Albumin-based inflammation scores with integrated thrombotic markers outperform conventional predictors for mortality in acute coronary syndrome and atrial fibrillation

Background Acute coronary syndrome (ACS) combined with atrial fibrillation (AF) is a critical situation with heightened inflammation. Albumin, an easily measured nutritional indicator, has been reported to reflect the overall inflammatory status of sepsis and cancer patients in recent years. However, the predictive value of albumin-based inflammation scores in ACS and AF patients' survival—a condition marked by intertwined thrombo-inflammatory pathways—remains unexplored. Purpose We aimed to develop several albumin-based inflammation scores, compare them with traditional inflammation markers, and evaluate their prognostic value in patients with ACS and AF. Methods We conducted a prospective observational study and consecutively recruited 1153 patients with AF combined with ACS (2017-2019, median follow-up 34 months; 160 all-cause deaths). The study endpoint was all-cause mortality. Receiver operating characteristic (ROC) curve with area under the curve (AUC) was used to compare the predict value and cut-off point of inflammation scores. Three multivariable adjusted Cox proportional hazards models, progressively adjusted for sex, age, body mass index, blood pressure, heart rate, type of AF and ACS, comorbidities and treatments, were performed to evaluate the associations between the scores and the endpoints. Results We developed three albumin-based inflammation scores: C-reactive protein-to-albumin ratio (CAR), Creatinine to albumin ratio (CrAR), and D-dimer-to-albumin ratio (DAR). Compared with traditional markers of inflammation, including C-reactive protein (CRP) and erythrocyte sedimentation rate (ESR), we found that CAR, CrAR and DAR had better predictive power, with DAR being the best (0.735 vs. ESR 0.666, P<0.01, Figure 1A). The optimal cut-off values of CAR, CrAR, and DAR based on the Youden index were 0.165, 2.306 and 0.012 respectively. Kaplan-Meier analysis showed that patients with higher CAR (P<0.01, Figure 1B), CrAR (P<0.01, Figure 1C) and DAR (P<0.01, Figure 1D) had a significantly worse prognosis. After adjustment for confounders, multivariable Cox models showed that higher CAR, CrAR, and DAR were independently associated with increased risk of all-cause mortality (Table 1). Restricted cubic splines showed non-linear associations between these albumin-based inflammation scores and overall survival (Figure 1E-G). Conclusions Albumin-based inflammation scores, particularly DAR integrating thrombotic activity, offer superior risk stratification over conventional inflammatory markers in ACS and AF patients. These simple, cost-effective tool scores may provide valid indicators for inflammatory and prognostic assessment of ACS and AF patients in primary care settings.