A review argues for combining viral markers to predict sustained surface-antigen loss — while conceding that a single cheap test remains the right tool in primary care.

Chronic hepatitis B is treatable and not curable. The virus parks a circular form of its genome, called cccDNA, inside the nuclei of liver cells, where current drugs cannot reach it — so treatment suppresses replication indefinitely rather than clearing the infection, and stopping usually means relapse.

What the field aims at instead is a functional cure: sustained loss of hepatitis B surface antigen, or HBsAg, from the blood. The virus is still present and the immune system is holding it down well enough that the antigen disappears, which is associated with substantially better long-term outcomes. Few patients get there.

A review in Portal Hypertension & Cirrhosis surveys the markers used to predict who will, and proposes a framework for using them together.

The research team was led by Professor Qishui Ou of the Department of Laboratory Medicine at the National Regional Medical Center, Binhai Campus of the First Affiliated Hospital, Fujian Medical University, with Professor Zhen Xun of the same department.

Why prediction is the problem

“Functional cure remains the most realistic treatment goal for chronic hepatitis B because complete elimination of the virus is not yet possible,” Ou said.

The clinical difficulty is that the two main treatments work differently and suit different patients. Nucleos(t)ide analogues suppress replication reliably, are well tolerated, and rarely produce HBsAg loss; pegylated interferon-alpha can produce it, works in a minority, and carries considerable side effects. Knowing in advance who is likely to respond is what makes offering interferon reasonable rather than a gamble.

On the markers themselves, the review’s assessment is that quantitative HBsAg is the most useful single predictor — baseline levels and early declines during interferon therapy both predict sustained clearance.

That makes mechanistic sense: the quantity of surface antigen in blood reflects how much virus-producing material the liver still holds, and a level that falls early under treatment indicates the immune response is gaining ground.

The argument for panels, and the argument against

The review’s central claim is that combined multi-marker models generally discriminate better than single markers in large cohorts.

“Our review demonstrates that integrating multiple viral markers offers a more accurate and practical approach for predicting which patients are most likely to achieve sustained HBsAg loss,” Ou said, “allowing clinicians to make more informed treatment decisions.”

The additional markers named are HBV DNA, hepatitis B core-related antigen, HBV RNA and quantitative anti-HBc, and each measures something the others do not. HBV DNA tracks circulating virus; HBcrAg and HBV RNA are closer proxies for transcriptional activity of the cccDNA that treatment cannot reach; quantitative anti-HBc reflects the host’s immune engagement rather than the virus. A panel spanning viral load, viral activity and host response has more independent information in it than any one of those alone.

What is notable is that the review does not therefore recommend panels everywhere. Its own assessment is that single quantitative HBsAg remains the most cost-effective screening tool in primary care with limited testing — which is a more useful conclusion than a blanket endorsement, and the kind of distinction reviews often fail to draw. Discriminating better is not the same as being worth the cost in every setting.

The framework proposed covers four decision points: pretreatment stratification, treatment monitoring, timing of nucleos(t)ide analogue discontinuation, and post-treatment consolidation. The third of those is the hardest question in practice, since stopping therapy carries a real risk of flare, and deciding when a patient is safe to stop is precisely where a predictive marker would earn its place.

The review also says host factors — immune status, sex, age and prior treatment — should be weighed alongside the viral markers, and points to potential in artificial intelligence, high-sensitivity assays and multidimensional models.

“As standardized assays and multimarker prediction models become more widely available, clinicians will be better equipped to personalize treatment strategies and identify patients most likely to achieve functional cure,” Xun said. “These advances have the potential to increase functional cure rates while reducing unnecessary treatment costs and supporting global hepatitis B elimination efforts.”

What the release does not contain

The most significant absence is numerical. The release reports no study results of any kind — no predictive accuracies, no areas under the curve, no sample sizes, no effect sizes, and no figure for how much better a panel performs than quantitative HBsAg alone.

For a review whose entire claim is comparative, that gap matters. “Generally discriminate better” could describe an improvement that changes clinical decisions or one too small to justify the additional assays, and a reader cannot tell which from the release. The distinction between those two is the whole question of whether to adopt a panel.

There is also the general limitation of the format. A narrative review selects and synthesises published studies according to the authors’ judgement rather than a pre-specified protocol, which makes it useful for framing a field and weaker than a systematic review as evidence about effect sizes. The framework offered is an expert proposal, not a validated algorithm, and the release does not report that it has been tested prospectively in any cohort.

Nothing here bears on any individual’s treatment. Decisions about starting, continuing or stopping hepatitis B therapy are the province of a hepatologist with the patient’s own marker levels and history in front of them.

One administrative discrepancy: the release says the review was published on 27 June 2026 while its metadata gives 27 July 2026. Funding came from the National Natural Science Foundation of China, a Fujian Provincial Health Technology Project, and an Innovation of Science and Technology programme in Fujian Province. The authors declare no financial or non-financial conflicts of interest related to the review.

---

Notes

1. “Advances in Viral Markers for Predicting Functional Cure of Chronic Hepatitis B.” Portal Hypertension & Cirrhosis, 2026. DOI: 10.1002/poh2.70063.
2. Chinese Medical Journals Publishing House Co., Ltd. “New viral marker strategies improve prediction of functional cure in chronic hepatitis B.” EurekAlert!, 10 August 2026. https://www.eurekalert.org/news-releases/1139446.

Photo by Pranidchakan Boonrom on Pexels.com