Association between left ventricular ejection fraction, mortality and use of mechanical circulatory support in patients with non-ischaemic cardiogenic shock
Jonas Sundermeyer,Caroline Kellner,Benedikt N Beer,Lisa Besch,Angela Dettling,Letizia Fausta Bertoldi,Stefan Blankenberg,Jeroen Dauw,Zouhir Dindane,Dennis Eckner,Ingo Eitel,Tobias Graf,Patrick Horn,Joanna Jozwiak-Nozdrzykowska,Paulus Kirchhof,Stefan Kluge,Axel Linke,Ulf Landmesser,Peter Luedike,Enzo Lüsebrink,Nicolas Majunke,Norman Mangner,Octavian Maniuc,Sven Möbius Winkler,Peter Nordbeck,Martin Orban,Federico Pappalardo,Matthias Pauschinger,Michal Pazdernik,Alastair Proudfoot,Matthew Kelham,Tienush Rassaf,Clemens Scherer,Paul Christian Schulze,Robert H G Schwinger,Carsten Skurk,Marek Sramko,Guido Tavazzi,Holger Thiele,Luca Villanova,Nuccia Morici,Ralf Westenfeld,Ephraim B Winzer,Dirk Westermann,Benedikt Schrage
DOI: https://doi.org/10.1007/s00392-023-02332-y
Abstract:Background: Currently, use of mechanical circulatory support (MCS) in non-ischaemic cardiogenic shock (CS) is predominantly guided by shock-specific markers, and not by markers of cardiac function. We hypothesise that left ventricular ejection fraction (LVEF) can identify patients with a higher likelihood to benefit from MCS and thus help to optimise their expected benefit. Methods: Patients with non-ischaemic CS and available data on LVEF from 16 tertiary-care centres in five countries were analysed. Cox regression models were fitted to evaluate the association between LVEF and mortality, as well as the interaction between LVEF, MCS use and mortality. Results: N = 807 patients were analysed: mean age 63 [interquartile range (IQR) 51.5-72.0] years, 601 (74.5%) male, lactate 4.9 (IQR 2.6-8.5) mmol/l, LVEF 20 (IQR 15-30) %. Lower LVEF was more frequent amongst patients with more severe CS, and MCS was more likely used in patients with lower LVEF. There was no association between LVEF and 30-day mortality risk in the overall study cohort. However, there was a significant interaction between MCS use and LVEF, indicating a lower 30-day mortality risk with MCS use in patients with LVEF ≤ 20% (hazard ratio 0.72, 95% confidence interval 0.51-1.02 for LVEF ≤ 20% vs. hazard ratio 1.31, 95% confidence interval 0.85-2.01 for LVEF > 20%, interaction-p = 0.017). Conclusion: This retrospective study may indicate a lower mortality risk with MCS use only in patients with severely reduced LVEF. This may propose the inclusion of LVEF as an adjunctive parameter for MCS decision-making in non-ischaemic CS, aiming to optimise the benefit-risk ratio.