Evaluation of the utility of inhaled nitric oxide therapy in ECPELLA management
M Tanaka, T Unoki, N Yoneda, T Inamori, T Ishida, T Sato, Y Konami, H Suzuyama, M Inoue, E Horio, K Kodama, K Nakao, T Sakamoto, K Okumura, J KoyamaAbstract
Background
Veno-arterial extracorporeal membrane oxygenation (VA-ECMO) provides haemodynamic support in severe cardiogenic shock (CS) but may increase left ventricular afterload and worsen pulmonary congestion. ECPELLA (VA-ECMO plus a microaxial flow pump) enables concomitant circulatory support and left ventricular unloading. Inhaled nitric oxide (iNO) selectively reduces pulmonary vascular resistance and may facilitate VA-ECMO weaning by decreasing right ventricular afterload.
Purpose
To assess whether iNO initiated at the start of VA-ECMO weaning in ECPELLA-treated severe CS is associated with higher weaning success and improved 30-day outcomes, and to explore its haemodynamic effects.
Methods
In this single-centre retrospective study, 175 consecutive patients with severe CS treated with ECPELLA between August 2018 and December 2025 were screened. Since iNO became available from December 2021, groups were constructed according to iNO exposure at VA-ECMO weaning. During the iNO-available period, weaning was generally initiated when pulmonary artery catheter-derived cardiac output (PAC-CO) reached ≥1.0 L/min, and iNO was started at 40 ppm at weaning initiation (final decision by the treating physician). To minimise selection bias, patients who never achieved PAC-CO ≥1.0 L/min (n=34) and, in the iNO-available period, those not receiving iNO at weaning initiation (n=9) were excluded. The final groups were iNO (n=81) and non-iNO (n=51). Primary endpoints were VA-ECMO weaning rate, VA-ECMO duration, and 30-day survival. Secondary analyses assessed changes in oxygenation and haemodynamics from immediately before iNO initiation to 24 h after initiation in the iNO group.
Results
Baseline characteristics, including age, sex, cause of CS, comorbidities, and extracorporeal cardiopulmonary resuscitation, were similar between groups. VA-ECMO weaning was higher in the iNO group (91% vs 71%, p=0.004), while VA-ECMO duration was comparable (9.6 [6.9–10.3] vs 7.1 [5.5–10.8] days, p=0.27). In the iNO group, the PaO2/FiO2 ratio significantly improved 24 h after initiation, whereas other haemodynamic variables showed no significant change. Thirty-day survival did not differ (48% vs 55%, p=0.79). Among patients successfully weaned from VA-ECMO, survival was lower in the iNO group (53% vs 75%, p=0.03). Furthermore, landmark analysis after VA-ECMO weaning suggested a plateau of mortality hazard after almost 14 days in the non-iNO group, while risk persisted through 30 days in the iNO group.
Conclusion
In ECPELLA-treated severe CS, iNO at VA-ECMO weaning was associated with higher weaning success but not improved 30-day survival. Prospective studies are warranted to define optimal patient selection for iNO in this setting.Study flowchartKaplan-Meier Curve for 30-day Mortality