Development and Validation of a Risk Model for Primary Graft Failure after Heart Transplantation and Comparison to the RADIAL Risk Score

Abstract

Listen

Translate article

Purpose Primary graft failure (PGF) is the leading cause of early mortality following heart transplantation (HTx). Many risk factors and scores for the development of PGF have been defined, e.g. the RADIAL score. Since then a new ISHLT consensus paper was published and took new risk factors and pathological consideration into account. Therefore, the aim of this study is to develop and validate a new risk score based on the ISHLT Transplant Registry Database (TRD). Methods The ISHLT TRD (time-period 2005-June 2017) was used. Pediatric pts. (n = 6944) and pts. who underwent repeat-HTx or multi-organ transplantation (n = 2134) were excluded. Pts. with complete available data on donor/recipient age, sex, height, weight and recipient mean PA, PCW and CO were included. A dataset of 23,004 pts. was used including PGF (within first 30 d after HTx) as the primary cause of graft failure, death, or retransplant in 644 pts. Since RA pressure was not recorded in the ISHLT TRD, a modified RADIAL risk score excluding RA pressure was applied. At first, 173 variables were defined as potential predictors of PGF. Based on the year of HTx, the final database was split in a derivation (17229 pts.; 2005-2014; 532 PGF cases) and validation cohort (5775 pts.; 2015-2017; 112 PGF cases),followed by the application of a LASSO algorithm for variable selection. A logistic regression model was built by using the prior selected variable. The performance of the model (measured by AUC) was assessed in the validation and derivation dataset, and compared to the performance of the modified RADIAL score. Based on the regression coefficients, a point-based score was built. Results A model including 27 predictors was derived in the derivation cohort. The AUC of this model was calculated with 0.69. After correcting for overoptimism, the AUC was 0.67. The AUC in the validation cohort was similar with 0.72 (p Conclusion Based on the ISHLT TRD we developed a new score model with a higher predictive value than the RADIAL score. This new model has the advantage of including donor and recipient characteristics, and might therefore improve the outcome by early treatment of primary graft failure in patients at risk.