Prediction of survival after lung transplantation at one year (SALTO cohort) using information available at different key time-points.

Abstract

Lung transplantation is the final treatment option for end-stage lung disease. In this study, we evaluated the individual risk of 1-year mortality at each stage of the lung transplantation process. This was a retrospective analysis of patients undergoing bilateral lung transplantation between January 2014 and December 2019 in three French academic centers. Patients were randomly divided into development and validation cohorts. Three multivariable logistic regression models of 1-year mortality were applied (A) at recipient registration, (B) the graft allocation, and (C) after surgery. The 1-year mortality was predicted for individual patients assigned to three risk groups at time points A-C. The study population consisted of 478 patients with a mean (SD) age of 49.0 (14.3) years. The 1-year mortality rate was 23.0%. There were no significant differences in patient characteristics between the development (n = 319) and validation (n = 159) cohorts. The models analyzed recipient, donor, and intraoperative variables. The discriminatory power (area under the receiver operating characteristic curve) was 0.67 (0.62-0.73), 0.70 (0.63-0.77), and 0.82 (0.77-0.88), respectively, in the development cohort, 0.74 (0.64-0.85), 0.76 (0.66-0.86) and 0.87 (0.79 - 0.95), respectively, in the validation cohort. Survival rates were significantly different among the low- (< 15%), intermediate- (15%-45%), and high-risk (> 45%) groups in both cohorts. Risk prediction models allow estimation of the 1-year mortality risk of individual patients during the lung transplantation process. These models may help caregivers identify high-risk patients at times A-C, and reduce the risk at subsequent time-points.