Background: Temporary hemodynamic support using a microaxial left ventricular assist device (MLVAD) in patients (pts) with cardiogenic shock (CS) has been increasingly used for acute management in recent years. However, the invasive approach with a temporarily implanted circulatory support device may be associated with adverse events such as hemolysis, bleeding, limb ischemia and other complications. The overall prognostic impact of MLVAD has recently been questioned but little is known about the influence of MLVAD-associated complications on in-hospital mortality. 

Aim: To investigate the incidence of Impella-CP® related complications and their influence on in-hospital mortality in pts who underwent MLVAD placement due to cardiogenic shock.

Design: We retrospectively analyzed all pts who received hemodynamic mechanical support due to CS using the percutaneous Impella-CP® MLVAD between 2014 and 2021 in our institution with regard to predictors of in-hospital mortality and compared the outcome of pts with or without device-related complications, including a subgroup analysis in pts after cardiopulmonary resuscitation (CPR).

Results: 189 consecutive pts who underwent percutaneous placement of MLVAD due to CS were included. Mean age was 65.4 +/-11.8 years. Most common cause of cardiogenic shock was ST elevation myocardial infarction (91 patient, 48.1%). Mean duration of hemodynamic support with the pump was 13 +/- 51.7 hours. In-hospital mortality was 63.0% (119 pts). Main cause of death was refractory cardiogenic shock alone or accompanied with other forms of shock (hemorrhagic, septic). Patients who died due to more than one form of shock had significantly more complications (Table 2). Device-related complication occurred in 67 patients (35.4%). There was no statistically significant association between MLVAD-related complications and overall mortality (Table 2). Most common complications were hemolysis (23 pts, 12.1%), access site bleeding (21 pts, 11.1%), lower limb ischemia (11 pts, 5.9%) and other major bleeding (15 pts, 7.9%). Age at procedure ≥ 65 yrs was associated with a significantly higher mortality (77pts, 64.7%, p-value 0.003). Pts with out-of-hospital cardiac arrest (OHCA) (60 patients, 31.7%) had a higher mortality (45 pts, 75%, p = 0.033) whereas MLVAD-related complications occurred in 22 pts (37.7%) in this group, which not significantly different from the total pts group. Similarly, pts with in-hospital cardiac arrest (IHCA) (38 pts, 20.1%) had a higher mortality (30 pts, 78.9%) when compared to all pts (p-value 0.033) whereas device-related complications were observed less frequently in this group (8 patients, 21.1%, p = 0.035). In accordance, pts with extreme acidosis (defined as arterial pH <7) and severe hyperlactatemia (defined as serum lactate > 100 mg/dl) had no increase in complications but a higher mortality (11 pts, 100% and 16 pts, 88.9%. p 0.002 and 0.003 respectively). 

Conclusions: Despite a high complication rate associated with MLVAD use in pts with CS there is no correlation to in-hospital mortality. Patients with combined shock, including hemorrhagic and septic shock, have significantly more complications. Patients with an age ≥ 65 years, OCHA or IHCA have a higher mortality but no increase in MLVAD-associated complications.