Loss of HBsAg antigen during treatment with entecavir or lamivudine

Loss of HBsAg antigen during treatment with entecavir or lamivudine in nucleoside-naïve HBeAg-positive patients with chronic hepatitis B



Journal of Viral Hepatitis Aug 6 2009 Early view


R. G. Gish 1 , T.-T. Chang 2 , C.-L. Lai 3 , R. de Man 4 , A. Gadano 5 , F. Poordad 6 , J. Yang 7 , H. Brett-Smith 7 and R. Tamez 7

1 Division of Hepatology and Complex GI, Physicians Foundation California Pacific Medical Center, San Francisco, CA, USA ; 2 National Cheng Kung University Medical College, Tainan, Taiwan ; 3 Department of Medicine, University of Hong Kong, Hong Kong SAR, China ; 4 Erasmus Medical Centre, University Hospital Rotterdam, Netherlands ; 5 Hospital Italiano, Hepatologia, Buenos Aires, Argentina ; 6 Department of Hepatology and Liver Transplantation, Cedars-Sinai Medical Center, Los Angeles, CA, USA ; and 7 Bristol-Myers Squibb Company, Research and Development, Princeton, NJ, USA

Correspondence to Robert G. Gish, Division of Hepatology and Complex GI, Physician Foundation California Pacific Medical Center, 2340 Clay Street, San Francisco, CA 94115-1932, USA. E-mail: gishr@sutterhealth.org



*Presented in part at the 57th Annual Meeting of the American Association for the Study of Liver Diseases, October 27–31, 2006.



"In conclusion, this retrospective analysis of data from a randomized, international phase III trial shows that the rate of HBsAg loss was approx. 5% in nucleoside-naïve HBeAg-positive patients treated with entecavir for up to 120 weeks. Parameters of response demonstrated a sequential pattern of ALT normalization, HBV DNA reduction to <300 copies/mL, followed by HBeAg loss, HBe seroconversion and HBsAg loss. Most patients who experienced HBsAg loss had achieved HBV DNA <300 copies/mL, ALT normalization and HBsAg loss by week-48 of treatment. Importantly, sustained suppression off-treatment of HBV DNA was demonstrated in those patients who exhibited HBsAg loss."



ABSTRACT



Summary. This retrospective analysis was conducted to describe the characteristics of nucleoside-naïve hepatitis B e antigen (HBeAg)-positive patients with chronic hepatitis B, who achieved hepatitis B surface antigen (HBsAg) loss during entecavir or lamivudine therapy. HBeAg-positive adults with chronic hepatitis B, elevated serum alanine aminotransferase, and compensated liver disease were randomized to double-blind treatment for up to 96 weeks with entecavir 0.5 mg/day or lamivudine 100 mg/day. HBsAg and hepatitis B virus (HBV) DNA were measured at regular intervals during and off-treatment follow-up. Through a maximum duration of 96 weeks on-treatment and 24 weeks off-treatment, HBsAg loss was confirmed in 18/354 (5.1%) patients treated with entecavir and 10/355 (2.8%) patients treated with lamivudine. Among the 28 patients with confirmed HBsAg loss, 27 (96%) achieved HBV DNA <300 copies/mL, and 27 (96%) achieved confirmed HBeAg loss. All entecavir recipients with HBsAg loss had HBV DNA <300 copies/mL. Caucasian patients, and those infected with HBV genotype A or D, were significantly more likely to lose HBsAg. This retrospective analysis of data from a randomized, global phase three trial shows that confirmed loss of HBsAg occurred in 5% of nucleoside-naïve HBeAg-positive patients treated with entecavir, and that HBsAg loss is associated with sustained off-treatment suppression of HBV DNA.





It is estimated that 400 million people worldwide are chronically infected with hepatitis B virus (HBV), as indicated by the presence of hepatitis B surface antigen (HBsAg) in serum for longer than 6 months [1]. These patients are at substantially increased risk of cirrhosis and hepatocellular carcinoma (HCC), diseases that lead to one in every 40 deaths worldwide—approx. 1 million each year [2]. The risk of cirrhosis and HCC increases in proportion to the serum level of HBV DNA in patients with chronic hepatitis B [3,4]. Sustained suppression of viral replication has been shown to improve clinical outcomes, and since eradication or 'cure' of HBV is not possible, virologic control becomes a primary goal of treatment [5–7].



Clearance of HBsAg has been associated with very low or undetectable HBV DNA [<300 copies/mL by polymerase chain reaction (PCR)], normalization of serum alanine aminotransferase (ALT) levels, improvement in liver histology, a decreased risk of HCC, and prolonged survival [8,9]. On the basis of long-term natural history studies, spontaneous HBsAg seroclearance has been estimated to occur at a rate of 0.5–1.7% per year [10–13]. Compared with Western patients who typically are infected with HBV later in life, HBsAg seroclearance is known to occur less frequently in Asian patients (<1%/year) [14], who typically acquire HBV at birth.



None of the currently available therapies for chronic hepatitis B induce HBsAg loss or seroconversion in the majority of patients treated. Interferon-based therapies increase the rate of HBsAg clearance above the spontaneous rate of seroconversion, although the numbers of subjects in these clinical trials have been small [5,15]. It is generally thought that nucleoside analogues do not influence HBsAg clearance [5]. However, suppression of HBV DNA replication by nucleoside analogues has been shown to significantly reduce quantitative serum levels of HBsAg, intrahepatic cccDNA, and total intracellular HBV DNA [16]. It has been suggested that effective suppression of HBV DNA replication for prolonged periods with potent nucleoside analogues may increase the rate of HBsAg loss.



Entecavir (Baraclude®, Bristol-Myers Squibb Company) is a potent guanosine analogue that leads to significantly greater virologic, histologic and biochemical improvements than lamivudine in nucleoside-naïve patients with chronic HBV [17,18]. Suppression of HBV DNA replication to undetectable levels was observed in 92% of treatment-naïve HBeAg-positive patients through 5 years of treatment with entecavir [19,20]. Moreover, resistance to entecavir is rare in nucleoside/nucleotide-naïve patients; overall, the rate of acquisition of resistance substitutions is approx. 1% after 5 years of continuous treatment with the drug [21].



The objective of this analysis is to describe the characteristics of patients who achieved a confirmed HBsAg loss after exposure to entecavir or lamivudine for up to 96 weeks on-treatment, and 24 weeks of off-treatment follow-up in a large, randomized, double-blind, multicentre, phase III clinical trial (ETV-022) [17].



Discussion



This analysis demonstrated that HBsAg loss occurs in a subset of patients with chronic hepatitis B during treatment with nucleoside antivirals, and that the rate exceeds that expected for spontaneous seroclearance. Overall, approx. 5% of entecavir-treated patients experienced loss of HBsAg during a maximum observation period of 120 weeks on and off treatment (median observation period of 92 weeks). The analysis of baseline characteristics and of the temporal pattern of responses provide useful information about the natural history of this process. Normalization of serum ALT usually occurred first, followed sequentially by suppression of HBV DNA to undetectable (<300 copies/mL), loss of serum HBeAg, and in the majority of s responders, HBe seroconversion (production of anti-HBe). Loss of HBsAg and seroconversion to anti-HBs was generally the last response observed; Loss of HBsAg and seroconversion to anti-HBs was generally the last response observed. In this analysis, 14 patients (50%) with HBsAg loss had conversion to anti-HBs by week-120.



An analysis of predictive factors showed that loss of HBsAg was most commonly associated with Caucasian race and infection with HBV genotype A or D. These observations are consistent with previous reports documenting lower rates of HBsAg loss and seroconversion in Asian patients, in whom genotypes B and C are more common as compared with Caucasian patients [22,23] and in whom vertical transmission is the predominant mode of transmission. In the present study, four Asian patients (three entecavir; one lamivudine) achieved HBsAg loss. In addition, HBeAg loss and HBe seroconversion at week-24 were also associated with HBsAg loss through 120 weeks and most patients (96%) had HBV DNA suppression to undetectable levels (<300 copies/mL by PCR).



The overall goal of therapy for HBV infection is sustained suppression of HBV DNA replication and remission of liver disease (AASLD guidelines) [6]. Historically, the use of HBe seroconversion as a marker of successful chronic hepatitis B treatment was based on an understanding that HBe seroconversion resulted in low HBV DNA levels and normalization of ALT with resolution of necroinflammatory disease. The most recently updated treatment guidelines from the Asian-Pacific Association for the Study of the Liver recommend stopping treatment not only after achieving HBE seroconversion, but also durable suppression of HBV replication [24]. Limited data are available regarding long-term off-treatment follow-up of patients on antiviral therapy who achieve HBe loss, HBe seroconversion, or both, along with suppression of HBV replication. In one report, 44% of adefovir-treated patients who achieved HBe seroconversion still had HBV DNA <1000 copies/mL at their last follow-up (median follow-up off-treatment was 143 weeks) [25]. A study of telbivudine demonstrated sustained (52 weeks) HBe seroconversion, but HBV DNA levels were not analysed among those who had HBe seroconversion [26].



In contrast with HBeAg loss and HBe seroconversion, where HBV DNA levels may modestly increase, HBsAg loss generally signals persistent loss of serum HBV DNA and has been shown to lead to improved survival, as well as diminished risk of hepatocellular carcinoma in patients with cirrhosis [14]. Clinical trials of lamivudine and adefovir have shown the rate of HBsAg loss to be around 2%, and 0%, respectively, at 1 year, and after 5 years, 2% of patients demonstrated HBsAg loss and HBs seroconversion with adefovir [27–29]. In a study of tenofovir, 6% of patients demonstrated HBsAg loss by week-96 [30]. However, these analyses have not reported the relationship between HBsAg loss and sustained suppression of HBV DNA with the nucleotide analogues. In one study of pegylated interferon, 61% of patients who exhibited HBsAg loss had HBV DNA <400 copies/mL after 1 year [31]. The current analysis showed that among patients with confirmed HBsAg loss, all patients maintained HBV DNA suppression to <104 copies/mL, and most to 300 copies/mL at the week-24 off-treatment visit.



A recent report suggested that seroclearance of HBsAg observed during lamivudine therapy may not indicate viral clearance, because it may be caused by a point mutation in the S-gene, which results in failure to detect HBsAg [32]. In this report, a P120A mutation in the S-gene was demonstrated in six out of 11 patients who experienced HBsAg seroclearance. All patients had detectable serum HBV DNA either by a standardized quantitative test or by PCR. Santantonio et al. [33] have suggested that the hypothesis of S-gene mutant selection should be limited to those patients with persistent viremia despite having achieved HBsAg loss. This observation suggests that the HBsAg loss observed in the current study reflects control of viremia rather than S-gene mutation.



Several recent reports have attempted to quantify HBsAg titers as predictive of long-term HBsAg loss and response to treatment [34,35]. However, at the time of the current study, investigation into the role of quantitative HBsAg titers was at its infancy and, therefore, not included in the analysis. Furthermore, the tests used to measure quantitative HBsAg, mainly the Abbott Architect® (Abbott Laboratories, Abbott Park, IL, USA) HBsAg and the Elecsys® (Roche Diagnostics, Penzberg, Germany) HBsAg II assays, were not available at the time that this study was conducted and, are only beginning to be validated for this purpose [36,37].



In conclusion, this retrospective analysis of data from a randomized, international phase III trial shows that the rate of HBsAg loss was approx. 5% in nucleoside-naïve HBeAg-positive patients treated with entecavir for up to 120 weeks. Parameters of response demonstrated a sequential pattern of ALT normalization, HBV DNA reduction to <300 copies/mL, followed by HBeAg loss, HBe seroconversion and HBsAg loss. Most patients who experienced HBsAg loss had achieved HBV DNA <300 copies/mL, ALT normalization and HBsAg loss by week-48 of treatment. Importantly, sustained suppression off-treatment of HBV DNA was demonstrated in those patients who exhibited HBsAg loss.



Results



A total of 709 HBeAg-positive patients were treated with entecavir 0.5 mg/day (n = 354) or lamivudine 100 mg/day (n = 355) for 52–96 weeks. Through a maximum duration of 120 weeks of observation (up to 96 weeks on-treatment plus up to 24 weeks off-treatment follow-up; mean and median duration of observation = 84.8 and 92.1 weeks, respectively), HBsAg loss was confirmed in a total of 28 patients: 18 (5.1%) entecavir recipients and 10 (2.8%) lamivudine recipients. Among all treated patients, HBe seroconversion was confirmed during the same observation period in 202 (110 entecavir- and 92 lamivudine-treated) patients. Of the 202 patients who underwent HBe seroconversion, 8/110 (7%) entecavir- and 8/92 (9%) lamivudine-treated patients also experienced confirmed HBsAg loss.



A comparison of baseline characteristics between patients with and without HBsAg loss revealed that Caucasian patients and those infected with HBV genotypes A or D were more likely to experience confirmed HBsAg loss compared with those who did not experience HBsAg-loss (Table 1). Analysis of baseline serum HBV DNA and ALT, Knodell necroinflammatory score, age and a history of treatment with interferon did not show an association with HBsAg loss.



By week-24, more patients with confirmed HBsAg loss through 120 weeks had achieved HBV DNA <104 or <300 copies/mL (86% and 46% respectively), compared with patients who did not achieve confirmed HBsAg loss (63% and 33%, respectively) (Table 2). The study also confirmed that HBsAg loss was associated with HBeAg loss and HBe seroconversion at week-24 (50% and 43%, respectively, vs 10% and 10%, respectively, among patients without confirmed HBsAg). At the time that HBsAg loss was documented, all patients had HBV DNA <104 copies/mL, and most had achieved HBV DNA <300 copies/mL (89%), ALT normalization (93%), or had experienced HBeAg loss (86%). The proportion of patients achieving these end points increased during continued follow-up, with 96% eventually achieving HBV DNA <300 copies/mL, 96% achieving confirmed HBeAg loss, and 86% achieving confirmed HBe seroconversion. Among the 28 patients with confirmed HBsAg loss, 14 patients (15%) had confirmed HBs seroconversion, which was sustained at the last observation.



Figure 1 shows the temporal distribution of each end point by patient. In the majority of patients with confirmed HBsAg loss, therapeutic end points were achieved in a sequential manner. ALT normalization generally preceded suppression of HBV DNA to below the limit of detection, which, in turn, preceded HBeAg loss and HBe seroconversion, and, subsequently, HBsAg loss and HBs seroconversion. Among the 24 subjects with both HBsAg and HBeAg loss, the median time to the first confirmed response was: 26.3 weeks for HBV DNA <300 copies/mL, 25.5 weeks for HBeAg loss, 31.8 weeks for HBe seroconversion and 60.4 weeks for HBsAg loss.



Among the 18 entecavir recipients with confirmed HBsAg loss, the majority achieved other relevant clinical end points at some point during the observation period. All patients achieved suppression of HBV DNA to <300 copies/mL; 17 of 18 patients had ALT normalization; 17 of 18 patients had confirmed HBeAg loss; and 16 of 18 patients had confirmed HBeAg seroconversion. The single patient who did not experience confirmed HBeAg loss (patient ETV-16, Fig. 1) also failed to achieve confirmed HBe or HBs seroconversion. Suppression of HBV DNA to <300 copies/mL in this patient was documented at week-36, and ALT normalization at week-52. HBsAg loss occurred at the week-96 visit, at which time the patient discontinued therapy. During off-treatment follow-up, HBeAg loss was documented once but not confirmed; this patient became HBsAg-positive at the 24 week off-treatment visit, with an HBV DNA <104 copies/mL. A second entecavir patient did not have a confirmed HBe seroconversion (patient ETV-17, Fig. 1) during the observation period. This patient had HBV DNA <300 copies/mL at week-24 and ALT normalization at week-28. This patient discontinued treatment after 81 weeks, and had HBeAg loss, HBsAg loss and HBs seroconversion at week-5 of off-treatment observation. Although an HBV DNA sample was unavailable for this patient at the last observation, this patient's HBV DNA was <104 copies/mL at the previous observation (9 months prior).



Sixteen of the 18 entecavir-treated patients with confirmed HBsAg loss completed 24 weeks of off-treatment follow-up. At the 24 week off-treatment visit, the majority of patients sustained HBV DNA suppression (12/16 had HBV DNA <300 copies/mL; all had HBV DNA <104 copies/mL) and HBe seroconversion. Testing for e-antibody was indeterminate in two patients and negative in another. All three patients had HBV DNA <300 copies/mL. At this visit, patient ETV-10 became HBsAg positive but sustained HBeAg loss and HBe seroconversion. Therefore, a total of two patients (ETV-16 and ETV-10, Fig. 1) regained HBsAg during off-treatment observation.



Likewise, in the lamivudine-treatment group, the majority of those with HBsAg loss achieved other relevant clinical end points at some time during the observation period: nine of 10 patients achieved HBV DNA <300 copies/mL (all had HBV DNA <104 copies/mL); nine of 10 achieved ALT normalization; and all 10 experienced HBeAg loss, with eight having a confirmed HBe seroconversion. The first patient without HBe seroconversion (patient LVD-9, Fig. 1) had ALT normalization at week-16, HBV DNA <300 copies/mL at week-24, and HBeAg loss with HBs seroconversion at week-85. This patient completed 97 weeks of treatment and discontinued therapy, becoming HBsAg negative at week-9 of off-treatment observation. At the 24 week off-treatment visit, this patient had sustained HBV DNA suppression to <300 copies/mL; HBeAg loss, HBsAg loss and HBs seroconversion were sustained, but remained HBeAb negative. The second patient who did not achieve HBe seroconversion (patient LVD-10, Fig. 1), had ALT normalization at week-12, HBV DNA <104 copies/mL at week-26, HBeAg loss at week-49, and discontinued therapy at week-52 without ever achieving HBV DNA <300 copies/mL. By the week-24 off-treatment visit, this patient had confirmed HBsAg loss, HBs seroconversion, and was negative for anti-HBe despite previously having a positive and indeterminate result. This patient's HBV DNA remained <104 copies/mL. All lamivudine-treated patients with confirmed HBsAg loss completed 24 weeks of off-treatment follow-up. HBV DNA remained suppressed in the majority of patients (nine out of 10 had HBV DNA <300 copies/mL; all had HBV DNA <104 copies/mL), and HBeAg loss, HBe seroconversion, HBsAg loss and HBs seroconversion were sustained in all patients at 24 weeks off treatment.



Materials and methods



The complete study design and patient selection criteria for the trial have been published elsewhere [17]. Informed consent was obtained from each patient included in the study, and the study protocol conforms to the ethical guidelines of the 1975 Declaration of Helsinki as reflected in a priori approval by the institution's human research committee. Nucleoside-naïve, HBeAg-positive patients aged ≥16 years with serum HBV DNA ≥3 mEq/mL by Quantiplex™ (Chiron, Emeryville, CA, USA) branched-chain DNA assay, elevated serum ALT [1.3–10 times the upper limit of normal (ULN)], evidence for chronic hepatitis B on liver biopsy, and compensated liver disease (i.e. Child-Pugh class A) were eligible. All patients had chronic HBV infection that was documented with positive HBsAg for at least 24 weeks. Patients coinfected with hepatitis C or D or human immunodeficiency virus were excluded.



Eligible patients were randomized to receive at least 52 weeks of double-blind treatment with either entecavir (Baraclude®, Bristol-Myers Squibb, Princeton, NJ, USA) 0.5 mg once daily or lamivudine (Epivir-HBV or Zeffix®, GlaxoSmithKline UK, Uxbridge, UK) 100 mg once daily. According to protocol-defined criteria, patients were to discontinue study at week-52, if they achieved a protocol-defined response (HBV DNA <0.7 mEq/mL and HBeAg loss) at week-48. Patients who achieved HBV DNA <0.7 mEq/mL but not HBeAg loss could continue blinded therapy for up to 96 weeks and discontinue study therapy once a response was achieved. Patients achieving a response were monitored off treatment for up to 24 weeks.



Serum HBV DNA, HBeAg, HBsAg and ALT levels were measured at regular intervals during treatment and untreated follow-up. HBV DNA was centrally quantified by Roche Amplicor PCR assay (limit of detection 300 copies/mL) at weeks-24, 36, 48, 96 and at week-24 of off-treatment follow-up. HBeAg, anti-HBe antibody and HBsAg were measured by Abbott AxSYM microparticle enzyme immunoassay at weeks-12, 24, 36, 48, 60, 68, 84 and 96 during treatment. ALT was measured at local laboratories at 4-week intervals during treatment and follow-up. A confirmed end point was defined as two consecutive measurements meeting a particular criterion or at last observation.



All data analyses were descriptive and exploratory. Baseline characteristics were examined to determine if patients with confirmed HBsAg loss were similar to patients without confirmed HBsAg loss. Efficacy parameters at week-24, were also presented to examine the differences between patients who achieved confirmed HBsAg loss and those who did not. The times that efficacy end points (HBV DNA<300 copies/mL, loss of HBeAg, HBe seroconversion and ALT ≤1 × ULN) were achieved relative to HBsAg loss were plotted for each individual patient