The Value Of Ultrasonic Acoustic Structure Quantification In Grading Liver Fibrosis And Classification Of Cirrhosis

BackgroundViral hepatitis is caused by hepatitis B virus (HBV), characterized by inflammatory lesions inthe liver and multiple organ damage. In our country, hepatitis B has become one of the mostwidespread and hazardous infectious disease. Each year about350000patients die ofHBV-associated liver diseases, including end-stage liver cirrhosis, liver failure and livercancer. Therefore, reducing the incidence of HBV-associated with end-stage liver disease andmortality, and effectively monitoring in the whole process of the disease progression andblocking, solving related important scientific problems, is urgent. Liver fibrosis is thereversible transition stage between hepatitis and liver cirrhosis. Studies have shown that withearly diagnosis and elimination of pathogen, fibrosis can be absorbed gradually, which willreverse or decelerate the development of liver cirrhosis. Thus, accurately estimating thedegree of liver fibrosis in hepatitis patients is essential. However, the “fibrosis to cirrhosis”transition is slow and not easily detectable. Once reached end-stage liver cirrhosis, patientswill suffer dearly from gastrointestinal bleeding, hepatic encephalopathy and subsequentcomplications. So accurately estimating the development of liver cirrhosis is very valuable forunderstanding the disease progression and follow-up treatment.Ultrasound diagnosis of liver disease is one of the commonly used noninvasive examinationmethod. Conventional two-dimensional ultrasound can dynamically observed the change ofliver morphology. It is of some diagnostic value, but the diagnostic specificity is not high fordiffusive lesions without any morphological changes, such as liver fibrosis and early-stage ofliver cirrhosis. In recent years, with the development of the ultrasonographic elastography,instantaneous elasticity imaging (FibroScan), Acoustic Radiation Force Impulse imaging(ARFI) and Real-time Tissue Elastography technology (RTE) have been used to quantitate thedegree of liver fibrosis as new noninvasive acoustic examination methods. But FibroScan donot provide two-dimensional images to accurately indicate the potential location for livertissue sampling. On the other hand, ARFI and RTE do not provide entirely quantitativediagnosis. The principle of ultrasonic acoustic structure quantitative analysis (ASQ) isperforming chi-square test on acquired liver tissue raw echo signal (radio frequency), echosignal strength, the change of backscattering to determine degree of liver fibrosis. This studywas conducted on patients with chronic hepatitis B to compare ASQ results (analyzed fromconventional two-dimensional ultrasound) with pathology scores, in order to determine the value of ASQ in grading liver fibrosis and classification of cirrhosis.ObjectiveWe aim to explore the value of ultrasonic acoustic structure quantification in grading liverfibrosis and classification of cirrhosis, via comparing the results of ASQ and pathologicalexaminations, and seek effective methods for evaluation of liver fibrosis and classification ofcirrhosis.Method1、All cases were selected from250patients with surgery (July2012-October2012). Wefound liver fibrosis stage S0in14cases (no fibrosis, negative control group), liver fibrosisstage S1-S4for83cases (liver fibrosis); we identified114patients with liver cirrhosis and39patients as negative controls. All patients confirmed by post-operative pathologicalexamination.2、First, we performed routine ultrasound examination for the250cases, then we gatheredliver sonograms from different parts of the liver in ASQ mode. After that, we conducted ASQanalysis offline on dynamically stored images, focusing on region of interest, to get the Cm2average values. And finally, we used statistical analysis software, including SPSS13.0, SAS9.3and R2.15, to analyze the Spearman’s correlation test, Principal Component Analysis(PCA), Multivariate ANalysis Of VAriance (MANOVA) and the Receiver-OperatingCharacteristic curve (ROC curve).Result1、According to the results of Spearman’s correlation test, Cm2averages of ASQ exhibitedsignificant positive correlation (P<0.01) with pathological staging of liver fibrosis.2、PCA showed clear separation among different stages of liver fibrosis. MANOVA indicatesthat significant differences (P<0.01) exist among the5stages. Post hoc pairwise comparisonsindicated that the significant differences (P<0.05) can be found between any two groups(P=0.0277for S0vs S1, the rest of the P values were less than0.01).3、Though analysis of variance, there was no significant difference between four parts of liver(P=0.129). Therefore, just looking at the ROC curves for the right lobe below the rib, theresults are listed below. When S≥S1,Cm2Average1has AUC (area under curve)=0.969,sensitivity=92.8%, specificity=92.9%, criterion>110; Cm2Average2has AUC=0.917, sensitivity=81.9%, specificity=100.0%, criterion>133. When S≥S3Cm2Average1hasAUC=0.886, sensitivity=78.4%, specificity=86.7%, criterion>116; Cm2Average2hasAUC=0.867, sensitivity=78.4%, specificity=76.7%, criterion>140. When S≥S4,Cm2Average1has AUC=0.802, sensitivity=100.0%, specificity=64.4%, criterion>118;Cm2Average2has AUC=0.882, sensitivity=100.0%, specificity=71.3%, criterion>145.4、PCA and MANOVA showed statistically significant difference (P <0.01) among thenegative control group and3cirrhosis groups, including small, big and mixed nodularcirrhosis. Post hoc pairwise comparisons indicated that negative control group was differentfrom any1of the3cirrhosis groups, but among just the3cirrhosis groups, statisticalsignificances (P<0.01) only exist between small and mixed nodular cirrhosis. The differencesof “big vs mixed”(P=0.5529) and “small vs big”(P=0.1133) did not reach the level ofstatistical significance.Conclusion1. ASQ analysis provided objective and accurate results, which have become a promisingnoninvasive method in grading liver fibrosis.2. Our preliminary research indicated that ASQ analysis, as a new diagnostic technique, isalso an effective method for the clinical diagnosis of liver cirrhosis and it is certainlybeneficial for treatment and prognosis.