HBV treatment: how is the landscape changing?

Global health sector strategies aiming to eliminate viral hepatitis infections as a public health threat by 2030 have established ambitious goals for HBV, which include enhanced diagnosis and reduced mortality.4 However, high attrition rates at each step of the HBV care cascade are impeding progress, driven by intersecting societal, health system and individual factors. Recent estimates suggest that among people living with HBV (PLWHB), only 10%–13% are aware of their infection status and <10% of those eligible for NA treatment are receiving it.1 2 But how well is treatment ultimately working in the group who access care, and are offered NA agents?

The aim of long-term NA therapy is to suppress HBV DNA viral load (VL), reducing the risk of complications including cirrhosis, hepatocellular carcinoma and death. To date, HBV guidelines have only recommended NA therapy for selected individuals deemed at the highest risk of adverse outcomes based on HBV VL, liver enzymes and fibrosis stage (eg, guidelines published by the European Association for the Study of the Liver (EASL), American Association for the Study of Liver Diseases (AASLD), and Asian Pacific Association for the Study of the Liver (APASL)).5–7 However, the landscape is changing, with the WHO releasing new recommendations in 2024 that widen treatment eligibility,8 alongside other guidelines that likewise relax treatment criteria,9 such that treatment will be offered to a wider population.10 As these changes take effect, we argue that it becomes increasingly important to focus on understanding and improving treatment outcomes, particularly for groups who are under-represented by current evidence and in whom risk factors for poor treatment outcomes are amplified.

HBV viraemic suppression is not universally achieved in individuals prescribed NA agents. Although clinical trial data suggest robust outcomes on NA therapy, these are biased estimates, typically based on observations of carefully selected populations under regular clinical scrutiny, and therefore do not represent real-world circumstances and should not be generalised across settings. A recent estimate from the UK suggests that (even in well-resourced secondary/tertiary care centres) as many as 20%–30% of PLWHB are not virologically suppressed 1 year after first being prescribed NA therapy.11 The extent and clinical or public health impact of persistent viraemia is variable, and more detailed characterisation is undoubtedly needed to determine when, why and in whom NA agents do not deliver intended outcomes so that we can identify and tackle the underlying causes.

Thus while recognising that NA therapy is a fundamental cornerstone of interventions for HBV, we must avoid complacency. Increasing roll-out of diagnosis and therapy will have huge individual and population health benefits, but will also result in an increase in the absolute number of people who have persistent HBV viraemia despite having been offered NA therapy. Evaluating factors that influence treatment outcomes is needed to develop an evidence base that supports individual clinical management, informs public health approaches and underpins the collation of relevant, representative data.

Who is at risk of suboptimal treatment outcomes?

Our case report in the current BMJ STI issue Lumley et al 12 exemplifies some of the diverse factors that operate together to influence outcomes of HBV treatment (figure 1B) and illustrates why a holistic approach is needed to evaluate and tackle the causes of incomplete virologic response. Achieving HBV VL suppression in individuals and at a population level is dependent on a complex interplay between social, economic, logistic and biological parameters, which influence access, affordability, acceptability and efficacy of long-term treatment. However, the specific causes and impact of viraemia persisting on therapy have not been well explored, and there has been inadequate representation of individuals in key populations who are most at risk of suboptimal therapeutic outcomes. For instance, these might include vulnerable migrants, people engaging in transactional sex, marginalised members of the LGBTQ+ (lesbian, gay, bisexual, transgender, queer or questioning and other gender identities) community, people who inject drugs or suffer from alcohol dependency, people experiencing homelessness, those on low incomes, and those with competing physical and mental health concerns, perhaps especially pertinent to chronic coinfection with other blood-borne viruses.3 Experiences in different global settings and cultures have impact on beliefs and education, and may influence the start of treatment or treatment adherence. Traumatic experiences and discrimination within healthcare settings can greatly impact trust in medical care; individuals may have been advised against treatment or discharged from follow-up, and engagement can be damaged by suspicions that clinicians or services offer treatment based on financial incentives. Often these factors and experiences are interdependent, leading to a complex ‘syndemic’ challenge, for which healthcare systems are poorly resourced and designed.

What factors lead to inadequate drug exposure?

For a proportion of individuals for whom HBV viraemia is not suppressed on treatment, this is explained by the pragmatic view that they do not have sufficient drug exposure to derive a therapeutic effect (either because the levels of the drug in blood or tissue are too low, or not maintained consistently over time). Exploring this space is difficult, because the definition of therapeutic NA drug levels is not established for HBV management. Laboratory quantification of drug levels may help to establish if an individual is taking and absorbing therapy, but this is typically a binary assessment (concluding whether an NA drug is present or not). Routine drug level testing is not currently affordable or accessible and is not included in global treatment guidelines; further nuanced interpretation is currently limited due to gaps in data. What is an optimal NA level in peripheral blood (either serum or intracellular)? How does this measurement in blood correlate with liver penetration and antiviral activity? What is the effect of genotype and other viral polymorphisms (including specific resistance-associated mutations (RAMs)) on response to treatment? Addressing these questions may provide evidence that can help us to improve outcomes in the final stages of the care cascade (figure 1A).

Drug levels that are insufficient to mediate consistent virologic suppression are most likely to be a result of incomplete adherence (figure 1C, domain 1). Indefinite (or lifelong) HBV treatment is a challenge, as it is commonly advocated for individuals who are asymptomatic, but who may lack education and support that is needed to motivate daily treatment.13 At the same time, social and health systems factors create barriers such as poor accessibility to specialised clinical services (especially in rural areas), out-of-pocket costs (eg, direct costs of medication, or indirect costs associated with treatment such as disruption to work or domestic commitments to attend a clinic or pharmacy, costs of travel, payment for laboratory tests), inadequate access to information, or health beliefs that lead individuals to pursue traditional medicine.13 14 Psychological factors, such as disease-related anxiety, experience or expectation of stigma and discrimination, and absence of familial or community support may also impede adherence.14 15 Other possible contributors include physical barriers to taking the drug (eg, difficulty swallowing, high pill burden) and experience or anticipation of side effects (eg, gastrointestinal disturbance, headache, rash). Key populations are highly susceptible to these challenges, as they are more likely to have fragmented access to healthcare, low advocacy and awareness, experiences of stigma, limited financial resources, and competing health and social priorities.3 This puts them at risk of persistent HBV viraemia, adverse clinical outcomes and an increased risk of drug resistance.16

Inadequate drug levels can also reflect pharmacokinetic issues (figure 1C, domain 2). Severe malabsorption syndromes can potentially lead to reduced bioavailability of any orally ingested medication—although there is no robust evidence to suggest that this is a concern for tenofovir (TFV) based on evaluations of adults with HIV infection, even in the setting of chronic diarrhoea17 or gastrectomy.18 In contrast, there is some evidence to suggest that TFV clearance is modified by ethnicity, renal function and use of other antiretroviral drugs.19 For instance, there is an interaction between tipranavir, a protease inhibitor used to treat resistant HIV infection, and TFV, resulting in reduced TFV plasma levels (https://www.drugs.com).

Why does inadequate HBV suppression arise even in the presence of consistent exposure to NA drugs?

HBV viraemic suppression may not be achieved even when an NA agent is taken daily, absorbed well and not affected by drug–drug interactions. Considerations in this case include duration of treatment, pharmacodynamic factors, which include binding of the drug to its target (the HBV reverse transcriptase (RT)), viral escape (ie, drug resistance) and other host and viral factors that determine the immune response, metabolism and clearance of viraemia.

HBV viraemia can take several months to suppress, particularly in those who start treatment with high HBV DNA VL.20 The majority of individuals (57%–81%) achieve HBV suppression (HBV DNA<20 IU/mL or undetectable by PCR) after 1 year of TFV treatment, with the proportion of those virologically suppressed increasing over time with continued treatment (up to 88% at year 2 and 94% at year 3).11 21 On these grounds, a period of observation is reasonable to document the virologic response to treatment and to ensure that sufficient time has elapsed for suppression before any further investigations or interventions are considered (figure 1C, domain 3).

Variation in avidity (ie, the strength of drug binding to its target) is an explanation for altered drug susceptibility. Amino acid polymorphisms in HBV RT may induce structural changes, altering the conformation of the NA binding site. In general, NA agents are pangenotypic and work well to suppress viraemia irrespective of the genotype of infection (explaining why genotyping is not routinely undertaken in clinical practice). However, there is a suggestion that HBV genotypes could alter NA susceptibility due to natural lineage-related variations in the RT protein22 (figure 1C, domain 4). As this observation has not been consistently demonstrated in vitro or linked to treatment outcomes,23 24 further exploration is needed. The lack of routine clinical genotyping data contributes to a data blind spot.

The selection of specific RAMs during treatment can reduce or completely abrogate drug–target interaction (figure 1C, domain 5), either by decreasing the binding affinity of the drug to the RT and/or by enhancing the affinity of natural nucleotide substrates to the RT such that they outcompete the drug.25 The risk of selection of RAMs and subsequent virologic breakthrough may be amplified by treatment interruptions, which present a window of opportunity for the generation of new viral quasispecies, some of which carry RAMs that can become dominant when drug selection pressure is reapplied.26 Drug resistance is predictably selected by lamivudine,5 with the best described HBV resistance polymorphism, M204I/V, also conferring a degree of cross-resistance to other NA agents,27 setting the foundations for resistance to ETV, and potentially also TFV. Although both these agents (in particular, TFV) have high genetic barriers to resistance, decreased drug efficacy and virologic breakthrough has been reported.25 26 In a modest number of cases where this has been explored, clinically significant TFV resistance is dependent on the selection of multiple RAMs,12 which may also include secondary (or compensatory) changes.

Finally, other host factors such as coinfections, immunological and metabolic differences may influence HBV suppression. Host genetic polymorphisms have been associated with response to interferon therapy.28 While it is less explored whether host genetic (including immune or metabolic) factors associate with NA treatment response, there is evidence that factors modulating the immune system influence viraemic persistence. For instance, in a South African cohort, prescribed regimens including TFV and lamivudine in the context of HIV/HBV coinfection and immunosuppression (with CD4⁺ T-cell counts<200 cells/µL) was significantly associated with persistent HBV viraemia.29 This situation exemplifies an environment in which risk factors coincide, including social disadvantage, immunological dysfunction, exposure to multiple NA agents and barriers to care including stigma, collectively increasing the risk of the selection of drug resistance.

What interventions are required in cases of incomplete virologic suppression?

Strategies to improve responses to NA therapy must recognise the interplay between risk factors for adverse outcomes, taking into account the social, clinical, and molecular landscapes and recognising the influences between these domains; for example, considering how disadvantaged populations experience many barriers to care which could put them at increased risk of developing drug resistance.

Healthcare systems and providers need to be equipped and trained to identify and improve modifiable barriers to consistent treatment, including providing optimisation and support for adherence and addressing competing challenges. A particular focus on vulnerable groups is strategically crucial to mitigate the potential risk of harm due to lack of virologic suppression in people receiving NA treatment.3 Provision of education, peer support and decentralised care are likely to improve retention in the care cascade. Active case finding by specialist inclusion health teams has demonstrated successful re-engagement among some highly vulnerable populations, such as people seeking asylum and people experiencing homelessness,30 and peer support has been effective in improving engagement in care. Other innovative solutions, such as an ‘HBV passport’ that facilitates movement between clinics, may be beneficial to enhance linkage to services, with consistent access to treatment, and monitoring. Measures such as appointment reminder messages in the appropriate language, arranging interpreters for appointments and avoiding automated discharge, may also improve retention in the care cascade.

A scale-up of access to laboratory services that can support the generation and interpretation of drug resistance data is an ongoing aspiration for the HBV field, but careful integration with existing services, accounting for limited resources, is required to ensure that such activities do not distract from the other priority areas of diagnosis and treatment. Published guidelines already provide advice on switching therapy or combining drug agents, which can be implemented alongside addressing access and adherence challenges.

What are the key research questions to be addressed?

At a population level, enhanced clinical and epidemiological data are required to estimate the prevalence and impact of persistent HBV viraemia in individuals who are prescribed NA treatment. Such data are already routinely collected in many healthcare systems, but there would be significant benefits in enhanced multicentre collation (eg, through frameworks such as the Health Informatics Collaborative in the UK: https://hic.nihr.ac.uk/viral-hepatitis). Enhanced efforts and resources are needed to expand global reach and to improve representation of key populations. As the treatment landscape changes, with more people offered NA therapy and the potential for new agents to emerge in clinical practice, data collection to document outcomes is particularly crucial. Implementation research is needed to investigate, refine and document best-practice strategies for optimising retention in care through improvements in healthcare systems, with investment in data collection to document the health benefits and cost-effectiveness of interventions.

There are also evidence gaps to be filled in laboratory science. An understanding of optimal drug levels, supported by therapeutic windows for NA agents in the treatment of HBV infection, would support the use of therapeutic drug monitoring in selected situations. Experience and infrastructure to support this are already in place through laboratory provision for monitoring HIV therapy. To determine specific outcomes in HBV, drug levels in liver tissue are evidently important; TFV and ETV are widely distributed through body tissues, so quantifying drug levels in peripheral blood, urine or hair may be a surrogate for liver levels,31 but this has not been clearly established.

The specific combinations of RAMs associated with clinically significant drug resistance and associations between viral factors (eg, genotype) and host factors (eg, genetics, immunosuppression) and drug absorption, avidity, clearance and resistance are yet to be elucidated. An improved understanding of RAMs and cross-resistance patterns will be crucial to the development of evidence-based guidelines for managing individuals with inadequate viraemic suppression.