A Clinical Study On The Establishment Of A Nomogram For Donor Liver Quality Assessment Based On Metabolite Analysis

Background.Liver transplantation is the only effective treatment for end-stage liver disease,but its complex postoperative complications are still thorny problems.The quality of the donor liver,the systemic condition of the recipient,the anastomosis technique and the perioperative treatment plan are the four important factors that determine the incidence and severity of postoperative complications.With the gradual standardization of the operation procedures of liver transplantation and the increasing maturity of perioperative treatment,the technical bottleneck and donor-recipient management are no longer the main problems perplexing the transplantation team.However,under the background that the supply of liver source exceeds the demand for a long time,various transplantation centers have resorted to the expansion of acceptance criteria to increase the utilization rate of donor liver,resulting in uneven quality of donor livers[2].If the donor liver quality is poor,even if there are no complications related to surgical techniques and treatment options,the recipient is still at risk of complications such as EAD.EAD is a common early complication after liver transplantation,with an incidence ranging from 10.8%to 36.3%[3]and largely depends on the quality of the donor liver.EAD means a 10.7-fold increase in the risk of recipient death and a 7.4-fold increase in the risk of donor liver function loss[4].Recipients with EAD are more likely to experience acute rejection,septicemia and other complications,resulting in prolonged ICU and total hospital stay,and a sharp increase in medical costs,which seriously affect the prognosis of patients.In view of the lack of unified standards and unavoidable defects in all kinds of donor liver quality assessment methods reported in the existing literature,it is an important task for every transplantation center to explore a rapid,objective and quantifiable method for donor liver quality assessment.Objective.The contents of biomarkers representing different tissue of donor liver were studied by metabonomic analysis to construct and validate a donor liver quality assessment model.At the same time,we also analyzed the relationship between the changes of these biomarkers and the ischemia time of donor liver.Methods.According to the acceptance criteria designed by the study,100 cases of orthotopic liver transplantation in our hospital from June 2018 to August 2020 were included,of which the first 69 cases were used in the training group for model construction,and the latter 31 cases were used as external cohorts for model validation.The samples of group A and B were respectively collected at the time of acquisition and after reperfusion.Taking the occurrence of EAD as the grouping variable,the contents of hyaluronic acid（Hyaluronic acid,HA）,bile lactate dehydrogenase（Lactate dehydrogenase,LDH）and lysophosphatidylcholine（Lysophosphatidylcholine,LPC）in group B were analyzed to construct a nomogram model to evaluate the quality of donor liver.The prediction accuracy and discriminant ability of the model are tested by AUC,and the consistency between the predicted value of the nomogram model and the actual observation is verified by the calibration curve.Then verify it externally with the verification group data.After the model was constructed,the regression relationship between the changes of traumatic biomarkers and pathomorphological indexes and ischemia time in group A and B was analyzed to enhance the clinical application value of the model.Results.The incidence rate of EAD in the training cohort was 39%,and 29%in the validation group.The results of univariate analysis showed that the contents of HA in perfusion fluid,LDH in bile and LPC in liver tissue of donor liver were three independent risk factors of EAD,which could be used to evaluate the quality of donor liver and were further incorporated into Logistic regression analysis and nomogram model construction.Based on the existing sample size and data,we can realize that the incidence of EAD was more than 60%when the total score was beyond 76 after model establishment and correction,which indicated poor quality of donor liver.The AUC of the training set was 0.96,and the AUC of the verification cohort was 0.97.Both of them were more than 0.7.The calibration curves of the two groups showed good consistency,suggesting that the nomogram model had good stability,sensitivity,specificity and accuracy,and the prediction performance was outstanding.In addition,the results also showed that the pathomorphological score of donor liver in group A ranged from0 to 2,which basically showed normal microscopic pathomorphology,while the pathomorphological score of donor liver in group B was higher than that in group A,but it only showed mild to moderate ischemia-reperfusion injury.The pathomorphological scores of EAD group were higher than those of non-EAD group,and the difference was statistically significant,but it was also limited to mild to moderate ischemia-reperfusion injury.The changes of the contents of three kinds of traumatic biomarkers in groups A and B were positively correlated with the time of cold ischemia.Based on the existing sample size and data,the best regression equations were obtained as follows:YLPC=-7554+22.64×t-0.1692×X₀+0.000345×t×X₀;YHA=36.03+0.8427×X0+1.878×t-0.0007731×X0×t;YLDH=-0.0006564+0.8949×X0+0.02051×t （Y is the detected value of group B,X0 is the detected value of group A,and T is the total ischemia time）.Conclusions:1.Based on the measurement and analysis of three biomarkers representing the key sites of donor liver,we constructed a nomogram model which can guide clinicians to evaluate the quality of donor liver and predict the EAD of early postoperative complications.After AUC detection,calibration curve fitting and external cohort validation,the nomogram showed good prediction performance and discriminant ability,and had good stability,sensitivity,specificity and accuracy.The incidence of EAD was more than 60%when the total score was beyond 76.2.The establishment of three regression equations about the changes of biomarkers with ischemia time,so that the transplant physician can detect the corresponding samples and estimate the donor liver state after reperfusion,which enhances the clinical application value of the line diagram model.