Invasive hemodynamic ramp test stratifying short-term fatal event risk after weaning of impella microaxial flow pump

Background Clinical predictors of successful Impella weaning remain unclear. We aimed to evaluate the predictive value of invasive hemodynamic parameters measured during a ramp test for fatal events following Impella weaning. Methods We prospectively evaluated patients who underwent Impella weaning after treatment for cardiogenic shock between April 2023 and October 2024. A ramp test was performed before decannulation, during which the Impella support level was gradually reduced from P-8 to P-2 at 5-minute intervals under continuous monitoring with a pulmonary artery catheter (PAC). Hemodynamic parameters obtained from the PAC, including pulmonary artery pulsatility index (PAPI), pulmonary artery wedge pressure (PAWP), and cardiac power output (CPO), were assessed at each P-level. The primary endpoint was a composite of all-cause mortality or the need for an implantable left ventricular assist device. Results A total of 43 patients (median age: 62 years; 74% male) underwent the ramp test. Among them, 18 patients (41%) had acute myocardial infarction, and 23 patients (53%) were supported with veno-arterial extracorporeal membrane oxygenation (VA-ECMO). Primary events occurred in 11 patients (26%). There was no significant difference in absolute PAPI at P-2 (1.8 vs. 1.5, P = 0.3) or in the change in PAPI from P-8 to P-2 (+6% vs. -5%, P = 0.4) between patients with and without primary events. Similarly, absolute PAWP at P-2 (15 mmHg vs. 16 mmHg, P = 0.3) and the change in PAWP from P-8 to P-2 (+14% vs. +13%, P = 0.4) did not differ significantly between the two groups. Notably, absolute CPO at P-2 was lower in patients with primary events than in those without (0.56 W vs. 0.77 W, P = 0.02). Additionally, a greater reduction in CPO from P-8 to P-2 was observed in patients with primary events than in those without (-11.4% vs. ±0%, P = 0.004). Both absolute CPO at P-2 (odds ratio [OR]: 1.9 per -0.1 W, area under the curve [AUC]: 0.75, optimal cutoff < 0.6 W, P = 0.04) and a reduction in CPO from P-8 to P-2 (OR: 2.6 per -10%, AUC: 0.79, optimal cutoff < -10%, P = 0.01) had strong predictive value for primary events. A risk prediction model incorporating these cutoff values demonstrated a high predictive value for primary events (AUC: 0.82) (Figure). Conclusions Absolute CPO at the lowest support level (P-2) and dynamic changes in CPO during the ramp test from P-8 to P-2 effectively stratified the risk of fatal events after Impella weaning. Continuous CPO trend monitoring during the rapid reduction of Impella support may provide critical guidance for determining safe weaning strategies.