3014Phenomapping to detect non-invasively patients with combined postcapillary pulmonary hypertension

Abstract

Background: By essence, pulmonary hypertension (PH) is dichotomized in preand postcapillary. The combination of pre- and postcapillary PH in one patient is called “combined-postcapillary PH”. Nowadays, detection of this phenotype is impossible unless right heart catheterization is performed. Purpose: Using dense phenotypic data (phenomapping), the purpose of this work was to characterize phenotypically and noninvasively patients with combined-postcapillary PH. Methods: 82 consecutive patients (mean age 67.8±11.6 y, males 47%) from a referral center were included between 2009 and 2017. Inclusion criteria were as follows: mean pulmonary artery pressure (PAP) ≥25 mmHg and wedge pressure (WP) >15 mmHg. Combined-postcapillary PH patients were those with a diastolic pressure gradient ≥7 mmHg and/or pulmonary vascular resistance (PVR) >3 WU. Phenotypic domains were clinical, laboratory, imaging, hemodynamic and functional respiratory continuous variables. Parameters were imputed with 10 eigenvectors for missing values, filtered if the Spearman correlation coefficient was >0.6, standardized to mean±SD of 0±1, grouped using agglomerative hierarchical clustering and participants were separated. Phenogroups were compared for phenotypic domains (p<0.05 was considered significant) and Cox regression was used for survival analysis at 2 years follow-up. Results: Mean PAP, WP, cardiac index and PVR were 39.6±9.8, 23.2±5.2, 2.96±0.77 and 3.21±1.77 respectively. Among the 65 phenotypic domains entering the model, 31 were selected from which, heatmap was drawn (fig 1) and Penalized Model-Based Clustering identified 3 phenogroups. Combinedpostcapillary PH was found in 90% of phenogroup 3 remarkable by older age females (p<0.01), more frequent and uncontrolled systemic hypertension, more severe functional status, more frequent atrial fibrillation, super-normal ejection fraction, smaller ventricular cavities and more elevated tricuspid regurgitation velocity (above 4 m/s) compared to phenogroups 1 and 2. Compared to phenogroup 1, two-years all-causes mortality (fig 2) was not increased in phenogroup 3 (Hazard Ratio: 0.79 [0.2 - 3.14], p=0.73) but in phenogroup 2, characterized by younger males with lower ejection fraction (48.8), intermediate PAP and PVR with only 27% combined-PH forms (Hazard Ratio: 3.74 [1.24 - 11.34], p=0.02). Conclusion: From heterogeneous clinical presentation, statistical learning algorithms applied to dense phenotypic data allowed to identify non-invasively a population at high risk of combined-postcapillary PH. Phenomapping could also be applied to characterize PH patients with high mortality risk. (Figure Presented).