Fabrice Ivanes1, Christelle Cantrelle2, Thibaud Genet3, Claude Le Feuvre4, Camille Legeai2, Carine Jasseron2, Isabelle Pipien2, Eric Epailly5, Olivier Bastien2, Denis Angoulvant6, Richard Dorent2. 1. Service de Cardiologie, Centre Hospitalier Régional Universitaire de Tours, Tours, France; EA 4245 T2I "Transplantation, Immunologie, Inflammation" et Fédération Hospitalo-Universitaire SUPORT, Université de Tours, Tours, France. Electronic address: fabrice.ivanes@univ-tours.fr. 2. Agence de la Biomédecine, La Plaine Saint Denis, France. 3. Service de Cardiologie, Centre Hospitalier Régional Universitaire de Tours, Tours, France. 4. Institut de Cardiologie, Centre Hospitalier Universitaire La Pitié-Salpêtrière, Paris, France. 5. Service de chirurgie cardiaque, nouvel hôpital civil, Centre Hospitalier Universitaire de Strasbourg, Strasbourg, France. 6. Service de Cardiologie, Centre Hospitalier Régional Universitaire de Tours, Tours, France; EA 4245 T2I "Transplantation, Immunologie, Inflammation" et Fédération Hospitalo-Universitaire SUPORT, Université de Tours, Tours, France.
Abstract
BACKGROUND: Allograft shortage might be overcome by the use of hearts from expanded-criteria donors (ECD) but their estimated high-risk of coronary artery disease (CAD) results in a limited utilization of these hearts for transplantation. We aimed to determine if performing coronary angiography (CA) in ECD enhances cardiac procurement and to develop a predictive model estimating their probability of absence of CAD. METHODS: We retrospectively used the French National Transplant Registry CRISTAL and considered all donors aged 45 to 70 with ≥ 1 organ harvested between March 2012 and June 2014 to derive a high-risk donor population. Of 515 donors with ≥ 1 CAD risk factor and no obvious contraindication for cardiac procurement, 230 underwent CA. Coefficients estimated by multivariate logistic regression models were used to evaluate the impact of CA on procurement and build the predictive model. RESULTS: Among CA donors, 133 had CAD, 53 (23%) with at least one stenosis ≥ 50%. Predictors of cardiac graft offer were female gender, age below 60, no cardiac arrest, no intravenous adrenaline/dobutamine requirement and no treated hypercholesterolemia. CA increased the probability of procurement by 9% (p = 0.028). Female gender, non-vascular cause of death, absence of diabetes and BMI ≥ 25 kg/m2 (p < 0.05) were associated with a normal CA and used for the prediction model. The area under the ROC curve of the model was 0.70. Specificity for the highest quartile was 82%. CONCLUSION: Performing CA in ECD enhances cardiac procurement. When CA is not feasible, we defined a clinical score allowing accurate estimation of normal CA probability.
BACKGROUND: Allograft shortage might be overcome by the use of hearts from expanded-criteria donors (ECD) but their estimated high-risk of coronary artery disease (CAD) results in a limited utilization of these hearts for transplantation. We aimed to determine if performing coronary angiography (CA) in ECD enhances cardiac procurement and to develop a predictive model estimating their probability of absence of CAD. METHODS: We retrospectively used the French National Transplant Registry CRISTAL and considered all donors aged 45 to 70 with ≥ 1 organ harvested between March 2012 and June 2014 to derive a high-risk donor population. Of 515 donors with ≥ 1 CAD risk factor and no obvious contraindication for cardiac procurement, 230 underwent CA. Coefficients estimated by multivariate logistic regression models were used to evaluate the impact of CA on procurement and build the predictive model. RESULTS: Among CA donors, 133 had CAD, 53 (23%) with at least one stenosis ≥ 50%. Predictors of cardiac graft offer were female gender, age below 60, no cardiac arrest, no intravenous adrenaline/dobutamine requirement and no treated hypercholesterolemia. CA increased the probability of procurement by 9% (p = 0.028). Female gender, non-vascular cause of death, absence of diabetes and BMI ≥ 25 kg/m2 (p < 0.05) were associated with a normal CA and used for the prediction model. The area under the ROC curve of the model was 0.70. Specificity for the highest quartile was 82%. CONCLUSION: Performing CA in ECD enhances cardiac procurement. When CA is not feasible, we defined a clinical score allowing accurate estimation of normal CA probability.