https://doi.org/10.1007/s00392-024-02526-y
1Universitätsklinikum Würzburg Medizinische Klinik und Poliklinik I Würzburg, Deutschland; 2Universitätsmedizin der Johannes Gutenberg-Universität Mainz Präventive Kardiologie und Medizinische Prävention Mainz, Deutschland; 3Universitätsmedizin der Johannes Gutenberg-Universität Mainz Kardiologie 1, Zentrum für Kardiologie Mainz, Deutschland; 4Universitätsklinikum Würzburg Deutsches Zentrum für Herzinsuffizienz Würzburg, Deutschland
Despite advances in understanding the molecular pathophysiology of atrial fibrillation (AF), data on the interaction between clinical risk factors and molecular alterations remain limited. Using proteomic and bioinformatic methodologies, we aimed to i) characterize the protein profile of AF, ii) identify relevant signaling pathways linked to AF and iii) estimate the diagnostic and prognostic utility of a newly derived AF score in a cohort of heart failure (HF) patients.
Methods:
Data from the MyoVasc Study, a prospective cohort of HF (n=3.289), were analyzed. Subjects with active cancer (n=262) and/or missing protein data (n=89) were excluded. Baseline and biannual follow-up (FU) comprised deep clinical phenotyping and biobanking. All subjects with prior or primary diagnosis of AF at baseline were identified, individuals without diagnosis of AF served as controls. The AF polygenic risk score (AF-PRS) was calculated. Using proximity extension assay technology (Olink®), 536 proteins were analyzed. Differentially expressed proteins were identified and a novel protein score for AF was derived by supervised machine-learning. The association of the AF protein score with prevalent AF was studied by multivariable logistic regression analyses adjusting for age, sex, cardiovascular risk factors (RFs), humoral markers, and HF status. The prognostic utility of the AF protein score was analyzed using multivariable Cox regression models. Bioinformatic methods were applied for pathway analysis.
Results:
Prevalence of AF in the study sample was 28% (n=812; 69.9% men; mean age 68.9±9.3 years). Subjects with AF had a higher burden of cardiovascular RFs and comorbidities compared to controls. Between the AF subsample and controls, the frequency of symptomatic HF (78% vs 46%) and the AF-PRS were different (both p<0.0001). In total, 145 proteins associated with AF were identified using supervised machine learning. Coagulation factor VII (F7), fibroblast growth factor 23 (FGF-23) and tartrate-resistant acid phosphatase type 5 (TR-AP) were ranked among the best markers based on the importance metric ratio λ. Proteins with the most pronounced expression changes (Cohen’s d) showed a notable gradient of effect size from group paroxysmal to group permanent AF. The AF protein score predicted AF independent of the clinical profile (OR 3.09, 95%CI 2.56–3.73, p<0.0001; AUC 0.81). This effect persisted across the HF phenotypes. A weak association between protein score and AF-PRS confirms a significant added value of the protein score. Worsening of HF (HR 1.27; 1.09–1.47) and cardiac death (HR 1.47; 1.22–1.78) were predicted by the AF protein score independent of sex, age, cardiovascular RFs and comorbidities. Bioinformatic analysis unveiled significant regulation of the semaphorin-plexin signaling pathway as well as positive regulation of aldosterone secretion and inflammatory signaling through interleukins. External validation is currently performed in the Gutenberg Health Study (n=15010).
Conclusion:
We identified novel protein markers and signaling pathways associated with AF independent of the clinical profile. Proteomic profiling combined with bioinformatic pathway analysis emerges as a promising approach to advance our understanding of the molecular mechanisms underlying AF within the context of HF, thus providing a basis for the development of biomarkers and novel antiarrhythmic therapies.