45% (9 out of 20 patients) in Child-Pugh stage A and B. Similarly, 82% (23 out of 28 patients) with MELD scores >25 had RAI compared to 54% with MELD scores <20. None of biochemical parameters were predictive of RAI on logistic regression analysis. Three-month mortality rate was not significantly different in patients with or without adrenal insufficiency (44% vs 28%, p = 0.11).

Conclusion: The present study showed RAI to be common in noninfected decompensated cirrhotic patients, but did not predict 3-month mortality. There were no other predictive factors in those with RAI. Hence, in patients with cirrhosis without infection, the clinical utility of routine adrenal function testing needs further elucidation.

Keywords: Child–Turcotte–Pugh stage; adrenal insufficiency; chronic liver disease; cirrhosis; hepatic encephalopathy; hepatoadrenal syndrome; model for end-stage liver disease (MELD) score; variceal bleeding.