Background: Autonomic dysfunction (AD) is a common finding in heart failure (HF). Heart rate recovery (HRR) is a measure derived from cardiopulmonary exercise testing (CPET) based on early parasympathetic reactivation. Parasympathetic withdrawal, a key component of AD, is reflected by impaired HRR. Prior research identified AD and reduced HRR as risk factor for all-cause mortality in HF. However, data explaining underlying mechanisms linking AD to HF outcome is very limited.
Methods: In the present study, data of the MyoVasc study (NCT04064450; N=3,289), a prospective cohort study on chronic HF, were analyzed. Participants were examined in a highly standardized setting including biobanking and CPET. HRR60 was defined as the difference between maximum heart rate and the heart rate at 60 seconds after termination of exercise in CPET. Biosamples were used for high-throughput proteomic profiling of N=358 proteins per individual via Immuno-PCR technology (Olink, Uppsala, Sweden). Elastic net regularized regression with HRR60 as the dependent variable adjusted for age, sex, clinical profile, and medication was calculated to identify proteomic signatures associated with HRR60. STRING analysis and Gene Ontology enrichment analysis were used to identify functional protein clusters. Based on the proteomic signature identified, a proteomic score was generated by weighting the measures of selected proteins by their beta coefficients estimated with the regression model predicting HRR60. The score was applied to the full sample of the MyoVasc cohort (N=3,289) to analyze its relationship to mortality and HF-specific outcome (i.e. worsening of HF). Information on clinical outcome was derived from structured follow-up with subsequent validation via source data and independent adjudication of clinical endpoints.
Results: The analysis sample for the identification of the proteomic signature of HRR60 comprised 1,557 individuals (mean age: 62.9±11.2 years, 33.0% females). A total number of 39 unique proteins associated with HRR60 were identified. Subsequent enrichment analysis identified three different clusters based on the function of proteins involved in the following processes: extracellular matrix and structure organization, immune cell signaling and inflammation, hemostasis and coagulation. In multivariable Cox regression analysis, the proteomic score of HRR60 was identified as a strong predictor for all-cause mortality (hazard ratio (HR) 2.28 (95% confidence interval (CI) 1.94-2.69), P<0.0001), cardiac death (HR 2.25 (95%CI 1.69-3.01), P<0.0001), and worsening of HF (HR 1.74 (95%CI 1.51-2.00), P<0.0001) independent of age and sex. After additional adjustment for clinical profile and medication the proteomic score was confirmed to be a strong predictor for all-cause mortality (HR 1.58, (95%CI 1.50-1.79), P<0.0001), cardiac death (HR 1.45 (95%CI 1.15-1.82), P<0.0001), and worsening of HF (HR 1.24 (95%CI 1.11-1.39), P=0.0002).
Conclusion: With this targeted proteomics approach, clusters of functional protein groups associated with HRR60were identified allowing further mechanistic insights in the relationship of autonomous dysfunction and HF. A protein score based on the protein signature of HRR60 was a strong predictor of HF-outcome indicating clinical relevance.