Clinical Value Of MELD Prediction System And Clinical Study Of Direct Assessment Hepatic Fibrosis Degree For LT Patient

Objectives: To investigate the clinical application value of using computer-assisted digital image analysis method for quantitative assessment of hepatic fibrosis, quantitative assessment of hepatic fibrosis of LT patients will be performed and the correlation among this method, the MELD score method and the Ishak pathological grade will be evaluated. And a criterion will be built for evaluating the applicability of direct assessment of hepatic fibrosis using MELD.At the same time, evaluating the different outcome among digital image analysis with transplantation (OLT) patients three months survival term. And the rationality was tested for MELD scores in donor livers allocation for cirrhotic patients on waiting lists for OLT.Methods: Quantitative hepatic fibrosis evaluation of the 58 patients who underwent OLT because of end-stage liver disease from February 2006 to September 2006 in west china hospital sichuan university was performed by using computer-assisted digital image analysis. Pathological diagnosis according to the Ishak's hepatic fibrosis standard was also performed. MELD scores were calculated using the original formula based on the clinical examination data collected on the admission day. The correlations among the image analysis method, Ishak grading and MELD scoring method were analyzed using the Spearman's rank correlation analysis. The linear relationship between the MELD scores and the degree of hepatic fibrosis shown from linear regression analysis was used to define the reference criterion.58 patients were followed up in three months, the outcome included the survival and death. Area under of receiver operating characteristic curve (ROC) was used to evalute the standard and compare the value of digital image analysis with MELD scores with CTP scores for predicting the prognosis, Kaplan-Meier survival curves were made using the cutoffs identified by means of ROC. The different survival rates were analyzed by log-rank test.Results: The hepatic fibrosis area ratios of the 58 patients' liver specimen were between 23.2% and 88.4% which were calculated by computer-assisted digital image analysis, and the average was 56.7%. The MELD scores on the admission day were between 11 and 38, and the average was 22.85±9.32. The semi-quantitative Ishak classification showed that there were 0, 2, 7, 12, 18, 12, and 7 cases in each of the 7 grades respectively, with higher the grade the higher ration of hepatic fibrosis area ratio and the higher the MELD score. Spearman rank correlation test indicated that there was statistical significance (P < 0.01) of the correlation among these three methods. Linear regression analysis showed there was a linear relationship between the MELD scores and the degree of hepatic fibrosis, so the MELD scoring system can be directly used to assess the severity of hepatic fibrosis, for which a reference range was defined.Following up showed that 10 patients died in three months, hepatic fibrosis area ratio with MELD scores and CTP scores for non-survivors (71.93%±13.11%, 29.55±8.12, 11.72±2.04) were higher than those for survivors (39.62%±19.47%, 17.26±6.33, 7.81±2.29) significantly (P<0.01); Area under the ROC (c-statistic) of MELD and digital image analysis for three months (0.843,0.822) was significantly (P<0.05) different from that of CTP (0.742), Kaplan- Meier survival curves who made using the cutoffs showed both MELD scores and CTP scores was clearly discriminated(P<0.05) between patients who survived and those who died in short term; among the evaluation index of diagnostic test, sensitivity and specificity for MELD scores and CTP scores were 90%,80% vs 70%,75%, respectively, It indicated both MELD scores and CTP scores could predict prognosis in short term for LT patients, and MELD scores were superior to CTP scores.Conclusions: Computer-assisted digital image analysis can evaluate objectively hepatic fibrosis degree and Hepatic fibrosis degree can be evaluated directly by MELD , an objective reference criteria had been formulated and worth using in clinical setting. MELD scores and CTP scores could efficiently predict prognosis in short term for LT patients, and MELD scores were superior to CTP scores. MELD is also an objectively predictive system for short term survival for patients on waiting lists for OLT.