Pulmonary vein stenosis: Severity and location predict survival after surgical repair

Abstract Objectives Pulmonary vein characteristics that influence survival after repair of stenosis have not been defined. We sought to develop a predictive model relating postrepair survival to preoperative pulmonary vein characteristics on computed tomography and magnetic resonance imaging. Methods Patients who underwent pulmonary vein stenosis repair (1990-2012) with preoperative computed tomography and magnetic resonance imaging were reviewed. We measured pulmonary vein short and long cross-sectional diameters at the left atrial junction (downstream), vein bifurcation (upstream), and narrowest point, and calculated the total cross-sectional area indexed for body surface area. The relationship between pulmonary vein dimensions and survival was related via risk-adjusted parametric hazard analyses. Results Of 145 patients who underwent surgical repair, 31 had preoperative computed tomography and magnetic resonance imaging and were analyzed. Surgical repairs were sutureless (n = 30) or pericardial patch reconstruction (n = 1). Mean follow-up was 4.28 ± 4.2 years. In-hospital mortality was 9.7%; unadjusted survival was 75% ± 7%, 69% ± 8%, and 64% ± 7% at 1, 3, and 5 years, respectively. Median downstream total cross-sectional area indexed for body surface area was 163 mm 2 /m 2 , upstream total cross-sectional area indexed for body surface area was 263 mm 2 /m 2 , and total cross-sectional area indexed for body surface area at maximal stenosis, localized at the left atrial junction in approximately two thirds of patients, was 163 mm 2 /m 2 . Smaller upstream total cross-sectional area indexed for body surface area ( P  = .030) and greater number of stenotic pulmonary veins ( P  = .0069) were associated with increased early ( P  = .059). Conclusions Smaller upstream or downstream total cross-sectional area indexed for body surface area negatively influenced survival. Early survival seemed especially poor for patients with a greater number of stenotic veins and upstream pulmonary vein involvement. The total cross-sectional area indexed for body surface area measurements can help to inform prognosis and stratify patients for enrollment in clinical trials of agents directed at pulmonary vein pathology.