Cardiogenic shock complicating acute myocardial infarction; mechanical circulatory support use in the Netherlands

Abstract Introduction Mortality rates in cardiogenic shock complicating acute myocardial infarction (AMICS) linger around 40-50%. Despite the lack of evidence, the use of mechanical circulatory support (MCS) in this population is increasing rapidly worldwide. Purpose This study aimed to give insights in MCS use in the Netherlands and to assess how the use of MCS is associated with mortality. Methods Data from all patients undergoing percutaneous coronary intervention (PCI) in the Netherlands is prospectively registered in the Netherlands Heart Registration (NHR) database. On top of the obligatory registration, additional data were retrieved for all patients with cardiogenic shock undergoing PCI between 2017 and 2021 in 14 Dutch hospitals. The additional variables included the use of different types of MCS. Multivariate logistic regression was used to test the association between MCS and mortality in adjustment for confounders. Results In total, we included 2328 patients, with a mean age of 66.4 ± 12.3 years and of whom 72.9% were male. Of all patients, 23.6% received a form of MCS (11.5% an IABP, 4.5% an Impella, 3.1% VA-ECMO, 4.2% a combination of devices and 0.3% other). Compared to patients who did not receive MCS, MCS supported patients were younger (64.2 vs. 67.1 years, p <0.001), presented after an out-of-hospital cardiac arrest (OHCA) less often (33.3% vs. 39.9%, p = 0.008), had complaints for over three hours more often (50.3% vs. 38.8%, p = 0.001) and had a lower mean arterial pressure (MAP) (74.9 vs. 78.6 mmHg, p <0.005) and higher heartrate (88.2 vs 81.8 beats per minute, p <0.001) at admission. Also, mean admission lactate (6.4 vs. 5.2, p <0.001), glucose (13.4 vs. 11.8, p <0.001) and creatinine levels (104 vs. 99, p <0.001) were higher in MCS-receiving patients. Additionally, patients who received MCS support more often suffered from multivessel disease (68.7% vs. 58.4%, p <0.001) and chronic total occlusions (CTO) (3.5% vs. 1.9%, p = 0.027) and were treated with inotropes prior to PCI more often (54.9% vs. 46.2%, p <0.001). The overall 30-day mortality was higher in the MCS cohort than in the no-MCS cohort, namely 54.4% vs. 34.3% (OR = 2.28, 95%CI = 1.88-2.78). In adjustment for age, OHCA, complaints duration, admission MAP, heartrate, lactate, glucose and creatinine, the presence of multivessel disease and CTO and inotrope use before PCI, MCS was independently associated with 30-day mortality (adjusted OR = 2.23, 95% CI = 1.54-3.23). Conclusion MCS was deployed in 23.6% of the patients in this cohort. Despite the results of the IABP-SHOCK II trial, IABP was the most frequently used device. In this cohort, MCS was independently associated with 30-day mortality. However, residual confounding is probably present. Besides randomised controlled trials guiding clinical decision making regarding MCS use and timing, future research should also focus on investigating barriers to implement the results of previous trials.KM-curve of unadjusted survival