CA 125 in the diagnosis and risk stratification of acute heart failure

Abstract

Abstract Background Recent evidence confirms the elevation of CA 125 in non-tumor processes such as acute heart failure (AHF). However, the utility of this novel biomarker for diagnosis, prognosis, and therapy guidance in AHF remains unclear. Purpose To investigate the potential of CA 125 for diagnosis, prognosis and therapy guidance in unselected AHF patients presenting with acute dyspnea to the emergency department (ED). Methods We quantified CA 125 in a blinded fashion among patients presenting with acute dyspnea to the ED in a multicenter diagnostic study. Final diagnosis of AHF including AHF-phenotype was centrally adjudicated by two independent cardiologists. To further characterize CA 125's potential in AHF correlations with established biochemical and imaging markers were assessed. Diagnostic accuracy for AHF was quantified by the area under the receiver operating characteristic curve (AUC). All-cause mortality within 360 days was the prognostic endpoint. Results Among 470 patients eligible for this analysis, 268 (57.0%) had adjudicated AHF. CA 125 concentrations at presentation were significantly higher among AHF patients vs. patients with other final diagnoses (45.8 U/ml [interquartile range (IQR), 18.5–110.3] vs. 16.2 U/ml [IQR, 9.6–31.6], p<.001). Patients with worsening heart failure had significant higher CA 125 levels compared to other heart failure phenotypes (p=.018). There was a significant positive correlation of CA 125 and high-sensitivity cardiac troponin T and NTproBNP and a significant negative correlation of CA 125 and left ventricular ejection fraction (correlation coefficients 0.204, 0.220, −0.331, respectively; all ps<.001). CA 125's AUC for AHF was significantly lower compared to NTproBNP's in the overall population (0.72, 95% confidence interval (CI) 0.67–0.76 vs. 0.93, 95% CI 0.90–0.95, p<.001, Figure 1) and in predefined subgroups according to age, gender and renal function. Among 268 AHF patients, 84 (31.3%) died within 360 days of follow-up. CA 125 plasma concentrations above the median indicated increased risk of all-cause mortality (hazard ratio 2.06, 95% CI 1.31–3.24; p=.002, Figure 2). CA 125's prognostic accuracy for 360-days mortality was comparable with NT-proBNP's and high-sensitivity cardiac troponin T's. CA 125 did not independently predict all-cause mortality at 360 days when used in validated multivariable regression models and had no interactions with medical therapies at discharge. Conclusion CA 125 may aid physicians in the risk stratification and rapid triage of patients with suspected AHF. Funding Acknowledgement Type of funding sources: Foundation. Main funding source(s): Swiss National Science FoundationSwiss Heart Foundation Figure 1. ROC curve comparisonFigure 2. Kaplan-Meier curve 360 days mortality