Photo Credit: Gilnature

An artificial neural network (ANN) algorithm that uses donor and recipient variables can predict early graft failure in adult-to-adult living donor liver transplantation (ALDLT), according to results published in Liver Transplantation. Roberto Ivan Troisi, MD, PhD, MSc, and colleagues developed a risk prediction model for early graft failure (within 3 months). They used 2,073 patients with ALDLT to train the ANN and conducted cross and external validation. Factors associated with early graft failure included graft type, graft weight, extent of hospitalization, and liver disease severity. The model had area under the receiver operating characteristic curve values of 0.69 in cross validation, 0.70 in the independent test set, and 0.68 in external validation. The decision curve analysis showed a positive net benefit of the model, with an estimated net decrease of five to 15 early graft failures per 100 patients with ALDLT. The model also stratified long-term graft survival (P<0.001), which ranged from 87% in the low-risk group to 60% in the high-risk group.