- Email: [email protected]
Role of HBsAg decline in patients with chronic hepatitis B HBeAg-negative and E genotype treated with pegylated-interferon
Antiviral Research 136 (2016) 32e36
Contents lists available at ScienceDirect
Antiviral Research journal homepage: www.elsevier.com/locate/antiviral
Role of HBsAg decline in patients with chronic hepatitis B HBeAgnegative and E genotype treated with pegylated-interferon Lucio Boglione*, Jessica Cusato, Giuseppe Cariti, Giovanni Di Perri, Antonio D'Avolio Unit of Infectious Diseases, University of Turin, Department of Medical Sciences, Amedeo di Savoia Hospital, Turin, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 May 2016 Received in revised form 7 September 2016 Accepted 23 October 2016 Available online 26 October 2016
Treatment options for patients with chronic hepatitis B (CHB) and hepatitis B e antigen (HBeAg)-negative are pegylated interferon alfa-2a (PEG-IFN) for 48 weeks or nucleos(t)ide analogues (NAs). The choice of patients with higher chance of sustained response (SR) to PEG-IFN can be made with pre-treatment and on-treatment factors; recent studies evidenced the role of early drop of serum hepatitis B surface antigen (HBsAg) as predictor of SR. The aim of this study was the evaluation of early decrease of HBsAg on the SR in HBeAg-negative patients with E genotype. A retrospective analysis was performed on 72 patients affected by HBeAg-negative CHB with E genotype, treated for 48 weeks with PEG-IFN. HBsAg and HBV-DNA kinetics were evaluated. Decline of HBsAg (>0.5 logIU/mL) and HBV-DNA (2 logIU/mL) at 12 weeks was described according to observed chance of SR. After 96 weeks of follow-up, SR was observed in 10 patients (13.9%); HBsAg loss 6 (8.3%), HBsAg seroconversion in 3 (4.2%). No patients with HBsAg decline 0.5 log IU/mL and HBV-DNA<2 logIU/mL achieved SR (negative predictive value, NPV 100%). In multivariate analysis were significantly associated with SR the combined decline of HBsAg and HBV-DNA at 12 weeks (OR ¼ 35.336; 95% CI: 4.668e112.226; p < 0.001) and the HBsAg7500 IU/mL at 24 weeks (OR ¼ 51.824; 95% CI: 9.692e134.144; p < 0.001). Combining the HBsAg and HBV-DNA decline at 12 weeks we can identify patients without chance of SR who may stop PEG-IFN treatment. Stopping rule at 24 weeks using HBsAg7500 IU/mL is strong predictor of SR in HBeAg-negative patients with E genotype. © 2016 Elsevier B.V. All rights reserved.
Keywords: Hepatitis B treatment PEG-IFN HBV genotype qHBsAg HBeAg-negative
Current treatment options for chronic hepatitis B (CHB) are pegylated interferon alfa (PEG-IFN) or nucleos(t)ide analogues (NAs); even if the optimal endpoint is the hepatitis B surface antigen (HBsAg) loss, this result is difficult to achieve and sustained response (SR) is the most attainable goal (EASL, 2012). However, the use of PEG-IFN is limited in the hepatitis e antigen (HBeAg)-negative patients for the low rate of virological response and the sideeffects associated with the subcutaneous administration (Zoulim and Perrillo, 2008). Therefore, NAs were the most used treatments in HBeAg-negative patients, but the HBsAg decrease is very poor (Boglione et al., 2013; Chevaliez et al., 2013; Wong et al., 2013)
List of abbreviations: HBV, Hepatitis B virus; NA, nucleos(t)ide analogues; CHB, chronic hepatitis B; PEG-INF, pegylated interferon; HBsAg, hepatitis B surface antigen; ALT, alanine aminotransferase; HBeAg, hepatitis e antigen; SVR, sustained virological response; UNL, upper normal level; qHBsAg, quantitative HBsAg; IQR, inter-quartile range. * Corresponding author. E-mail address: [email protected] (L. Boglione). http://dx.doi.org/10.1016/j.antiviral.2016.10.011 0166-3542/© 2016 Elsevier B.V. All rights reserved.
and the HBsAg loss is difficult to achieve. The PEG-IFN therapy could be optimized by selection of patients with best chance of response using baseline predictive factors (Krishnamoorthy and Mutimer, 2015), while host genetic factors as interleukin (IL)28B were analyzed in some studies but without reliable results (Boglione et al., 2014b; Galmozzi et al., 2014; Lampertico et al., 2013). Other promising predictive marker of HBsAg loss seems to be the serum level of interferon-inducible protein 10 (IP10), but has been studied only during NAs therapy (Papatheodoridis et al., 2014). The reduction of HBsAg during PEGIFN has been examined as predictive factor of SR; a quantitative HBsAg (qHBsAg) decline of at least 0.5 and 1 log IU/mL after 12 and 24 weeks evidenced 90% and 97% of negative predictive value (NPV) on SR in a cohort of 48 HBeAg-negative patients with mixed genotypes (Moucari et al., 2009). Using both HBsAg and HBV-DNA decline after 12 weeks we can identify the patients without chance of response (Rijckborst et al., 2010a); the validation of this stopping rule in patients with D
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
genotype allows to treatment interruption in non-responders (Rijckborst et al., 2012); however, in the PARC trial other genotypes were present (Rijckborst et al., 2012) and also in B and C genotypes Peng et al. demonstrated the useful of this stopping rule (Peng et al., 2012). An additional stopping rule could be the HBsAg level>7500 IU/mL after 24 weeks of treatment: the SR rate in patients with HBsAg level7500 IU/mL was 16%, while 6% if HBsAg was above this cut-off (Lampertico et al., 2011). Few data are currently available about the treatment in E genotype; in our previous study we reported a poor rate of SR in this genotype (Boglione et al., 2014a). This retrospective analysis was performed in our entire cohort of patients affected by CHB with E genotype and treated with PEG-IFN at our centre from April 2008 to January 2014. Eligible patients for this study were all CHB with E genotype, age 18 years or older, HBeAg-negative, naïve for previous treatment. From an entire cohort of 122 eligible subjects, we excluded 11 patients with previous treatment, 5 treated with non-pegylated interferon, 13 with incomplete treatment (drop-out or adverse events) or follow-up for any reason, 17 for unavailability of qHBsAg during the treatment and follow-up, 4 without available informed consent: therefore 72 patients were finally included in this analysis. This study was conducted in accordance with the guidelines of the Declaration of Helsinki and the principles of good pharmacoepidemiology practices (ISPE, 2008) and was approved by our local Ethic Committee as “HBV-GEN”. All included patients were treated with PEG-IFN alpha-2a with starting dose of 180 mg/weekly for 48 weeks. The follow-up time was 48 weeks after treatment completion. The end points of therapy were established according to EASL Clinical Practice Guidelines (EASL, 2012): end-of-treatment (ET) virological response was defined as HBV-DNA HBV-DNA <2000 IU/ mL (10,000 copies/mL) at the end of therapy; the sustained response (SR) was defined as HBV-DNA <2000 IU/mL (10,000 copies/mL) at the end of follow-up. Serological response was defined as HBsAg loss with or without anti-HBs; the virological relapse was defined as HBV-DNA >2000 IU/mL (10,000 copies/mL) after previous ET virological response. Baseline characteristics of study population were described in Table 1. Virological and serological response were reported in Table 2; ET response was observed in 21 patients (29%); SR in 10 patients (13.9%); virological relapse was reported in 11 (15.%); patients who did not obtain ET or SR (non-responders) were 51 (70%). HBsAg loss was seen in 6 patients (8%) and 3 with antiHBs seroconversion (4%). In Fig. 1A were represented the virological and serological outcomes: virological response decreased from 29% at ET to 20% after 1 year and 13.9% after 2 years. Relapser rate increased from 8% to 15% after 1 and 2 years; HBsAg loss increased from 1.3% at ET from 4% to 8% after 1 and 2 years.
Table 1 Baseline characteristics of the study patients. Patients overall Gender: male, n (%) Age (yr) Route of transmission, n (%)
BMI (kg/m2) Liver stiffness (kPa) Alanine aminotransferase (ALT) [IU/mL] HBV-DNA [LogIU/mL] HVL (HBV-DNA>8 log IU/mL), n (%) qHBsAg [LogIU/mL]
N ¼ 72 62 (86.1) 30.3 [27e38] Unknown: 23 (31.9) Sexual: 9 (12.5) Familiary history: 40 (55.6) 22 [20.1e23.8] 9.6 [6.9e23.8] 77.5 [66e96.7] 5.3 [4.8e6.3] 6 (8.3) 4.1 [3.8e4.2]
33
Table 2 Treatment outcomes and adverse events. Outcomes End-of treatment response Sustained response Relapser Non-responders HBsAg loss Anti-HBsAg seroconversion
N ¼ 72 21 (29.2) 10 (13.9) 11 (15.2) 51 (70.8) 6 (8.3) 3 (4.2)
Adverse events Grade 1 leucopenia Grade 1 anemia Grade 2 anemia Flu-like syndrome Fatigue Weight loss Psychiatric symptoms (anxiety or depression) PEG-IFN alfa-2a reduction dose to 135 mg
12 (16.6) 18 (25) 8 (11.1) 28 (38.8) 22 (30.5) 14 (19.4) 4 (5.5) 6 (8.3)
In Table 2 were described the side-effects and adverse events (AE) observed during the treatment: grade 1 anemia was the most common AE, flu-like syndrome the most referred side-effect. PEGIFN dose reduction to 135 mg/weekly was performed in 6 patients for intolerance, anemia or leucopenia. The qHBsAg values were reported in Fig. 1B; patients who gained the SR showed higher HBsAg decrease from baseline in comparison to non-responders or relapsers. In all time-points the qHBsAg leves in SR patients were significantly higher than in nonresponders or relapsers (p < 0.001). The HBV-DNA decline was described in Fig. 1C; patients who relapsed showed similar decrease during the treatment but recurrence of viremia in the follow-up. Non-responders patients did not showed any significant decline in qHBsAg or HBV-DNA. In the treatment algorithm (Fig. 2) was reported the chance of SR in patients treated with PEG-IFN according to HBsAg and HBVDNA decline after 12 weeks of therapy. 40 patients did not have any decline in HBsAg and HBV-DNA: in this group the chance of SR was 0% (NPV on SR 100%). In 16 patients with no HBsAg decline and HBV-DNA decrease only 1 (1.4%) gained the SR. In 16 patients with HBsAg decline, the SR rate was 9.7% and 2.7% with or without HBV-DNA drop. Using the HBsAg cut-off >7500 IU/mL at 24 weeks in the 32 patients with a chance of response we have found 14 patients (43.7%) with HBsAg >7500 IU/mL, six of whom had ET response but subsequently relapsed; 18 patients (56.2%) had HBsAg level 7500 IU/mL and 15 of whom gained ET response and 10 the SR (55.5%) with only 5 relapsers. Furthermore, all 6 patients who lost the HBsAg had a level 7500 IU/mL at 24 weeks of treatment. In univariate analysis the baseline HBV-DNA<5.3 log IU/mL (OR ¼ 14.250, 95% CI 1.696e119.702, p ¼ 0.014), HBV-DNA<8 log IU/ mL (OR ¼ 8.429, 95% CI 1.419e50.071, p ¼ 0.019), the combined decline of HBsAg and HBV-DNA and qHBsAg7500 IU/mL were significantly associated with SR (OR ¼ 22.333, 95% CI 2.983e78.382, p < 0.001; OR ¼ 29.500, 95% CI 3.189e114.063, p < 0.001, respectively). In multivariate analysis were predictive on SR the combined HBsAg and HBV-DNA decline after 12 weeks (OR ¼ 35.336, 95% CI 4.668e112.226, p < 0.001) and qHBsAg7500 IU/mL (OR ¼ 51.824, 95% CI 9.629e134.144, p < 0.001) (Table 3). PEG-IFN therapy leads to higher rates of SR and HBsAg loss compared with NAs (Marcellin et al., 2009; Rijckborst et al., 2010b), but the unfortunately the baseline predictive factors in HBeAgnegative patients were difficult to apply in order to exclude patients with lower chance of response (Bonino et al., 2007): in particular have not been defined the optimal cut-off for ALT, HBV-
34
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
Fig. 1. A Virological and serological outcomes of treatment in the study population. B Kinetics of HBsAg during treatment and follow-up in patients with SR, relapse and nonresponse. C Kinetics of HBV-DNA during treatment and follow-up in patients with SR, relapse and non-response.
Fig. 2. Chances of SR according to HBsAg and HBV-DNA decline at 12 weeks of treatment.
DNA, HBsAg and age to identify patients with higher response rates to PEG-IFN (Marcellin et al., 2004). HBsAg decrease was a strong predictor of SR and HBsAg loss as observed in previous studies (Moucari et al., 2009; Rijckborst et al., 2010a). The IP-10 could be an interesting novelty tool, but has not been yet applied during PEG-IFN treatment (Papatheodoridis et al., 2014). The optimal strategy should be the early discontinuation of treatment in patients without any chance of SR. The lack of decrease of HBsAg and <2 log IU/mL of HBV-DNA allowed to identify a group of patients (19%) who may stopped the therapy because they have not likely of SR (0%, NPV ¼ 100%). These patients
were mainly affected by D genotype; this stopping rule named “PARC rule” was recently included in the european clinical guidelines (EASL, 2012). An interesting recent study by Goulis et al. (Goulis et al., 2015) performed on 107 HBeAg-negative patients with D genotype (PERSEAS study) evaluated a double stopping rule using PARC rule at 12 weeks of treatment and an additional step at 24 weeks considering a decline of HBsAg 10% from baseline: in the 24 patients with decrease 10%, 50% gained the SR, while only one of 15 (6.6%) with no decline obtained the SR. Applying this treatment algorithm the discontinuation rate changes from 17% at 12 weeks to 49% at 24 weeks.
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
35
Table 3 Univariate and multivariate analysis of baseline and on-treatment factors associated with sustained response. Factors Univariate analysis Age >30 years Gender: M vs F Liver stiffness >9.6 kPa HBV-DNA <5.3 log IU/mL HBV-DNA<8 log IU/mL qHBsAg<4.1 log IU/mL ALT>77 IU/mL YqHBsAg>0.5 log IU/mL and YHBV-DNA2 log IU/mL at 12 weeks qHBsAg7500 IU/mL at 24 weeks Multivariate analysis HBV-DNA <5.3 log IU/mL HBV-DNA<8 log IU/mL qHBsAg<4.1 log IU/mL YqHBsAg>0.5 log IU/mL and YHBV-DNA2 log IU/mL at 12 weeks qHBsAg7500 IU/mL at 24 weeks
The HBsAg decline differs significantly in patients who gained the SR in comparison to non-responder or relapse; the SR rate in our study was 13.9%, similar to SR observed in the PegBeLiver trial in the arm of 48 weeks treated patients (12% after 1 year of followup), but lower than observed in the study of Goulis et al. (Goulis et al., 2015) (23.2%) or Peng et al. (Peng et al., 2012) (31%); in this study, however, the SR was evaluated after 24 weeks of follow-up and patients had only B or C genotypes. In our previous retrospective analysis on 39 CHB patients with E genotype treated with PEG-IFN the global SR was 7.6%, but this cohort included also HBeAg-positive patients, and subjects who interrupted the treatment for side-effects or lost in the follow-up or treated with nonpegylated interferon. Surprisingly, the HBsAg loss rate was similar than reported in the study of Goulis et al. (Goulis et al., 2015) (8.3% vs 9.5%, respectively), but this finding should be accepted with caution due to follow-up time of 2 years only and the overall lower number of patients with SR. With the treatment algorithm (Fig. 2) it would be conceivable stop the PEG-IFN therapy in the 40 patients (55.5%) without any chance of SR; this rate is greater than reported in previous studies (ranging from 17% to 19% (Goulis et al., 2015; Rijckborst et al., 2012), but this finding is explicable with the applied qHBsAg cut-off >0.5 log IU/mL decline at 12 weeks of treatment that is more selective than generic “any HBsAg decline” reported in the PARC rule; we have chosen this cut-off evaluating the NPV ¼ 90% published in the study by Moucari et al. (Moucari et al., 2009) and considering that an early stop at 12 weeks should be preferred for not extend an unnecessary treatment in non-responder patients. However, the optimal HBsAg decline cut-off has not been established: for example, in the study of Peng et al. (Peng et al., 2012) a cut-off >0.6 log IU/mL at 12 weeks is predictive for SR with an OR ¼ 24.8 (95% CI ¼ 2.1e281.4, p ¼ 0.009). Performing the evaluation of HBsAg level >7500 IU/mL at 24 weeks reported by Lampertico et al. (Lampertico et al., 2011) we observed that in 18 patients with values 7500 IU/mL, 15 obtained the ET-response (83.3%) and 10 the SR (55.5%). In our multivariate analysis the combined decline at 12 weeks (>0.5 log IU/mL of qHBsAg and 2 log IU/mL of HBV-DNA) and the qHBsAg7500 IU/ mL at 24 weeks resulted as strong predictive factors for SR (OR ¼ 35.336 and 51.824 respectively). Considering the poor response to PEG-IFN observed in the E genotype, the combined stopping rules at 12 and 24 weeks can help to select the patients with higher chance of SR, convincing them to remain in therapy, and stopping the PEG-IFN in non-responders with switch to NAs. With this rationale approach of a first-line therapy using PEG-IFN and NAs in non-responders we could
OR
95% CI
P value
0.586 1.528 0.525 14.250 8.429 2.647 2.335 22.333 29.500
0.150e2.283 0.172e13.567 0.102e2.702 1.696e119.702 1.419e50.071 0.680e10.307 0.760e12.434 2.983e78.382 3.189e114.063
0.441 0.703 0.441 0.014 0.019 0.160 0.498 <0.001 <0.001
7.809 4.767 2.630 35.336 51.824
1.505e14.339 1.223e18.923 0.313e22.073 4.668e112.226 9.692e134.144
0.605 0.705 0.373 <0.001 <0.001
improve the cost-effectiveness in HBeAg-negative patients with E genotype. However, some aspects should be clarified: this study was the first homogeneous study in patients with HBeAg-negative CHB and E genotype, but is retrospective and with limited number of enrolled subjects; therefore, perspective studies are needed to confirm the role of early discontinuation using these stopping rules; besides, optimal HBsAg cut-off should be identified and validated to optimize the NPV of stopping rule and the choice of patients who may obtain the SR. Conflicts of interest The authors disclose no conflicts. Funding This study was not supported. References Boglione, L., Cusato, J., Cariti, G., Di Perri, G., D'Avolio, A., 2014a. The E genotype of hepatitis B: clinical and virological characteristics, and response to interferon. J. Infect. 69, 81e87. Boglione, L., Cusato, J., De Nicolo, A., Cariti, G., Di Perri, G., D'Avolio, A., 2014b. Role of CYP27B1þ2838 promoter polymorphism in the treatment of chronic hepatitis B HBeAg negative with PEG-interferon. J. Viral Hepat. 22, 318e327. Boglione, L., D'Avolio, A., Cariti, G., Gregori, G., Burdino, E., Baietto, L., Cusato, J., Ghisetti, V., De Rosa, F.G., Di Perri, G., 2013. Kinetics and prediction of HBsAg loss during therapy with analogues in patients affected by chronic hepatitis B HBeAg negative and genotype D. Liver Int. 33, 580e585. Bonino, F., Marcellin, P., Lau, G.K., Hadziyannis, S., Jin, R., Piratvisuth, T., Germanidis, G., Yurdaydin, C., Diago, M., Gurel, S., Lai, M.Y., Brunetto, M.R., Farci, P., Popescu, M., McCloud, P., 2007. Predicting response to peginterferon alpha-2a, lamivudine and the two combined for HBeAg-negative chronic hepatitis B. Gut 56, 699e705. Chevaliez, S., Hezode, C., Bahrami, S., Grare, M., Pawlotsky, J.M., 2013. Long-term hepatitis B surface antigen (HBsAg) kinetics during nucleoside/nucleotide analogue therapy: finite treatment duration unlikely. J. Hepatol. 58, 676e683. EASL clinical practice guidelines, 2012. Management of chronic hepatitis B virus infection. J. Hepatol. 57, 167e185. Galmozzi, E., Vigano, M., Lampertico, P., 2014. Systematic review with metaanalysis: do interferon lambda 3 polymorphisms predict the outcome of interferon-therapy in hepatitis B infection? Aliment. Pharmacol. Ther. 39, 569e578. Goulis, I., Karatapanis, S., Akriviadis, E., Deutsch, M., Dalekos, G.N., RaptopoulouGigi, M., Mimidis, K., Germanidis, G., Drakoulis, C., Triantos, C., Zintzaras, E., Bakalos, G., Papatheodoridis, G., 2015. On-treatment prediction of sustained response to peginterferon alfa-2a for HBeAg-negative chronic hepatitis B patients. Liver Int. 35, 1540e1548. ISPE, 2008. Guidelines for good pharmacoepidemiology practices (GPP). Pharmacoepidemiol Drug Saf. 17, 200e208. Krishnamoorthy, T.L., Mutimer, D., 2015. Hepatitis B: encouraging the use of interferon. Curr. Opin. Infect. Dis. 28, 557e562.
36
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
Lampertico, P., Vigano, M., Cheroni, C., Facchetti, F., Invernizzi, F., Valveri, V., Soffredini, R., Abrignani, S., De Francesco, R., Colombo, M., 2013. IL28B polymorphisms predict interferon-related hepatitis B surface antigen seroclearance in genotype D hepatitis B e antigen-negative patients with chronic hepatitis B. Hepatology 57, 890e896. Lampertico, P., Vigano, M., Di Costanzo, G., et al., 2011. A response guided approach to peginterferon alfa-2a therapy based on HBsAg levels at weeks 12 and 24 improves response rates in HBeAg-negative, genotype D chronic hepatitis B patients. Hepatology 54, 1021Ae1022A (Suppl.). Marcellin, P., Bonino, F., Lau, G.K., Farci, P., Yurdaydin, C., Piratvisuth, T., Jin, R., Gurel, S., Lu, Z.M., Wu, J., Popescu, M., Hadziyannis, S., 2009. Sustained response of hepatitis B e antigen-negative patients 3 years after treatment with peginterferon alpha-2a. Gastroenterology 136, e2161e2164, 2169-2179. Marcellin, P., Lau, G.K., Bonino, F., Farci, P., Hadziyannis, S., Jin, R., Lu, Z.M., Piratvisuth, T., Germanidis, G., Yurdaydin, C., Diago, M., Gurel, S., Lai, M.Y., Button, P., Pluck, N., 2004. Peginterferon alfa-2a alone, lamivudine alone, and the two in combination in patients with HBeAg-negative chronic hepatitis B. N. Engl. J. Med. 351, 1206e1217. Moucari, R., Mackiewicz, V., Lada, O., Ripault, M.P., Castelnau, C., MartinotPeignoux, M., Dauvergne, A., Asselah, T., Boyer, N., Bedossa, P., Valla, D., Vidaud, M., Nicolas-Chanoine, M.H., Marcellin, P., 2009. Early serum HBsAg drop: a strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. Hepatology 49, 1151e1157. Peng, C.Y., Lai, H.C., Li, Y.F., Su, W.P., Chuang, P.H., Kao, J.T., 2012. Early serum HBsAg level as a strong predictor of sustained response to peginterferon alfa-2a in
HBeAg-negative chronic hepatitis B. Aliment. Pharmacol. Ther. 35, 458e468. Papatheodoridis, G., Goulis, J., Manolakopoulos, S., Margariti, A., Exarchos, X., Kokkonis, G., Hadziyiannis, E., Papaioannou, C., Manesis, E., Pectasides, D., Akriviadis, E., 2014. J. Hepatol. 60 (1), 62e68. Rijckborst, V., Hansen, B.E., Cakaloglu, Y., Ferenci, P., Tabak, F., Akdogan, M., Simon, K., Akarca, U.S., Flisiak, R., Verhey, E., Van Vuuren, A.J., Boucher, C.A., ter Borg, M.J., Janssen, H.L., 2010a. Early on-treatment prediction of response to peginterferon alfa-2a for HBeAg-negative chronic hepatitis B using HBsAg and HBV DNA levels. Hepatology 52, 454e461. Rijckborst, V., Hansen, B.E., Ferenci, P., Brunetto, M.R., Tabak, F., Cakaloglu, Y., Lanza, A.G., Messina, V., Iannacone, C., Massetto, B., Regep, L., Colombo, M., Janssen, H.L., Lampertico, P., 2012. Validation of a stopping rule at week 12 using HBsAg and HBV DNA for HBeAg-negative patients treated with peginterferon alfa-2a. J. Hepatol. 56, 1006e1011. Rijckborst, V., ter Borg, M.J., Cakaloglu, Y., Ferenci, P., Tabak, F., Akdogan, M., Simon, K., Raptopoulou-Gigi, M., Ormeci, N., Zondervan, P.E., Verhey, E., van Vuuren, A.J., Hansen, B.E., Janssen, H.L., 2010b. A randomized trial of peginterferon alpha-2a with or without ribavirin for HBeAg-negative chronic hepatitis B. Am. J. Gastroenterol. 105, 1762e1769. Wong, D.K., Seto, W.K., Fung, J., Ip, P., Huang, F.Y., Lai, C.L., Yuen, M.F., 2013. Reduction of hepatitis B surface antigen and covalently closed circular DNA by nucleos(t)ide analogues of different potency. Clin. Gastroenterol. Hepatol. 11, 1004e1010 e1001. Zoulim, F., Perrillo, R., 2008. Hepatitis B: reflections on the current approach to antiviral therapy. J. Hepatol. 48 (Suppl. 1), S2eS19.
Contents lists available at ScienceDirect
Antiviral Research journal homepage: www.elsevier.com/locate/antiviral
Role of HBsAg decline in patients with chronic hepatitis B HBeAgnegative and E genotype treated with pegylated-interferon Lucio Boglione*, Jessica Cusato, Giuseppe Cariti, Giovanni Di Perri, Antonio D'Avolio Unit of Infectious Diseases, University of Turin, Department of Medical Sciences, Amedeo di Savoia Hospital, Turin, Italy
a r t i c l e i n f o
a b s t r a c t
Article history: Received 28 May 2016 Received in revised form 7 September 2016 Accepted 23 October 2016 Available online 26 October 2016
Treatment options for patients with chronic hepatitis B (CHB) and hepatitis B e antigen (HBeAg)-negative are pegylated interferon alfa-2a (PEG-IFN) for 48 weeks or nucleos(t)ide analogues (NAs). The choice of patients with higher chance of sustained response (SR) to PEG-IFN can be made with pre-treatment and on-treatment factors; recent studies evidenced the role of early drop of serum hepatitis B surface antigen (HBsAg) as predictor of SR. The aim of this study was the evaluation of early decrease of HBsAg on the SR in HBeAg-negative patients with E genotype. A retrospective analysis was performed on 72 patients affected by HBeAg-negative CHB with E genotype, treated for 48 weeks with PEG-IFN. HBsAg and HBV-DNA kinetics were evaluated. Decline of HBsAg (>0.5 logIU/mL) and HBV-DNA (2 logIU/mL) at 12 weeks was described according to observed chance of SR. After 96 weeks of follow-up, SR was observed in 10 patients (13.9%); HBsAg loss 6 (8.3%), HBsAg seroconversion in 3 (4.2%). No patients with HBsAg decline 0.5 log IU/mL and HBV-DNA<2 logIU/mL achieved SR (negative predictive value, NPV 100%). In multivariate analysis were significantly associated with SR the combined decline of HBsAg and HBV-DNA at 12 weeks (OR ¼ 35.336; 95% CI: 4.668e112.226; p < 0.001) and the HBsAg7500 IU/mL at 24 weeks (OR ¼ 51.824; 95% CI: 9.692e134.144; p < 0.001). Combining the HBsAg and HBV-DNA decline at 12 weeks we can identify patients without chance of SR who may stop PEG-IFN treatment. Stopping rule at 24 weeks using HBsAg7500 IU/mL is strong predictor of SR in HBeAg-negative patients with E genotype. © 2016 Elsevier B.V. All rights reserved.
Keywords: Hepatitis B treatment PEG-IFN HBV genotype qHBsAg HBeAg-negative
Current treatment options for chronic hepatitis B (CHB) are pegylated interferon alfa (PEG-IFN) or nucleos(t)ide analogues (NAs); even if the optimal endpoint is the hepatitis B surface antigen (HBsAg) loss, this result is difficult to achieve and sustained response (SR) is the most attainable goal (EASL, 2012). However, the use of PEG-IFN is limited in the hepatitis e antigen (HBeAg)-negative patients for the low rate of virological response and the sideeffects associated with the subcutaneous administration (Zoulim and Perrillo, 2008). Therefore, NAs were the most used treatments in HBeAg-negative patients, but the HBsAg decrease is very poor (Boglione et al., 2013; Chevaliez et al., 2013; Wong et al., 2013)
List of abbreviations: HBV, Hepatitis B virus; NA, nucleos(t)ide analogues; CHB, chronic hepatitis B; PEG-INF, pegylated interferon; HBsAg, hepatitis B surface antigen; ALT, alanine aminotransferase; HBeAg, hepatitis e antigen; SVR, sustained virological response; UNL, upper normal level; qHBsAg, quantitative HBsAg; IQR, inter-quartile range. * Corresponding author. E-mail address: [email protected] (L. Boglione). http://dx.doi.org/10.1016/j.antiviral.2016.10.011 0166-3542/© 2016 Elsevier B.V. All rights reserved.
and the HBsAg loss is difficult to achieve. The PEG-IFN therapy could be optimized by selection of patients with best chance of response using baseline predictive factors (Krishnamoorthy and Mutimer, 2015), while host genetic factors as interleukin (IL)28B were analyzed in some studies but without reliable results (Boglione et al., 2014b; Galmozzi et al., 2014; Lampertico et al., 2013). Other promising predictive marker of HBsAg loss seems to be the serum level of interferon-inducible protein 10 (IP10), but has been studied only during NAs therapy (Papatheodoridis et al., 2014). The reduction of HBsAg during PEGIFN has been examined as predictive factor of SR; a quantitative HBsAg (qHBsAg) decline of at least 0.5 and 1 log IU/mL after 12 and 24 weeks evidenced 90% and 97% of negative predictive value (NPV) on SR in a cohort of 48 HBeAg-negative patients with mixed genotypes (Moucari et al., 2009). Using both HBsAg and HBV-DNA decline after 12 weeks we can identify the patients without chance of response (Rijckborst et al., 2010a); the validation of this stopping rule in patients with D
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
genotype allows to treatment interruption in non-responders (Rijckborst et al., 2012); however, in the PARC trial other genotypes were present (Rijckborst et al., 2012) and also in B and C genotypes Peng et al. demonstrated the useful of this stopping rule (Peng et al., 2012). An additional stopping rule could be the HBsAg level>7500 IU/mL after 24 weeks of treatment: the SR rate in patients with HBsAg level7500 IU/mL was 16%, while 6% if HBsAg was above this cut-off (Lampertico et al., 2011). Few data are currently available about the treatment in E genotype; in our previous study we reported a poor rate of SR in this genotype (Boglione et al., 2014a). This retrospective analysis was performed in our entire cohort of patients affected by CHB with E genotype and treated with PEG-IFN at our centre from April 2008 to January 2014. Eligible patients for this study were all CHB with E genotype, age 18 years or older, HBeAg-negative, naïve for previous treatment. From an entire cohort of 122 eligible subjects, we excluded 11 patients with previous treatment, 5 treated with non-pegylated interferon, 13 with incomplete treatment (drop-out or adverse events) or follow-up for any reason, 17 for unavailability of qHBsAg during the treatment and follow-up, 4 without available informed consent: therefore 72 patients were finally included in this analysis. This study was conducted in accordance with the guidelines of the Declaration of Helsinki and the principles of good pharmacoepidemiology practices (ISPE, 2008) and was approved by our local Ethic Committee as “HBV-GEN”. All included patients were treated with PEG-IFN alpha-2a with starting dose of 180 mg/weekly for 48 weeks. The follow-up time was 48 weeks after treatment completion. The end points of therapy were established according to EASL Clinical Practice Guidelines (EASL, 2012): end-of-treatment (ET) virological response was defined as HBV-DNA HBV-DNA <2000 IU/ mL (10,000 copies/mL) at the end of therapy; the sustained response (SR) was defined as HBV-DNA <2000 IU/mL (10,000 copies/mL) at the end of follow-up. Serological response was defined as HBsAg loss with or without anti-HBs; the virological relapse was defined as HBV-DNA >2000 IU/mL (10,000 copies/mL) after previous ET virological response. Baseline characteristics of study population were described in Table 1. Virological and serological response were reported in Table 2; ET response was observed in 21 patients (29%); SR in 10 patients (13.9%); virological relapse was reported in 11 (15.%); patients who did not obtain ET or SR (non-responders) were 51 (70%). HBsAg loss was seen in 6 patients (8%) and 3 with antiHBs seroconversion (4%). In Fig. 1A were represented the virological and serological outcomes: virological response decreased from 29% at ET to 20% after 1 year and 13.9% after 2 years. Relapser rate increased from 8% to 15% after 1 and 2 years; HBsAg loss increased from 1.3% at ET from 4% to 8% after 1 and 2 years.
Table 1 Baseline characteristics of the study patients. Patients overall Gender: male, n (%) Age (yr) Route of transmission, n (%)
BMI (kg/m2) Liver stiffness (kPa) Alanine aminotransferase (ALT) [IU/mL] HBV-DNA [LogIU/mL] HVL (HBV-DNA>8 log IU/mL), n (%) qHBsAg [LogIU/mL]
N ¼ 72 62 (86.1) 30.3 [27e38] Unknown: 23 (31.9) Sexual: 9 (12.5) Familiary history: 40 (55.6) 22 [20.1e23.8] 9.6 [6.9e23.8] 77.5 [66e96.7] 5.3 [4.8e6.3] 6 (8.3) 4.1 [3.8e4.2]
33
Table 2 Treatment outcomes and adverse events. Outcomes End-of treatment response Sustained response Relapser Non-responders HBsAg loss Anti-HBsAg seroconversion
N ¼ 72 21 (29.2) 10 (13.9) 11 (15.2) 51 (70.8) 6 (8.3) 3 (4.2)
Adverse events Grade 1 leucopenia Grade 1 anemia Grade 2 anemia Flu-like syndrome Fatigue Weight loss Psychiatric symptoms (anxiety or depression) PEG-IFN alfa-2a reduction dose to 135 mg
12 (16.6) 18 (25) 8 (11.1) 28 (38.8) 22 (30.5) 14 (19.4) 4 (5.5) 6 (8.3)
In Table 2 were described the side-effects and adverse events (AE) observed during the treatment: grade 1 anemia was the most common AE, flu-like syndrome the most referred side-effect. PEGIFN dose reduction to 135 mg/weekly was performed in 6 patients for intolerance, anemia or leucopenia. The qHBsAg values were reported in Fig. 1B; patients who gained the SR showed higher HBsAg decrease from baseline in comparison to non-responders or relapsers. In all time-points the qHBsAg leves in SR patients were significantly higher than in nonresponders or relapsers (p < 0.001). The HBV-DNA decline was described in Fig. 1C; patients who relapsed showed similar decrease during the treatment but recurrence of viremia in the follow-up. Non-responders patients did not showed any significant decline in qHBsAg or HBV-DNA. In the treatment algorithm (Fig. 2) was reported the chance of SR in patients treated with PEG-IFN according to HBsAg and HBVDNA decline after 12 weeks of therapy. 40 patients did not have any decline in HBsAg and HBV-DNA: in this group the chance of SR was 0% (NPV on SR 100%). In 16 patients with no HBsAg decline and HBV-DNA decrease only 1 (1.4%) gained the SR. In 16 patients with HBsAg decline, the SR rate was 9.7% and 2.7% with or without HBV-DNA drop. Using the HBsAg cut-off >7500 IU/mL at 24 weeks in the 32 patients with a chance of response we have found 14 patients (43.7%) with HBsAg >7500 IU/mL, six of whom had ET response but subsequently relapsed; 18 patients (56.2%) had HBsAg level 7500 IU/mL and 15 of whom gained ET response and 10 the SR (55.5%) with only 5 relapsers. Furthermore, all 6 patients who lost the HBsAg had a level 7500 IU/mL at 24 weeks of treatment. In univariate analysis the baseline HBV-DNA<5.3 log IU/mL (OR ¼ 14.250, 95% CI 1.696e119.702, p ¼ 0.014), HBV-DNA<8 log IU/ mL (OR ¼ 8.429, 95% CI 1.419e50.071, p ¼ 0.019), the combined decline of HBsAg and HBV-DNA and qHBsAg7500 IU/mL were significantly associated with SR (OR ¼ 22.333, 95% CI 2.983e78.382, p < 0.001; OR ¼ 29.500, 95% CI 3.189e114.063, p < 0.001, respectively). In multivariate analysis were predictive on SR the combined HBsAg and HBV-DNA decline after 12 weeks (OR ¼ 35.336, 95% CI 4.668e112.226, p < 0.001) and qHBsAg7500 IU/mL (OR ¼ 51.824, 95% CI 9.629e134.144, p < 0.001) (Table 3). PEG-IFN therapy leads to higher rates of SR and HBsAg loss compared with NAs (Marcellin et al., 2009; Rijckborst et al., 2010b), but the unfortunately the baseline predictive factors in HBeAgnegative patients were difficult to apply in order to exclude patients with lower chance of response (Bonino et al., 2007): in particular have not been defined the optimal cut-off for ALT, HBV-
34
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
Fig. 1. A Virological and serological outcomes of treatment in the study population. B Kinetics of HBsAg during treatment and follow-up in patients with SR, relapse and nonresponse. C Kinetics of HBV-DNA during treatment and follow-up in patients with SR, relapse and non-response.
Fig. 2. Chances of SR according to HBsAg and HBV-DNA decline at 12 weeks of treatment.
DNA, HBsAg and age to identify patients with higher response rates to PEG-IFN (Marcellin et al., 2004). HBsAg decrease was a strong predictor of SR and HBsAg loss as observed in previous studies (Moucari et al., 2009; Rijckborst et al., 2010a). The IP-10 could be an interesting novelty tool, but has not been yet applied during PEG-IFN treatment (Papatheodoridis et al., 2014). The optimal strategy should be the early discontinuation of treatment in patients without any chance of SR. The lack of decrease of HBsAg and <2 log IU/mL of HBV-DNA allowed to identify a group of patients (19%) who may stopped the therapy because they have not likely of SR (0%, NPV ¼ 100%). These patients
were mainly affected by D genotype; this stopping rule named “PARC rule” was recently included in the european clinical guidelines (EASL, 2012). An interesting recent study by Goulis et al. (Goulis et al., 2015) performed on 107 HBeAg-negative patients with D genotype (PERSEAS study) evaluated a double stopping rule using PARC rule at 12 weeks of treatment and an additional step at 24 weeks considering a decline of HBsAg 10% from baseline: in the 24 patients with decrease 10%, 50% gained the SR, while only one of 15 (6.6%) with no decline obtained the SR. Applying this treatment algorithm the discontinuation rate changes from 17% at 12 weeks to 49% at 24 weeks.
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
35
Table 3 Univariate and multivariate analysis of baseline and on-treatment factors associated with sustained response. Factors Univariate analysis Age >30 years Gender: M vs F Liver stiffness >9.6 kPa HBV-DNA <5.3 log IU/mL HBV-DNA<8 log IU/mL qHBsAg<4.1 log IU/mL ALT>77 IU/mL YqHBsAg>0.5 log IU/mL and YHBV-DNA2 log IU/mL at 12 weeks qHBsAg7500 IU/mL at 24 weeks Multivariate analysis HBV-DNA <5.3 log IU/mL HBV-DNA<8 log IU/mL qHBsAg<4.1 log IU/mL YqHBsAg>0.5 log IU/mL and YHBV-DNA2 log IU/mL at 12 weeks qHBsAg7500 IU/mL at 24 weeks
The HBsAg decline differs significantly in patients who gained the SR in comparison to non-responder or relapse; the SR rate in our study was 13.9%, similar to SR observed in the PegBeLiver trial in the arm of 48 weeks treated patients (12% after 1 year of followup), but lower than observed in the study of Goulis et al. (Goulis et al., 2015) (23.2%) or Peng et al. (Peng et al., 2012) (31%); in this study, however, the SR was evaluated after 24 weeks of follow-up and patients had only B or C genotypes. In our previous retrospective analysis on 39 CHB patients with E genotype treated with PEG-IFN the global SR was 7.6%, but this cohort included also HBeAg-positive patients, and subjects who interrupted the treatment for side-effects or lost in the follow-up or treated with nonpegylated interferon. Surprisingly, the HBsAg loss rate was similar than reported in the study of Goulis et al. (Goulis et al., 2015) (8.3% vs 9.5%, respectively), but this finding should be accepted with caution due to follow-up time of 2 years only and the overall lower number of patients with SR. With the treatment algorithm (Fig. 2) it would be conceivable stop the PEG-IFN therapy in the 40 patients (55.5%) without any chance of SR; this rate is greater than reported in previous studies (ranging from 17% to 19% (Goulis et al., 2015; Rijckborst et al., 2012), but this finding is explicable with the applied qHBsAg cut-off >0.5 log IU/mL decline at 12 weeks of treatment that is more selective than generic “any HBsAg decline” reported in the PARC rule; we have chosen this cut-off evaluating the NPV ¼ 90% published in the study by Moucari et al. (Moucari et al., 2009) and considering that an early stop at 12 weeks should be preferred for not extend an unnecessary treatment in non-responder patients. However, the optimal HBsAg decline cut-off has not been established: for example, in the study of Peng et al. (Peng et al., 2012) a cut-off >0.6 log IU/mL at 12 weeks is predictive for SR with an OR ¼ 24.8 (95% CI ¼ 2.1e281.4, p ¼ 0.009). Performing the evaluation of HBsAg level >7500 IU/mL at 24 weeks reported by Lampertico et al. (Lampertico et al., 2011) we observed that in 18 patients with values 7500 IU/mL, 15 obtained the ET-response (83.3%) and 10 the SR (55.5%). In our multivariate analysis the combined decline at 12 weeks (>0.5 log IU/mL of qHBsAg and 2 log IU/mL of HBV-DNA) and the qHBsAg7500 IU/ mL at 24 weeks resulted as strong predictive factors for SR (OR ¼ 35.336 and 51.824 respectively). Considering the poor response to PEG-IFN observed in the E genotype, the combined stopping rules at 12 and 24 weeks can help to select the patients with higher chance of SR, convincing them to remain in therapy, and stopping the PEG-IFN in non-responders with switch to NAs. With this rationale approach of a first-line therapy using PEG-IFN and NAs in non-responders we could
OR
95% CI
P value
0.586 1.528 0.525 14.250 8.429 2.647 2.335 22.333 29.500
0.150e2.283 0.172e13.567 0.102e2.702 1.696e119.702 1.419e50.071 0.680e10.307 0.760e12.434 2.983e78.382 3.189e114.063
0.441 0.703 0.441 0.014 0.019 0.160 0.498 <0.001 <0.001
7.809 4.767 2.630 35.336 51.824
1.505e14.339 1.223e18.923 0.313e22.073 4.668e112.226 9.692e134.144
0.605 0.705 0.373 <0.001 <0.001
improve the cost-effectiveness in HBeAg-negative patients with E genotype. However, some aspects should be clarified: this study was the first homogeneous study in patients with HBeAg-negative CHB and E genotype, but is retrospective and with limited number of enrolled subjects; therefore, perspective studies are needed to confirm the role of early discontinuation using these stopping rules; besides, optimal HBsAg cut-off should be identified and validated to optimize the NPV of stopping rule and the choice of patients who may obtain the SR. Conflicts of interest The authors disclose no conflicts. Funding This study was not supported. References Boglione, L., Cusato, J., Cariti, G., Di Perri, G., D'Avolio, A., 2014a. The E genotype of hepatitis B: clinical and virological characteristics, and response to interferon. J. Infect. 69, 81e87. Boglione, L., Cusato, J., De Nicolo, A., Cariti, G., Di Perri, G., D'Avolio, A., 2014b. Role of CYP27B1þ2838 promoter polymorphism in the treatment of chronic hepatitis B HBeAg negative with PEG-interferon. J. Viral Hepat. 22, 318e327. Boglione, L., D'Avolio, A., Cariti, G., Gregori, G., Burdino, E., Baietto, L., Cusato, J., Ghisetti, V., De Rosa, F.G., Di Perri, G., 2013. Kinetics and prediction of HBsAg loss during therapy with analogues in patients affected by chronic hepatitis B HBeAg negative and genotype D. Liver Int. 33, 580e585. Bonino, F., Marcellin, P., Lau, G.K., Hadziyannis, S., Jin, R., Piratvisuth, T., Germanidis, G., Yurdaydin, C., Diago, M., Gurel, S., Lai, M.Y., Brunetto, M.R., Farci, P., Popescu, M., McCloud, P., 2007. Predicting response to peginterferon alpha-2a, lamivudine and the two combined for HBeAg-negative chronic hepatitis B. Gut 56, 699e705. Chevaliez, S., Hezode, C., Bahrami, S., Grare, M., Pawlotsky, J.M., 2013. Long-term hepatitis B surface antigen (HBsAg) kinetics during nucleoside/nucleotide analogue therapy: finite treatment duration unlikely. J. Hepatol. 58, 676e683. EASL clinical practice guidelines, 2012. Management of chronic hepatitis B virus infection. J. Hepatol. 57, 167e185. Galmozzi, E., Vigano, M., Lampertico, P., 2014. Systematic review with metaanalysis: do interferon lambda 3 polymorphisms predict the outcome of interferon-therapy in hepatitis B infection? Aliment. Pharmacol. Ther. 39, 569e578. Goulis, I., Karatapanis, S., Akriviadis, E., Deutsch, M., Dalekos, G.N., RaptopoulouGigi, M., Mimidis, K., Germanidis, G., Drakoulis, C., Triantos, C., Zintzaras, E., Bakalos, G., Papatheodoridis, G., 2015. On-treatment prediction of sustained response to peginterferon alfa-2a for HBeAg-negative chronic hepatitis B patients. Liver Int. 35, 1540e1548. ISPE, 2008. Guidelines for good pharmacoepidemiology practices (GPP). Pharmacoepidemiol Drug Saf. 17, 200e208. Krishnamoorthy, T.L., Mutimer, D., 2015. Hepatitis B: encouraging the use of interferon. Curr. Opin. Infect. Dis. 28, 557e562.
36
L. Boglione et al. / Antiviral Research 136 (2016) 32e36
Lampertico, P., Vigano, M., Cheroni, C., Facchetti, F., Invernizzi, F., Valveri, V., Soffredini, R., Abrignani, S., De Francesco, R., Colombo, M., 2013. IL28B polymorphisms predict interferon-related hepatitis B surface antigen seroclearance in genotype D hepatitis B e antigen-negative patients with chronic hepatitis B. Hepatology 57, 890e896. Lampertico, P., Vigano, M., Di Costanzo, G., et al., 2011. A response guided approach to peginterferon alfa-2a therapy based on HBsAg levels at weeks 12 and 24 improves response rates in HBeAg-negative, genotype D chronic hepatitis B patients. Hepatology 54, 1021Ae1022A (Suppl.). Marcellin, P., Bonino, F., Lau, G.K., Farci, P., Yurdaydin, C., Piratvisuth, T., Jin, R., Gurel, S., Lu, Z.M., Wu, J., Popescu, M., Hadziyannis, S., 2009. Sustained response of hepatitis B e antigen-negative patients 3 years after treatment with peginterferon alpha-2a. Gastroenterology 136, e2161e2164, 2169-2179. Marcellin, P., Lau, G.K., Bonino, F., Farci, P., Hadziyannis, S., Jin, R., Lu, Z.M., Piratvisuth, T., Germanidis, G., Yurdaydin, C., Diago, M., Gurel, S., Lai, M.Y., Button, P., Pluck, N., 2004. Peginterferon alfa-2a alone, lamivudine alone, and the two in combination in patients with HBeAg-negative chronic hepatitis B. N. Engl. J. Med. 351, 1206e1217. Moucari, R., Mackiewicz, V., Lada, O., Ripault, M.P., Castelnau, C., MartinotPeignoux, M., Dauvergne, A., Asselah, T., Boyer, N., Bedossa, P., Valla, D., Vidaud, M., Nicolas-Chanoine, M.H., Marcellin, P., 2009. Early serum HBsAg drop: a strong predictor of sustained virological response to pegylated interferon alfa-2a in HBeAg-negative patients. Hepatology 49, 1151e1157. Peng, C.Y., Lai, H.C., Li, Y.F., Su, W.P., Chuang, P.H., Kao, J.T., 2012. Early serum HBsAg level as a strong predictor of sustained response to peginterferon alfa-2a in
HBeAg-negative chronic hepatitis B. Aliment. Pharmacol. Ther. 35, 458e468. Papatheodoridis, G., Goulis, J., Manolakopoulos, S., Margariti, A., Exarchos, X., Kokkonis, G., Hadziyiannis, E., Papaioannou, C., Manesis, E., Pectasides, D., Akriviadis, E., 2014. J. Hepatol. 60 (1), 62e68. Rijckborst, V., Hansen, B.E., Cakaloglu, Y., Ferenci, P., Tabak, F., Akdogan, M., Simon, K., Akarca, U.S., Flisiak, R., Verhey, E., Van Vuuren, A.J., Boucher, C.A., ter Borg, M.J., Janssen, H.L., 2010a. Early on-treatment prediction of response to peginterferon alfa-2a for HBeAg-negative chronic hepatitis B using HBsAg and HBV DNA levels. Hepatology 52, 454e461. Rijckborst, V., Hansen, B.E., Ferenci, P., Brunetto, M.R., Tabak, F., Cakaloglu, Y., Lanza, A.G., Messina, V., Iannacone, C., Massetto, B., Regep, L., Colombo, M., Janssen, H.L., Lampertico, P., 2012. Validation of a stopping rule at week 12 using HBsAg and HBV DNA for HBeAg-negative patients treated with peginterferon alfa-2a. J. Hepatol. 56, 1006e1011. Rijckborst, V., ter Borg, M.J., Cakaloglu, Y., Ferenci, P., Tabak, F., Akdogan, M., Simon, K., Raptopoulou-Gigi, M., Ormeci, N., Zondervan, P.E., Verhey, E., van Vuuren, A.J., Hansen, B.E., Janssen, H.L., 2010b. A randomized trial of peginterferon alpha-2a with or without ribavirin for HBeAg-negative chronic hepatitis B. Am. J. Gastroenterol. 105, 1762e1769. Wong, D.K., Seto, W.K., Fung, J., Ip, P., Huang, F.Y., Lai, C.L., Yuen, M.F., 2013. Reduction of hepatitis B surface antigen and covalently closed circular DNA by nucleos(t)ide analogues of different potency. Clin. Gastroenterol. Hepatol. 11, 1004e1010 e1001. Zoulim, F., Perrillo, R., 2008. Hepatitis B: reflections on the current approach to antiviral therapy. J. Hepatol. 48 (Suppl. 1), S2eS19.