Decoding WHO functional class in pulmonary hypertension: optimal integration of conventional and speckle-tracking echocardiographic parameters

Abstract Aims Functional assessment of pulmonary hypertension (PH) patients using the World Health Organization Functional Class (WHO-FC) is crucial for treatment guidance but requires objective quantification. This cross-sectional, single-center study aimed to identify conventional and 2D-speckle tracking echocardiographic parameters of right ventricular (RV) and right atrial (RA) function—individually and in predictive models—that best differentiate WHO-FC in PH patients. Methods and Results Sixty-one PH patients (mean age 61 years, 61% female, 87% precapillary PH) underwent right heart catheterization and echocardiography. WHO-FC distribution: I (n=25), II (n=25), III/IV (n=11). RV function was assessed using fraction of area change (FAC), tricuspid annular plane systolic excursion (TAPSE), global longitudinal strain (GLS), free wall longitudinal strain (FWLS), and myocardial work indices: global work index (GWI), global constructive work (GCW), global wasted work (GWW), and global work efficiency (GWE). RA function was evaluated via reservoir strain (RA RS), conduit strain (RA CDS), and contractile strain (RA CONS). RV-arterial coupling was analyzed by indexing RV parameters to echocardiographically estimated RV systolic pressure (sPAP): TAPSE/sPAP, RV GLS/sPAP, and RV FWLS/sPAP. RA-ventricular coupling was assessed by indexing RA strain parameters to sPAP: RA RS/sPAP, RA CDS/sPAP, and RA CONS/sPAP. Most RV and RA parameters progressively declined across WHO-FC (I → II → III/IV), with stronger correlations when normalized to sPAP (Figure 1). Univariate ordinal regression identified 14 echocardiographic predictors of WHO-FC (Figure 2), refined to 7 after collinearity analysis . In multivariate analysis, RV GCW was the only independent predictor (OR = 0.58, 95% CI: 0.39–0.87, p = 0.008). Hierarchical model comparisons (Models A, B, C) showed that adding RA CDS/sPAP to RV GCW significantly improved model fit (p = 0.008), highlighting RA function’s role in WHO-FC stratification (Figure 2). Adding TAPSE/sPAP provided only borderline improvement (p = 0.08). Model D (RV GLS/sPAP, -2LL = 88.280, R² = 0.536) significantly outperformed Model B (RV GCW + RA CDS/sPAP, p = 0.018) but showed no significant difference from the full multiparametric Model C (p = 0.111). Although Model C exhibited the highest accuracy (73.8%) and F1 score (72.3%), Model D had comparable classification performance (accuracy 72.1%, F1 score 69.7%). Conclusions (i) WHO-FC progression reflects significant RV and RA dysfunction, quantifiable via 2D speckle-tracking; (ii) Both RV and RA strain parameters correlate better with WHO-FC when normalized to RV afterload; (iii) RV GLS/sPAP is the strongest individual WHO-FC predictor but shows high collinearity; (iv) RV GCW is the only independent WHO-FC predictor; (v) The best multiparametric model includes RV GCW, RA CDS/sPAP, and TAPSE/sPAP, yet RV GLS/sPAP alone provides superior fit with comparable classification performance.