Pulmonary hypertension is defined by invasive pulmonary hemodynamics, yet reliable diagnostic and prognostic biomarkers remain limited. The authors used nuclear magnetic resonance (NMR) spectroscopy to profile circulating metabolites and lipoprotein-related parameters, aiming to identify molecular signatures that associate with hemodynamic severity and prognosis in pulmonary arterial hypertension (PAH).
The primary objective was to explore NMR-derived measures for associations with clinically relevant hemodynamic parameters and to evaluate prognostic properties in a PAH cohort. Based on observed associations with hemodynamics, the investigators defined a composite index, PHIHDL, derived from high-density lipoprotein (HDL)–related measures.
This is a retrospective analysis of patients enrolled in the prospective GRaz Pulmonary Hypertension-Metabolism (GRAPH-M) registry. Inclusion required a complete diagnostic workup including right heart catheterization and available baseline serum samples for NMR analysis.
The study included 100 patients with PAH treated at the pulmonary hypertension clinic of LKH University Hospital, Medical University of Graz, between 2011 and 2021. Baseline demographics and hemodynamics were reported as follows: mean age 61 ± 15 years, female:male ratio 2.5, body mass index 26 ± 7 kg/m2. Hemodynamic means were mPAP 41 ± 16 mmHg, PAWP 8.8 ± 3.2 mmHg, and PVR 8.0 ± 4.9 Wood units. Median survival in the cohort was 8.0 years and 46 patients died during follow-up.
All subjects provided written informed consent and the study protocol was approved by the Ethics Committee of the Medical University of Graz (EK: 23-408 ex 10/11).
Circulating metabolites and lipoprotein-related parameters were quantified by NMR spectroscopy. Among the measured variables, a cluster of 12 HDL-related measures showed consistent and significant inverse associations with pulmonary hemodynamic indices.
From these HDL-associated measures the investigators derived PHIHDL, defined in the manuscript as the reversed scaled average of those 12 HDL particle metrics. The score was constructed specifically because these HDL measures demonstrated the strongest relationships to baseline pulmonary hemodynamics in the cohort.
Details on exact individual HDL metrics used in the cluster and specific mathematical scaling steps were reported in the source; if additional computational specifics or formulae are required, those were provided in the full data resource linked by the authors.
PHIHDL showed inverse associations with pulmonary hemodynamic parameters. The identified cluster of HDL-related measures correlated negatively with parameters representing pulmonary vascular load, consistent with lower HDL-related signals in patients with worse hemodynamics.
Reported cohort hemodynamics were: mean pulmonary arterial pressure 41 ± 16 mmHg, pulmonary arterial wedge pressure 8.8 ± 3.2 mmHg, and pulmonary vascular resistance 8.0 ± 4.9 WU. PHIHDL was derived expressly from measures that associated with these and related clinical indices.
During follow-up, 46 of 100 patients died; median survival in the cohort was 8.0 years. In survival analyses adjusted for age and sex, PHIHDL was associated with all-cause mortality with a hazard ratio (HR) of 2.96 (95% CI 1.52–5.70).
The analysis indicates that lower PHIHDL (reflecting the reversed scaled average of the HDL cluster) identifies patients at higher risk of death over the observed follow-up period. The authors present PHIHDL as a pulmonary hemodynamics–based metabolomic score with prognostic relevance.
The prognostic association of PHIHDL persisted after adjustment for established clinical risk scores. Specifically, PHIHDL provided prognostic information independent of COMPERA 2.0 and REVEAL Lite2, two commonly used PAH risk stratification tools. The source reports independence from these scores but does not provide further comparative metrics in this summary beyond the adjusted hazard ratio.
The authors report that all data produced are available online at Zenodo (DOI provided). Ethical approval and written informed consent were obtained; the Ethics Committee of the Medical University of Graz approved the protocol (EK: 23-408 ex 10/11).
Funding sources listed include the Austrian Science Fund, the Austrian Research Promotion Agency, regional government and city funding, and the European Regional Development Fund. Several authors disclosed competing interests (honoraria, travel support, advisory roles) related to pharmaceutical companies; other authors declared no conflicts.
In this PAH cohort of 100 patients, an NMR-derived composite index based on a cluster of 12 HDL-related measures (PHIHDL) correlated inversely with pulmonary hemodynamics and independently predicted all-cause mortality (age- and sex-adjusted HR 2.96, 95% CI 1.52–5.70). The authors conclude that PHIHDL provides prognostic information beyond established clinical risk scores COMPERA 2.0 and REVEAL Lite2. The source includes data and code availability statements and ethical approvals. Further details on the individual HDL components and scaling methodology are available in the full dataset referenced by the authors.