Terry, J.G.

Abstract Rationale: Prior studies report association between pulmonary artery (PA) dilation on computed tomography (CT) imaging and pulmonary hypertension (PH). These studies predate expansion of PH hemodynamic diagnostic criteria and do not report patient mortality. As such, little is known about the presence of vascular changes in mild PH or their association with survival. We sought to investigate these relationships in individuals without significant parenchymal lung disease. Methods: We obtained CT images for subjects with group 1 or 2 PH, disease comparators, and healthy controls enrolled in the Pulmonary Vascular Disease Phenomics (PVDOMICS) study. We used Osirix imaging software to measure the main, left, and right PA (MPA, LPA, and RPA) and MPA:aorta (PA:Ao) ratio by diameter and cross-sectional area at locations consistent with prior studies. We evaluated associations with mortality and mean PA pressure (mPAP) using multivariable Cox and linear regression models adjusting for age, sex, and BMI. Descriptive statistics are presented as mean ± SD. Results: We identified 558 subjects (59 controls, 138 comparators, and 361 PH [267 group 1, 94 group 2]). 79 subjects had mPAP < 21 mmHg, 39 had mPAP 21-24 mmHg, and 371 had mPAP ≥ 25 mmHg. Increased MPA, LPA, and RPA diameter and area, PA:Aodia, and PA:Aoarea were associated with decreased survival (all p < 0.001) after multivariable adjustment. First versus third quantiles of all measurements were associated with mortality, with adjusted hazard ratio of 2.99 [95% CI 1.77, 5.05] for MPA diameter (2.84 vs 3.82 cm), 3.74 [2.07, 6.74] for MPA area (7.26 vs 12.74 cm2) and 5.24 [2.58, 10.66] for LPA area (4.55 vs 7.03 cm2). All measures significantly increased alongside PH severity from mPAP < 20 to 21-24 to ≥ 25 mmHg (see figure 1 for MPA diameter). All measures positively correlated with mPAP, including MPA diameter (ρspearman = 0.55) and PA:Aodia (ρ = 0.56). After multivariable adjustment, Q1 to Q3 increments in MPA diameter and PA:Aodia were associated with mPAP increase of 12.5 [10.7, 14.4] and 13.9 [11.9, 15.8] mmHg, respectively. BMI showed no correlation with PA size (ρ = 0.076 for MPA area, -0.051 for PA:Aodia). Conclusions: PA size increases alongside PH severity using updated hemodynamic criteria and is associated with incremental increases in PA pressure and mortality. LPA dilation demonstrates particularly strong association with mortality. PA size does not correlate with BMI. These novel findings enhance our understanding of vascular changes across the PH disease spectrum.