Advanced hemodynamic and cluster analysis for identifying novel RV function subphenotypes in patients with pulmonary hypertension

Alexandra M. Janowski, Keeley S. Ravellette, Michael Insel, Joe G.N. Garcia, Franz P. Rischard, Rebecca R. Vanderpool

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Background: Quantifying right ventricular (RV) function is important to describe the pathophysiology of in pulmonary hypertension (PH). Current phenotyping strategies in PH rely on few invasive hemodynamic parameters to quantify RV dysfunction severity. The aim of this study was to identify novel RV phenotypes using unsupervised clustering methods on advanced hemodynamic features of RV function. Methods: Participants were identified from the University of Arizona Pulmonary Hypertension Registry (n = 190). RV-pulmonary artery coupling (Ees/Ea), RV systolic (Ees), and diastolic function (Eed) were quantified from stored RV pressure waveforms. Consensus clustering analysis with bootstrapping was used to identify the optimal clustering method. Pearson correlation analysis was used to reduce collinearity between variables. RV cluster subphenotypes were characterized using clinical data and compared to pulmonary vascular resistance (PVR) quintiles. Results: Five distinct RV clusters (C1-C5) with distinct RV subphenotypes were identified using k-medoids with a Pearson distance matrix. Clusters 1 and 2 both have low diastolic stiffness (Eed) and afterload (Ea) but RV-PA coupling (Ees/Ea) is decreased in C2. Intermediate cluster (C3) has a similar Ees/Ea as C2 but with higher PA pressure and afterload. Clusters C4 and C5 have increased Eed and Ea but C5 has a significant decrease in Ees/Ea. Cardiac output was high in C3 distinct from the other clusters. In the PVR quintiles, contractility increased and stroke volume decreased as a function of increased afterload. World Symposium PH classifications were distributed across clusters and PVR quintiles. Conclusions: RV-centric phenotyping offers an opportunity for a more precise-medicine-based management approach.

Original languageEnglish (US)
Pages (from-to)755-770
Number of pages16
JournalJournal of Heart and Lung Transplantation
Volume43
Issue number5
DOIs
StatePublished - May 2024

Keywords

  • RV function
  • RV subphenotypes
  • cardiopulmonary hemodynamics
  • machine learning
  • pressure-volume loop
  • pulmonary hypertension
  • right heart failure

ASJC Scopus subject areas

  • Surgery
  • Pulmonary and Respiratory Medicine
  • Cardiology and Cardiovascular Medicine
  • Transplantation

Fingerprint

Dive into the research topics of 'Advanced hemodynamic and cluster analysis for identifying novel RV function subphenotypes in patients with pulmonary hypertension'. Together they form a unique fingerprint.

Cite this