Background: Pulmonary arterial hypertension (PAH) is a progressive disease with lung transplantation as the only option for those patients refractory to medical therapy. Although several equations have been developed to predict PAH patient survival, it is unclear whether they can predict survival for patients awaiting transplantation.
Methods: Data were analyzed on 827 patients listed since 1991 on the Scientific Registry of Transplant Recipients. Overall survival and survival for patients listed prior to and after January 1, 2006 was estimated using the Kaplan-Meier (K-M) method and compared with predicted survival from the pulmonary hypertension connection (PHC) and lung allocation system (LAS) equations. A new equation using a novel model selection algorithm for correlated covariates and missing data was developed using clinical factors and variables in the LAS score. Model validation statistics were calculated and averaged across 500 bootstrap resamples within each of 5 imputation data sets. K-M with 95% confidence intervals and receiver-operator characteristic (ROC) curves assessed model performance.
Results: PHC predicted overall survival but underestimated and overestimated survival for those listed pre- and post-2006, respectively. The best model included baseline 6-minute walk distance (6MWD), invasive cardiac output and resting oxygen requirement (O2). Factors associated with 1-year waitlist survival included: resting O2 amount; invasive hemodynamics; 6MWD; and functional class. The new equation by ROC analysis outperformed the LAS and PHC equations.
Conclusions: Current prediction models overestimate survival for PAH patients listed for transplant in the LAS era. This new survival equation can help guide clinicians caring for PAH patients with progression of disease requiring transplant.
Keywords: lung allocation score; lung transplant; pulmonary arterial hypertension; survival prediction model; waitlist.
Copyright © 2013 International Society for Heart and Lung Transplantation. Published by Elsevier Inc. All rights reserved.