Lipidomics for diagnosis and prognosis of pulmonary hypertension

Abstract

Pulmonary hypertension (PH) is a severe hemodynamic, progressive condition associated with high morbidity and mortality where early and less invasive diagnostics could crucially improve management. There is a need for biomarkers in PH that are functional, diagnostic, and prognostic. We used a broad metabolomics approach with machine learning analysis and specific free fatty acid (FFA)/lipid-ratios to develop diagnostic and prognostic PH biomarkers. In a training cohort of 74 PH patients, 30 disease controls without PH, and 65 healthy controls, we identified diagnostic and prognostic markers that were validated in an independent cohort of 64 subjects. Markers based on lipophilic metabolites were more robust than those based on hydrophilic metabolites. FFA/lipid-ratios provided excellent diagnostic accuracy for PH with an AUC of up to 0.89 and 0.90 in the training and the validation cohorts, respectively. The ratios provided age-independent prognostic information and a combination of a ratio with established clinical scores increased the hazard ratio (HR) for FPHR4p and COMPERA2 from 2.5 to 4.3 and from 3.3 to 5.6, respectively. Pulmonary arteries (PA) of idiopathic PAH (IPAH) lungs show lipid accumulation and altered expression of lipid homeostasis-related genes that may explain this accumulation. Our functional studies in PA endothelial and smooth muscle cells have shown that increased FFA levels caused excessive proliferation and PA endothelial barrier dysfunction, both hallmarks of pulmonary artery hypertension (PAH). In conclusion, lipidomic changes in PH provide novel diagnostic and prognostic biomarkers and may point to new metabolic therapy targets.