Breath acetone concentration: too heterogeneous to constitute a diagnosis or prognosis biomarker in heart failure? A systematic review and meta-analysis.

Introduction. Exhaled breath acetone (ExA) has been investigated as a biomarker for heart failure (HF). Yet, barriers to its use in the clinical field have not been identified. The aim of this systematic review and meta-analysis was to assess the ExA heterogeneity and factors of variability in healthy controls (HC), to identify its relations with HF diagnosis and prognostic factors and to assess its diagnosis and prognosis accuracy in HF patients.Methods. A systematic search was conducted in PUBMED and Web of Science database. All studies with HC and HF patients with a measured ExA were included and studies providing ExA's diagnosis and prognosis accuracy were identified.Results. Out of 971 identified studies, 18 studies involving 833 HC and 1009 HF patients were included in the meta-analysis. In HC, ExA showed an important heterogeneity (I2= 99%). Variability factors were fasting state, sampling type and analytical method. The mean ExA was 1.89 times higher in HF patients vs. HC (782 [531-1032] vs. 413 [347-478] ppbv;p< 0.001). One study showed excellent diagnosis accuracy, and one showed a good prognosis value. ExA correlated with New York Heart Association (NYHA) dyspnea (p< 0.001) and plasma brain natriuretic peptide (p< 0.001). Studies showed a poor definition and reporting of included subjects.Discussion. Despite the between-study heterogeneity in HC, the evidence of an excellent diagnosis and prognosis value of ExA in HF from single studies can be extended to clinical populations worldwide. Factors of variability (ExA procedure and breath sampling) could further improve the diagnosis and prognosis values of this biomarker in HF patients.