Long-term PAGE-B data sharpen the discussion around HCC risk after HBsAg clearance
The exchange around the long-term PAGE-B cohort puts two issues at the center of chronic hepatitis B care: how to interpret hepatocellular carcinoma risk over time, and how to think about HBsAg clearance in patients maintained on potent nucleos(t)ide analogue therapy. The manuscript’s most consequential contribution is its long follow-up window. It described the first large cohort to assess outcomes in chronic hepatitis B patients treated with entecavir or tenofovir disoproxil fumarate for more than 10 years.
That duration matters because hepatitis B risk does not unfold on a short clinical clock. Viral suppression changes the pace of liver injury, but it does not erase the biology that drives carcinogenesis in every patient. Fibrosis, prior inflammatory burden, and the residual architecture of chronic liver disease remain relevant long after serum HBV DNA has been controlled. A cohort that extends beyond a decade offers a more credible view of what long-term management can and cannot achieve.
The response to Dr Xiao and colleagues also makes clear that methodological scrutiny is part of the value of this dataset. Specific data points and risk stratification were raised as issues that may need clarification, and those questions are exactly the sort that emerge when a cohort is used to inform bedside risk assessment. In chronic hepatitis B, the temptation is always to turn one risk score into a universal rule. PAGE-B was never meant to do that. It is a stratification tool, and its performance should be interpreted in the same clinical context in which it was developed and validated: patients receiving long-term antiviral therapy whose residual cancer risk remains clinically important despite treatment.
A single-sentence reality anchors the whole discussion: antiviral suppression lowers risk, but it does not abolish it.
Long-term follow-up is especially relevant when evaluating HCC incidence after functional viral control. Hepatocellular carcinoma can arise from pathways that persist after treatment begins, including cumulative inflammation, regenerative pressure, and the effects of established liver damage before suppression was achieved. In practice, this is why HCC surveillance does not disappear when HBV treatment is working. The clinical question shifts from whether cancer can still occur to how to identify which patients remain at meaningful risk over time.
The clarification exchange also reflects a broader issue in hepatology research: risk stratification becomes more complex as treatment duration lengthens. Early in therapy, baseline variables may dominate. Over years of follow-up, the story can change as some patients clear surface antigen, some remain stably suppressed, and some accumulate enough background liver disease to remain vulnerable despite virologic control. Any long-term cohort that attempts to map those transitions is inevitably tested by the assumptions built into its analytic framework. That is not a flaw unique to this study; it is a central challenge in every long-horizon hepatitis B dataset.
Long follow-up is what allows a cohort to separate transient treatment effects from durable clinical behavior. In chronic hepatitis B, that distinction is essential. Potent agents such as entecavir and tenofovir disoproxil fumarate suppress replication effectively, but the major outcomes that matter to patients and clinicians are delayed: cirrhosis progression, decompensation, HCC, and the rare but important state of HBsAg loss. A cohort that runs past 10 years is one of the few ways to observe those endpoints without compressing them into a short and potentially misleading snapshot.
The manuscript’s emphasis on being the first large cohort of this duration gives the work a practical importance that goes beyond the debate over individual data points. It provides a longer horizon for clinicians who must decide how aggressively to follow patients who appear well controlled on therapy but still carry residual oncologic risk. It also reminds readers that HBsAg clearance, while biologically meaningful, should not automatically be treated as the end of cancer surveillance in every patient. The long-term behavior of the liver remains the decisive issue.
A second single-sentence truth follows from that logic: HBsAg loss is important, but it is not a complete substitute for risk assessment.
The discussion around risk stratification therefore has real bedside implications. If a patient with chronic hepatitis B has been treated with ETV or TDF for years, the clinician still needs to know how to interpret long-term prognosis, whether a risk score remains informative, and whether biomarker changes alter surveillance intensity. These questions do not have abstract relevance. They shape hepatology clinic decisions about imaging follow-up, counseling, and how reassuring a favorable laboratory pattern should be when the history of chronic viral liver injury remains in the background.
The response also signals a familiar truth in liver disease research: the value of a cohort is not only in the answers it gives, but in the precision of the questions it forces investigators to confront. Long-term outcome studies often draw scrutiny because the longer the follow-up, the more exposed the analysis becomes to changes in baseline risk, treatment era, and the interpretation of biomarker evolution. In that sense, methodological debate is not a distraction from the science. It is part of how the field learns to use the data responsibly.
For clinicians managing chronic hepatitis B, the main takeaway from this exchange is straightforward. Long-term suppression with entecavir or tenofovir disoproxil fumarate remains foundational, but the residual risk of hepatocellular carcinoma still has to be tracked over years, not months. The first large cohort extending beyond a decade gives the field a more durable platform for that conversation, even as the interpretation of specific data points and risk categories continues to be refined.