Risk stratification in patients with residual pulmonary hypertension after pulmonary endarterectomy

Background: Few studies addressed the issue of risk stratification in patients with residual pulmonary hypertension (PH) after pulmonary endarterectomy (PEA). This study tested the potential added value of parameters that have not been included in existing risk models.

Methods: We evaluated 546 consecutive patients with chronic thromboembolic pulmonary hypertension who underwent PEA and were followed-up for a median period of 58 months.

Results: Among the 242 with residual PH, 27 died and had 127 a clinical worsening event. At univariable analysis, the parameters associated with poor survival were pulmonary vascular resistance (PVR) ≥425 dyn·s·cm^-5 (p ≤ 0.001), mean pulmonary artery pressure (mPAP) ≥38 mmHg (p = 0.003) and pulmonary artery compliance (CPA) ≤1.8 ml/mmHg (p = 0.014). In the bivariable models including either PVR or mPAP as first parameter, the addition of CPA was not statistically significant. The parameters associated with poor clinical worsening were CPA ≤1.8 ml/mmHg (p < 0.001), PVR ≥425 dyn·s·cm^-5 (p = 0.002), arterial oxygen tension (PaO2) ≤ 75 mmHg (p = 0.003), mPAP ≥38 mmHg (p = 0.008). In a multivariable analysis which included PVR ≥425 as the first parameter, the addition of both CPA ≤1.8 ml/mmHg and of PaO2 ≤ 75 mmHg significantly improved prognostic stratification (Harrel's C of the model = 0.64, p < 0.001). Noticeably, the lower tertile of the model's predictor index identified a subgroup of 91 patients who had an event rate numerically similar to that of patients without residual PH.

Conclusions: Risk stratification in residual PH can be refined if CPA and PaO2 are considered in association with standard hemodynamic parameters.