Supplementary Material for: Correlation between bilirubin elimination with the cytokine adsorber Cytosorb® and mortality in critically ill patients with hyperbilirubinemia

Introduction: Hyperbilirubinemia is often the first evidence for any kind of liver disorder and over one third of all patients on intensive care units (ICU) show elevated bilirubin concentrations. In critically ill patients, high concentrations of serum bilirubin are correlated with a poor outcome. Therapies to lower bilirubin concentrations are often just symptomatically and their effect on the patients’ outcome is hardly evaluated. Therefore, this study investigates whether the extracorporeal elimination of bilirubin with the cytokine adsorber Cytosorb® (CS) reduces mortality in patients with hyperbilirubinemia. Methods: Patients with bilirubin concentrations > 10 mg/dl at the ICU were screened for evaluation from 2018 to 2020. Patients with kidney replacement therapy (KRT) and older than 18 years were included. Patients with continuously decreasing bilirubin concentrations after liver transplantation or other liver support systems (i.e. Molecular Adsorbents Recirculating System (MARS®), Advanced organ support (ADVOS)) were excluded. CS therapy was used in clinical routine and was indicated by the treating physicians. Statistical analysis was performed with IBM SPSS statistics utilizing a multivariate model. Primary outcome measure was the effect of CS on the 30-day-mortality. Results: Data from 82 patients (mean Simplified Acute Physiology Score (SAPS) II: 74 points, mean bilirubin: 18 mg/dl, mean lactate: 3.7 mmol/l) were analyzed. There were no significant differences in patients with and without CS treatment. The multivariate model showed no significant effect of CS therapy (p=0.402) on the 30-day-mortality. In addition, a significant effect of bilirubin concentration (p =0.274) or MELD score (p=0.928) on the 30-day-mortality could not be shown. In contrast, lactate concentration (p =0.001, b=0.044) and SAPS II (p =0.025, b=0.008) had significant impact on 30-day mortality. Conclusion: The use of CS in patients with hyperbilirubinemia did not result in a significant reduction of 30-day mortality. Randomized and controlled studies with mortality as primary outcome measure are needed in the future to justify the use.