Exercise haemodynamics in pulmonary hypertension - a prospective pressure-volume loop study on right ventricular adaptation and prognosis.

The haemodynamic response to exercise is prognostic in pulmonary hypertension (PH). However, little is known about right ventricular (RV) adaptation in this context. We analysed the patterns and prognostic relevance of RV adaptation to exercise in PH. We prospectively analysed 46 patients with PH and 19 disease controls with invasive exclusion of PH. All underwent three-dimensional echocardiography, pressure-volume catheterization, and right heart catheterization at rest and during stepwise exercise on a semi-supine ergometer. Patients with PH were classified as homeometric if they had increased RV end-systolic elastance and preserved RV-pulmonary arterial coupling (end-systolic/arterial elastance) during exercise (18 patients); otherwise, they were classified as heterometric (28 patients). The mean pulmonary arterial pressure/cardiac output (mPAP/CO) slope was similar in the homeometric and heterometric groups (8.8 [6.5-13.1] vs. 8.6 [4.8-18.8] mmHg·min/L), and lower in disease controls (2.1 [1.1-4.0] mmHg/L). Multivariable logistic regression identified systolic pulmonary arterial pressure change during exercise (ΔsPAP) (odds ratio [OR] 0.93, 95% confidence interval [CI] 0.87-0.99; p = 0.019) and peak exercise cardiac index (OR 0.42, 95% CI 0.18-0.97; p = 0.042) as key differentiators of homeometric/heterometric adaptation. Heterometric adaptation was significantly associated with clinical worsening and all-cause mortality (log-rank p = 0.0006 and p = 0.0246, respectively) and independently predicted clinical worsening (hazard ratio [HR] 6.52, 95% CI 2.16-19.63; p = 0.001); the HR for all-cause mortality was 6.96 (95% CI 0.87-55.66; p = 0.067). Pulmonary hypertension can present with two RV patterns under stress: homeometric with increased contractile reserve and heterometric with poorer outcome. While the mPAP/CO slope does not differentiate the two, ΔsPAP and peak cardiac index offer potential for RV adaptation pattern identification and thus prognostication.