Impact of Metabolic Associated Steatotic Liver Disease on Antiviral Therapy Outcomes on Chronic Hepatitis B Patients Receiving Antiviral Therapy.

1
Introduction
Chronic hepatitis B (CHB) remains a formidable global health challenge, affecting over 250 million individuals and contributing to approximately 800 000 deaths annually due to cirrhosis and hepatocellular carcinoma (HCC) [1]. The advent of nucleus(t)ide analogs (NAs), such as entecavir, tenofovir alafenamide, and tenofovir disoproxil, has revolutionized CHB management by effectively suppressing viral replication, reducing liver inflammation, and slowing disease progression [2]. Despite these advances, a significant proportion of CHB patients still progress to severe liver complications, highlighting the need to identify and address additional risk factors that may undermine therapeutic efficacy.
In recent years, the rising global prevalence of metabolic‐associated steatotic liver disease (MASLD)—previously known as nonalcoholic fatty liver disease (NAFLD)—has emerged as a critical concern in the context of CHB [3]. MASLD, characterized by hepatic steatosis in the absence of significant alcohol consumption, is closely linked to metabolic syndrome components such as obesity, type 2 diabetes, hypertension, and dyslipidemia [4]. These metabolic abnormalities drive liver inflammation and fibrosis, creating a synergistic effect when combined with chronic viral hepatitis [5]. Emerging evidence suggests that MASLD not only exacerbates liver damage in CHB patients but may also attenuate the effectiveness of NA therapies, which primarily target viral replication rather than metabolic pathways [6].
The interaction between MASLD and CHB is complex and multifaceted. Several studies have reported an increased risk of HCC in CHB patients with concurrent MASLD, even when viral replication is suppressed, likely due to metabolic‐driven hepatocarcinogenesis [7]. Additionally, metabolic risk factors such as diabetes have been shown to amplify HCC risk in this population [8]. However, conflicting data exist, with some studies suggesting that hepatic steatosis might reduce HBV replication and disease severity, potentially offering a protective effect in certain contexts [9]. These discrepancies underscore the urgent need for comprehensive research to clarify the combined impact of MASLD and CHB on treatment outcomes.
To address this gap, our study integrates two complementary approaches: a large‐scale real‐world analysis using the TriNetX US Collaborative Network and a detailed clinical cohort study. The TriNetX analysis provides robust, population‐level insights into the long‐term outcomes of CHB patients on NA therapy, comparing those with and without MASLD. In parallel, the clinical cohort offers granular data on hepatic steatosis and fibrosis, measured through controlled attenuation parameter (CAP) and liver stiffness, respectively, in a well‐characterized group of CHB and MASLD patients. We also employ tools such as the albumin–bilirubin (ALBI) grade and fibrosis‐4 (FIB‐4) index to assess liver function and fibrosis severity, providing a multidimensional view of disease progression [10]. By examining liver‐related complications, survival rates, and fibrosis burden, this study aims to elucidate the detrimental effects of MASLD on NA therapy outcomes in CHB patients and advocate for integrated management strategies that address both viral and metabolic drivers of liver disease.

2
Methods
2.1
Study Design
This study integrated a retrospective analysis using the TriNetX US Collaborative Network, covering 68 healthcare organizations, with a clinical cohort study to evaluate MASLD's effect on CHB patients receiving long‐term NA therapy with undetectable serum HBV DNA.

2.2
TriNetX Cohort
CHB patients receiving long‐term NA therapy with undetectable serum HBV DNA were divided into CHB‐MASLD‐NA (CHB with MASLD) and CHB‐non MASLD‐NA (CHB without MASLD) groups. MASLD was defined using ICD‐10 codes for fatty liver (K76.0) or nonalcoholic steatohepatitis (K75.81) plus metabolic comorbidities (e.g., obesity [E66], type 2 diabetes [E11], hypertension [I10], and dyslipidemia [E78.5]). CHB was identified via ICD‐10 code B18.1. Propensity score matching (1:1) balanced cohorts for age, sex, ethnicity, ALBI/FIB‐4 parameters, and comorbidities. Before matching, the CHB‐MASLD‐NA cohort had 5600 patients and the CHB alone‐NA cohort had 11 021. After 1:1 matching, each cohort comprised 4761 patients. IRB approval number: N202601100.

2.3
Clinical Cohort
The clinical cohort included 201 patients (64 CHB and 137 MASLD controls). The 64 CHB patients on long‐term NA therapy with undetectable serum HBV DNA, stratified by CAP into four groups: S1 (215–252, mild steatosis, n = 13), S2 (252–296, moderate steatosis, n = 17), S3 (> 296, severe steatosis, n = 20), and CHB patient without MASLD (CHB alone, n = 14). A comparison group of 137 MASLD patients without HBV, also stratified by CAP (MASLD‐non CHB, S1: n = 13, S2: n = 27, S3: n = 97), was included. Liver stiffness was measured via transient elastography, and fibrosis stages (F0–F4) were compared across MASLD‐non CHB, CHB alone‐NA, and CHB‐MASLD‐NA groups. Liver stiffness values < 6 kPa were considered normal, while 8 and 12.5 kPa served as cut‐offs for F3 and F4 fibrosis, respectively. Metabolic parameters (BMI, triglycerides, uric acid, and HbA1c), HBsAg, IFN‐γ, and TNF‐α levels were assessed. The study protocol was approved by the Institutional Review Board of Taipei Medical University (approval numbers: N201608026, N201605071, and N201805009), with informed consent obtained from all participants in accordance with the Declaration of Helsinki. The study was also registered on ClinicalTrials.gov (NCT02953600, NCT02875392, and NCT04715776).

2.4
Outcome Measures
TriNetX: Primary outcomes were HCC (ICD‐10: C22.0) and cirrhosis (ICD‐10: K74) incidence over 10 years, with survival probability as a secondary outcome.
Clinical Cohort: Outcomes included liver stiffness, fibrosis stage distribution, and metabolic trends across CAP groups.

2.5
Statistical Analysis
TriNetX: Kaplan–Meier analysis estimated survival, with Cox models calculating hazard ratios (HRs) for liver complications, adjusting for cohort, sex, age, and HBV DNA. Risk ratios (RR), odds ratios (OR), and risk differences evaluated outcome incidence (p < 0.05).
Clinical Cohort: Descriptive statistics summarized liver stiffness, fibrosis stages, and metabolic parameters. Comparative analyses assessed trends across CAP groups and disease categories.

3
Results
3.1
TriNetX Cohort: Baseline Characteristics After Propensity Score Matching
Before matching, the CHB‐MASLD‐NA cohort had 5600 patients and the CHB alone‐NA cohort had 11 021. After 1:1 matching, each cohort comprised 4761 patients, balanced for age, sex, ethnicity, ALBI/FIB‐4 parameters, and comorbidities, but CHB‐MASLD‐NA showed higher BMI (29.2 vs. 23.7, p < 0.001), triglyceride (152.7 vs. 111.6 mg/dL, p = 0.004), total cholesterol (180.7 vs. 168.0 mg/dL, p = 0.002), and HbA1c (6.4% vs. 5.7%, p = 0.026), which are indicators associated with MASLD (Table 1).

3.2
Survival Analysis and Risk Assessment for TriNetX Cohort After Matching
In the propensity score–matched cohort, 10‐year overall survival (OS) was compared between CHB patients with and without MASLD using Kaplan–Meier analysis. The B‐MASLD‐NA cohort exhibited a lower survival probability of 61.45% compared to 79.30% in the B‐non MASLD‐NA cohort (log‐rank test: χ² = 66.911, p < 0.001).
The hazard ratio (HR) in the CHB‐MASLD‐NA cohort was 1.747 (95% CI 1.526–2.001, p < 0.001), indicating a 1.747‐fold increased risk compared to the CHB alone‐NA cohort. This highlights the detrimental impact of MASLD on long‐term survival outcomes in CHB patients receiving NA therapy (shown in Figure 1 and Table 2).

3.3
Independent Predictors in the Progressive Development of HCC or Cirrhosis in CHB Patients With or Without MASLD
To delineate independent risk factors for the development of HCC or cirrhosis, a Cox proportional hazards model was conducted via the TriNetX platform, analyzed data from 5600 patients in the CHB‐MASLD‐NA cohort and 11 021 patients in the CHB alone‐NA cohort (Table 3). Consistent with the survival analysis, the CHB‐MASLD‐NA was identified as a significant risk factor compared to CHB alone, with an HR of 1.547 (95% CI: 1.453–1.648, p < 0.001). Gender also significantly contributed to risk stratification, male was associated with a comparable risk increase (HR: 1.546, 95% CI: 1.451–1.647, p < 0.001), and advancing age showed a consistent association with disease progression (HR: 1.028, 95% CI: 1.025–1.030, p < 0.001). Notably, HBV DNA viral load was not a significant predictor of HCC or cirrhosis in this model (HR: 1.000, p = 0.590). This suggests that metabolic and demographic factors may play a more dominant role in this population.

3.4
Impact of Steatosis Severity on Cirrhosis Progression From Clinical Cohort
Insights from the TriNetX cohort suggest that MASLD‐related factors, rather than HBV DNA levels, may play a more dominant role in the progressive development of HCC or cirrhosis in CHB patients receiving NA therapy. However, because the TriNetX database does not allow for a granular analysis of the relationship between fibrosis severity and MASLD, we utilized a clinical cohort to further investigate the correlations between fibrosis and steatosis. The clinical cohort comprised 64 CHB patients undergoing NA therapy with undetectable serum HBV DNA, who were stratified according to MASLD criteria into two groups: CHB‐MASLD‐NA (n = 50) and CHB alone‐NA (n = 14). Among the 50 CHB‐MASLD‐NA patients, liver stiffness varied with steatosis severity (Table 4). Mean liver stiffness values increased from 7.931 kPa in mild steatosis (S1) to a peak of 9.853 kPa in moderate steatosis (S2), before slightly decreasing to 9.150 kPa in severe steatosis (S3). For reference, LS values < 6 kPa were considered normal, while 8 and 12.5 kPa served as cut‐offs for F3 and F4 fibrosis, respectively. Metabolic markers rose with CAP: BMI from 25.240 to 29.148 kg/m², triglycerides (TG) from 113.778 to 126.286 mg/dL, HbA1c from 5.072 to 6.231, and uric acid (UA) from 5.577 to 6.360 mg/dL. HBsAg levels decreased with steatosis severity from 2.722 to 2.539 log IU/mL. In the MASLD‐non CHB group (n = 137), liver stiffness was lower: 5.262 kPa (S1), 4.952 kPa (S2), and 6.228 kPa (S3), despite similar CAP levels. Fibrosis distribution showed CHB‐MASLD‐NA patients had the highest advanced fibrosis rate (28.0% at F4), compared to 14.3% for CHB alone‐NA and 4.38% for MASLD‐non CHB (shown in Figure 2). Our findings emphasize that in CHB patients with viral suppression, MASLD‐related factors, rather than HBV DNA levels, serve as the predominant drivers for the development of cirrhosis.

3.5
Increased Steatosis Severity Promotes Antiviral Cytokine Production but Accelerates Pro‐inflammatory Liver Damage in CHB Patients Receiving NA Therapy From Clinical Cohort
Interestingly, our data reveal that as CAP severity increases, HBsAg levels tend to decline from 2.722 ± 0.220 log IU/mL in the S1 group to 2.539 ± 0.163 log IU/mL in the S3 group, suggesting that steatosis may actually facilitate HBV clearance (Table 4).
In chronic HBV infection, HBV‐specific T cells play a pivotal role in balancing viral clearance and liver damage. HBV‐specific IFN‐γ producing CD4 T cells are crucial immune cells that recognize HBV antigens and release interferon‐gamma (IFN‐γ) to help clear the infection, representing an effective immune response, unlike dominant TNF‐α producing CD4 cells in chronic cases, which can worsen liver damage [11, 12]. Data from patients categorized by CAP levels reveal that as steatosis severity increases from S1 to S3, there is a significant rise in cytokine production. Specifically, IFN‐γ levels increase from 3.291 ± 0.392 ng/mL (S1) to 8.168 ± 1.031 ng/mL (S3), while TNF‐α levels rise from 1.909 ± 0.144 ng/mL (S1) to 6.987 ± 1.359 ng/mL (S3). These findings indicate that both the potential for viral clearance and the extent of liver injury intensify as steatosis worsens (Table 4 and Figure 3A).
Furthermore, we examined the relationship between hepatic steatosis severity and the relative concentration fold‐change in IFN‐γ and TNF‐α levels, normalized to the means in CAP S1 group. Our analysis reveals that as CAP levels progress to S3, the fold‐increase of the pro‐inflammatory cytokine TNF‐α rises to 3.726 ± 0.724 fold, which is significantly sharper than the 2.128 ± 0.268 observed for IFN‐γ (Figure 3B). This disproportionate escalation suggests that while the immune system attempts to clear the virus more vigorously in patients with high CAP scores, the resulting severe hepatic inflammation and liver damage become the dominant pathological features.

4
Discussion
This study provides compelling evidence that MASLD significantly aggravates the prognosis of CHB patients receiving long‐term NA therapy with undetectable serum HBV DNA, supported by evidence from both a large real‐world dataset and a well‐characterized clinical cohort. The TriNetX analysis revealed a striking 1.547‐fold increased risk of adverse liver outcomes, including cirrhosis and HCC, in CHB‐NA patients with MASLD (HR 1.547, 95% CI 1.453–1.648, p < 0.001). Survival outcomes were equally concerning, with a 10‐year survival probability dropping to 61.45% in the CHB‐MASLD‐NA cohort compared to 79.30% in CHB alone‐NA patients (p < 0.001). These findings are mirrored by the clinical cohort, which demonstrated that CHB patients with MASLD exhibited significantly higher liver stiffness—peaking at 9.853 kPa in the CAP S2 group (252–296)—compared to 4.952 kPa in MASLD‐non CHB patients at the same steatosis level. Furthermore, the CHB‐MASLD‐NA group had a twofold higher advanced fibrosis rate (28.0%) than CHB alone‐NA patients (14.3%) and a sixfold increase over MASLD‐non CHB patients (4.38%), underscoring the synergistic burden of these conditions.
The clinical cohort's findings highlight the additive fibrogenic effect of HBV in the presence of MASLD, even when viral replication is suppressed by NA therapy. The elevated liver stiffness in CHB‐MASLD‐NA patients compared to MASLD‐non CHB patients at matched CAP levels suggests that HBV drives additional inflammatory and fibrotic pathways, potentially through mechanisms such as oxidative stress and immune activation [13, 14]. Interestingly, HBsAg levels in the CHB‐MASLD‐NA group decreased with increasing CAP (from 2.722 to 2.539 log IU/mL), supporting prior observations that hepatic steatosis may suppress HBV replication [15]. The suppressive effect of steatosis on HBV DNA is clinically misleading, as the Cox model yielded an HR of 1.000 for HBV DNA regarding the risk of HCC or cirrhosis (Table 3). Therefore, viral load no longer predicts outcomes, as metabolic dysfunction has replaced viral replication as the chief determinant of disease progression. While this study did not directly characterize T‐cell subsets, the observed imbalance in cytokine profiles may offer a mechanistic explanation. Our analysis of cytokine trends relative to steatosis severity revealed that TNF‐α levels increased at a significantly steeper rate than IFN‐γ as steatosis progressed from the CAP S1 means (Table 4 and Figure 3). This interplay between steatosis and HBV is ultimately deleterious, as the drive toward hepatic inflammation and fibrosis overrides the benefits of lower viral loads. This aligns with the TriNetX data, where HBV DNA viral load was not a significant predictor of adverse outcomes (HR 1.000, p = 0.590), while MASLD emerged as a key independent risk factor (HR 1.547, p < 0.001).
This observation resonates with the “viral‐metabolic paradox” recently characterized by Abu Baker et al., who reported that while CHB‐MASLD patients often exhibit lower viral loads and higher rates of HBeAg seroconversion, they paradoxically face a significantly higher risk of cirrhosis [16]. Furthermore, Van Hees et al. provided longitudinal evidence clarifying the impact of MASLD on medication efficacy; they demonstrated that concurrent steatotic liver disease blunts the decline in liver stiffness typically induced by NA therapy, with many patients failing to achieve fibrosis regression despite viral suppression [17]. Although our clinical cohort predominantly represents an Asian population, the consistency of these findings with the multi‐ethnic populations in the US‐based TriNetX network and the cohorts from Israel and Europe suggests that MASLD acts as a universal driver of fibrosis progression across different ethnicities. Together, these data reinforce that NA therapy alone, while effective for viral control, is insufficient to reverse the metabolic‐driven fibrosis, explaining the persistently elevated liver stiffness and poor survival outcomes observed in our study.
The metabolic burden in CHB‐MASLD‐NA patients further exacerbates their risk profile. In the clinical cohort, increasing CAP levels were associated with progressive elevations in BMI (25.240–29.148 kg/m²), triglycerides (113.778–126.286 mg/dL), HbA1c (5.072–6.231), and uric acid (5.577–6.360 mg/dL). These findings corroborate the metabolic profiles observed in the TriNetX cohort, where patients in the CHB‐MASLD‐NA group similarly exhibited significantly higher BMI, triglycerides, total cholesterol, and HbA1c levels compared to those with CHB alone. These metabolic derangements likely contribute to hepatic inflammation and fibrosis through pathways such as insulin resistance and lipid accumulation, which are well‐documented drivers of hepatocarcinogenesis in MASLD [18]. The survival disparity observed in the TriNetX cohort—nearly 18% lower in CHB‐MASLD‐NA patients—may be amplified by these metabolic comorbidities, which compound the liver's inflammatory milieu and accelerate disease progression.
From a clinical perspective, these findings have profound implications for the management of CHB patients with MASLD. The TriNetX data's high incidence of HCC and cirrhosis in CHB‐MASLD‐NA patients (19.7% vs. 10.0%) and the clinical cohort's elevated fibrosis burden underscore the need for vigilant monitoring of liver function and fibrosis [10]. Regular screening with non‐invasive tools like transient elastography, as used in our clinical cohort, can facilitate early detection of advanced fibrosis and HCC, enabling timely intervention. Moreover, the data advocate for a paradigm shift in treatment strategies. While NA therapy effectively controls viral replication, it does not address the metabolic drivers of liver disease in MASLD. Integrating antiviral therapy with targeted metabolic interventions—such as weight management, lipid‐lowering therapies, and glycemic control—is critical to mitigate disease progression [19]. Lifestyle modifications, including dietary changes and increased physical activity, should be emphasized as part of patient education to enhance NA therapy efficacy and reduce complications [20].
Looking forward, our findings suggest potential avenues for therapeutic innovation. The high fibrosis burden in CHB‐MASLD patients highlights the need for combination therapies that address both viral and metabolic factors. Emerging agents targeting steatosis and inflammation, such as FXR agonists or PPAR agonists, could be explored in conjunction with NAs to improve outcomes in this high‐risk population [21]. Additionally, the observation that HCC can develop in MASLD patients without significant fibrosis, as noted in prior studies, calls for tailored screening approaches that account for metabolic risk factors beyond traditional fibrosis staging [21, 22]. Future research should focus on prospective studies with larger cohorts, incorporating histologic data and detailed metabolic profiling to further elucidate the mechanisms underlying the MASLD‐CHB interaction. Such studies could also explore the role of genetic and epigenetic factors in driving disease progression, offering insights into personalized treatment strategies.
Limitations in this study. The TriNetX analysis, while robust, is retrospective and may be subject to selection bias and diagnostic variability across healthcare organizations. The clinical cohort, although insightful, is limited by its small sample size (n = 64 for CHB) and lack of longitudinal outcome data, which restricts causal inferences. Neither dataset assessed lifestyle factors such as diet or physical activity, which are known to influence MASLD and CHB progression [20]. The lack of direct analysis regarding T‐cell subsets. Future research should include a detailed characterization of these immune cells to better understand how metabolic dysfunction disrupts the balance between pro‐inflammatory and antiviral responses. Additionally, the absence of histologic data limits the precision of fibrosis staging, a common challenge in real‐world studies [16]. Despite these constraints, the integration of real‐world and clinical data provides a comprehensive view of MASLD's impact on CHB, offering actionable insights for clinical practice.
Clinical Implications. Clinicians treating CHB patients with MASLD should not rely solely on HBV DNA suppression. Instead, they should actively incorporate metabolic risk stratification (e.g., regular CAP/FibroScan monitoring) and targeted metabolic interventions, specifically the management of BMI, lipid profiles (triglycerides), and glycemic status (HbA1c), into the standard of care. Given the observed cytokine imbalance, future care should incorporate immunophenotyping to elucidate the role of HBV‐specific T‐cell exhaustion or polarization within the MASLD microenvironment.
In conclusion, MASLD significantly exacerbates liver‐related outcomes and reduces survival in CHB patients, despite effective HBV suppression with NA therapy, driven by increased fibrosis and metabolic burden. The synergy between viral and metabolic factors underscores the need for a holistic management approach that combines antiviral therapy with targeted interventions for metabolic risk factors. By addressing both components, clinicians can better mitigate the risk of liver complications and improve long‐term outcomes for this vulnerable population.

Author Contributions
Conceptualization: Sheng‐Jie Shiue and Ming‐Shun Wu. Collecting and interpreting data: Tung‐Han Ho, Sheng‐Wei Cheng, Li‐Wei Wu, Tze‐Sian Chan, Chun‐Nan Chen, Chao‐Ling Cheng, Yan Kang Lee, Hsien‐Yao Yang, Han‐Shiang Shiue, and Ming‐Shun Wu. Formal analysis: Hsin‐Yi Lin, Sheng‐Jie Shiue, Kuo‐Feng Leng, Han‐Shiang Shiue, and Ming‐Shun Wu. Methodology: Sheng‐Jie Shiue and Ming‐Shun Wu. Drafting the manuscript: Sheng‐Jie Shiue, Han‐Shiang Shiue, and Ming‐Shun Wu. Review and editing the manuscript: Shih‐Ping Huang, Hsin‐Yi Lin, and Ming‐Shun Wu. All authors have read and agreed to the published version of the manuscript.

Ethics Statement
The TriNetX Cohort:TriNetX operates in compliance with the Health Insurance Portability and Accountability Act (HIPAA), the U.S. federal law that safeguards the privacy and security of healthcare data. Additionally, TriNetX is certified under the ISO 27001:2022 standard and maintains a comprehensive Information Security Management System (ISMS) to ensure the protection of healthcare data and adherence to the HIPAA Security Rule. Further information and documentation are available on the official TriNetX website. Verbal informed consent was obtained from the parent/legal guardian of participants prior to the study. This consent procedure and the clinical cohort protocol were reviewed and approved by the Institutional Review Board of Taipei Medical University, approval numbers: N202601100, N201608026, N201605071, and N201805009, with informed consent obtained from all participants in accordance with the Declaration of Helsinki.

Conflicts of Interest
The authors declare no conflicts of interest.

Data Availability Statement
The data that support the findings of this study are available from the corresponding author upon reasonable request.