Investigating the Efficiency of APRI, FIB-4, AAR and AARPRI as Noninvasive Markers for Predicting Hepatic Fibrosis in Chronic Hepatitis B Patients in Bangladesh

2.1. Study Design

From January 2017 to May 2018, this study was carried out in Dhaka Hospital, Bangladesh. All recruited patients were at that point beginning consistent follow-up visits in HBV out-patient center. Laboratory examinations and analytic liver biopsy were performed as a standard strategy for distinguishing proof of their fibrosis and consequent decision of antiviral treatment. Patients were chosen in view of biochemical examinations which were in accordance with biopsy results.

In this study, a total of 1254 CHB people were examined and confirmed by monitoring of HBsAg titer by ELISA technique and measuring of HBV DNA copies by using Smart Cycler II (HBV DNA PCR). 104 patients were grouped as an internal validation cohort. Also, to validate the overall performance of non-invasive biomarkers, we included 109 CHB patients as an external validation cohort from Rajshahi Medical Hospital, Bangladesh between September 2017 and April 2018.

Inclusion and exclusion criteria utilized were the same for all the recruited patients in the study.

2.2. Ethical Clearance

All recruited patients gave written consent for approval. The examination was performed in agreement with the latest version of Declaration of Helsinki for medical research and it was permitted by the Ethics Committee of Dhaka Medical College and Hospital and Rajshahi Medical College and Hospital, Bangladesh.

2.3. Inclusion Criteria

CHB was characterized as constant HBsAg positive for over a half year and it was confirmed by PCR. Naïve HBeAg-negative CHB patients were selected after assessing their liver biopsy results obtained from 2 principal histopathologists.

2.4. Exclusion Criteria

Patients who showed antagonistic activity to HBV infection were not included in the study in addition to the Immune tolerant patients characterized as HBeAg positive with relentless typical ALT levels.

2.5. Laboratory Tests

Hematological profiles of hepatitis B Patients including Hb, WBC, RBC, Red Distribution Width (RDW) and Platelet Count (PLT) were determined using the automated analyzer (ADVIA 2120i, Siemens Health care Diagnostics, Deerfield, IL). Biochemical parameters including ALT and AST levels were measured using ARCHITECT i2000SR (Abbott, IL). HBsAg was detected by ELISA and HBV DNA copy levels were measured using Smart Cycler II PCR equipment (USA, Detection limit: 300 IU/mL).

2.6. Liver Biopsy

Liver biopsy was executed using Ultrasound, Siemens Company, Germany. In this biopsy, a minimal of 1.6 cm of liver tissue was considered for diagnosis. Each specimen section was fixed, paraffin-embedded, and fixed in formalin with hematoxylin-eosin and reticular fiber staining. Histopathological examination was cross-checked by two pathologists.

Liver fibrosis stage was classified with METAVIR scoring system. Liver fibrosis stages are mentioned below:

F0: No Fibrosis.

F1: Portal Fibrosis without septa.

F2: Portal Fibrosis with few septa

F3: Numerous septa without cirrhosis.

F4: Cirrhosis.

2.7. Model Calculation

FIB-4, AAR, APRI, RPR, platelet count, were used to predict the liver fibrosis. The following formulas were used to calculate the investigated non-invasive markers.

FIB-4= (Age(Y) ×AST (U/L))/ (Platelet count (109/L) ×√ALT (U/L))

APRI= {AST (ULN)/Platelet count (109/L)} ᵡ100

AAR= (AST/ALT) ratio

AARPRI= AAR/ {Platelet count (109/L)/150}

2.8. Statistical Analysis

Normality tests were performed by Kolmogorov–Smirnov test. Mean ± standard deviation and percentages were used to represent continuous and categorical data. t-test and U-test were carried out for comparison of continuous parameters. The correlation between the four noninvasive fibrosis scores and METAVIR fibrosis scores was analyzed using Pearson’s test. The diagnostic ability of the four investigated fibrosis scores was assessed by ROC curve analysis and calculating the area under the curve using Z test. Cut off values were determined by Youden index (sensitivity + specificity − 1). All tests were 2-tailed and P value < 0.05 was considered statistically significant. Data analysis was performed using the SPSS measurable bundle, form 16.0 (SPSS, Chicago, IL, USA) or R, variant 3.0 (http://www.r-project.org/).

3.1. General Features of CHB Patients

General features of the recruited CHB patients are presented in Table 1. The results, as shown in Table 1 indicate that there were no significant differences of mean age, FIB-4, APRI, AAR, and AARPRI among the training (79.8% men), external cohort (83.5% men), and internal cohort (78.8% men). There were no significant differences between the training external and internal cohorts regarding fibrosis stages using METAVIR scoring.

3.2. Correlation of the APRI, FIB-4, AAR and AARPRI Scores with METAVIR Fibrosis Stages

The relationship between METAVIR fibrosis stages and the non-invasive fibrosis scores is shown in Table 2. The METAVIR scores were positively correlated with FIB-4 (r=0.764, p<0.001), AAR (r=0.397, p<0.001), and AARPRI (r=0.628, p<0.001).Interestingly, APRI score showed the highest correlation with METAVIR (r=0.850, p<0.001).

Fig. (1) shows in training cohort, the relationship between non-invasive markers and liver fibrosis stages of HBV people. The non-invasive markers of FIB-4(ρ=0.610,p<0.001) and APRI (ρ=0.759, p<0.001) score showed the strongest (positive) correlation with fibrosis stages according to METAVIR score.

3.3. Diagnostic Performances of FIB-4, APRI, AAR, AARPRI for Fibrosis

To assess the diagnostic performance of the blood biomarkers, we performed ROC analysis, results are shown in Fig. (2). In the training cohort Fig. (2A), the area under curve (AUC) of APRI was 0.767 (0.722-0.812), which was considerably higher than that of FIB-4 (AUC=0.631, 0.581-0.681, P<0.001) while those of AAR and AARPRI were (AUC=0.484, 0.420-0.547, P>0.05), AARPRI (AUC=0.531, 0.474-0.588, P>0.05) respectively. On the other hand, the external cohort Fig. (2B) and the internal cohort Fig. (2C) showed a similar array of results. Relevant to diagnostic performance, the AUC with 95% CI was 0.847 (0.780-0.914, P<0.001), 0.707 (0.622-0.792, P<0.001), 0.568 (0.468-0.669, P=0.197), 0.594 (0.498-0.690, P=0.075) for APRI, FIB-4, AAR, and AARPRI respectively in external cohort Fig. (2B). In internal cohort, the AUC (95% CI) for APRI, FIB-4, AAR, and AARPRI was 0.775 (0.693-0.857, P<0.001), 0.653 (0.565-0.742, P<0.001), 0.542 (0.435-0.650, P=0.417), and 0.571 (0.468-0.674, P=0.190) one-to-one Fig. (2C).

3.4. The Optimal Cut Off Values for FIB-4, APRI, AAR and AARPRI Scores

The optimal cut off values for the assessment of liver fibrosis in CHB are shown in Table 3. In the training cohort, an optimal FIB-4 cut-off value of 1.159 produced a sensitivity of 83%, a specificity of 87% while the optimum cut-off value of APRI was 0.512 with a sensitivity and specificity of 84% and 67% respectively. The optimum cut-off value of AAR and AARPRI was 0.411 and 0.365 respectively and the corresponding sensitivity and specificity were 63%, 45%, 83%, and 72% respectively.