Study design and setting

A cross-sectional case-control study was conducted in the gastroenterology departments of Al-Ramadi, Baghdad, and Al-Yarmouk Teaching Hospitals in Iraq, between November 2024 and June 2025.

Ethical considerations

The study protocol was approved by the Scientific and Ethical Committee of the Iraqi Board for Medical Specializations (Ref. 2023-HEP-11). Written informed consent was obtained from all participants in accordance with the principles of the Declaration of Helsinki.

Participants

Treatment-naive adults with documented HBsAg seropositivity for a minimum of 6 months constituted the study group. Exclusion criteria encompassed co-infection with hepatitis C, D, and/or HIV, autoimmune hepatitis, metabolic dysfunction-associated steatotic liver disease with controlled attenuation parameter (CAP) > 290 dB/m, HCC, prior antiviral therapy, pregnancy, and immunosuppressive conditions. Healthy controls comprised HBsAg-negative volunteers with normal hepatic biochemistry. HBeAg-negative, anti-HBe-positive participants were further classified, in accordance with EASL and WHO criteria [12, 13], into three phases: inactive carriers (HBV DNA < 2,000 IU/mL with persistently normal ALT), the indeterminate “grey-zone” (HBV DNA 2,000–20,000 IU/mL and/or intermittently elevated ALT), and HBeAg-negative chronic active hepatitis (HBV DNA > 20,000 IU/mL with elevated ALT) [14]. This classification was subsequently related to viral load and liver stiffness.

Sample size calculation

Using OpenEpi software, a minimum sample of 84 CHB patients was calculated to achieve 80% statistical power at a significance level of 0.05 and an effect size of 0.5. Accordingly, 90 patients and 100 controls were enrolled to permit adequately powered subgroup analyses.

Clinical and laboratory assessments

Demographic and clinical data were systematically collected, and blood pressure and anthropometric parameters were recorded. Fasting venous blood samples were obtained for determination of complete blood counts, hepatic enzymes, total bilirubin, serum albumin, creatinine, prothrombin time, fasting glucose, and lipid profile. HBV serological markers were assayed by electrochemiluminescence immunoassay using the Cobas e601 analyzer, and HBV DNA was quantified by real-time polymerase chain reaction using the Sansure Biotech platform with a lower limit of detection of 10 IU/mL. Serum complement components C3 and C4 were measured by nephelometry.

Cytokine measurements

Serum concentrations of IL-6, IL-8, IL-10, IFN-γ, TNF-α, and TGF-β were quantified in duplicate using high-sensitivity enzyme-linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN, USA), with an intra-assay coefficient of variation of < 5%.

miR extraction and quantification

Total RNA was isolated from 200 μL serum aliquots using the miRNeasy Serum/Plasma Kit (Qiagen, Hilden, Germany), with exogenous synthetic cel-miR-39 (1.6 × 10⁸ copies) added as a spike-in normalization control. Expression levels of miR-122 and U6 were quantified using miScript RT and SYBR Green PCR Kits on a Rotor-Gene Q thermocycler. Relative expression was calculated using the 2^(−ΔΔCt) method.

Transient elastography

Transient elastography was performed using the FibroScan 530 Compact System by certified operators blinded to clinical and laboratory data. A minimum of ten valid measurements were acquired per subject, with an interquartile range (IQR)-to-median ratio of < 30% required for result validity. Liver stiffness measurements (LSMs) were stratified into fibrosis stages as follows: F0 (< 5.3 kPa), F1 (5.3–6.9 kPa), F2 (7.0–9.4 kPa), F3 (9.5–12.4 kPa), and F4 (≥ 12.5 kPa), adapted from previously validated transient elastography thresholds [15, 16]. CAP values were simultaneously recorded.

Non-invasive fibrosis scores

APRI = [(AST/ULN)/platelets (10⁹/L)] × 100; FIB-4 = [age × AST]/[platelets × $\sqrt{\mathrm{A}\mathrm{L}\mathrm{T}}$] [4].

Statistical analysis

Statistical analyses were conducted using MedCalc version 20 and SPSS version 28.0. Distributional normality was assessed using the Shapiro-Wilk test. Continuous variables were expressed as median (IQR) or mean ± standard deviation, as appropriate. Intergroup comparisons were performed using the independent samples t-test or Mann-Whitney U test for continuous variables and the chi-square test for categorical data. Multi-group comparisons were conducted using one-way analysis of variance (ANOVA) or the Kruskal-Wallis test according to data distribution. Correlation analyses were performed using Spearman rank correlation coefficients. Multivariable logistic regression was employed to identify independent predictors of significant fibrosis (F ≥ 2) and advanced fibrosis (F3–F4). The composite biomarker score was derived as the β-weighted linear predictor of a multivariable binary logistic regression model in which significant fibrosis (F ≥ 2) served as the dependent variable and log-transformed IL-6, TNF-α, and miR-122 were entered as covariates; the resulting continuous score was then evaluated by receiver operating characteristic (ROC) analysis, and an optimal cut-off was selected by the Youden index. Pairwise comparisons of area under the ROC curve (AUC) were performed using the DeLong test. A two-sided P-value < 0.05 was considered statistically significant.

Study population characteristics

A total of 90 CHB patients (54% male; median age, 35 years) and 100 age- and sex-matched healthy controls (52% male; median age, 34 years) were enrolled. Among CHB patients, 50 (56%) were HBeAg-positive. Compared with controls, CHB patients exhibited significantly elevated ALT (46 vs. 22 IU/L), AST (42 vs. 21 IU/L), and LSM (7.4 vs. 4.3 kPa; all P < 0.001), along with lower platelet counts (230 vs. 275 × 10⁹/L, P < 0.001). Median HBV DNA was 6.1 log₁₀ IU/mL. No statistically significant differences were observed between groups with respect to age, sex, body mass index, or CAP (Table 1).

Distribution of fibrosis and inflammation by HBeAg status

The distribution of fibrosis stages was as follows: F0, 22 patients (24%); F1, 20 (22%); F2, 18 (20%); F3, 15 (17%); and F4, 15 (17%). Among HBeAg-positive patients (n = 50), fibrosis stages F0 through F4 were observed in 8, 9, 10, 11, and 12 patients, respectively. In contrast, HBeAg-negative patients (n = 40) demonstrated the following distribution: F0, 14; F1, 11; F2, 8; F3, 4; and F4, 3 patients. Necroinflammatory activity grades S0, S1, S2, and S3 were documented in 24 (27%), 28 (31%), 22 (24%), and 16 (18%) patients, respectively. HBeAg-positive patients demonstrated significantly higher LSM than HBeAg-negative patients (9.8 ± 5.2 vs. 7.5 ± 4.8 kPa; P = 0.04), with a correspondingly greater proportion of advanced fibrosis [F3–F4: 23/50 (46%) vs. 7/40 (18%); P = 0.004] (Figure 1; Table S1). Among the 40 HBeAg-negative, anti-HBe-positive patients, 20 (50%) were categorized as inactive carriers, 11 (28%) as indeterminate “grey-zone,” and 9 (22%) as HBeAg-negative chronic active hepatitis. Median HBV DNA and liver stiffness rose progressively across these phases (inactive, 5.0 kPa; grey-zone, 7.1 kPa; active hepatitis, 9.8 kPa), and significant fibrosis (F ≥ 2) was confined predominantly to the grey-zone and active-hepatitis groups, whereas inactive carriers exhibited predominantly F0–F1 disease.

Display full size

Figure 1. Distribution of fibrosis stages by hepatitis B e-antigen (HBeAg) status.

Serum cytokine levels across fibrosis stages

Serum concentrations of IL-6, IL-8, TNF-α, and TGF-β demonstrated a progressive escalation across fibrosis stages F0 through F4 (P for trend < 0.001 for all biomarkers, Table 2). Serum IL-6 at stage F4 was 5.8-fold higher than at F0 (798 vs. 138 pg/mL). Correspondingly, serum IL-8, TNF-α, and TGF-β concentrations increased from 45 pg/mL, 310 pg/mL, and 1.8 ng/mL at F0 to 135 pg/mL, 620 pg/mL, and 6.4 ng/mL at F4, respectively. Conversely, serum IL-10 and IFN-γ exhibited peak concentrations at the moderate inflammation stage (S2) followed by a decline at S3, consistent with a state of immunoregulatory exhaustion (Figure 2). Cytokine concentrations stratified by HBeAg status are presented in Table S1.

Display full size

Figure 2. Biphasic trajectory of interleukin-10 (IL-10) and interferon-gamma (IFN-γ) across necroinflammatory stages (S0 to S3) in 90 patients with chronic hepatitis B (CHB). Both cytokines demonstrated peak expression at the moderate inflammation stage (S2), with IL-10 reaching 78.5 pg/mL (73% increase relative to S0 baseline of 45.5 pg/mL) and IFN-γ reaching 110.4 pg/mL (37% increase relative to S0 baseline of 80.5 pg/mL). Subsequently, cytokine concentrations declined during severe inflammation (S3), with IL-10 decreasing to 65.2 pg/mL (17% reduction from S2) and IFN-γ declining to 95.1 pg/mL (14% reduction from S2). This non-linear biphasic pattern contrasts with the monotonic escalation observed for pro-inflammatory cytokines [IL-6, tumor necrosis factor-alpha (TNF-α), transforming growth factor-beta (TGF-β)].

Expression of miR-122 and its association with fibrosis progression

Circulating miR-122 expression was significantly downregulated in CHB patients relative to healthy controls, with relative fold-change values of 1.0000 and 0.0440 in controls and F0, respectively. A progressive decline in miR-122 expression was observed with advancing fibrosis stages, with fold-change values of 0.0070 (F1), 0.0010 (F2), 0.0003 (F3), and 0.0002 (F4) relative to controls (Figure 3). miR-122 expression demonstrated significant inverse correlations with IL-6 (ρ = −0.62), TNF-α (ρ = −0.60), and TGF-β (ρ = −0.45), the full biomarker correlation matrix is provided in Table S2. Suppression of miR-122 below a threshold of ≤ 0.001 was associated with 6.4-fold higher odds of advanced fibrosis [odds ratio (OR) 6.4; 95% confidence interval (CI): 2.8–14.5; P < 0.001]. miR-122 demonstrated superior diagnostic performance for CHB compared with conventional aminotransferases, achieving an AUC of 0.89 versus 0.69 for ALT and 0.71 for AST. A miR-122 threshold of ≤ 0.001 yielded optimal sensitivity and specificity.

Display full size

Figure 3. Stepwise suppression of microRNA-122 (miR-122) across fibrosis stages (bar chart of median relative miR-122 expression (2^−ΔΔCt, log₁₀ scale) in controls and across stages F0–F4).

Diagnostic performance of the biomarkers for significant fibrosis

For the detection of significant fibrosis (F ≥ 2), IL-6 at a threshold of ≥ 420 pg/mL yielded an AUC of 0.87 (95% CI: 0.79–0.93), with a sensitivity of 84% and specificity of 81%, significantly outperforming APRI (P = 0.02). TNF-α at a threshold of ≥ 480 pg/mL achieved an AUC of 0.83 (95% CI: 0.74–0.90), with 78% sensitivity and 79% specificity, also significantly superior to APRI (P = 0.04). Among individual biomarkers, miR-122 at a threshold of ≤ 0.001 demonstrated the highest discriminative accuracy, with an AUC of 0.89 (95% CI: 0.81–0.94), 86% sensitivity, and 85% specificity, significantly exceeding the performance of APRI (AUC 0.78; sensitivity 72%; specificity 74%) for detecting significant fibrosis (Table 3).

Composite biomarker score outperforms conventional indices

A composite score integrating these three biomarkers demonstrated superior predictive accuracy for significant fibrosis, with an AUC of 0.93 (95% CI: 0.87–0.97), significantly outperforming both APRI (AUC 0.78; sensitivity 72%; specificity 74%) and FIB-4 (AUC 0.74; 95% CI: 0.64–0.83; sensitivity 75%; specificity 72%; P = 0.002 for pairwise comparisons). At an optimal composite score threshold of ≥ 1.7, the model achieved 91% sensitivity, 88% specificity, and 89% overall accuracy, representing an AUC gain of 0.15–0.19 over conventional indices (Figure 4).

Display full size

Figure 4. ROC curves comparing the composite score with APRI and FIB-4 for significant fibrosis (F ≥ 2). ANOVA: analysis of variance; APRI: aspartate aminotransferase/platelet ratio index; AUC: area under the receiver operating characteristic curve; FIB-4: fibrosis-4 index; ROC: receiver operating characteristic; Se: sensitivity; Sp: specificity.

Independent predictors of advanced fibrosis

Multivariable logistic regression analysis was performed to identify independent predictors of advanced fibrosis, adjusting for age, sex, viral load, ALT, and platelet count. Independent predictors of advanced fibrosis (F3–F4) included HBeAg positivity (OR 4.2; 95% CI: 2.0–8.9; P < 0.001), miR-122 suppression ≤ 0.001 (OR 6.4; 95% CI: 2.8–14.5; P < 0.001), IL-6 elevation ≥ 420 pg/mL (OR 3.9; 95% CI: 1.7–8.8; P = 0.001), age per 10-year increment (OR 1.6; 95% CI: 1.2–2.2; P = 0.003), and platelet count per 50 × 10⁹/L increment (OR 0.7; 95% CI: 0.5–0.9; P = 0.01) (Table 4).

Correlation with non-invasive fibrosis scores

APRI demonstrated moderate correlations with IL-6 (ρ = 0.56, P < 0.001), TNF-α (ρ = 0.48, P < 0.001), and miR-122 (ρ = −0.52, P < 0.001). FIB-4 exhibited comparatively weaker correlations with these biomarkers. Notably, neither APRI nor FIB-4 demonstrated the capacity to capture the biphasic immunological trajectories of IL-10 and IFN-γ observed across necroinflammatory stages.

Mechanistic insights

The inverse correlation observed between miR-122 and pro-inflammatory cytokine concentrations suggests a mechanistic interrelationship. miR-122 modulates hepatic lipid metabolism and toll-like receptor signaling cascades [9]. The progressive depletion of miR-122 may be attributable to hepatocyte apoptosis and concomitant stellate cell activation, potentially establishing a self-perpetuating positive feedback loop of fibrogenesis. Among the cytokines, IL-6 showed the strongest correlation with liver stiffness (ρ = 0.72), and TGF-β was also positively correlated (ρ = 0.65; Table S2), consistent with its well-characterized role as a principal mediator of extracellular matrix deposition [7].

The biphasic expression patterns of IL-10 and IFN-γ, characterized by peak expression at moderate necroinflammatory activity followed by subsequent decline at severe inflammation, are suggestive of immunoregulatory exhaustion, a recognized hallmark of chronic viral infections [17, 23]. Mechanistically, this pattern may reflect an early expansion of regulatory T cells and monocyte-derived IL-10 that restrains necroinflammation during moderate injury, followed at severe stages by progressive exhaustion of effector CD8⁺ T cells (characterized by up-regulation of PD-1, CTLA-4, and TIM-3 and by declining IFN-γ output), together with contraction of the regulatory compartment. The resulting collapse of both effector and regulatory arms would leave fibrogenic signaling (TGF-β, IL-6) unopposed, providing a plausible immunopathological explanation for accelerated fibrosis at stage S3. Clinically, this biphasic signature is not captured by APRI or FIB-4 and may identify a window in which patients with waning IL-10/IFN-γ have already transitioned toward high-risk disease and merit prioritized antiviral therapy; it also nominates immune-checkpoint modulation as a candidate strategy for restoring immunoregulatory competence, a hypothesis requiring dedicated mechanistic study.

Strengths and limitations

The present study possesses several methodological strengths, including prospective enrollment with standardized protocols, comprehensive multi-analyte biomarker profiling, and universal application of transient elastography for fibrosis staging. However, several limitations warrant acknowledgment. The cross-sectional design precludes establishment of causal relationships; longitudinal validation is currently underway. The healthy control group, although demographically matched, may not be fully representative of the general population. Fibrosis staging by transient elastography was not corroborated by histological assessment. Moreover, the composite score was developed and evaluated within a single cohort without an independent internal or external validation set, so its apparent performance may be optimistic and requires confirmation in a separate population. Because transient elastography, rather than liver biopsy, was used as the fibrosis reference standard, it does not constitute a true histological gold standard and may itself misclassify fibrosis in the presence of active necroinflammation. The geographic restriction of the study population to Iraqi centers limits the generalizability of these findings to other HBV genotypes and ethnic populations. The study did not assess therapeutic outcomes, and future investigations are warranted to determine whether biomarker-guided treatment strategies confer clinical benefit.

Future directions

Future investigations should encompass prospective multicenter studies to validate the composite score across diverse populations and HBV genotypes. Longitudinal studies conducted during antiviral therapy may facilitate early identification of treatment responders. Incorporation of emerging biomarkers, such as hepatitis B core-related antigen and serum HBV RNA [24], into the predictive model may further enhance its prognostic accuracy. Elucidation of the regulatory mechanisms governing miR-122 expression may reveal novel therapeutic targets for antifibrotic intervention.

Conclusions

HBeAg positivity, elevated pro-inflammatory cytokine concentrations, and suppressed miR-122 expression constitute an integrated molecular triad that correlates with fibro-inflammatory severity in CHB. This three-biomarker panel may complement elastography for fibrosis assessment, particularly where elastography is unavailable, and may inform therapeutic decision-making regarding antiviral intervention. Prospective validation of this triad is warranted to determine whether biomarker-guided therapeutic modulation confers clinical benefit in CHB patients.