Controlled Therapy Interruption Improves Host Immune Control of Chronic Hepatitis B Virus Infection in HBeAg-Positive Patients

Abstract

Prolonged viral suppression with oral antiviral drugs allows partial immune reconstitution. Controlled therapy interruption (CTI), by leveraging secondary immune response, proposes further augmentation in chronic hepatitis B virus (HBV) infection. Transient treatment interruptions (TIs) at months 0, 1, and 3 during otherwise continuous oral antiviral therapy allow viremic bursts, simulating autovaccination. Four weekly injections of Hepatitis B Immunoglobulin are given before the second and third TI to simulate prime boosting, which specifically amplifies the immune response. Fourteen patients (10 males; four controls, four HBeAg positive, and six anti-HBe positive) aged 28–46 years were studied. The period between TI and reappearance of viremia, time to relapse (TTR) (weeks) estimated immune control. The other endpoints included reduction in serum HBsAg IU/mL and loss of HBeAg. TTR after the first TI was significantly shorter in HBeAg-positive patients, indicating low baseline immunity. TTR increased significantly after the second and subsequent TI in all four HBeAg-positive patients. One patient persistently lost HBeAg. Mean HBsAg levels fell significantly in three of four patients after the second TI. In contrast, in the anti-HBe-positive group, TTR was unchanged after all three TI. Furthermore, no significant changes in HBsAg levels were detected after the second or subsequent TIs. No significant differences in adverse events were noted between groups. HBeAg-positive patients have low baseline levels of host immune control against HBV. CTI consistently boosts this immunity. CTI did not influence immunity in anti-HBe-positive patients.

Introduction

Chronic persistent viral infections are characterized by high rates of replication that induce tolerance through relentless stimulation of host immune cells. T cells chronically stimulated by high antigen load become exhausted, leading eventually to immune paralysis (30,31). This induced immune paralysis can be partially reversed by prolonged suppression of viral replication (2). Approximately, 50% of HBeAg-positive subjects seroconvert to anti-HBe within 5 years of continuous viral suppression with potent oral antiviral therapy (2,13). Augmentation of the partial immune reconstitution could lead to long-term protective immunity.

One possible approach to providing the needed boost is to leverage the vaccine theory to provoke a powerful secondary host immune response (6,22). However, conventional vaccines given alone or concomitantly with antiviral therapy in chronic hepatitis B virus (HBV) infection have not proved effective in providing long-term protective immunity (25). The authentic autologous virus could prove a more effective immunogen.

The controlled therapy interruption (CTI) protocol consists of continuous drug-induced aviremia, with brief intervals of treatment interruption (TI), which allows for short bursts of low-level virion release. This briefly exposes the host to autologous viral proteins and DNA, simulating autovaccination. The first therapy interruption represents the priming vaccine dose. Hepatitis B virus immunoglobulin (HBIg) is given before the second and subsequent TI cycles, which simulate booster doses. This addition of HBIg confers several potential advantages.

HBV-HBIg complexes improve the antigen presentation cell response. A prototype therapeutic vaccine comprising HBsAg-HBIg immune complexes has been studied extensively. Some, although not all, suggested benefit (28,33 –35).

Drug withdrawal during treatment of chronic HBV liver disease has been associated with hepatitis flares, occasionally leading to hepatic decompensation and even death (12). The flares are most likely due to de novo infection of uninfected cells. HBIg has been used successfully to prevent de novo infection of donor draft during liver transplantation in patients with chronic HBV infection (4,17).

Altering the context of antigen presentation between the priming (first antigen dose) and booster (subsequent but altered antigen complex) dose is an example of prime–boost strategy (32). Prime boosting significantly increases development of effector memory (Tem) T cell (1) and induces a more effective long-term protective immunity. In the current scenario, unbound virus released in cycle #1 acts as priming and virus complexed with HBIg as a booster in subsequent cycles.

Autovaccination or structured therapy interruption has been tried using highly active antiretroviral therapy (HAART) in patients with human immunodeficiency virus (HIV) disease. The results have been mixed in primary HIV infection (20,27) and largely disappointing results in chronic infection (11,14,15). The risk of development of drug resistance has not been significantly different from continuous treatment (21). In contrast to chronic HBV infection, chronic HIV infection is associated with irreversible depletion of CD4⁺ cells that may not allow for immune reconstitution after autovaccination. Thus, we hypothesize that the effect of autovaccination in chronic HBV infection is more likely to mimic primary HIV infection.

This article describes preliminary results in two groups of chronic HBV-infected patients: HBeAg positive and anti-HBe positive, subjected to the CTI protocol. A separate control group of chronic HBV-infected subjects, not subjected to the CTI protocol, was studied concomitantly.

Materials and Methods

Patients

A total of 14 patients (controls n = 4; HBeAg positive n = 4; anti-HBe positive n = 6) aged 28–46 years were included in the study. The group demographics are summarized in Tables 1, 2, and 3. The majority of patients were males (n = 10). Ten were of Asian descent, three were Caucasians, and one African American.

Table 1.

Data on the Group Demographics: Control

ID	Age/sex	E or E+T	eAg/eAb	Ethnicity	Genotype
LG 102	32/M	E	Anti-HBe	Caucasian	NK
YFC 108	32/F	E+T	HBeAg	Asian	NK
XRL 113	35/M	E+T	HBeAg	Asian	C
ETC 114	33/M	E+T	Anti-HBe	Asian	NK

Table 2.

Data on the Group Demographics: HBeAg-Positive Group

ID	Age/sex	E or E+T	eAg/eAb	Ethnicity	Genotype
TB 103	28/F	E	HBeAg	Caucasian	NK
ER 109	38/M	E+T	HBeAg	Asian	C
YL 111	46/F	E+T	HBeAg	Asian	B
CL 112	31/M	E+T	HBeAg	Asian	B

Table 3.

Comprises Data on the Group Demographics: Anti-HBe-Positive Group

ID	Age/sex	E or E+T	eAg/eAb	Ethnicity	Genotype
TL 101	44/M	E	Anti-HBe	Asian	NK
LJL 104	34/M	E+T	Anti-HBe	Asian	NK
MJ 105	37/F	E	Anti-HBe	Afro-American	E
YGL 106	39/M	E+T	Anti-HBe	Asian	B
ZXC 107	30/M	E	Anti-HBe	Asian	B
LG 115	32/M	E	Anti-HBe	Caucasian	NK

For inclusion into the study, HBV viral load was required to be undetectable. Patients were initially all started on entecavir alone for HBV DNA suppression. In eight patients, HBV DNA was still above the limit of detection after more than 3 months of continuous therapy. Tenofovir was added and led to complete virus suppression in all eight. Thus, only three anti-HBe-positive patients and one control patient received entecavir alone, without tenofovir.

Control group patients received standard of care only. This consisted of complete viral suppression with entecavir ± tenofovir only. They were not subject to the CTI protocol.

Methods

Complete blood count (CBC) and differential urea and electrolytes, as well as liver function tests, were all measured through the routine hospital laboratory. HBV serological markers as well as HBV DNA levels in serum were also measured through the routine laboratory.

Samples for HBsAg and HBV DNA estimations were collected simultaneously at timed periods during the course of the study.

HBsAg assays were performed in sera stored at −80°C using an assay kit, ADVIA Centaur HBsAg Kit (Siemens Medical Solutions Diagnostics, Tarrytown, NY) according to the manufacturer's instructions. Samples from each individual patient were assayed in the same batch to exclude interassay variation. The HBsAg concentration in individual samples collected during each cycle was averaged out, and the mean values for each of the cycles #1, #2, and #3 were recorded. Thus, the mean HBsAg concentrations could be compared between cycles.

HBIg (Nabi-HB, Biotest Pharmaceuticals Corporation) was kindly supplied by the Biotest Pharmaceuticals Corporation (Boca Raton, FL). Each study patient received 300 U of HBIg intramuscularly weekly, for four consecutive weeks, immediately before the second and subsequent TI.

Description of study protocol—CTI

Patients were followed for at least 3 months after inclusion, with monthly measurement of HBV DNA to ensure continued aviremia. The first interruption CTI, cycle #1, was at month 0.

Thereafter, serum HBV DNA concentrations were measured at weekly intervals for 1 month, then biweekly for 1 month, and then monthly. The time (weeks) between therapy interruption and the first measurement of detectable virus in plasma was estimated and defined as time to relapse (TTR). In this study, we have arbitrarily assigned TTR as a measure of immune control, since it intuitively reflects the host's ability to control de novo cell reinfection. The therapy was restarted as soon as virus became detectable in the serum. The TTR at the end of cycle #1 is a measure of “baseline” host immune control, TTR1.

The antiviral therapy was continued and serum HBV DNA levels monitored at monthly intervals until aviremia. One month after this treatment-induced aviremia, four weekly injections of HBIg (300 U) were administered by intramuscular injection.

The second TI started immediately after the fourth injection, marking the start of cycle #2. The serum was monitored for HCV RNA at timed intervals, exactly as with cycle #1. The time (weeks) between therapy interruption and the first measurement of detectable virus in plasma, in cycle #2, was determined, TTR2.

Treatment was restarted with the reappearance of viremia and continued until drug-induced aviremia was again observed. However, this time aviremia was maintained for at least 3 months before the HBIg injections were started. After the fourth injection, therapy was interrupted for the third time.

The time (weeks) between TI and the first measurement of detectable virus in plasma was estimated, TTR3. When the virus reappeared, the therapy was restarted.

Results

No significant changes were observed in any of the CBC indices during the study.

CTI was unassociated with any adverse changes in the CBC indices.

Tests of renal function also appeared unaffected during the period of the study.

Minor instances of transaminase elevation (<10% from baseline) were noted on occasions following therapy interruption, but no hepatitis flares or significant increases in serum bilirubin were noted.

Baseline host immune control

The TTR is defined as the time (weeks) between therapy interruption and the earliest reappearance of detectably quantifiable virus in the serum. It is a measure of host ability to control virus appearance in serum. The average HBsAg concentration in the samples taken during each of the three cycles was the other surrogate estimate of host immune control.

The mean TTR at baseline or cycle #1 (Tables 4 and 5) was significantly shorter in the HBeAg patients, 1 ± 0.43 weeks than in the anti-HBe group, 8 ± 4.6 weeks (p = 0.01).

Table 4.

Data on Time to Relapse, the Elapsed Period (Weeks) Between Stopping Antiviral Treatment and First Appearance of Virus in Serum: HBeAg-Positive TTR

ID	E or E+T	TTR1	TTR2	TTR3
TB 103	E+T	1.0	9.0	11.0
ER 109	E+T	1.0	3.5	ND
YWL 111	E+T	2.0	2.0	ND
CL 112	E+T	1.0	2.0	ND
	Mean ± SD	1 ± 0.43	4 ± 3.01 (p = 0.09)

Data estimate the host immune control of viremia.

ND, not determined; SD, standard deviation; TTR, time to relapse.

Table 5.

Comprises Data on Time to Relapse, the Elapsed Period (Weeks) Between Stopping Antiviral Treatment and First Appearance of Virus in Serum: Anti-HBe Antibody-Positive TTR

ID	E or E+T	TTR1	TTR2	TTR3
TL 101	E	9.0	6.0	12
LJL 104	E+T	6.0	9.0	10^*
MJ 105	E	2.0	4.0	2.0
YGL 106	E+T	4.0	10.0	12.0
ZXC 107	E	14.0	11.0	11.0
LG 110	E	14.0	13.0	NA
	Mean ± SD	8 ± 4.6	9 ± 3.0	10 ± 4.1 (p = 0.09)

Data estimate the host immune control of viremia.

Last HBeAg.

NA, not applicable.

The average of the HBsAg concentrations in samples collected during each of the three cycles is shown in Tables 6 and 7. The mean HBsAg concentration during cycle #1 or baseline in HBeAg-positive patients, 1086 + 558 IU/mL was significantly higher than in the anti-HBe-positive group, 172.5 + 152.2 IU/mL (p = 0.0045, two-tailed t-test).

Table 6.

Comprises Data on HBsAg Levels in Serum During Controlled Therapy Interruption: HBeAg-Positive Group

ID	Cycle #1	Cycle #2	p-Value
TB 103^*	301 ± 127	232 ± 109	0.27
ER 109	1075 ± 665	420 ± 317	0.042
YWL 111	1510 ± 451	350 ± 211	0.0024
CL 112	1458 ± 142	558 ± 100	0.002
Mean ± SD	1086 ± 0.43	390 ± 184

Last HBeAg.

CTI, controlled therapy interruption.

Table 7.

Comprises Data on HBsAg Levels in Serum During Controlled Therapy Interruption: Anti-HBe Antibody Positive

ID	Cycle #1	Cycle #2	p-Value	Cycle #3	p-Value
TL 101	266 ± 67	241 ± 150	0.33	154 ± 90	0.003
LJL 104	8.2 ± 5.6	19.2 ± 30.6	0.21	8.4 ± 5.1	0.21
MJ 105	384 ± 224	554 ± 409	0.16	521 ± 89	0.39
YGL 106	266 ± 74	217 ± 41	0.048	246 ± 103	0.33
ZXC 107	72 ± 27	91 ± 41	0.114	118 ± 132	0.197
LG 110	41 ± 33	62 ± 2	0.1	154 ± 90	N/A

The shorter TTR and the higher levels of HBsAg at baseline in HBeAg patients compared to anti-HBe suggest that the host immune control of HBV is significantly less in HBeAg patients compared to anti-HBe patients.

Autovaccination-induced secondary host immune control

TTR increased in all four HBeAg-positive patients from 1 ± 0.43 weeks at baseline to 4 ± 3.01 weeks (p = 0.09) after the first TI (Table 4). In contrast, in the anti-HBe group, CTI had no significant effect on TTR between any of the three cycles. At baseline, TTR was 8 ± 4.6 weeks. After subsequent CTI cycles, it was unchanged, 9 ± 3.0 weeks and 10 ± 4.1 weeks, respectively (Table 5).

Mean HBsAg concentrations in HBeAg-positive patients during cycles #1 and #2 are shown in Tables 6 and 7. The mean serum HBsAg concentration during cycle #2 fell in all four patients compared to cycle #1. The reduction in HBsAg concentration was statistically significant in all but patient #1. Interestingly, the baseline HBsAg concentration in this patient was 70–80% lower than the other three. Furthermore, this patient lost HBeAg at the end of the study.

In contrast, HBsAg levels were largely unchanged between all three cycles on all, but two, occasions in the anti-HBe patients. There were small but significant reductions in patient TL 101 between cycles #2 and #3 and patient YGL 106 between cycles #1 and #2 (Tables 6 and 7).

In the control group, HBV DNA levels drawn at the baseline did not change significantly from levels at the end of the study. No significant differences in adverse events were noted between treated and control patients.

Discussion

In models of chronic, persistent viral infection, potent suppression of the characteristically high-viral replication rates partially reverses the T-cell paralysis and can allow for modest generation of T-cell memory (2,29,30). This could explain the observation that long-term suppression of HBV replication with oral antiviral drugs in patients with HBeAg-positive chronic HBV infection can lead to durable loss of e-antigen and seroconversion to anti-HBe. Approximately, 50% of HBeAg-positive patients seroconvert over 5 years during antiviral therapy (5,7). CTI seeks to leverage vaccine theory to further augment the partially reconstituted host immune response induced by prolonged oral antiviral therapy, allowing progression to durable, long-term protective immunity (6).

CTI simulates autovaccination with autologous virus. In between prolonged durations of complete viral suppression, controlled TI exposes the host immune system to brief bursts of low-level viremia. HBIg is given before the second and subsequent cycles of TIs. HBIg binds available HBsAg in the extracellular fluid space.

The putative HBV-HBIg complexes confer several advantages, as follows. HBIg in combination with HBsAg vaccine is significantly more effective than either agent alone in invoking host response. The combination is universally effective in preventing vertically transmitted HBV in neonates (3). There is also evidence that, although conflicting, HBsAg-HBIg complex can act as a therapeutic vaccine in chronic hepatitis B patients (28,33 –35).

The combination of HBIg and oral antiviral therapy has also proved to be invaluable after orthotopic liver transplantation. HBIg with antiviral therapy has reduced both HBV recurrence and the risk of drug resistance, even in patients with active HBV replication (4).

The prime–boost strategy involves priming with antigen delivered by one vector and then selectively boosting this immunity with the same antigen delivered in the context of a second distinct vector (heterologous boosting). It is associated with increased numbers of antigen-specific T cells with higher avidity and increased efficacy against pathogen challenge and specifically with the generation of high levels of T-effector memory cells in animal models (1). Asynchronous timing of HBIg injection has been exploited within the context of CTI, as a heterologous “prime–boost” strategy. The first TI cycle is delivered without antibody. The second and subsequent cycles deliver HBV in the context of the HBIg vector. The development of strategies that emphasize cellular immune responses against intracellular pathogens such as HBV has been a major challenge. Several studies have successfully highlighted the power of prime–boost strategies to elicit protective cellular immunity against a variety of pathogens (10,18,23,26). Finally, a potential complication of TI with the oral antiviral agents is the development of hepatitis flares (8). The mechanism for hepatitis flares in the context of stopping antiviral treatment for HBV is unclear. Heightened host immunity against an exponentially increasing mass of infected cells is considered the most likely explanation. The presence of HBIg in the extracellular space traps virus and should limit de novo infection and infected cell mass. Interestingly, no patient demonstrated hepatitis flares after any of the TIs. However, given the relatively small numbers of treated patients, it is not clear that this is related to HBIg because none developed a flare during cycle #1 when no HBIg had been given.

TTR is the duration between TI and appearance of measurable viremia in the serum. We believe this to be a surrogate measure of host immune control of HBV. The mean TTR after TI in cycle #1 or baseline, 1 ± 0.43 weeks was significantly shorter in the HBeAg-positive group compared to 8 ± 4.6 weeks in the anti-HBe group. Thus, HBeAg-positive patients infected with HBV appear to be significantly more immune tolerant of the virus than anti-HBe-positive patients.

The data on HBsAg concentration in the two groups also support the conclusion that HBeAg patients are more immune tolerant of HBV. Mean HBsAg concentration at the baseline ranged from 301 to 1510 IU/mL in HBeAg patients. This is much higher than in anti-HBe-positive patients, 8.2–384 IU/mL. As a group, in clinical practice, hepatitis B viral loads are higher, patients are more infectious, and the probability of developing hepatocellular cancer is higher in HBeAg-positive patients (9,19,24).

The more rapid reappearance of HBV DNA and the higher concentrations of HBsAg in sera of patients with HBeAg-positive infection can be explained by either greater replication fitness and excretion of virus or less effective clearance from plasma by the host. The ability of CTI to alter TTR in HBeAg-positive patients argues more strongly for an effect on the host immunity than a change in the virus. Our data, therefore, suggest that CTI partially reverses immune tolerance in HBeAg-positive patients and that it has little influence in the anti-HBe group. TTR in the anti-HBe group was unchanged even after over two cycles of TI. HBsAg concentration followed the same pattern as the TTR and demonstrated no significant change between the three cycles.

The levels of circulating viral antigens (e-antigen and HBs antigen) in the HBeAg-positive patients are significantly higher than the anti-HBe group. These higher levels likely contribute to T-cell exhaustion and greater immune tolerance to HBV at the baseline (36,37). Prolonged reduction of the viral antigenic load appears to have induced some immune restitution in the HBeAg-positive patients. However, it does not appear to have much effect in the anti-HBe patients. The anergy to HBV in anti-HBe-positive patients appears to be unaffected by CTI. The mechanism(s) underlying the difference between these two groups are unclear.

Autovaccination or structured therapy interruption has been tried using HAART in patients with HIV disease with good results in primary HIV, but largely disappointing results in chronic infection (11,14,15).

In contrast to chronic HBV infection, chronic HIV infection is associated with irreversible depletion of CD4⁺ cells that may not allow for immune reconstitution after autovaccination.

We hypothesize that the effect of autovaccination in chronic HBV infection mimics the more favorable outcome in primary HIV infection (20,27) because CD4⁺ cells are better preserved in both cases.

Sera and peripheral blood mononuclear cell (PBMC) samples were collected and stored during the course of the study. These samples could prove invaluable in helping us to explore the differences between the two populations of patients and aid our understanding of the mechanisms that underlie viral persistence. For example, virus-specific T cells frequently hyperexpress the programmed death 1 (PD-1) molecule. Antiviral suppression of HBV replication may be inversely correlated with PD-1 levels and return of T-cell function. More recent studies also implicate T cell and mucin domain containing protein-3 (TIM-3) in T-cell exhaustion. The stored sera and PBMC offer a unique opportunity to correlate immune recovery in chronically infected subjects with respect to PD-1 levels, TIM-3 levels, HBV protein levels, and T-cell reactivity by studying central and effector memory phenotypes, pentamer staining, and reactivity to HBV antigens by EliSpot.

Although the combination of HBIg with CTI has not proved particularly effective in boosting host immunity in chronic anti-HBs-positive HBV infection, compounds currently in development may offer improvement. One example may be the oral Toll-like receptor-7 agonist. This compound induces prolonged suppression of HBV in chronically infected chimpanzees (16).

In summary, our data suggest that baseline HBV immune control, as determined by TTR or serum HBsAg concentration, is significantly lower for HBeAg-positive patients compared to their anti-HBe counterpart. CTI with HBIg partially reverses and improves host immune control in HBeAg-positive patients, but has little influence in anti-HBe patients. The mechanism(s) underlying the differential effects of CTI in HBeAg-positive versus anti-HBe-positive patients are unclear.

Footnotes

Acknowledgments

This work was supported by an Investigator Initiated Support Grant from the Bristol-Myers Squibb, BMS AI 463-197. Hepatitis B Immunoglobulin was supplied by the Biotest Pharmaceuticals Corporation. The study was performed under the auspices of the Harvard-Thorndike Clinical Research Center, Beth Israel Deaconess Medical Center, Boston.

Author Disclosure Statement

No competing financial interests exist.

References

Badovinac

, Messingham

, Jabbari

, et al. Accelerated CD8+ T-cell memory and prime-boost response after dendritic-cell vaccination. Nat Med, 2005; 11:748–756.

Barber

, Wherry

, Masopust

, et al. Restoring function in exhausted CD8 T cells during chronic viral infection. Nature, 2006; 439:682–687.

Beasley

, Hwang

, Szmuness

, et al. HBIG prophylaxis for perinatal HBV infections—final report of the Taiwan trial. Dev Biol Stand, 1983; 54:363–375.

Buchanan

, and Tran

. Current status of liver transplantation for hepatitis B virus. Clin Liver Dis, 2011; 15:753–764.

Buti

, Tsai

, Petersen

, et al. Seven-year efficacy and safety of treatment with tenofovir disoproxil fumarate for chronic hepatitis B virus infection. Dig Dis Sci, 2015; 60:1457–1464.

Campos

, and Godson

. The effectiveness and limitations of immune memory: understanding protective immune responses. Int J Parasitol, 2003; 33:655–661.

Chan

, Chan

, Hui

, et al. Effects of tenofovir disoproxil fumarate in hepatitis B e antigen-positive patients with normal levels of alanine aminotransferase and high levels of hepatitis B virus DNA. Gastroenterology, 2014; 146:1240–1248.

Chang

, Jeng

, Liaw

. Clinical events after cessation of lamivudine therapy in patients recovered from hepatitis B flare with hepatic decompensation. Clin Gastroenterol Hepatol, 2015; 13:979–986.

Chen

, and Yang

. Natural history of chronic hepatitis B REVEALed. J Gastroenterol Hepatol, 2011; 26:628–638.

10.

De Filette

, Soehle

, Ulbert

, et al. Vaccination of mice using the West Nile virus E-protein in a DNA prime-protein boost strategy stimulates cell-mediated immunity and protects mice against a lethal challenge. PLoS One, 2014; 9:e87837.

11.

Fagard

, Oxenius

, Gunthard

, et al. A prospective trial of structured treatment interruptions in human immunodeficiency virus infection. Arch Intern Med, 2003; 163:1220–1226.

12.

Flink

, Sprengers

, Hansen

, et al. Flares in chronic hepatitis B patients induced by the host or the virus? Relation to treatment response during Peg-interferon {alpha}-2b therapy. Gut, 2005; 54:1604–1609.

13.

Jacobson

. Therapeutic options for chronic hepatitis B: considerations and controversies. Am J Gastroenterol, 2006; 101 Suppl 1:S13–S18.

14.

Jacobson

, Pat Bucy

, Spritzler

, et al. Evidence that intermittent structured treatment interruption, but not immunization with ALVAC-HIV vCP1452, promotes host control of HIV replication: the results of AIDS Clinical Trials Group 5068. J Infect Dis, 2006; 194:623–632.

15.

James

. Structured treatment interruption: important controlled trial in monkeys. AIDS Treat News, 2000; 2–3.

16.

Lanford

, Guerra

, Chavez

, et al. GS-9620, an oral agonist of Toll-like receptor-7, induces prolonged suppression of hepatitis B virus in chronically infected chimpanzees. Gastroenterology, 2013; 144:1508–1517.

17.

Levstik

, Wong

, Greanya

, and Yoshida

. The role of hepatitis B immunoglobulin in hepatitis B related liver transplantation: Canadian Transplant Centre Position Paper. Ann Hepatol, 2011; 10:441–444.

18.

, Arevalo

, Chen

, et al. T-cell-mediated cross-strain protective immunity elicited by prime-boost vaccination with a live attenuated influenza vaccine. Int J Infect Dis, 2014; 27:37–43.

19.

Liaw

. Natural history of chronic hepatitis B virus infection and long-term outcome under treatment. Liver Int, 2009; 29 Suppl 1:100–107.

20.

Miller

. Structured treatment interruptions in antiretroviral management of HIV-1. Curr Opin Infect Dis, 2001; 14:29–37.

21.

Papasavvas

, Kostman

, Mounzer

, et al. Randomized, controlled trial of therapy interruption in chronic HIV-1 infection. PLoS Med, 2004; 1:e64.

22.

Pichichero

. Booster vaccinations: can immunologic memory outpace disease pathogenesis?. Pediatrics, 2009; 124:1633–1641.

23.

Ranasinghe

, and Ramshaw

. Genetic heterologous prime-boost vaccination strategies for improved systemic and mucosal immunity. Expert Rev Vaccines, 2009; 8:1171–1181.

24.

van Schalkwyk

, Nourmoussavi

, Massey

, et al. Missed opportunities for prevention of perinatal transmission of hepatitis B: a retrospective cohort study. Can J Gastroenterol Hepatol, 2014; 28:525–528.

25.

Vandepapeliere

, Lau

, Leroux-Roels

, et al. Therapeutic vaccination of chronic hepatitis B patients with virus suppression by antiviral therapy: a randomized, controlled study of co-administration of HBsAg/AS02 candidate vaccine and lamivudine. Vaccine, 2007; 25:8585–8597.

26.

Villarreal

, Talbott

, Choo

, et al. Synthetic DNA vaccine strategies against persistent viral infections. Expert Rev Vaccines, 2013; 12:537–554.

27.

Walensky

, Goldie

, Sax

, et al. Treatment for primary HIV infection: projecting outcomes of immediate, interrupted, or delayed therapy. J Acquir Immune Defic Syndr, 2002; 31:27–37.

28.

Wen

, Guo

, Zhang

, et al. Enhanced immunogenicity in mice with hepatitis B vaccine complexed to human hepatitis B immunoglobulin. Chin Med J (Engl), 1994; 107:741–744.

29.

Wherry

, and Ahmed

. Memory CD8 T-cell differentiation during viral infection. J Virol, 2004; 78:5535–5545.

30.

Wherry

, McElhaugh

, and Eisenlohr

. Generation of CD8(+) T cell memory in response to low, high, and excessive levels of epitope. J Immunol, 2002; 168:4455–4461.

31.

Wherry

, Puorro

, Porgador

, and Eisenlohr

. The induction of virus-specific CTL as a function of increasing epitope expression: responses rise steadily until excessively high levels of epitope are attained. J Immunol, 1999; 163:3735–3745.

32.

Woodland

. Jump-starting the immune system: prime-boosting comes of age. Trends Immunol, 2004; 25:98–104.

33.

, Wang

, Shen

, et al. Results of a phase III clinical trial with an HBsAg-HBIG immunogenic complex therapeutic vaccine for chronic hepatitis B patients: experiences and findings. J Hepatol, 2013; 59:450–456.

34.

, Zhao

, Guo

, et al. A randomized controlled phase IIb trial of antigen-antibody immunogenic complex therapeutic vaccine in chronic hepatitis B patients. PLoS One, 2008; 3:e2565.

35.

Yao

, Zheng

, Zhou

, et al. Therapeutic effect of hepatitis B surface antigen-antibody complex is associated with cytolytic and non-cytolytic immune responses in hepatitis B patients. Vaccine, 2007; 25:1771–1779.

36.

, Cox

, and Zajac

. T-cell exhaustion: characteristics, causes and conversion. Immunology, 2010; 129:474–481.

37.

Zinselmeyer

, Heydari

, Sacristan

, et al. PD-1 promotes immune exhaustion by inducing antiviral T cell motility paralysis. J Exp Med, 2013; 210:757–774.