ABSTRACT
INTRODUCTION: Cytokine adsorption using the CytoSorb® adsorber has been proposed in various clinical settings including sepsis, ARDS, hyperinflammatory syndromes, cardiac surgery or recovery after cardiac arrest. The aim of this analysis is to provide evidence for the efficacy of the CytoSorb® adsorber with regard to mortality in various settings. METHODS: We searched PubMed, Cochrane Library database and the database provided by Cytosorbents™ (01.1.2010-29.5.2022). We considered randomized controlled trials and observational studies with control groups. The longest reported mortality was defined as the primary endpoint. We computed risk ratios and 95%-confidence intervals and used DerSimonian and Lairds random effects model. We analysed all studies combined and divided them into the subgroups: sepsis, cardiopulmonary bypass surgery (CPB), other severe illness, SARS-CoV-2 infection and recovery from cardiac arrest. The meta-analysis was registered in advance (PROSPERO: CRD42022290334). RESULTS: Of an initial 1295 publications, 34 studies were found eligible, including 1297 patients treated with CytoSorb® and 1314 controls. Cytosorb® intervention did not lower mortality (RR [95%-CI]: all studies 1.07 [0.88; 1.31], sepsis 0.98 [0.74; 1.31], CPB surgery 0.91 [0.64; 1.29], severe illness 0.95 [0.59; 1.55], SARS-CoV-2 1.58 [0.50; 4.94]). In patients with cardiac arrest, we found a significant survival advantage of the untreated controls (1.22 [1.02; 1.46]). We did not find significant differences in ICU length of stay, lactate levels, or IL-6 levels after treatment. Of the eligible 34 studies only 12 were randomized controlled trials. All observational studies showed moderate to serious risk of bias. INTERPRETATION: To date, there is no evidence for a positive effect of the CytoSorb® adsorber on mortality across a variety of diagnoses that justifies its widespread use in intensive care medicine.