In-Hospital Cardiac Arrest in Impella-Supported Cardiogenic Shock: Insights From the J-PVAD Registry.

Abstract

Impella-treated cardiogenic shock (CS) includes distinct arrest presentations. In-hospital cardiac arrest (IHCA) during CS may reflect progressive in-hospital hemodynamic collapse, but outcomes after Impella support remain unclear. This study compared 30-day outcomes across four cardiac arrest presentations and identified mortality-associated factors among patients with IHCA before Impella implantation. Overall, 8,577 J-PVAD patients with CS treated with Impella were classified as IHCA (n=1,924), no cardiac arrest (n=4,595), out-of-hospital cardiac arrest (OHCA) with return of spontaneous circulation (ROSC; n=1,321), or OHCA without ROSC (n=737). The primary outcome was 30-day mortality from first Impella initiation. Multivariable Cox models and restricted cubic splines were used. Thirty-day mortality was 48.1% in IHCA, 28.7% in no cardiac arrest, 41.9% in OHCA with ROSC, and 62.5% in OHCA without ROSC (P < 0.001). Compared with no cardiac arrest, IHCA was associated with higher mortality (adjusted HR, 1.54; 95% CI, 1.41-1.69; P <0.001) and higher rates of major bleeding, Impella-related bleeding, limb ischemia, ischemic stroke, and sepsis. Within IHCA, older age, lactate >4.0 mmol/L, creatinine >1.5 mg/dL, longer arrest time, and venoarterial extracorporeal membrane oxygenation use were associated with higher mortality, whereas myocarditis and electrical defibrillation were associated with lower mortality. Spline analyses showed nonlinear associations for lactate, creatinine, and arrest time. IHCA before Impella implantation identified a distinct, high-risk CS phenotype with high mortality and a substantial complication burden. Metabolic, renal, and low-flow injury may support early risk stratification and clinical decision-making.