BACKGROUND: Acute-on-chronic heart failure (ACHF) currently represents the leading etiology of cardiogenic shock (CS). We aimed to assess the prognostic value of history of heart failure (HF) in patients with refractory CS as well as its effect on the benefit of moderate hypothermia (MH) (33-34 °C). METHODS: Of the 334 patients included in the HYPO-ECMO trial, 321 (96 %) had available HF history information, among whom 65 (20 %) had prior HF. Inverse probability weighting (IPW) was used to compare ACHF patients and de-novo HF (DNHF) patients. Primary outcome was all-cause mortality at day 30. Main secondary outcomes were mortality and the composite of death, heart transplant, escalation to left ventricular assist device, or stroke up to day 180. RESULTS: At 30 days, 26 patients (40.0 %) died in the ACHF group versus 122 patients (47.7 %) in the DNHF group (crude risk difference (RD), -7.7 % [-21.0 to 5.7] p = 0.26
IPW RD, -11.6 % [-24.8 to 1.6] p = 0.084). Mortality (IPW RD, -13.7 % [-27.1 to -0.2], p = 0.047) and the composite outcome (IPW RD, -19.5 % [-32.9 to -6.1], p = 0.004) were significantly lower at day 180 in the ACHF group. Patients randomized to MH tended to have a lower risk for the primary outcome (RD -10.9 %, [-23.1 to 1.2], p = 0.078) and a significant reduction in composite outcome (p <
0.05 at each timepoint) in the DNHF group but not in the ACHF group, despite the absence of a significant interaction (p >
0.05). CONCLUSIONS: In VA-ECMO-treated CS, ACHF was associated with comparable 30-day survival but lower 180-day mortality and morbidity-mortality. In this exploratory post-hoc analysis, MH appeared to be associated with improved outcomes in DNHF patients only. CLINICALTRIALS: gov Identifier: NCT02754193.