Predictive Utility of Lung Allocation Score for Retransplantation Outcomes

Abstract

Listen

Translate article

Treatment of primary graft failure after lung transplantation (LTx) may include retransplantation (rLTx). The number of rLTx cases has doubled since implementation of the Lung Allocation Score in 2005. The Lung Allocation Score was intended to predict LTx outcomes, but its predictive utility has not been assessed in rLTx. We investigated whether 1-year outcomes of LTx and rLTX were equally well predicted by the Lung Allocation Score. Recipients of LTx and rLTx aged 18 years or more were identified in 2005 to 2015 United Network for Organ Sharing data. The Lung Allocation Score was entered in multivariable logistic regression models of 1-year retransplant-free survival. Areas under the receiver-operating characteristics curve summarized model predictive value. We examined whether the Lung Allocation Score and its components were differentially associated with outcomes of LTx and rLTx. There were 16,837 LTx and 765 rLTx cases meeting inclusion criteria. Crude 1-year retransplant-free survival rates were 86% after LTx compared with 74% after rLTx. On univariate analysis, both LTx and rLTx cohorts showed poor predictive utility of the Lung Allocation Score (area under the curve 0.55 and 0.57, respectively; difference by transplant type, p= 0.307). Neither the Lung Allocation Score nor its components was differentially associated with LTx compared with rLTx outcomes. The Lung Allocation Score achieved comparable, but poor, predictive utility for 1-year outcomes of primary LTx and rLTx. We found no evidence that Lung Allocation Score components should be weighted differently for rLTx candidates.