The Impact of Currently Licensed Therapies on Viral and Immune Responses in Chronic Hepatitis B

Considerations for Future Novel Therapeutics

Upkar S. Gill; Patrick T. F. Kennedy


J Viral Hepat. 2019;26(1):4-15. 

In This Article

Viral and Immune Aspects of Therapy

The drug development pipeline in HBV is rapidly advancing, and thus, we are on the cusp of major change in the treatment of CHB. It is likely that many of these strategies may require combination therapy with NAs and/or Peg-IFNα, and therefore, current therapies may constitute a central component of any future treatment regimen.[43] In this regard, the optimization of currently licensed therapies still remains important. Both NAs and Peg-IFNα have shown some ability to restore immune function in CHB. A number of studies, albeit limited, have investigated the role of current therapies in viral and immune responses, to determine whether these can be harnessed to deliver better treatment outcomes.


Pegylated interferon-α can offer sustained immune control in a proportion of CHB patients and lead to HBsAg loss and seroconversion at higher rates than that seen with NAs. IFN is an innate immune cytokine; it induces ISG's encoding antiviral proteins and activates immune cells. A recent study from the woodchuck model showed the induction of a T/NK cell signature in the liver correlating with treatment outcome, highlighting that it may have a more predominant role in immune modulation, rather than an antiviral mechanism.[44] Although the decline in HBV DNA may be slow with Peg-IFNα, circulating virus decreases as does HBsAg in a cohort of patients. These markers, however, may not be ideal surrogates for the viral kinetics of the intrahepatic compartment. More recently, hepatitis B core-related antigen (HBcrAg), which can be measured in the blood, has been proposed as a more accurate surrogate of the intrahepatic milieu (cccDNA and intrahepatic viral replication) than HBsAg. HBcrAg was recently demonstrated to reflect cccDNA in HBeAg-negative disease and thus may be a better determinant of the viral dynamics in patients treated with Peg-IFNα. In line with this, hepatitis B core antibody levels have also been shown to correlate with HBV DNA and HBsAg seroclearance.[45] In addition, HBV RNA can also be measured in the serum and levels are thought to reflect intrahepatic cccDNA.[46] Validation of these markers along with their correlation with immune responses in patients treated with Peg-IFNα, or other novel immune modulators, requires further study.

Interferon is known to activate the innate immune response. Micco et al[26] demonstrated with a 48-week course of Peg-IFNα that there was potent expansion of activated (HLA-DR+, Ki67+ and TRAIL+) CD56bright NK cells, and recovery of their antiviral potential, (IFNγ production), in HBeAg-negative disease, findings subsequently confirmed in a HBeAg-positive cohort.[27] Peg-IFNα, following administration, induces a rapid upregulation of the IFN signalling pathway, as marked by increases in serum cytokines, IL-16, IL-6 and CXCL10. Therapy with Peg-IFNα alone, however, does not result in a rapid decline of viral load, thus highlighting its predominant immune modulatory action.[47] In a recent study, patients treated with IFN; NKp30+ NK cells were found to be associated with HBV control, with IL-15 contributing to the upregulation of functional antiviral NK cells. Interestingly, in nonresponders to IFN, NKp30+ NK cells were found to be dysfunctional with an expansion of the inhibitory receptor NKG2A.[48] In the same cohort of patients, an expansion of CD3brightCD56+ T cells (innate-like T cells) expressing high levels of NKG2A and low CD8 was associated with nonresponse to IFN. These nonresponders had increased levels of TIM3+ CD3brightCD56+ T cells, which negatively correlated with IFNγ production contributing to the dysfunction of these cells and potentially contributing to poor responses to IFN.[49] Further evaluation of innate-like T cell populations, such as MAIT cells, during IFN therapy is required to determine whether their function can be recovered by reduced expression of inhibitory molecules PD-1 and CTLA-4, which contribute to the dysfunctional immune response in CHB.[50] KIR genotyping has also been studied where the combination of genes encoding KIR3DS1 and HLA-B Bw4-80Ile synergistically predicted sustained responders to Peg-IFNα.[51] Toll-like receptors (TLRs) have been studied in Peg-IFNα treated patients where a favourable response is associated with elevated levels of TLR2 and TLR2 associated IL-6 production at baseline, indicating that an inflammatory phenotype is more likely to be associated with a favourable treatment response.[52] Such markers and genotyping might facilitate patient selection for treatment with Peg-IFNα, which would be relatively simple to undertake.

Although only proven in small studies, IFNα therapy may have a more negative effect on HBV-specific adaptive immunity. Despite the fact that the cytokine can increase T cell survival, boost viral antigen presentation and trigger IL-12 production, which might directly rescue the function of exhausted T cells,[53] HBV-specific T cell responses in treated patients are inhibited by Peg-IFNα therapy.[26,54] A recovery of HBV-specific T cell function is only observed after Peg-IFNα therapy cessation in treatment responders,[55] and these patients also exhibit increased HBV-specific T-helper cell proliferation.[56] Of note, the expression of inhibitory checkpoint markers (PD-1, Lag-3 and CTLA-4) does not change on T cells during Peg-IFNα therapy.[26,47] Peg-IFNα leads to an expansion of IL-10 producing T-regs in nonresponders, which may contribute to HBV persistence. In addition, γδ T cells have been shown to decrease during Peg-IFNα treatment, although the effector phenotype (CD27-CD45RA+) and production of cytokines from these cells are enhanced in patients with sustained response, but not in nonresponders.[57] The negative effects of IFNα on virus-specific T cell responses have been detected in studies of LCMV infected mice, where T cell responses were restored by inhibiting the effects of virus-induced IFNα rather than by treating the infection with it.[58,59] As NK cells can negatively regulate HBV-specific T cells, the mechanisms of HBV-specific T cell inhibition during IFNα therapy might be mediated by NK cells.[60] However, this is controversial since recent studies in animal models demonstrated that type I IFN may protect T cells from NK cell-mediated attack[61,62] and thus merits further investigation. As Peg-IFNα and NAs act differentially on the immune response, the rationale for re-evaluating combination or sequential treatment is required for future therapeutic approaches, which are discussed further here.

Nucleos(t)ide Analogues

The 3rd generation NAs have excellent rates of viral suppression, but have little impact on HBsAg levels. Detailed evaluation of the intrahepatic viral repertoire during NA therapy is limited. Studies have shown that serum HBV DNA correlates with intrahepatic HBV DNA mainly in treatment naïve patients, but this is less clear in patients undergoing NA therapy. For example, after many years of therapy, serum HBV DNA may be undetectable, but the intracellular/intrahepatic viral HBV DNA only appears to reduce by 1–2 logs.[63] Thus, a lack of serum viraemia does not reflect the intrahepatic viral DNA. The decline in HBsAg levels during NA therapy is slow, and any correlation between HBsAg and cccDNA is unlikely to be statistically significant as the regulation of HBsAg expression is complex. It is now recognized that factors other than the quantity of cccDNA in infected hepatocytes, its transcriptional regulation, and the possibility that envelope proteins could be expressed from viral sequences integrated into the host genome also contribute to HBsAg levels.[64] As HBcrAg may correlate with serum HBV DNA and intrahepatic cccDNA,[65] as shown in patients treated with NA with a decline in intrahepatic cccDNA, it is possible that HBcrAg may better reflect the intrahepatic compartment than HBsAg.[66]

The rapid decline of HBV DNA secondary to NAs, and the inhibition of HBV DNA polymerase function, may allow for the restoration of IFN signalling; however, the data to support this are lacking. The increment of ISGs has only been demonstrated in the peripheral compartment of HBeAg-positive patients treated with TDF.[47] The effect of NAs on T cell responses has shown a recovery of CD4 and CD8 T cell function.[67–71] During the initial phases of NA treatment, CD4 T cell responses are stronger than those from CD8 T cells, and the presence of new and expanded clonotypes inversely associate with the decline of viral antigen, demonstrating that a broad T cell expansion is critical in HBeAg control.[72] It is important to note that following NA administration, T cell recovery is often partial and not uniform in all treated subjects.[69] The functional recovery of antiviral immunity is likely to be dependent on the ability of NAs to reduce liver inflammation, marked by a reduction in serum transaminase levels. These events are linked with the reduction in a number of immunological suppressive factors (eg, IL-10, arginase and T-reg frequencies),[73] impacting T cell recovery. Importantly, studies have demonstrated a restoration of the balance of Th17/T-regs with reductions in IL-10 and TGF-β upon viral suppression.[74,75]

In addition to a robust HBV-specific T cell response, critical in eliminating HBV infected hepatocytes, NK cells have also shown importance in HBV pathogenesis. A number of studies describe NK cell dysfunction in CHB patients, compared with healthy controls.[27,76,77] NA monotherapy does not appear to restore antiviral NK cell function, and further adjunct therapy is likely required for this innate boosting.[27,76] Such findings have also been confirmed in the intrahepatic compartment, where viral suppression resulted in limited changes in NK cell function.[78] However, during early NA treatment, in patients with raised transaminases, LdT demonstrated an expansion of CD56bright NK cells via upregulation of IL-15 and NKG2D, which may be important in viral control;[79] however, larger studies are required to evaluate NK cell phenotype and function along with KIR genotyping. NAs have been shown to improve T cell function, although the data regarding nonconventional T cells are sparse. One study reported on the presence of activated MAIT cells in CHB, with functional recovery upon viral suppression.[80] With regard to antigen-specific adaptive immune responses, Boni et al reported on the recovery of HBV-specific T cells, following expansion, in patients on long-term NA therapy, revealing that T cell function may be improved with HBV DNA suppression. This recovery is more pronounced in those who clear HBsAg during NA therapy. Regardless, these data reveal that T cell function may be improved with reduction in viral load.[69] This reduction in viraemia by NAs provides an ideal window for reconstitution of the antigen-specific immune response, which may be important in future therapeutic strategies for HBV.

Antigen-specific T cell recovery, albeit partial, is possible with NAs, but the impact on antiviral NK function is still inferior.[76] In HBV, the role of NK cells has generated some controversy, described as a "double edged sword," whether they provide a pathogenic or protective role continues to be debated.[60] A robust antiviral NK cell response is important for cytolysis of HBV infected hepatocytes;[81] however, NK cells also have a regulatory role, causing deletion of HBV-specific T cells when in close contact.[82] The interaction of TRAIL+ and NKG2D+ NK cells with T cells expressing the receptor, TRAIL-R2 and/or NKG2D ligands leads to T cell apoptosis, which, in vitro, can be partially prevented by blockade of these pathways.[82,83] The data on whether the phenotype of NK cells is altered, with viral load reduction, are limited. A recent report demonstrated an inverse correlation with an "activatory" NK cell phenotype (HLA-DR+, CD38+, Ki67+, TRAIL+, NKG2D+) and the proportion of HBV-specific T cells in patients undergoing NA therapy.[84] The regulatory role of NK cells and the interaction with T cells may of course be a protective homoeostatic mechanism for the liver microenvironment, where NK cells govern T cell-mediated immune pathology. However, in an attempt to curb liver damage by downregulating bystander T cells, HBV-specific T cells are also dampened. These concepts, however, require further elucidation, ideally with focused "on-treatment" studies of the liver compartment and to determine the impact of therapy on tissue-resident immunity.[85–88]

Combination/Sequential/add-on Therapies

With their differential action on the immune response, NAs and Peg-IFNα used in combination, "add-on" or sequentially may generate additive or synergistic effects and could be important in future therapeutic strategies. Recent clinical studies have shown the combination or addition of Peg-IFNα to NAs results in greater HBeAg seroconversion rates in HBeAg-positive patients along with greater declines and loss of HBsAg.89–91 Thus, their use in the clinical setting requires further consideration, while the outcomes of such studies performed to date have been extensively reviewed elsewhere.[92]

Addition of Peg-IFNα, in a cohort of patients virally suppressed on NAs, induced the activation of DC's, expansion of CD56bright NK cells and increased the frequency of Th1/Th17-orientated HBV-specific T cells.[93] Notably, these effects were not associated with improved clinical outcomes. Viral load reduction is able to maintain the immune stimulatory effects of Peg-IFNα, when administered in combination or sequence, compared to Peg-IFNα alone, which implies there may be beneficial outcomes with add-on or combination therapies.[47] This has been further demonstrated where Peg-IFNα add-on was employed in patients virally suppressed with ETV, resulting in a reduction in T-reg frequencies with an increase in NKG2C+ NK cells and TLR-2+ CD14 monocytes, which was associated with treatment response.[94] In the same cohort, the expansion of CD56bright NK cells expressing activatory receptors NKp30 and NKp46 along with TRAIL and IFNγ correlated with HBsAg decline with potential cccDNA clearance through TRAIL-induced cytolysis, demonstrating the importance of Peg-IFNα for immune modulation and HBV clearance.[95] Similarly, in patients primed with Peg-IFNα prior to viral suppression, the maintenance of expanded functional CD56bright NK cells has been shown to correlate with treatment response.[27] It is noteworthy that the recovery of nonconventional T cells (iNKT and γδ T) was limited despite significant declines in HBsAg in a cohort of patients undergoing combination therapy.[96]

In a study of combination Peg-IFNα with ADV, those patients achieving HBsAg loss demonstrated increased frequency of TRAIL+, IFNγ+ NK cells at the end of treatment. This indicates that NK cells may play a role in the clearance of HBsAg with this therapeutic approach.[97] In the same patients with HBsAg loss, T cells with broad antiviral capacity could be expanded.[98] Interestingly, in this study, baseline levels of HBsAg/anti-HBs immune complexes were higher in patients that went on to lose HBsAg, which may also be a factor in selecting patients for such combination therapy approaches.[99] Peg-IFNα add-on therapy has also been shown to increase TNFα-monofunctional HBV-S and core-specific CD4 T cell numbers, which may contribute to viral control.[100] Along with the direct analysis of immune cells, microRNAs (miRNA) have been reported to be regulated in liver disease. miRNA-155 is a key regulator of innate and adaptive immunity, and the higher expression of miRNA-155 at baseline was associated with improved treatment response and NK cell function.[101] Further studies of miRNA's and their role in HBV and the immune response are warranted and important in future therapeutic strategies. The type-III interferon Peg-IFNλ in combination with NA therapy has also been shown to induce robust innate and adaptive immune responses, where NK cell polyfunctionality along with recovery of HBV-specific T cells was observed in patients with enhanced HBV DNA and HBsAg decline.[102] This highlights the importance of immune modulation with viral suppression as key elements for HBV therapy.

The schedule of therapies remains crucial, and this may need revisiting to optimize patients for future trials, especially if the treatment pool is widened. The concept of combination or IFN based therapies still needs further investigation in larger studies with parallel analysis of both the innate and adaptive immune response. Adjusting the sequence in which therapies are combined is potentially important in altering treatment outcome, but may also prime patients for future clinical trials, which is a key avenue of exploration.