A Novel Risk Scoring System For Predicting Primary Graft Dysfunction After Orthotopic Heart Transplantation

Introduction Despite advances in graft survival following orthotopic heart transplantation (OHT), primary graft dysfunction (PGD) remains a major problem. PGD - defined as ventricular dysfunction occurring within 24 hours after OHT not associated with a discernible cause - is the leading cause of death in the first 30 days post OHT. The etiology of PGD is multifactorial including donor age, recipient inotropic support and amiodarone use. Hypothesis We would be able to identify novel donor and recipient factors associated with increased incidence of PGD. Methods We analyzed patients undergoing OHT from January 2014 to December 2019. Patients without 1-year follow up data were excluded. PGD was diagnosed based on the 2013 ISHLT consensus definition. Donor and recipient characteristics were analyzed using chi-square or Fisher exact test as appropriate, 2-tailed, with p<0.05 considered to be statistically significant. Multivariate analysis was performed to evaluate for predictors of PGD, and results were used to develop a PGD risk score. Results 122/143 OHT recipients had 1 year followup. PGD occurred in 36 (29.5%) patients. Patient proportions with donor-recipient height mismatch>20% (0 vs 0), weight mismatch>20% (34% vs 42%, p=0.32), and sex mismatch (13.8% vs 10.5%, p=0.76) were similar for PGD and non-PGD groups. Independent predictors of PGD were donor pH<7.05 (OR 12.7, 95%CI 3.4-47.6, p=0.01), pre-transplant recipient amiodarone use (OR 8.68, 95%CI 2.42-31.3, p<0.01), pre-transplant recipient milrinone use (OR 5.04, 95% CI 1.4-17.6, p=0.01), and PRA Class 2 >50% (OR 13.85, 95%CI 3.3-58, p=0.01). Based on beta coefficients in the regression model, we allotted 1 point to milrinone use, 2 to amiodarone use, and 3 points each to donor pH<7.05 and Class 2 PRA >50%. The derived risk score demonstrated a powerful association with PGD occurrence (per point increase OR 2.0; p<0.01). Despite significantly higher post-transplant ECMO use in patients with PGD (80.6% vs 18.5%, p=0.01), there were no significant differences in in-hospital mortality (5.6% vs 3.5%, p=0.63), 1-year mortality (11.1% vs 14%, p=0.77), rejection events (27.8% vs 36%, p=0.41) and dialysis at 1 year (8.3% vs 2.3%, p=0.15). Conclusion We identified novel factors associated with the development of PGD, including recipient milrinone use and elevated class 2 PRA, while confirming prior well-validated risk factors such as amiodarone use and donor acidemia. A PGD risk score incorporating these factors powerfully predicts PGD occurrence, and may assist in future donor utilization and identification of recipients at risk for PGD.