Using multiparametric approach, pulmonary arterial hypertension (PAH) treatment is goal-oriented to get an estimated 1-year mortality below 5 %. Yet, transthoracic echocardiography (TTE) interest in a goal-oriented treatment strategy remains sparse.

We aimed to investigate whether the non-invasive serial assessment of several parameters by TTE may help to achieve this goal.

Between 2002 and 2019, all consecutive PAH patients were included and treated in a single PAH centre, using the current guidelines. As recommended, patients were reassessed within a year for PAH risk stratification. In addition to regular parameters, thirteen right ventricular TTE indices and their changes were either measured. Primary composite outcome was 3-years all-cause mortality or pulmonary transplantation. We used supervised machine learning with conditional inference tree, which provides more information than Cox Proportional analysis to link TTE parameters to risk stratification and to determine cut-off values from hierarchy of multiple covariates. Next, survival analysis was performed using Kaplan–Meier curves (log-rank test).

In total, 126 patients were included (63% female, mean age 59±18 y). At baseline, 70% were in NYHA functional class 3 or 4. Mean pulmonary arterial pressure, 6-min walking distance and NT-pro-BNP levels averaged 45±17mmHg, 290±136m and 2127±3438ng/L, respectively. From conditional inference tree, TAPSE increase>0mm (P<0.01), tricuspid S wave increase>1cm/s (P<0.01) and RV end-diastolic area decrease>2 cm² (P<0.01) at 1-year were of prognostic significance, even after adjustment to their baseline value. One, two and 3-year mortality rate with TAPSE improvement is 3.6%, 18.2%, 29.2%, respectively (Fig. 1).

One-year TAPSE improvement (>0mm) under vasodilatation therapy in PAH should be a new goal-oriented treatment strategy to get an estimated 1-year mortality below 5%, especially when regular parameters are challenging.