Effect of immunomodulation on post-cardiac arrest brain injury: a systematic review and meta-analysis.
Resuscitation · 2026-04-11 · Meta-analysis
Abstract
PURPOSE: To evaluate the effect of immunomodulatory therapeutics on mortality, neurological outcomes, and inflammatory biomarkers in patients resuscitated following cardiac arrest. METHODS: Studies from MEDLINE, Embase, CINAHL, and the Cochrane Central Register of Controlled Trials (CENTRAL) were included from database inception up to June 2025. All randomized controlled trials using immunomodulators in adults following both in- and out-of-hospital cardiac arrest were included. Assessed outcomes included functional neurological outcome, mortality at 30 days, and inflammatory biomarkers levels. We pooled estimates of effect size using random effect model and report relative risks with 95% CI. We assessed risk of bias using the Cochrane Risk of Bias 2 tool and rated the certainty of evidence using GRADE methodology. RESULTS: The search yielded 19,558 studies of which 2351 were duplicates. Fifty-eight full texts were reviewed and 11 studies plus 3 sub-studies was included for a total of 2605 patients. Of these clinical trials, immunomodulatory interventions consisted of corticosteroids (82%), interleukin-6 receptor blockade (9%), and calcineurin inhibition (9%). Immunomodulatory therapies were not associated with an increase in favorable neurological outcome (RR1.04 [0.85, 1.28]) nor a decrease in mortality (RR0.97 [0.91, 1.03]). A trend toward decreased IL-6 levels at 24 h following glucocorticoid administration was observed. The overall quality of evidence was limited by heterogeneous interventions and timing. CONCLUSION: Across all studies, immunomodulating therapies were not associated with significant increases in favorable neurologic outcomes or survival. Between-study heterogeneity in intervention dosage, outcomes, timing of assessment, and biomarker profiling limited direct comparison and supports the need for future trials. Prospero registration: CRD420251068980.