Validation of SAVE score and the value of serum lactate in patients admitted for cardiogenic shock treated with VA-ECMO

Abstract

Abstract Introduction Veno-Arterial extracorporeal membrane oxygenation (VA-ECMO) is a treatment option to provide circulatory and pulmonary support to patients with cardiogenic shock. However, a risk profile assessment is essential for an adequate selection of patients for this type of therapy. The aim of this study was to 1) validate the SAVE score in a Portuguese cohort of patients treated with VA-ECMO due to cardiogenic shock; 2) evaluate the prognostic impact of the maximum serum lactate level pre-VA-ECMO implantation; 3) assess the ability of lactate to improve risk stratification by the SAVE score. Methods We conducted a single-center retrospective analysis of patients treated with VA-ECMO due to cardiogenic shock from 2017 until 2022. Variable assessments were considered before VA-ECMO implantation. The primary outcome analyzed was in-hospital mortality. Results A total of 61 patients were included (52±12 years, 74% male, 40% with acute myocardial infarction, 54% with an ejection fraction <20%, and 74% in SCAI stage D pre-implantation). Overall, 38 (62%) died during hospitalization. The mean SAVE score was -1.7±7.2 points, and the median maximum serum lactate before ECMO implantation was 5.8 [2.7; 11.8] mmol/L. SAVE score showed a statistically significant association (1±5 vs. -4±7; OR 0.872 [0.789; 0.964] per each point increase; p=0.008) and good discriminative power (AUC 0.710; p=0.006) to predict in-hospital mortality. When grouping patients according to this score, 25 (41%) were classified as SAVE risk class I or II, and 36 (59%) as SAVE class III, IV, or V. Maximum serum lactate before VA-ECMO implantation also showed a significant association (4.5±3.1 vs. 9.6±6.6; OR 1.264 [1.066;1.498] per each 1mmol/l increase; p=0.007) and good discriminative power (AUC 0.771; p=0.001) to predict the primary outcome. The best lactate cut-off to identify high mortality risk was 5 mmol/l with a sensitivity of 74% and a specificity of 76%. Additionally, taking into account this threshold, lactate significantly enhanced the SAVE score group stratification, with a net reclassification improvement of 36.7% (p=0.021). Conclusion In this cohort of patients, the SAVE score was significantly associated with in-hospital mortality. Maximum serum lactate before VA-ECMO implantation was a strong predictor of in-hospital mortality and significantly improved SAVE score risk stratification.