Abstract 10064: Pulmonary Hypertension in Patients with Connective Tissue Disease-associated Interstitial Lung Disease and Vascular Changes on Computed Tomography

Abstract

Introduction: Pulmonary hypertension (PH) is a complication in patients with connective-tissue disease-associated interstitial lung disease (CTD-ILD) associated with mortality. In Idiopathic Pulmonary Fibrosis, increased vascular volume has been reported to be associated with poor outcomes. We sought to elicit whether similar changes could be observed in patients with CTD-ILD and pulmonary hypertension. Methods: We identified patients with a known diagnosis of CTD-ILD who had undergone chest CT, transthoracic echocardiography (TTE), and right heart catherization (RHC) using the Mass General Brigham Research Patient Data Repository (IRB #2018P000419). The electronic health records (EHR) were reviewed to confirm the diagnosis of CTD-ILD. A 3D reconstruction of the pulmonary vasculature using the Chest Imaging Platform ( www.chestimagingplatform.com ) to compute the total vascular volume was performed on thin sliced CT scans closest to the RHC. Vessels were separated into arterial and venous components using a deep learning approach. Statistical analysis was performed with R 4.01. Values are shown as medians and interquartile ranges and were compared using Wilcoxon Rank-Sum tests. Results: A total of 104 patients with CTD-ILD were identified: Pre-capillary PH (n=43), Isolated Post-capillary PH (n=9), Combined Pre- and Post-capillary PH (n=8), and PH Absent (n=44). There was no difference in patient age or lung volume by CT scan. On TTE, RVSP was elevated in PH (52 mmHg, IQR 41-64 mmHg vs 34 mm Hg, IQR 27-49 mmHg, p < 0.0001), as was the arterial total vascular volume (118 mL, IQR 85-152 ml vs 100 mL, IQR 77-126 ml, p = 0.009) with no trend for significance for venous total vascular volume. Addition of total arterial volume remained a statistically significant predictor of presence of PH (p = 0.01) when added to logistic regression models predicting PH by TTE RVSP, with improvement in model AIC (106.1 to 97.9) and AUC (0.75 to 0.80) of the model. Conclusions: In this retrospective study of patients with CTD-ILD and available imaging and hemodynamics, patients with PH have increased arterial vascular volume compared to those without PH. Information derived from CT images routinely acquired in this population could be used to improve screening and phenotyping.