Multimodal model for predicting exercise-induced pulmonary hypertension validated by invasive exercise hemodynamics:...

Area of Science:

• Cardiology
• Pulmonary Medicine
• Medical Diagnostics

Background:

• Exercise-induced pulmonary hypertension (EiPH) is an early-stage pulmonary vascular disease that is difficult to diagnose noninvasively.
• Patients with borderline resting hemodynamics pose a particular challenge for identifying EiPH.

Purpose of the Study:

• To develop and validate a multimodal non-invasive model for predicting invasively confirmed EiPH.
• The model integrates clinical characteristics, exercise echocardiography, and cardiopulmonary exercise testing (CPET).

Main Methods:

• A prospective cohort study enrolled patients with exercise limitation or increased tricuspid regurgitation velocity.
• Patients underwent clinical evaluation, stress echocardiography, CPET, and invasive right heart catheterization.
• A logistic regression model with elastic net regularization was developed and validated using cross-validation and subgroup analyses.

Main Results:

• The multimodal Clinical+Echo+CPET model achieved an AUC of 0.951, demonstrating strong discriminative ability.
• The model showed good calibration with an accuracy of 0.871 and a Brier score of 0.128.
• The multimodal model outperformed unimodal models and maintained performance across different age and sex groups.

Conclusions:

• A non-invasive multimodal model integrating clinical indicators, exercise echocardiography, and CPET reliably identifies EiPH.
• This approach provides a valuable tool for the early detection of pulmonary vascular disease.