Background: The DanGer-Shock trial recently demonstrated that active left ventricular (LV) unloading improves outcomes in patients with cardiogenic shock (CS) due to acute ST-elevation myocardial infarction (STEMI). This analysis evaluated the effects of active LV unloading in patients with CS due to heart failure (HF-CS).
Methods: Data of patients with HF-CS from 16 tertiary-care centers in five countries collected between 2010 and 2021 (NCT03313687) were analyzed. Patients without STEMI but fulfilling the other DanGer-Shock enrollment criteria (cardiogenic shock, left ventricular ejection fraction <45%, lactate ≥2.5 mmol/L, no prolonged cardiopulmonary resuscitation) were included. The primary outcome 30-day mortality and in-hospital complications were compared in patients receiving active LV unloading (Impella only or ECMELLA) and in patients without LV unloading using adjusted Cox and logistic regression models.
Results: Of 154 patients with HF-CS meeting DanGer-Shock-like enrollment criteria, 53 (34.4%) were treated with active LV unloading. The 30-day mortality was 31.6% vs. 41.5% in those with vs. without active LV unloading (adjusted hazard ratio 0.828, 95% CI 0.668–0.996, p=0.045, Figure1). However, patients with active LV unloading presented with a higher likelihood of in-hospital complications including moderate bleeding (adjusted odds ratio (OR) 2.43, 95% CI 1.88–3.14, p<0.01), severe bleeding (OR 2.84, 95% CI 1.89–4.26, p<0.01), and interventions due to bleeding (OR 4.70, 95% CI 3.22–6.85, p<0.01), as well as the need for renal replacement therapy (OR 2.90, 95% CI 2.30–3.65, p<0.01) and the occurrence of sepsis (OR 1.99, 95% CI 1.54–2.57, p<0.01). In contrast, active LV unloading was associated with a lower likelihood of ischemic stroke (OR 0.46, 95% CI 0.22–0.93, p=0.03, Figure 2).
Conclusions: These hypothesis-generating data suggest that active left ventricular unloading could improve outcomes in patients with CS due to heart failure without prolonged resuscitation. Randomized trials with sufficient follow-up are needed to further evaluate active left ventricular unloading in this population.
Figure 1.
Kaplan-Meier estimates for 30-day all-cause mortality in patients with vs. without active left ventricular unloading. Hazard ratio (HR) adjusted by age, sex, lactate, pH, prior cardiopulmonary resuscitation, and Society for Cardiovascular Angiography & Interventions (SCAI) class. CI, confidence interval.
Figure 2.
Association between active left ventricular (LV) unloading and safety endpoints. To evaluate the association between active LV unloading and secondary endpoints, logistic regression models were applied, adjusted for age, sex, lactate and pH at baseline, prior cardiopulmonary resuscitation (<10 min), and Society for Cardiovascular Angiography & Interventions (SCAI) class.
