Cardiogenic shock (CS) is the deadliest manifestation of acute heart failure, with persistently high mortality rates and a lack of recent therapeutic breakthroughs. Accurate risk prediction is crucial in clinical decision-making and the design of future clinical trials. We aimed to validate the CLIP score, a biomarker-based risk score comprising cystatin C, lactate, interleukin-6 and NT-proBNP, for predicting mortality in acute coronary syndrome (ACS) related CS, and to compare its predictive value with the previously published CardShock risk score. The study is a post hoc analysis of the CardShock Study, a prospective, observational European multicentre study on CS. The CLIP score was calculated 12h after hospital admission, and its ability to predict 90-day mortality was assessed using are under the curve (AUC) of the receiver-operating characteristics (ROC) curve analysis. The discriminative ability of the CLIP score was compared with the CardShock risk score by comparing the AUC's. The cohort was dichotomized into low and high risk groups by the optimal cut-off value derived from the ROC analysis of the CLIP score. Kaplan-Meier curves were constructed to evaluate risk stratification when combining the CLIP and CardShock risk scores. The cohort (n=121) comprised 77% (n=93) men and the median age was 67years (IQR 61-76). A total of 21% (n=25) of the patients had non-ACS related CS. The CLIP score demonstrated appropriate predictive accuracy for 90-day mortality (AUC 0.84, 95% CI 0.77-0.91), comparable with the CardShock risk score (AUC 0.77 [95% CI 0.69-0.85]; P=0.064 for comparison). A CLIP score cut-off of 0.28 stratified patients into high risk (65% mortality) and low risk (16% mortality) groups. In addition, incorporating the CLIP score enhanced risk stratification in all CardShock risk score categories. The CLIP score, calculated within 12h of hospital admission, accurately predicted 90-day mortality in CS and complemented the CardShock risk score. The biomarker-based score has potential utility in dynamic mortality risk assessment and could inform clinical management and trial design.
Read full abstract