Can elastography predict early allograft dysfunction or loss after liver transplantation? A prospective study of diagnostic accuracy.

Abstract

Early allograft dysfunction after liver transplantation is an important cause of allograft loss. There is a lack of data on applicability of liver stiffness measurement (LSM) by elastography in early post-liver transplantation that can predict these complications. Our aim was to evaluate the diagnostic accuracy of elastography to predict early allograft dysfunction or loss after liver transplantation. Sixty-one liver transplant patients were prospectively enrolled and underwent daily LSM over the first seven post-operative days. Early allograft dysfunction and loss occurred in 27 (44.2%) and 17 (27.8%) patients, respectively. C-statistic of each day of LSM showed high accuracy and no significant difference between them. For early allograft dysfunction, LSM>2.39m/s on the first day had c-statistics=0.83, sensitivity=0.41, specificity=0.97, positive predictive value=0.92, negative predictive value=0.67, positive likelihood ratio=13.85 to rule in, while a LSM<1.65m/s had c-statistics=0.83, sensitivity=0.96, specificity=0.50, positive predictive value=0.60, negative predictive value=0.94, negative likelihood ratio=0.07 to rule out. For early allograft loss, LSM>2.25m/s had c-statistics=0.93, sensitivity=0.76, specificity=0.98, positive predictive value=0.93, negative predictive value=0.91, positive likelihood ratio=33.65 to rule in, while a LSM<1.75m/s had c-statistics=0.93, sensitivity=0.94, specificity=0.94, positive predictive value=0.50, negative predictive value =0.97, negative likelihood ratio=0.09 to rule out. These results were internally validated with the bootstrap method. No methods (Olthof criteria, MELD, MELD-5, MEAF, BAR, biomarkers) performed better than elastography to predict early allograft loss. Elastography performed over the first week after liver transplantation can accurately predict early dysfunction and allograft loss.