Microaxial pump support in cardiogenic shock and high-risk PCI: experience from a Portuguese level II centre

Microaxial pump support is increasingly used in cardiogenic shock (CS) and high-risk PCI (HR-PCI), yet real-world mixed-indication data from Portugal remain limited, particularly from secondary-level institutions.

To describe clinical characteristics, complications, and outcomes of Impella therapy in CS and HR-PCI, assessing procedural growth and prognostic markers in a Secondary Centre (Level II).

All consecutive Impella cases performed between 2020 and 2025 were retrospectively analysed. Patients were classified as CS or HR-PCI by clinical indication. Demographics, ventricular function, vasoactive regimen, renal injury, haemolysis, bleeding, device-related complications, inter-hospital transfer and mortality were retrieved from institutional records. Annual procedural volume and the vasoactive–inotropic index were correlated with survival.

A total of 43 patients were analysed (CS = 51 %, HR-PCI = 49 %). Annual procedural volume increased progressively. After the publication of the DANGER Shock trial (April 2024), Impella use in AMI-related CS increased markedly. In the CS group, left-ventricular ejection fraction below 30 % was present in 86 %, and right-ventricular dysfunction in 27 %. Swan-Ganz monitoring was used in 23 %. Catecholaminergic support consisted mainly of noradrenaline plus dobutamine in 82 %, with adrenaline added in 9 %. The median vasoactive–inotropic index was higher among non-survivors, correlating with early mortality. LDH-defined haemolysis occurred in 77 %, and acute kidney injury secondary to haemolysis in 27 %, with no patient requiring renal replacement therapy. The mean haemoglobin decrease was 2.3 g/dL. Clinically relevant bleeding (BARC ≥ 3) occurred in 41 %, predominantly femoral-access related. Device events comprised mal-rotation (5 %), forward migration requiring repositioning (18 %); no exteriorisation or infection occurred. Transfers to tertiary hubs were required in 27 %. Mortality reached 41 % (within 7 days), mainly from refractory shock, one haemorrhagic and none septic. In HR-PCI, support duration was short (≈ 2 h) with immediate explantation. Major bleeding (BARC ≥ 3) occurred in 24 %, with a mean haemoglobin decrease of 1.1 g/dL. LDH-defined haemolysis was identified in 10%. No transfers or in-hospital deaths occurred. Overall survival improved progressively, paralleling experience and broader adoption of early Impella implantation after DANGER Shock.

In our center, microaxial pump support was feasible and safe across indications. CS patients exhibited severe ventricular dysfunction, frequent haemolysis and higher bleeding burden, whereas HR-PCI cases had brief, complication-sparse support. A higher vasoactive–inotropic index correlated with early mortality. The year-on-year improvement in survival underscores the importance of structured shock-network collaboration and early, protocol-driven Impella implantation in secondary centres.