Insights into pulmonary remodelling in heart failure with preserved ejection fraction through transpulmonary biomarker gradients

Background Pulmonary hypertension (PH) is common in heart failure with preserved ejection fraction (HFpEF), driven in many patients by a combination of elevated left atrial pressure and pulmonary vascular remodelling. Whether local biomarker release contributes to pulmonary remodelling in HFpEF is unknown. Purpose We, therefore, aimed to compare the transpulmonary inflammatory biomarker release profile in healthy control subjects and HFpEF patients and to investigate the contribution of biomarkers to adverse pulmonary vascular remodelling as reflected via invasive haemodynamics. Methods Thirty-four participants (20 HFpEF, 14 healthy controls) underwent exercise hemodynamic testing and concurrent target organ blood sampling (pulmonary artery & systemic arterial). A panel of 157 inflammatory biomarkers were analysed via OLink proteomics using proximity extension assay technology and subsequent polymerase chain reaction. Transpulmonary gradient (TPG) was calculated as [pulmonary artery]-[systemic arterial]. Results HFpEF patients were older (70±9 vs 53±8 years, p<0.001) with higher BMI (34±9 vs 26±9 kg/m², p=0.004). While no healthy controls had PH, 55% of HFpEF patients did. Pulmonary vascular resistance was higher (1.95±0.83 vs 0.93±0.50WU, p<0.001), and pulmonary artery compliance (PAC) lower (4.2±1.7 vs 7.1±2.4ml/mmHg, p<0.001) in HFpEF at rest as well as at peak exercise. From 157 biomarkers, 22 demonstrated significantly different TPGs between groups, with 12 demonstrating association with invasive haemodynamics, Figure 1. Principal coordinate analysis revealed significantly different transpulmonary biomarker release profiles between HFpEF and controls (p=0.002), Figure 2. Correlations between PAC and the TPG of follistatin, agouti-related protein, lipoprotein lipase, and thrombomodulin were significantly different (p<0.05) between groups, as were correlations for exercise mean pulmonary artery pressure and TPG of thrombomodulin, leptin, ADAMTS13, adrenomedullin, CCL11, decorin, GDF-2, and PD-L2 (all p<0.05). TPGs for lipoprotein lipase, adrenomedullin, leptin, agouti-related protein, thrombomodulin, ADAMTS13, decorin and PD-L2 were also associated with differences in exercise pulmonary vasodilatory capacity between HFpEF and controls, as assessed by ΔPVR or ΔPAC (all p<0.05). Conclusion We identified a significantly altered transpulmonary flux profile of inflammatory biomarkers in HFpEF compared to healthy controls. These locally generated biomarkers demonstrated association with adverse pulmonary remodelling in HFpEF and may thereby be targets for therapeutic intervention.