15 February 2020 by Christopher Rhodes

The combination of metabolomic and miRNA diagnostic scores to identify pulmonary hypertension and sub-classify pulmonary arterial hypertension

The patient journey from first symptoms to a diagnosis of pulmonary hypertension (PH) typically takes 1-3 years, and despite increased awareness, therapeutic advances and improved outcomes, this has remained unchanged over the last 20 years. PH is characterised by disturbed metabolic function and altered levels of circulating miRNA such as miR-150 have been associated with pulmonary arterial hypertension (PAH, group 1 PH) and survival. In systemic sclerosis, screening algorithms to identify patients earlier can improve outcome but remain resource intensive. Early diagnosis and classification of PH and PAH could be assisted by simple non-invasive testing of circulating metabolites and miRNA. 

Serum from 1521 patients with PH and disease comparators (chronic thromboembolic disease and patients referred to PH clinics who had PH ruled out) and healthy controls were collected from 3 UK expert centres (Imperial, Papworth and Sheffield) and analysed for 1523 metabolites (Metabolon) and 554 miRNAs (MiRXES). In a discovery-validation design, molecules with classifying profiles in PH and PAH patients were identified, and diagnostic models constructed. Prognosis was also assessed in PAH patients divided by miR-150 levels. 

A LASSO model of 5 miRNA with age, sex, BMI and eGFR distinguished PAH from other PH groups and disease comparators. Metabolomic profiles were able to differentiate all PH groups from disease comparators and healthy controls in univariate analyses and LASSO modelling achieved an AUC of 0.84. The combination of these models was able to distinguish both PH from controls and PAH from other PH groups. Low miR-150 levels identified PAH patients with poor survival outcomes. 

High-throughput profiling of circulating metabolites and miRNA enabled construction of models capable of distinguishing PH patients from relevant comparator groups and differentiate subgroups of PH. Prospective studies validating the clinical utility of these models in clinical settings early in the diagnostic pathway are required 

Key Contributors

Christopher J Rhodes, Timothy Jatkoe, John Wharton, Dennis Wang, Niamh Errington, Charles Bridges, Yiu-Lian Fong, Mark Toshner, Luke S Howard, Allan Lawrie, Martin R Wilkins Imperial College London, UK Jannsen, Johnson & Johnson Innovations, USA University of Sheffield, UK University of Cambridge, UK


Comments (0)

Our research platform is the world.

Through worldwide collaboration, we can begin to answer the question of a global disease.

Join the PVRI
standard-example-image.jpg