We read with interest the study by Hong et al. [1] investigating fibrosis risk in chronic hepatitis B (CHB) patients with metabolic dysfunction-associated steatotic liver disease (MASLD). The large biopsy-confirmed cohort provides valuable insights into dual liver disease burden. However, three methodological aspects warrant clarification to strengthen the conclusions.
First, the CHB control group included patients with steatosis <5% or >5% without cardiometabolic risk factors (CMRFs). This combines distinct populations: those without steatosis and those with steatosis lacking metabolic dysfunction. The authors note that steatosis grade did not correlate with fibrosis (r = −0.077, p = 0.016), yet subgroup analyses comparing these cohorts are absent. Prior work confirms that isolated steatosis (without CMRFs) may still influence fibrosis progression [2]. Re-analyzing fibrosis prevalence across steatosis subgroups (e.g., 0%, 1–5%, >5% without CMRFs) would clarify whether metabolic dysfunction, not steatosis alone, drives risk.
Second, multivariate analysis of significant fibrosis in CHB+MASLD patients omitted adjustment for steatosis severity (Table 3). While MASLD diagnosis requires ≥5% steatosis, higher grades (e.g., >33%) correlate with accelerated fibrosis in viral hepatitis [3]. The reported odds ratio for CMRF ≥3 (aOR = 3.234) might partially reflect unmeasured steatosis effects. Incorporating steatosis grade into the model would isolate the independent contribution of metabolic dysregulation.
Third, non-invasive tools (e.g., FIB-4) showed suboptimal performance for significant fibrosis (AUC = 0.679). The authors did not explore combining FIB-4 with liver stiffness measurement (LSM), which was strongly associated with fibrosis (aOR = 1.259). Transient elastography-based algorithms (e.g., Agile-3+) improve accuracy in metabolic liver disease [4]. Validating such models in this cohort could enhance clinical applicability.
To address these gaps, we suggest stratifying the CHB group by steatosis severity to quantify metabolic-independent effects, re-running multivariate models including steatosis grade as a covariate, and testing combined LSM/FIB-4 scores for fibrosis prediction.
These steps would refine risk stratification and align with emerging MASLD subphenotyping frameworks [5]. We commend the authors’ focus on early intervention and welcome their perspectives on these points.
Conflicts of InterestNone.
FundingThis study was supported by the National Natural Science Foundation of China (82274260).
