In-hospital versus out-of-hospital cardiac arrest define distinct phenotypes in acute myocardial infarction-related cardiogenic shock

Background Cardiac arrest (CA) is a major risk modifier in acute myocardial infarction–related cardiogenic shock (AMI-CS), yet presence of CA is typically reported as a binary variable without accounting for arrest location. Whether out-of-hospital (OHCA) and in-hospital cardiac arrest (IHCA) represent clinically distinct phenotypes with divergent management patterns and differential mechanisms of death remains unclear. Purpose To evaluate whether CA location in AMI-CS identifies biologically and therapeutically distinct phenotypes associated with differential management intensity and mode of death. Methods Adults with AMI-CS and CA were analysed from our shock registry, a multicentre registry of CS including patients treated between 2016–2022. AMI was defined as STEMI or NSTEMI. CA was categorised as OHCA or IHCA. Clinical characteristics, management strategies, and adjudicated causes of in-hospital death were compared. Results Among 376 patients, 147 (39%) had OHCA and 229 (61%) IHCA. OHCA patients were younger (64±13 vs 68±13 years; p<0.01), more frequently male (84% vs 65%; p<0.01), and more often presented with STEMI (80% vs 70%; p=0.03). Initial rhythm distribution differed significantly. Ventricular fibrillation was more frequent in OHCA (41% vs 25%; p<0.01), whereas ventricular tachycardia (8% vs 26%; p<0.01) and PEA (14% vs 22%; p=0.045) predominated in IHCA. Rearrest and CPR rates were similar. OHCA patients underwent less coronary angiography (63% vs 77%; p<0.01) and PCI (43% vs 54%; p=0.03) and received less MCS (47% vs 62%; p<0.01), but more frequently required mechanical ventilation (93% vs 86%; p=0.04). In-hospital mortality was high and similar in both groups (52% vs 51%; p=0.7). However, mechanisms of death differeted markedly. Neurologic withdrawal of care accounted for nearly half of deaths after OHCA (48% vs 10%; p<0.01), while refractory shock or multiorgan failure (48% vs 26%; p<0.01), and comorbidity-driven withdrawal of care (19% vs 5%; p<0.01) were more common in IHCA. Conclusion In AMI-CS, OHCA and IHCA identify distinct clinical phenotypes with similar overall mortality but fundamentally different modes of death; neurologic injury after OHCA and refractory shock after IHCA. Treating cardiac arrest as a single covariate may obscure important biological and therapeutic differences. These findings have implications for risk stratification and future CS study design.