OBJECTIVE: We aimed to evaluate the performance of the newly developed deep learning Radiomics of elastography (DLRE) for assessing liver fibrosis stages. DLRE adopts the radiomic strategy for quantitative analysis of the heterogeneity in two-dimensional shear wave elastography (2D-SWE) images.
DESIGN: A prospective multicentre study was conducted to assess its accuracy in patients with chronic hepatitis B, in comparison with 2D-SWE, aspartate transaminase-to-platelet ratio index and fibrosis index based on four factors, by using liver biopsy as the reference standard. Its accuracy and robustness were also investigated by applying different number of acquisitions and different training cohorts, respectively. Data of 654 potentially eligible patients were prospectively enrolled from 12 hospitals, and finally 398 patients with 1990 images were included. Analysis of receiver operating characteristic (ROC) curves was performed to calculate the optimal area under the ROC curve (AUC) for cirrhosis (F4), advanced fibrosis (≥F3) and significance fibrosis (≥F2).
RESULTS: AUCs of DLRE were 0.97 for F4 (95% CI 0.94 to 0.99), 0.98 for ≥F3 (95% CI 0.96 to 1.00) and 0.85 (95% CI 0.81 to 0.89) for ≥F2, which were significantly better than other methods except 2D-SWE in ≥F2. Its diagnostic accuracy improved as more images (especially ≥3 images) were acquired from each individual. No significant variation of the performance was found if different training cohorts were applied.
CONCLUSION: DLRE shows the best overall performance in predicting liver fibrosis stages compared with 2D-SWE and biomarkers. It is valuable and practical for the non-invasive accurate diagnosis of liver fibrosis stages in HBV-infected patients.
TRIAL REGISTRATION NUMBER: NCT02313649; Post-results.
BACKGROUND: We evaluated the performance of serum hyaluronan (HA), procollagen type III N-terminal peptide (PIIINP), type IV collagen (IVC), laminin (LN), alanine aminotransferase (ALT), and aspartate aminotransferase (AST), compared to transient elastography (FibroScan) in predicting significant liver fibrosis.
METHODS: We therefore determined 4 serum fibrosis markers, FibroScan and liver biopsy in 70 consecutive adult patients with chronic hepatitis B. According to a modified Scheuer scoring system, significant fibrosis was defined as fibrosis stage ≥S2. We compared serum fibrosis markers to histological staging and FibroScan results using Spearman correlation analysis and area under receiver operating characteristic (ROC) curves (AUROCs).
RESULTS: Of the 212 patients who had the results of FibroScans and four serum fibrosis markers for HBV, 70 had concurrent liver biopsy. Significant liver fibrosis was found in 24/70 patients. The serum levels of HA, PIIINP, IVC, LN, ALT, AST was all positively correlated with fibrosis stage of Liver biopsy. The coefficients with stages were respectively 0.468, 0.392, 0.538, 0.213, 0.350, 0.375. There was a significant difference between mild fibrosis (<S2) and significant fibrosis (≥S2), excluding LN, in the levels of these 5 serum makers (P < .05). AUROC for FibroScans and HA, PIIINP, IVC, LN, ALT, AST to correctly allocate patients to histological fibrosis stage ≥ S2 was 0.866, 0.784, 0.738, 0.827, 0.630, 0.713 and 0.728 respectively. Since LN shows the worst performance of the others. We decided to check the performance of the combination of HA, PIIINP, CIV, ALT, AST, excluding LN, to distinguish fibrosis stages. The index of the histological fibrosis stage ≥ S2, combining the 5 serum markers, significantly improved diagnostic performance (AUROC = 0.861) compared to the use of 5 serum markers alone in all HBV patients.
CONCLUSION: The combination of the 5 serum markers and FibroScan performed equally well in predicting significant fibrosis. The combination of the 5 serum markers is a reliable noninvasive method to predict significant liver fibrosis in patients with CHB. So, it provide another choice rather than FibroScan in predicting significant liver fibrosis.
BACKGROUND AND AIM: Fibrosis index based on four factors (FIB-4) and aspartate aminotransferase-platelet ratio (APRI) were validated with unsatisfactory efficiency. Routine hematology index red cell distribution width-platelet ratio (RPR) had been tried in liver fibrosis detection. This study tries to evaluate the stepwise application of FIB-4, RPR, and APRI in detecting chronic hepatitis B (CHB) fibrosis.
METHODS: A total of 246 compensated CHB patients who underwent liver biopsies, transient elastography, and routine blood tests including complete blood count were included. Dual cut-offs were determined to exclude or include cirrhosis diagnosis. Performance of stepwise combining routine biomarkers including RPR, FIB-4, and APRI were statistically analyzed.
RESULTS: The Metavir F0, F1, F2, F3, and F4 were identified in 2.4%, 22.0%, 32.1%, 24.0%, and 19.5% of the eligible patients, respectively. The area under receiver operating characteristics curves for detecting significant fibrosis and cirrhosis were 0.853 and 0.883 for transient elastography; 0.719 and 0.807 for FIB-4; 0.638 and 0.791 for RPR; 0.720 and 697 for APRI; and 0.618 and 0.760 for mean platelet volume-platelet ratio, respectively. The proportion of patient determined as cirrhosis or non-cirrhosis was 65.9% by transient elastography, 36.9% by FIB-4, 30.5% by RPR, and 19.5% by APRI, respectively. These numbers for determining significant fibrosis were 49.6%, 24.2%, 21.5%, and 23.6% in the same order. Detected by stepwise application of FIB-4, RPR, and APRI, 41.5% and 52.8% of patients could be determined the state of significant fibrosis and cirrhosis, respectively.
CONCLUSIONS: In source-limited settings without transient elastography, stepwise applying FIB-4, RPR, and APRI could free nearly half of CHB patients from liver biopsies in detecting significant fibrosis and cirrhosis.
BACKGROUND: The benefits of using serum markers to diagnose stages of liver disease in chronic hepatitis B (CHB) patients are controversial. We conducted a study to compare the clinical significance of four markers in evaluating liver inflammation and fibrosis in CHB patients.
METHODS: A total of 323 treatment-naive CHB patients who received a liver biopsy and routine laboratory testing were enrolled in our study. We used the Scheuer scoring system as a pathological standard for diagnosing liver inflammation and fibrosis. The diagnostic performance of the fibrosis index based on four factors (FIB-4), the aspartate transaminase to platelet ratio index (APRI), the gamma-glutamyl transpeptidase-to-platelet ratio (GPR), and the red cell distribution width-platelet ratio (RPR) were analyzed with receiver-operating characteristic curves (ROC).
RESULTS: No significant differences among the four indexes for diagnosing significant fibrosis (S ≥ 2) was found, while APRI and GPR were superior to FIB-4 and RPR in diagnosing moderate (G ≥ 2), severe (G ≥ 3) inflammation, and severe fibrosis (S ≥ 3). The AUROCs for diagnosing G ≥ 2 and G ≥ 3 were 0.732 and 0.861 for APRI, 0.726 and0.883 for GPR, 0.703 and0.705 for FIB-4, and 0.660 and 0.747 for RPR, respectively. The AUROCs for diagnosing S ≥ 2 and S ≥ 3 were0.724 and 0.799 for APRI, 0.714 and0.801 for GPR, 0.683 and0.730 for FIB-4, and 0.643 and 0.705 for RPR, respectively.
CONCLUSION: APRI and GPR were more effective than FIB-4 and RPR at diagnosing liver inflammation and fibrosis.
We aimed to evaluate the diagnostic accuracy of liver stiffness measurement (LSM) in 188 chronic hepatitis B (CHB) patients with alanine transaminase (ALT) ≤ twice the upper limit of normal (ULN). Liver fibrosis was staged using METAVIR scoring system. Define significant fibrosis as F2-F4, severe fibrosis as F3-F4, and cirrhosis as F4. To predict F2-F4, the AUROC of LSM was higher than that of APRI (0.86 vs 0.73, p = 0.001) and FIB-4 (0.86 vs 0.61, p < 0.001). To predict F4, the AUROC of LSM was also higher than that of APRI (0.93 vs 0.77, p = 0.012) and FIB-4 (0.93 vs 0.64, p < 0.001). Patients with ALT levels 1-2 ULN had higher cut-off values than patients with normal ALT levels for the diagnosis of F2-F4 (6.5 vs 6 kPa) and F4 (10.2 vs 7.8 kPa). Using cut-off values regardless of ALT levels, the diagnostic accuracy of LSM was 81% for F2-F4, and 89% for F4. Applying ALT-stratified cut-off values, the diagnostic accuracy of LSM was 82% for F2-F4, and 86% for F4. In conclusion, LSM is a reliable noninvasive test for the diagnosis of liver fibrosis. Applying ALT-stratified cut-off values did not enhance diagnostic accuracy of LSM in CHB patients with ALT ≤ 2 ULN.
It is of great significance to develop and evaluate noninvasive indexes predicting the level of liver fibrosis. The aim of this study was to comparatively evaluate gamma-glutamyl transpeptidase-to-platelet ratio (GPR) versus aspartate aminotransferase-to-platelet ratio index (APRI) and fibrosis index based on 4 factors (FIB-4) in predicting different levels of liver fibrosis of chronic hepatitis B (CHB) within the framework of HBeAg-positive and HBeAg-negative patients. A total of 1157 HBeAg-positive and 859 HBeAg-negative CHB patients were enrolled, among whom the pathological stage ≥S2, ≥S3, ≥S4 were defined as significant fibrosis, extensive fibrosis and cirrhosis, respectively. Receiver operating characteristic (ROC) curves were used to evaluate the performance of GPR, APRI and FIB-4 in predicting different levels of liver fibrosis. In HBeAg-positive patients, the area under ROC curves (AUROCs) of GPR in predicting extensive fibrosis and cirrhosis were both significantly larger than those of APRI (P = .0001 and P < .0001). In HBeAg-negative patients, the AUROCs of GPR in predicting significant fibrosis and cirrhosis were significantly larger than those of FIB-4 (P = .0006 and P = .0041). The AUROC of GPR in predicting extensive fibrosis was significantly larger than that of APRI and FIB-4 (P = .0320 and P = .0018). Using a cut-off of GPR > 0.500 as standard, the sensitivities and specificities of GPR in predicting significant fibrosis in HBeAg-positive patients were 59.6% and 81.2%, and for cirrhosis 80.9% and 63.8%, respectively; and those of HBeAg-negative patients were 60.3% and 78.3%, 84.5% and 66.1%, respectively. Regardless of HBeAg-positive or HBeAg-negative status, GPR had the best performance in predicting different levels of liver fibrosis.
AIM: The diagnostic performance of Fibroscan might be improved when combined with other serum fibrosis related markers. Previous study has demonstrated that S100A4 expression is associated with liver fibrosis in humans with hepatitis. This study aimed to clarify diagnostic accuracy of serum S100A4 levels for significant liver fibrosis in patients with chronic hepatitis B (CHB), and develop a combined algorithm of liver stiffness measurement (LSM) and S100A4 to predict significant liver fibrosis in CHB.
METHODS: One hundred and seventy-five CHB patients who had performed liver biopsy were consecutively included. We evaluated serum S100A4 levels, LSM values and other clinically-approved fibrosis scores.
RESULTS: Serum S100A4 level was higher in CHB patients with significant fibrosis, compared to those without [199.58 (33.31-1971.96) vs. 107.15 (2.10-1038.94), P<0.001]. Using receiver-operating characteristic (ROC) analyses, the area under the curves (AUC), sensitivity, specificity and accuracy of S100A4 were found to be 0.749, 62.7%, 75.9% and 0.70 for significant fibrosis (≥Stage 2), respectively. Although not superior to LSM, these results were better than the fibrosis index based on the 4 factor (FIB-4) and the aspartate aminotransferase-to-platelet ratio index (APRI) for significant fibrosis detection. An algorithm consisting of S100A4 and LSM was derived. The AUC, sensitivity, specificity and accuracy of model based on serum S100A4 level and LSM were 0.866, 86.6%, 77.8% and 0.79 for significant fibrosis detection, superior to those based on LSM alone (0.834, 76.1%, 80.7% and 0.76, P=0.041).
CONCLUSION: Serum S100A4 level was identified as a fibrosis marker of liver fibrosis in patients with CHB. Combining serum S100A4 with LSM improved the accuracy of transient elastography for hepatitis B significant fibrosis detection.
BACKGROUND/AIMS: Comparison of the accuracy of magnetic resonance elastography (MRE) and diffusion weighted imaging (DWI) for the diagnosis of liver fibrosis in patients with chronic hepatitis B (CHB).
METHODS: In this retrospective analysis, we investigated 63 patients with CHB and liver fibrosis. DWI was performed with both breath-hold (DWI-BH) and free-breathing (DWI-FB) sequences (b=0, 500). The mean liver stiffness and apparent diffusion coefficient (ADC) were calculated by drawing regions of interest maps. Fibrosis staging according to the METAVIR system was independently performed by an experienced pathologist. A receiver operating curve (ROC) analysis was conducted to determine the accuracy of MRE, DWI-BH and DWI-FB in the detection and stratification of liver fibrosis. The performance of the detection of significant fibrosis (≥F2), advanced fibrosis (≥F3), and cirrhosis (F4) was also evaluated by comparing areas under the ROC.
RESULTS: There was a moderate and significantly negative correlation between the ADC values and liver stiffness. The accuracies for the detection of ≥F2/≥F3/F4 stage fibrosis with DWI-FB, DWI-BH and MRE were 0.84/0.76/0.72, 0.72/0.83/0.79 and 0.99/0.99/0.98, respectively. The performance of MRE was significantly better than DWI-FB and DWI-BH. There were no significant differences between the performance of DWI-FB and DWI-BH.
CONCLUSIONS: MRE is more accurate than DWI for the detection and stratification of liver fibrosis in CHB.
OBJECTIVES: Noninvasive models have been established for the assessment of liver fibrosis in patients with chronic hepatitis B(CHB). However, the predictive performance of these established models remains inconclusive. We aimed to develop a novel predictive model for liver fibrosis in CHB based on routinely clinical parameters.
RESULTS: Platelets(PLT), the standard deviation of red blood cell distribution width(RDW-SD), alkaline phosphatase(ALP) and globulin were independent predictors of significant fibrosis by multivariable analysis. Based on these parameters, a new predictive model namely APRG(ALP/PLT/RDW-SD/globulin) was proposed. The areas under the receiver-operating characteristic curves(AUROCs) of APRG index in predicting significant fibrosis(≥F2), advanced fibrosis(≥F3) and liver cirrhosis(≥F4) were 0.757(95%CI 0.699 to 0.816), 0.763(95%CI 0.711 to 0.816) and 0.781(95%CI 0.728 to 0.835), respectively. The AUROCs of the APRG were significantly higher than that of aspartate transaminase(AST) to PLT ratio index(APRI), RDW to PLT ratio(RPR) and AST to alanine aminotransferase ratio(AAR) to predict significant fibrosis, advanced fibrosis and cirrhosis. The AUROCs of the APRG were also significantly higher than fibrosis-4 score (FIB-4) (0.723, 95%CI 0.663 to 0.783) for cirrhosis(P=0.034) and better than gamma-glutamyl transpeptidase(GGT) to PLT ratio(GPR) (0.657, 95%CI 0.590 to 0.724) for significant fibrosis(P=0.001).
MATERIALS AND METHODS: 308 CHB patients who underwent liver biopsy were enrolled. The diagnostic values of the APRG for liver fibrosis with other noninvasive models were compared.
CONCLUSIONS: The APRG has a better diagnostic value than conventionally predictive models to assess liver fibrosis in CHB patients. The application of APRG may reduce the need for liver biopsy in CHB patients in clinical practice.
PURPOSE: To assess the accuracy of the T1 relaxation time index on gadoxetic acid-enhanced magnetic resonance imaging (MRI) for staging liver fibrosis in chronic hepatitis B (CHB), in comparison and combination with the aspartate aminotransferase-to-platelet ratio index (APRI) and fibrosis-4 (FIB-4).
MATERIALS AND METHODS: A retrospective study of gadoxetic acid-enhanced T1 mapping and serum biochemical tests was performed on 126 CHB patients who underwent gadoxetic acid-enhanced 1.5T MRI, and the histological score used as the gold standard. The reduction rate of T1 relaxation time before and 20 minutes after gadoxetic acid injection (ΔT1 , ΔR1%), the contrast uptake rate (KHep ), APRI, and FIB-4 were calculated. The diagnostic efficacy of ΔT1 , ΔR1%, KHep , APRI, and FIB-4 for predicting stage 2 or greater (≥S2), stage 3 or greater (≥S3), and stage 4 (S4) was compared.
RESULTS: ΔT1 (r = -0.513, P < 0.001), ΔR1% (r = -0.626, P < 0.001), KHep (r = -0.527, P < 0.001), APRI (r = 0.519, P < 0.001), and FIB-4 (r = 0.476, P < 0.001) correlated significantly with fibrosis stages. Areas under the curves (AUCs) of ΔR1% for detecting ≥S2, ≥S3, and S4 were 0.849, 0.827, and 0.809, which were greater than that of APRI (0.763, 0.745, 0.787) and FIB-4 (0.727, 0.738, 0.772), but significant difference was found only in discriminating ≥S2 between ΔR1% and FIB-4 (P = 0.027). The combination of all five indices performed best, with AUC, sensitivity, and specificity of 0.860, 87.21%, and 72.50% for diagnosing ≥S2, 0.878, 82.81%, and 85.48% for ≥S3, and 0.867, 80.00%, and 83.95% for S4.
CONCLUSION: The gadoxetic acid-enhanced T1 relaxation time index appears to be superior to APRI and FIB-4 for predicting hepatic fibrosis. The combined use of gadoxetic acid-enhanced T1 mapping, APRI, and FIB-4 may be more reliable for staging liver fibrosis in CHB.
LEVEL OF EVIDENCE: 4 J. Magn. Reson. Imaging 2017;45:1186-1194.
We aimed to evaluate the performance of the newly developed deep learning Radiomics of elastography (DLRE) for assessing liver fibrosis stages. DLRE adopts the radiomic strategy for quantitative analysis of the heterogeneity in two-dimensional shear wave elastography (2D-SWE) images.
DESIGN:
A prospective multicentre study was conducted to assess its accuracy in patients with chronic hepatitis B, in comparison with 2D-SWE, aspartate transaminase-to-platelet ratio index and fibrosis index based on four factors, by using liver biopsy as the reference standard. Its accuracy and robustness were also investigated by applying different number of acquisitions and different training cohorts, respectively. Data of 654 potentially eligible patients were prospectively enrolled from 12 hospitals, and finally 398 patients with 1990 images were included. Analysis of receiver operating characteristic (ROC) curves was performed to calculate the optimal area under the ROC curve (AUC) for cirrhosis (F4), advanced fibrosis (≥F3) and significance fibrosis (≥F2).
RESULTS:
AUCs of DLRE were 0.97 for F4 (95% CI 0.94 to 0.99), 0.98 for ≥F3 (95% CI 0.96 to 1.00) and 0.85 (95% CI 0.81 to 0.89) for ≥F2, which were significantly better than other methods except 2D-SWE in ≥F2. Its diagnostic accuracy improved as more images (especially ≥3 images) were acquired from each individual. No significant variation of the performance was found if different training cohorts were applied.
CONCLUSION:
DLRE shows the best overall performance in predicting liver fibrosis stages compared with 2D-SWE and biomarkers. It is valuable and practical for the non-invasive accurate diagnosis of liver fibrosis stages in HBV-infected patients.
