Early hemodynamic and metabolic changes following extracorporeal hemadsorption in acute myocardial infarction-related cardiogenic shock: A retrospective, single-center observational cohort study

Background : Acute myocardial infarction-related cardiogenic shock (AMI-CS) remains associated with high mortality. We investigated early hemodynamic and metabolic changes following adjunctive extracorporeal hemadsorption in a real-world AMI-CS cohort. Methods : This retrospective single-center landmark analysis included adult intensive care unit (ICU) patients with AMI-CS classified as Society for Cardiovascular Angiography and Interventions (SCAI) stage C or higher between January 2017 and July 2025. Patients who died within 24 h of ICU admission were excluded; among patients receiving hemadsorption, only those treated for at least 24 h were included. Propensity score matching was performed using sex, prior cardiopulmonary resuscitation (CPR), temporary mechanical circulatory support (tMCS), continuous renal replacement therapy (CRRT), admission lactate, admission vasoactive-inotropic score (VIS), and admission Horovitz index. The principal analysis compared pretreatment values with those at 24 h after treatment initiation; corresponding pseudo-timepoints were assigned in matched controls. Serial assessments at 12 h, 24 h, and 48 h characterized VIS and lactate trajectories. Results : Of 479 screened patients with AMI-CS, 398 met eligibility criteria for the 24-h landmark analysis, including 57 treated with adjunctive hemadsorption and 341 controls. Propensity score matching yielded 35 matched pairs and improved balance across the prespecified covariates, although residual imbalance persisted for prior CPR, CRRT, admission lactate, and admission VIS. In the matched cohort, hemadsorption-treated patients showed early within-group reductions in VIS (56.2 to 23.0; P =0.009) and lactate (2.0 to 1.5 mmol/L; P <0.001) as well as declines in aspartate aminotransferase ( P =0.031) and lactate dehydrogenase ( P =0.018). In matched controls, lactate likewise decreased significantly, whereas VIS remained unchanged, and tissue-injury markers did not show a comparable early decline. Although VIS and lactate improved over time overall, no significant time-by-treatment interaction was observed over 48 h for VIS ( P =0.834) or lactate ( P =0.427). Exploratory subgroup and multivariable analyses revealed no evidence of a differential mortality benefit. In-hospital mortality was similar between matched groups (45.7% vs. 42.9%; P =1.000), and hemadsorption was not independently associated with mortality (odds ratio=1.05, 95% CI: 0.36–3.08; P =0.934). Conclusions : In this exploratory propensity score-matched landmark analysis of AMI-CS, hemadsorption-treated patients showed early within-group hemodynamic and metabolic changes, but subsequent VIS and lactate trajectories did not differ significantly from matched controls, and in-hospital survival was not improved. Residual imbalance and the observational design limit causal interpretation. These findings should be considered hypothesis-generating and require prospective validation.