Background
Converging evidence from proteogenomic analyses prioritises thrombospondin-2 (TSP2) as a potential biomarker for idiopathic or heritable pulmonary arterial hypertension (PAH). We aimed to assess TSP2 levels in different forms of pulmonary hypertension (PH) and to define its clinical phenotype.
Methods
Absolute concentrations of TSP2 were quantified in plasma samples from a prospective single-centre cohort study including 196 patients with different forms of PH and 16 disease controls (suspected PH, but normal resting pulmonary haemodynamics). In an unbiased approach, TSP2 levels were related to 152 clinical variables.
Results
Concentrations of TSP2 were increased in patients with PH versus disease controls (p<0.001 for group comparison). The discriminatory ability of TSP2 levels to distinguish between patients and controls was superior to that of N-terminal pro-brain natriuretic peptide (p=0.0023 for comparison of areas under the curve). Elevation of TSP2 levels was consistently found in subcategories of PAH, in PH due to lung disease and due to left heart disease. Phenotypically, TSP2 levels were robustly related to echocardiographic markers that indicate the right ventricular (RV) response to chronically increased afterload with increased levels in patients with impaired systolic function and ventriculoarterial uncoupling. Focusing on PAH, increased TSP2 levels were able to distinguish between adaptive and maladaptive RV phenotypes (area under the curve 0.87, 95% CI 0.76-0.98). Interpretation: The study indicates that plasma TSP2 levels inform on the presence of PH and associate with clinically relevant RV phenotypes in the setting of increased afterload, which may provide insight into processes of RV adaptability.