Evolution of Hepatitis B Serological Markers in HIV-Infected Patients Receiving Highly Active Antiretroviral Therapy

MAJOR ARTICLE HIV/AIDS Evolution of Hepatitis B Serological Markers in HIV-Infected Patients Receiving Highly Active Antiretroviral Therapy Wang-Huei Sheng, 1,3,6,a Jia-Horng Kao, 5,a Pei-Jer Chen, 5 Li-Ming Huang, 2 Sui-Yuan Chang, 4 Hsin-Yun Sun, 1,3 Chien-Ching Hung, 1,3 Mao-Yuan Chen, 1,3 and Shan-Chwen Chang 1,3 Departments of 1 Internal Medicine and 2 Pediatrics, National Taiwan University Hospital, Departments of 3 Internal Medicine and 4 Clinical Laboratory Sciences and Medical Biotechnology and 5 Graduate Institutes of Clinical Medicine, National Taiwan University College of Medicine, and 6 Department of Internal Medicine, Far-East Memorial Hospital, Taipei, Taiwan Background. Evolution of serological markers of hepatitis B virus (HBV) carriage or infection has rarely been investigated among human immunodeficiency virus (HIV) infected patients receiving highly active antiretroviral therapy (HAART). Methods. During the period 1997 2002, a total of 633 HIV-infected patients were tested for HBV serological markers at baseline, including hepatitis B surface antigen (HBsAg), antibody to HBsAg (anti-hbs ), antibody to hepatitis B core antigen (anti-hbc), hepatitis C virus (HCV) antibody (anti-hcv) antibody, HCV RNA level, and HBV DNA level, all of which were retested at least 1 year apart. Medical records were reviewed to identify clinical characteristics associated with evolution of these serological markers. Results. After a median duration of follow-up for 4.96 years, 161 patients (25.4%) had changes in HBV serological markers. Of 119 patients (18.8%) who tested positive for HBsAg at baseline, 6 (5.0%) developed anti- HBs, and 9 (7.6%) developed isolated anti-hbc. Of 270 patients (42.7%) who tested positive for anti-hbs, 18 (6.7%) lost anti-hbs. Of 179 patients (28.3%) in whom isolated anti-hbc had been detected, 73 (40.8%) developed anti-hbs, 18 (10.1%) lost all HBV markers, and 7 (3.9%) developed HBsAg. Of 65 patients (10.2%) who tested negative for all HBV markers, 13 (20%) developed anti-hbs, 13 (20%) developed isolated anti-hbc, and 4 (6.2%) developed HBsAg, indicating a high risk of HBV exposure. Patients in whom anti-hbc was detected at baseline were more likely to have acquired immunodeficiency syndrome ( P p.008). Multivariate analysis revealed that an increase in the CD4 cell count after the commencement of HAART was significantly associated with persistence or subsequent development of anti-hbs in patients with anti-hbs or anti-hbc at baseline, respectively. Conclusions. Periodic measurements of HBV serological markers in HIV-infected patients are recommended, because new HBV infections and changes of HBV serological markers are not uncommon in patients with improved immunity after commencement of HAART. Coinfection with hepatitis B virus (HBV) and HIV is common, because both viruses share similar transmission routes [1]. Chronic HBV infection has been shown Received 20 March 2007; accepted 13 July 2007; electronically published 21 September 2007. Presented in part: 14th Conference on Retroviruses and Opportunistic Infections, Los Angeles, California 25 28 February 2007 (abstract 941). a W.-H.S. and J.H.K. contributed equally to this article. Reprints or correspondence: Dr. Chien-Ching Hung, Dept. of Internal Medicine, National Taiwan University Hospital, 7 Chung-Shan South Rd., Taipei, Taiwan (hcc0401@ntu.edu.tw). Clinical Infectious Diseases 2007; 45:1221 9 2007 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2007/4509-0017$15.00 DOI: 10.1086/522173 to increase risk of virological failure of highly active antiretroviral therapy (HAART), development of acute hepatitis, hepatic decompensation, and liver-related mortality after initiation of HAART [2 4]. However, most studies have recognized HBV serological status as an invariant parameter (e.g., hepatitis B surface [HBs] antigenemia), without taking into account its evolution over time. In fact, HBV serological markers may change according to postinfection course, which is determined by host immunity and viral activities [5 10]. Among HIV-infected patients, several different HBV serological patterns may be encountered because of the patient s immunosuppression and increased risk of exposure to HBV. For example, presence of isolated HIV/AIDS CID 2007:45 (1 November) 1221

antibody to hepatitis B core antigen (anti-hbc) and persistent chronic HBs antigenemia have been observed more often in HIV-infected patients because of impaired host immunity against HBV [5, 11 13]. However, factors associated with evolution of serological markers of HBV in HIV-infected patients receiving HAART remain unclear. With HAART, immune reconstitution may increase the frequency of HBV reactivation in hepatitis B surface antigen (HBsAg) carriers or the development of antibody to HBsAg (anti-hbs) [14, 15]. Previous studies that assessed the evolution of HBV serological patterns in HIV-infected patients were limited by the inclusion of only injection drug users [5] or patients who were not receiving HAART [6 10] or by a small number of patients in an area with a low prevalence of HBV infection [11]. In this longitudinal follow-up study, we aimed to assess the evolution of HBV serological markers in HIV-infected patients receiving HAART in Taiwan, where the prevalence of HBV coinfection in HIVinfected patients is much higher than that in Western countries (22% vs. 6% 10%) [2, 3]. PATIENTS AND METHODS Patients. During the period from 1 January 1997 through 31 December, 2002, HIV-infected patients who were antiretroviral naive and underwent serological testing for HBV infection before the initiation of HAART at the National Taiwan University Hospital (Taipei) were eligible for this longitudinal follow-up study. Patients who underwent repeated HBV serological testing in 2003 and 2004, with tests taken 1 year apart, were included in the analysis. To better understand the evolution of HBV serological markers, patients who had a known history of HBV vaccination were excluded especially patients born after 1984, when universal HBV vaccination was launched in Taiwan [16]. Determination of HBV and hepatitis C virus (HCV) serological markers and data collection. Demographic, clinical, serological, and virological data were recorded using a standardized data collection form. Patients without any HBV serological markers noted in consecutive tests were encouraged to undergo HBV vaccination. At the end of the study, 20 of 35 patients without any HBV serological markers subsequently received HBV vaccine and, therefore, were censored on the date of HBV vaccination. HBsAg, anti-hbs, and anti-hbc were determined using an EIA (Abbott Laboratories). All HBV serological tests were repeated. If discrepancies of results occurred between the 2 tests, a third test was performed. Antibodies to HCV (anti-hcv) were determined using a third-generation EIA (AxSYM HCV III; Abbott Laboratories). Quantification of HBV DNA level, HIV RNA level, and CD4 lymphocyte count. HBV DNA was extracted from 200 mlof serum obtained from all patients at enrollment and from patients who tested positive for HBs antigenemia or isolated anti- HBc in subsequent tests (High Pure Viral Nucleic Acid Kit; Roche Molecular Biochemicals). The HBV DNA level was quantified using a real-time PCR with the LighterCycler instrument (Roche Molecular Biochemicals), in accordance with the manufacturer s instructions. The detection limit of HBV DNA was estimated to be 10 3 copies/ml. The plasma HIV RNA level was determined using a commercial kit (Roche Amplicor, version 1.5), which has a detection limit of 400 copies/ml (2.60 log 10 copies/ml). CD4 cell counts were determined using FACFlow (BD FACS Calibur; Becton Dickinson). Both the plasma HIV RNA level and the CD4 cell count were determined at enrollment, 1 month after initiation of HAART, and every 3 4 months thereafter. Definitions of isolated anti-hbc, virological and immunologic responses to antiretroviral therapy, and progression of HIV infection. Isolated anti-hbc was defined as an instance in which blood samples tested positive for anti-hbc but negative for both HBsAg and anti-hbs. Virological response to HAART was assessed by the proportion of patients who achieved an undetectable plasma HIV RNA level at 12 months, 24 months, and the last available assessment of the plasma HIV RNA level data. The analysis was performed on the basis of the on-treat principle. Virological failure was defined as failure to achieve an undetectable plasma HIV RNA level 6 months after the initiation of HAART or as a repeatedly detectable plasma HIV RNA level after initial suppression. Immunologic response was assessed by the increase in the CD4 cell count from baseline at 12 months, 24 months, and the last available assessment of CD4 cell count data and by the proportion of patients who achieved an increase in the CD4 cell count of 100 and 200 cells/ml or more during the follow-up period. Progression of HIV infection was defined as a relapse or new occurrence of an AIDS-defining opportunistic illness [17] 1 month after the initiation of HAART. Statistical analysis. All statistical analyses were performed using SPSS software, version 13.0 (SPSS). Categorical variables were compared using the x 2 test or Fisher s exact test, and noncategorical variables were compared using the Wilcoxon rank sum test. Point estimation for Poisson distribution was used for estimating the incidence and 95% CIs of changes of HBV serological markers. Multivariate analysis with the multiple logistic regression method and a stepwise forward and backward model was used to assess the impact of the variables (e.g., age, sex, risk behavior for HIV transmission, HCV coinfection, baseline CD4 cell count and plasma HIV RNA level, presence of AIDS-related diseases, use and duration of lamivudine and HAART, and CD4 cell count increase and undetectable plasma HIV RNA level after initiation of HAART) on the persistence or subsequent development of anti-hbs among patients with HBsAg, anti-hbs, and isolated anti-hbc detected at baseline. ORs and 95% CIs were calculated for logistic regression analyses. All tests were 2-tailed, and P values!.05 were 1222 CID 2007:45 (1 November) HIV/AIDS

considered to be statistically significant. Survival probabilities were estimated by the Kaplan-Meier method. The duration of observation for each patient was estimated from the date of the first HBV serological test to the date of the last HBV serological test. Equality of survival distributions was evaluated by the log-rank test. The duration of survival for patients was estimated from the date of enrollment to death, to last followup visit at this hospital and other designated hospitals in Taiwan, or to the end of this study (31 December 2005). RESULTS Characteristics of the population studied. During the period from January 1997 through December 2002, a total of 975 patients sought care for HIV infection at National Taiwan University Hospital; 342 patients were excluded from the study (figure 1), and 633 patients (583 male and 50 female patients) were enrolled. There were no significant differences in demographic characteristics and baseline clinical, virological, and immunologic data between patients who were enrolled and those who were excluded (data not shown). Sixty-three (18.4%) of the 342 excluded patients had tested positive for HBsAg. Demographic, virological, and immunologic characteristics of the study patients are shown in table 1. At baseline, 568 patients (89.7%) tested positive for at least 1 serological marker of HBV infection: 119 patients (18.8%) had detectable HBsAg, 270 (42.7%) had detectable anti-hbs, and 179 (28.3%) had isolated anti-hbc; 65 patients (10.2%) did not have any serological marker of HBV infection detected. Hepatitis B envelope antigen was present in 26.9% of patients with HBsAg (table 1). At baseline, the HBV DNA level was detectable in 80 patients (67.2%) with HBsAg present, in 2 patients (0.7%) with anti-hbs present, in 13 patients (7.3%) with isolated anti-hbc present, and in 3 patients (4.6%) without any HBV markers. Compared with patients who did not have any HBV markers, there were no significant baseline differences in terms of age, risk for HIV transmission, HCV seropositivity, and duration of HAART and lamivudine use in patients in the other 3 groups (table 1). However, there were more male patients who had any HBV serological markers ( P p.005.08), patients with isolated anti-hbc present had a greater level of immunosuppression, and patients with HBsAg present were more likely to have detectable HBV DNA ( P!.001). The statistical results were unchanged after we excluded the 3 patients with detectable HBV DNA levels from the group of patients without any markers. HBV serological markers and patient outcomes. The patients were observed for a median duration of 4.96 years (range, 1.07 7.97 years). After initiation of HAART, patients with HBsAg present and those with isolated anti-hbc present had less robust increases in the CD4 cell count, compared with patients who did not have any HBV serological markers (at 24 months after initiation of HAART, the values were 140 and 117 cells/ml vs. 180 cells/ml, respectively; P p.03 and.04, respectively), and were less likely to achieve an undetectable plasma HIV RNA level (at 24 months after initiation of HAART, the proportions of patients were 69.8% and 75.6% vs. 89.1%, respectively; P p.006 and.03, respectively). Moreover, patients with HBsAg present were more likely to develop virological failure (42.9% vs. 27.7%; P p.04) and to die (22.7% vs. 7.7%; P p.01) than were patients without any HBV serological markers. Evolution of HBV serological markers. A total of 161 patients Figure 1. Evolution of serological markers of hepatitis B virus (HBV) among HIV-infected patients enrolled during 1997 and 2002. anti-hbc, antibody to hepatitis B core antigen; anti-hbs, antibody to hepatitis B surface antigen; HBs, hepatitis B surface; HBsAg, hepatitis B surface antigen. HIV/AIDS CID 2007:45 (1 November) 1223

Table 1. Demographic, immunological, and virological characteristics of the HIV-infected patients who underwent hepatitis B virus (HBV) serological testing. Characteristic Group 1 (n p 119) Group 2 (n p 270) Group 3 (n p 179) Group 4 (n p 65) All patients (n p 633) P a Group 1 Group 2 Group 3 Age, years 34 (20 62) 33 (20 78) 36 (19 81) 33 (20 71) 34 (19 81).60.61.11 Male sex 114 (95.8) 252 (93.3) 163 (91.6) 54 (83.1) 583 (92.1).005.008.08 Risk group for HIV transmission Men who have sex with men 77 (64.7) 177 (65.6) 108 (60.3) 39 (60.0) 405 (64.0).53.40.96 Heterosexual persons 39 (32.8) 84 (31.1) 67 (37.5) 22 (33.8) 188 (29.7).88.66.61 Injection drug users 1 (0.8) 6 (2.2) 2 (1.1) 2 (3.1) 26 (4.1).29.69.29 HCV coinfection 9 (7.6) 26 (9.6) 15 (8.4) 4 (6.2) 54 (8.5).72.47.78 Baseline CD4 cell count, cells/ml 125 (2 1158) 155 (1 1202) 93 (0 877) 147 (3 1280) 132 (0 1280).57.82.04 CD4 cell count!200 cells/ml 68 (57.1) 149 (55.2) 121 (68.0) 35 (53.9) 373 (58.9).77.85.048 Baseline plasma HIV RNA level, b log 10 copies/ml 4.99 (3.27 6.21) 5.09 (3.55 6.51) 5.08 (3.24 6.72) 5.05 (3.36 6.57) 5.07 (3.24 6.72).26.39.19 Detectable HBV DNA level at baseline 80 (67.2) 2 (0.7) 13 (7.3) 3 (4.6) 98 (15.5)!.001.11.46 Baseline HBeAg presence 32 (26.9) 0 (0) 2 (1.1) 1 (1.5) 35 (5.5)!.001.61 Receipt of HAART containing lamivudine 113 (96.6) 260 (97.3) 161 (96.4) 61 (96.8) 585 (96.9).93.81.88 Duration of lamivudine use, years 4.59 (1.00 7.71) 4.39 (1.01 7.69) 4.53 (1.01 7.68) 4.19 (1.09 7.62) 4.45 (1.00 7.71).29.43.24 Duration of HAART use, years 4.80 (1.02 7.72) 4.81 (1.01 7.74) 4.75 (1.17 7.70) 4.74 (1.09 7.67) 4.77 (1.01 7.74).71.69.87 Laboratory data 12 months after initiation of HAART Increase in CD4 cell count, cells/ml 75 ( 428 to 418) 87 ( 647 to 1025) 105 ( 707 to 548) 63 ( 182 to 532) 91 ( 707 to 1025).52.27.15 CD4 cell count increase of 100 cells/ml 56 (47.1) 121 (44.8) 90 (50.3) 29 (44.6) 296 (46.8).75.97.43 Undetectable plasma HIV RNA level 74 (62.2) 202 (74.8) 128 (71.5) 47 (72.3) 451 (71.2).17.68.90 Laboratory data 24 months after initiation of HAART c Increase in CD4 cell count, cells/ml 140 ( 372 to 736) 183 ( 873 to 983) 117 ( 590 to 751) 180 ( 382 to 699) 156 ( 873 to 983).03.54.04 CD4 cell count increase of 100 cells/ml 68 (64.2) 163 (66.5) 88 (52.4) 41 (74.5) 360 (62.7).22.25.004 CD4 cell count increase of 200 cells/ml 37 (34.9) 122 (49.8) 62 (36.9) 26 (47.3) 247 (43.0).13.74.17 Undetectable plasma HIV RNA level 74 (69.8) 180 (73.5) 127 (75.6) 49 (89.1) 430 (74.9).006.01.03 Laboratory data at the end of the study Increase in CD4 cell count, cells/ml 147 ( 428 to 736) 183 ( 873 to 983) 170 ( 707 to 783) 151 ( 356 to 699) 168 ( 873 to 983).34.47.79 CD4 cell count increase of 100 cells/ml 76 (63.9) 169 (62.6) 91 (50.8) 41 (63.1) 377 (59.6).92.94.09 CD4 cell count increase of 200 cells/ml 40 (33.6) 124 (45.9) 63 (35.2) 31 (47.7) 258 (40.8).06.80.08 Undetectable plasma HIV RNA level 75 (63.0) 194 (71.9) 123 (68.7) 49 (75.4) 441 (69.7).09.57.31 Duration of observation, years 5.18 (1.12 7.78) 4.85 (1.07 7.81) 5.05 (1.21 7.94) 4.76 (1.27 7.97) 4.96 (1.07 7.97).13.43.22 Any virological failure 51 (42.9) 88 (32.6) 65 (36.3) 18 (27.7) 222 (35.1).04.45.21 Death d 27 (22.7) 18 (6.7) 25 (14.0) 5 (7.7) 75 (11.8).01.77.19 NOTE. Data are no. (%) of patients or median (range), unless otherwise indicated. Group 1 consisted of patients positive for hepatitis B surface antigen (HBsAg); group 2 consisted of patients positive for antibody to HBsAg ; group 3 consisted of patients positive for isolated antibody to hepatitis B core antigen; and group 4 consisted of patients who tested negative for HBV serological markers. HBeAg, hepatitis B envelope antigen; HCV, hepatitis C virus. a For comparison with group 4. b Plasma HIV RNA level was available for 98, 229, 148, and 51 patients in groups 1 4, respectively. c CD4 cell counts and plasma HIV RNA level were available for 106, 245, 168, and 55 patients in groups 1 4, respectively, at 24 months after the initiation of HAART. d There were 8, 1, 2, and 0 liver-related deaths among patients in groups 1 4. respectively. The causes of deaths of the 8 patients with HBsAg were cirrhosis of liver and hepatic failure ( n p 5), hepatocellular carcinoma ( n p 2), and fulminant hepatitis ( n p 1).

(25.4%) had changes of their HBV serological markers during follow-up (figure 1). Of the 119 patients with HBsAg present at baseline, 104 (87.4%) had persistent HBs antigenemia after follow-up for 5.18 years (range, 1.12 7.78 years); 6 patients (5.0%) became positive for anti-hbs, and 9 patients (7.6%) lost HBsAg and retained isolated anti-hbc. The incidence of HBsAg clearance and development of anti-hbs was 2.57 cases per 100 person-years (95% CI, 1.44 4.23 cases per 100 personyears) and 0.97 cases per 100 person-years (95% CI, 0.36 2.12 cases per 100 person-years), respectively. Of the 270 patients with anti-hbs present at baseline, 252 (93.3%) had persistent anti-hbs after follow-up for 4.85 years (range, 1.07 7.81 years); 18 patients (6.7%) lost anti-hbs, including 15 patients (5.6%) who developed isolated anti-hbc and 3 patients (1.1%) who lost all serological markers of HBV infection. Of the 179 patients with isolated anti-hbc at baseline, 81 (45.3%) had persistence of isolated anti-hbc, 73 (40.8%) subsequently became positive for anti-hbs, 18 (10.0%) lost all HBV markers, and 7 (3.9%) developed HBsAg after follow-up for 5.05 years (range, 1.21 7.94 years). The incidence of development of HBsAg was 0.77 cases per 100 person-years (95% CI, 0.31 1.58 per 100 person-years). Of the 65 patients without any HBV markers at baseline, the median follow-up duration was 4.76 years (range, 1.27 7.97 years). Thirty patients (46.2%) without any HBV markers at baseline were later found to be ever exposed to HBV: 4 (6.2%) developed HBsAg, 13 (20%) anti-hbs, and 13 (20%) isolated anti-hbc. The incidence of new HBV infection in HIV-infected patients without any HBV markers at baseline was 9.69 per 100 person-years (95% CI, 6.28 13.30 per 100 person-years). Outcome analysis of patients with and without changes in serological markers. Among patients with HBsAg present at baseline, the characteristics of 104 patients with persistent HBs antigenemia and 6 patients with subsequent development of anti-hbs were compared. There were no significant differences with regard to age, sex, risk factor for HIV transmission, baseline CD4 cell count and plasma HIV RNA level, increase in the CD4 cell count during HAART, and virological response to HAART between these 2 groups (data not shown). Among the patients with anti-hbs present at baseline, patients who subsequently lost anti-hbs during HAART had less robust increases in the CD4 cell count after the initiation of HAART, compared with patients who had persistent anti-hbs (40 vs. 93 cells/ml at 12 months [ P!.001], 73 vs. 200 cells/ml at 24 months [ P!.001], and 73 vs. 187 cells/ml at the end of study [ P!.001]) (table 2); in addition, these patients were less likely to achieve an undetectable plasma HIV RNA level after the initiation of HAART (44.4% vs. 77.0% at 12 months [ P p.002] and 44.4% vs. 73.8% at the end of the study [ P p.007]) and had a higher mortality rate (27.8% vs. 5.2%; P p.004). Twenty-four months after the initiation of HAART, an increase in the CD4 cell count of 100 cells/ml was the only independent factor associated with persistence of anti-hbs (adjusted OR, 5.02; 95% CI, 1.31 19.24; P p.02); male sex (adjusted OR, 2.34; 95% CI, 0.92 10.41; P p.06) and initial plasma HIV RNA level 5 log 10 copies/ml (adjusted OR, 0.29; 95% CI, 0.08 1.12; P p.07) were of borderline significance. Among the patients with isolated anti-hbc present at baseline, those who subsequently developed anti-hbs had a significantly greater increase in the median CD4 cell count, compared with patients who had persistence of isolated anti-hbc (223 vs. 83 cells/ml at 24 months after initiation of HAART [ P p.002] and 216 vs. 92 cells/ml at the end of study [ P p.004]); in addition, these patients had a higher likelihood of achieving an undetectable plasma HIV RNA level at the end of study (82.2% vs. 68.8%; P p.04) (table 3). Twenty-four months after the initiation of HAART, an increase in the CD4 cell count of 100 cells/ml was the only independent factor associated with the development of anti-hbs (adjusted OR, 4.65; 95% CI, 1.96 11.02; P p.001). Survival estimates among the 4 groups of patients are shown in figure 2A. Patients with detected HBsAg had a shorter duration of survival than did patients in the other 3 groups. Patients with isolated anti-hbc detected at baseline had a shorter duration of survival than did patients with anti-hbs present ( P p.02). Patients who had persistent HBsAg during followup had a worse duration of survival than did patients who had persistent anti-hbs ( P!.001), those who had persistent isolated anti-hbc ( P p.07), and those who subsequently developed anti-hbs ( P p.12) (figure 2B). DISCUSSION In this study, the prevalence of any exposure to HBV, which was defined as the presence of any serological marker of HBV infection (89.8%) and as chronic HBV infection (16.4%), in HIV-infected patients was similar to that of other Taiwanese adult populations, in which 90% of persons had exposure to HBV and 15% 20% were chronic carriers of HBV [18]. The higher HBV seroprevalence could be explained by the fact that, in Taiwan, most cases of HBV infection occur during the perinatal period or early childhood, before the implementation of the nationwide HBV vaccination program [18, 19]. The proportion of the patients in our study who had changes in HBV serological patterns (25.4%) was significantly higher than the proportions reported in other studies (12% and 16%, with median durations of follow-up of 2.2 years and 21 months, respectively) [5, 11]. Such changes of HBV serological markers after initiation of HAART mainly involved anti-hbs development in patients who had isolated anti-hbc and acquisition of HBV infection in those who had no HBV serological markers at baseline (45.3% and 18.6% of the 161 patients with changes of serological markers, respectively). The HBsAg clearance rate HIV/AIDS CID 2007:45 (1 November) 1225

Table 2. Comparison of patients with persistent antibody to hepatitis B surface antigen (anti-hbs) and those who subsequently lost anti-hbs among patients who initially tested positive for anti-hbs. Characteristics Patients with persistent anti-hbs (n p 252) Patients who lost anti-hbs (n p 18) P HCV coinfection 24 (9.5) 2 (11.1).69 Baseline CD4 cell count Median cells/ml (range) 157 (1 1202) 114 (3 868).27!200 cells/ml 141 (56.0) 8 (44.4).34 Baseline plasma HIV RNA level, a median log 10 copies/ml (range) 5.09 (3.55 6.51) 5.05 (3.75 6.19).49 HAART duration, median years (range) 4.81 (1.01 7.74) 4.53 (1.04 7.68).52 Receipt of lamivudine-containing HAART 243 (96.4) 17 (94.4).67 Duration of lamivudine use, median years (range) 4.43 (1.01 7.69) 4.35 (1.04 7.68).69 Laboratory data 12 months after initiation of HAART Increase in the CD4 cell count, median cells/ml (range) 93 ( 647 to 1025) 40 ( 441 to 469)!.001 Increase in the CD4 cell count 100 cells/ml 119 (47.2) 2 (11.1)!.001 Undetectable plasma HIV RNA level 194 (77.0) 8 (44.4).002 Laboratory data 24 months after initiation of HAART b Increase in the CD4 cell count, median cells/ml (range) 200 ( 873 to 983) 73 ( 400 to 184)!.001 Increase in the CD4 cell count 100 cells/ml 159 (68.5) 4 (30.8).01 Increase in the CD4 cell count 200 cells/ml 120 (51.7) 2 (15.4).02 Undetectable plasma HIV RNA level 172 (74.1) 8 (61.5).34 Laboratory data at the end of the study Increase in the CD4 cell count, median cells/ml (range) 187 ( 873 to 983) 73 ( 400 to 324)!.001 Increase in the CD4 cell count 100 cells/ml 164 (65.1) 5 (27.8).002 Increase in the CD4 cell count 200 cells/ml 122 (48.4) 2 (11.1).002 Undetectable plasma HIV RNA level 186 (73.8) 8 (44.4).007 Duration of observation duration, median years (range) 4.84 (1.07 7.81) 4.94 (1.11 7.71).65 Death 13 (5.2) 5 (27.8).004 NOTE. Data are no. (%) of patients, unless otherwise indicated. HCV, hepatitis C virus. a Plasma HIV RNA level was available for 213 and 16 patients with persistent anti-hbs and patients with subsequently undetectable anti-hbs antibodies, respectively. b CD4 cell count and plasma HIV RNA level were available for 232 and 13 patients with persistent anti-hbs and patients with subsequently undetectable anti- HBs antibodies, respectively, at 24 months after the initiation of HAART. among our patients (2.57 cases per 100 person-years) was low, even after immune reconstitution associated with HAART that contained lamivudine. The low HBsAg clearance rate is not surprising, because the overall incidence of HBsAg clearance has been low even in HIV-uninfected HBV-carrying adults [20]. Furthermore, the cumulative resistance rate of HBV to lamivudine is high in HIV-infected patients [21]. Despite recent advances in anti-hbv therapy, the seroconversion rate remains low, regardless of whether the agents are used in combination [22, 23]. Loss of hepatitis B immunity occurs in association with HIV infection [5, 10, 24]. Among HIV-infected patients who were not receiving HAART, loss of anti-hbs occurred in 13% 17% of patients with hemophilia [24] and 28% of men who have sex with men [25], compared with 0% of HIV-uninfected patients during observation periods of 2 and 3 years. Achieving an undetectable plasma HIV RNA level and greater gains in the CD4 cell count in association with HAART enables HIV- HBV coinfected patients to achieve HBsAg clearance, anti-hbe seroconversion, and suppression of HBV replication [25]. In this study, 193% of such patients who received HAART continued to have detectable anti-hbs, despite severe immunosuppression; these patients tended to have better virological and immunologic responses to HAART than did patients who subsequently lost anti-hbs. These data suggest that preservation of anti-hbs requires better adherence to HAART to achieve better immune reconstitution. Our results showed that determination of only HBsAg and anti-hbs data may miss patients who had isolated anti-hbc. In such patients in whom isolated anti-hbc is detected, determination of the HBV DNA level is important to clarify the status of their HBV infection [26, 27]. By determination of the HBV DNA level, we found that occult or window-period HBV infection occurred in 7.3% of the patients in our study who had isolated anti-hbc detected and in 4.6% of patients who did not have any HBV serological markers. The prevalence of 1226 CID 2007:45 (1 November) HIV/AIDS

Table 3. Comparison of patients with persistent isolated antibodies to hepatitis B core antigen (anti-hbc) and those who subsequently developed detectable anti-hbs among patients with initial isolated anti-hbc. Characteristics Patients who developed anti-hbs (n p 73) Patients with persistent isolated anti-hbc (n p 81) P HCV coinfection 6 (8.2) 7 (8.8).91 Baseline CD4 cell count Median cells/ml (range) 119 (2 877) 88 (0 786).12!200 cells/ml 49 (67.1) 55 (67.9).91 Baseline plasma HIV RNA level, a median log 10 copies/ml (range) 5.06 (3.24 6.68) 5.11 (3.27 6.72).13 HAART duration, median years (range) 4.78 (1.17 7.64) 4.55 (1.21 7.70).37 Receipt of lamivudine-containing HAART 68 (93.2) 75 (92.6).89 Duration of lamivudine use, median years (range) 4.54 (1.01 7.68) 4.49 (1.21 7.61).72 Laboratory data 12 months after initiation of HAART Increase in the CD4 cell count, median cells/ml (range) 117 ( 707 to 479) 9 ( 354 to 548).23 Increase in the CD4 cell count 100 cells/ml 41 (56.2) 7 (45.7).19 Undetectable plasma HIV RNA level 56 (76.7) 58 (71.6).47 Laboratory data 24 months after nictitation of HAART b Increase in the CD4 cell count, median cells/ml (range) 223 ( 590 to 567) 83 ( 551 to 751).002 Increase in the CD4 cell count 100 cells/ml 48 (72.7) 2 (41.6)!.001 Increase in the CD4 cell count 200 cells/ml 36 (54.5) 19 (24.7)!.001 Undetectable plasma HIV RNA level 52 (78.8) 62 (80.5).80 Laboratory data at the end of the study Increase in the CD4 cell count, median cells/ml (range) 216 ( 590 to 567) 92 ( 551 to 751).004 Increase in the CD4 cell count 100 cells/ml 52 (71.2) 35 (43.2)!.001 Increase in the CD4 cell count 200 cells/ml 39 (35.6) 21 (20.0)!.001 Undetectable plasma HIV RNA level 60 (82.2) 55 (68.8).04 Duration of observation duration, median years (range) 4.98 (1.21 7.77) 5.11 (1.22 7.94).48 Death 9 (12.3) 16 (19.8).21 NOTE. Data are no. (%) of patients, unless otherwise indicated. HCV, hepatitis C virus. a Plasma HIV RNA level was available for 68 and 60 patients with subsequently detectable anti-hbs and patients with persistent isolated anti-hbc, respectively. b CD4 cell counts and plasma HIV RNA level were available for 66 and 77 patients with subsequently detectable anti-hbs and patients with persistent isolated anti-hbc, respectively, at 24 months after the initiation of HAART. occult or window-period HBV infection in our patients was much higher than that (0.6%) in another large cohort study [28]; this can possibly be attributed to a higher background prevalence of chronic HBV infection in Taiwan and lower baseline CD4 cell counts in our patients (132 vs. 414 cells/ml [28]). The appropriate treatment for HIV-infected persons who have isolated anti-hbc remains unknown. HBV vaccination may be considered, because a recent study suggested that 160% of patients with isolated anti-hbc developed seroprotective anti-hbs titers [29]. However, determination of the HBV DNA level should be considered, to exclude the possibility of occult HBV infection in patients with isolated anti-hbc (as occurred in 7.3% in our patients) before HBV vaccine is to be administered. Of the patients in our study who did not receive HBV revaccination, 40.8% of those with isolated anti-hbc who achieved better virological and immunologic responses to HAART developed anti-hbs, suggesting that immune reconstitution may restore immunity against HBV. Among patients without any HBV serological markers, we observed a higher rate of HBV infection (9.69 cases per 100 person-years) than has been observed in other reports (0.2 0.3 cases per 100 person-years) [3 5, 11] and in a study of HIVuninfected persons in Taiwan [16]. The higher rates of HBV transmission may be associated with low CD4 cell counts and decreased cellular immunity against HBV infection by HIV infection after risky exposure. This finding highlights the importance of vaccination of all HIV-infected patients without HBV markers and of counseling against HBV infection. There were limitations to our study. First, not all patients received sequential HBV serological tests, and patients with shorter observation periods (i.e.,!1 year) were excluded from the study; thus, selection bias could not be avoided, although the characteristics of excluded patients were similar to those of enrolled patients. Second, our population is unique in that most patients with detected HBsAg were chronic HBV carriers since childhood, and the injection drug users and HIV-HCV coin- HIV/AIDS CID 2007:45 (1 November) 1227

Figure 2. A, Kaplan-Meier survival estimates among patients with hepatitis B surface antigen (HBsAg), antibody to HBsAg (anti-hbs), and isolated antibody to hepatitis B core antigen (anti-hbc) and without hepatitis B virus serological markers at baseline. Patients with HBsAg had shorter durations of survival than did those with anti-hbs, those with isolated anti-hbc, and those without HBV serological markers ( P!.001 to.001). There were no statistically significant survival differences between patients with anti-hbs and those without any HBV serological markers ( P p.99) or between patients with isolated anti-hbc and those without any HBV serological markers ( P p.16). However, patients with isolated anti-hbc at baseline had shorter survival durations than did those with anti-hbs ( P p.02). B, Kaplan-Meier survival estimates among patients with HBsAg persistence, patients with anti-hbs persistence, patients with isolated anti-hbc who subsequently developed anti-hbs, and patients who had persistent isolated anti-hbc noted in follow-up blood tests. Patients with HBsAg persistence had shorter durations of survival than did those with anti-hbs persistence ( P!.001), those with isolated anti- HBc who subsequently developed anti-hbs ( P p.12), and those with persistent isolated anti-hbc ( P p.07). Patients with anti-hbs persistence had longer durations of survival than did those with persisted isolated anti-hbc ( P p.02) and those who subsequently developed anti-hbs ( P!.001). fected patients accounted for only a small proportion of the study population. Thus, our findings might not be generalizable to other populations with different epidemiologic and demographic characteristics. Third, attempts to identify factors (e.g., baseline HBV DNA level) associated with changes in HBV serological markers after initiation of lamivudine-containing HAART may be limited by the small numbers of cases in certain groups (tables 2 and 3). Finally, we were not able to completely exclude the possibility that changes in HBV serological markers may have been caused by the variability of serological tests, although we repeated serological tests with each serum sample in our study. In conclusion, changes in HBV serological markers are not uncommon in a region where HBV infection is endemic, and repeated measurement of HBV markers (including anti-hbc) in patients with HIV infection should be considered. Better virological and immunologic responses to HAART are associated with preservation and development of anti-hbs. HBV vaccination should be routinely advised to patients without any HBV markers. Acknowledgments Financial support. The Centers for Disease Control, Department of Health, Executive Yuan, Taiwan. Potential conflicts of interest. All authors: no conflicts. References 1. Perrillo RP, Regenstein FG, Roodman ST. Chronic hepatitis B in asymptomatic homosexual men with antibody to the human immunodeficiency virus. Ann Intern Med 1986; 105:382 3. 2. Sheng WH, Chen MY, Hsieh SM, et al. Impact of chronic hepatitis B virus (HBV) infection on outcomes of patients infected with HIV in an area where HBV infection is hyperendemic. Clin Infect Dis 2004; 38:1471 7. 3. De Luca A, Bugarini R, Lepri AC, et al. Co-infection with hepatitis viruses and outcome of initial antiretroviral regimens in previously naïve HIV-infected subjects. Arch Intern Med 2002; 162:2125 32. 4. Thio CL, Seaberg EC Jr, Skolasky R, et al. HIV-1, hepatitis B virus, and risk of liver-related mortality in the multicenter cohort study (MACS). Lancet 2002; 360:1921 6. 5. Rodriguez-Mendez M, Gonzalez-Quintela A, Aguilera A, Barrio E. Prevalence, patterns, and course of past hepatitis B virus infection in intravenous drug users with HIV-1 infection. Am J Gastroenterol 2000; 95:1316 22. 6. Biggar RJ, Goedert JJ, Hoofnagle J. Accelerated loss of anti-body to hepatitis B surface antigen among immunodeficient homosexual men infected with HIV. N Engl J Med 1987; 316:630 1. 7. Drake JH, Parmley RTW, Britton HA. Loss of hepatitis B antibody in human immunodeficiency virus-positive hemophilia patients. Pediatr Infect Dis J 1987; 6:1051 4. 8. Laukamm-Josten U, Müller O, Bienzle U, et al. Decline of naturally acquired antibodies to hepatitis B surface antigen in HIV-1 infected homosexual men. AIDS 1988; 2:400 1. 9. Homann C, Krogsgaard K, Pedersen C, et al. High incidence of hepatitis B infection and evolution of chronic hepatitis B infection in patients with advanced HIV infection. J Acquir Immune Defic Syndr 1991;4: 416 20. 10. Kisker CT, Mahoney EM, Arkin S, et al. Changes in hepatitis B serologic 1228 CID 2007:45 (1 November) HIV/AIDS

titers in HIV+ and HIV children with haemophilia. Haemophilia 1999; 5:354 9. 11. Piroth L, Binquet C, Vergne M, et al. The evolution of hepatitis B virus serological patterns and the clinical relevance of isolated antibodies to hepatitis B core antigen in HIV infected patients. J Hepatol 2002; 36: 681 6. 12. Jilg W, Sieger E, Zachoval R, Schatzle H. Individuals with antibodies against hepatitis B core antigen as the only serological marker for hepatitis B infection: high percentage of carriers of hepatitis B and C virus. J Hepatol 1995; 23:14 20. 13. Nunez M, Puoti M, Camino N, Soriano V. Treatment of chronic hepatitis B in the human immunodeficiency virus-infected patient: present and future. Clin Infect Dis 2003; 37:1678 85. 14. Manegold C, Hannoun C, Wywiol A, et al. Reactivation of hepatitis B virus replication accompanied by acute hepatitis in patients receiving highly active antiretroviral therapy. Clin Infect Dis 2001; 32:144 8. 15. Siriaksorn S, Puthanakit T, Sirisanthana T, Sirisanthana V. Prevalence of protective antibody against hepatitis B virus in HIV-infected children with immune recovery after highly active antiretroviral therapy. Vaccine 2006; 24:3095 9. 16. Chen DS, Hsu NH, Sung JL, et al. A mass vaccination program in Taiwan against hepatitis B virus infection in infants of hepatitis B surface antigen-carrier mothers. JAMA 1987; 257:2597 603. 17. Centers for Disease Control and Prevention. Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. MMWR Morb Mortal Wkly Rep 1992; 41:1 19. 18. Chen DS, Sung JL, Lai MY. A seroepidemiologic study of hepatitis B virus infection in Taiwan. Taiwan Yi Xue Hui Za Zhi 1978; 77:908 18. 19. Hsu HY, Chang MH, Chen DS, Lee CY, Sung JL. Baseline seroepidemiology of hepatitis B virus infection in children in Taipei: a study just before mass hepatitis B vaccination program in Taiwan. J Med Virol 1986; 18:301 7. 20. Chen DS. Natural history of chronic hepatitis B virus infection: new light on an old story. J Gastroenterol Hepatol 1993; 8:470 5. 21. Benhamou Y, Bochet M, Thibault V, et al. Long-term incidence of hepatitis B virus resistance to lamivudine in human immunodeficiency virus-infected patients. Hepatology 1999; 30:1302 6. 22. Chang TT, Gish RG, de Man R, et al. A comparison of entecavir and lamivudine for HBeAg-positive chronic hepatitis B. N Engl J Med 2006; 354:1001 10. 23. Lai CL, Shouval D, Lok AS, et al. Entecavir versus lamivudine for patients with HBeAg-negative chronic hepatitis B. N Engl J Med 2006; 354:1011 20. 24. Drake JH, Parmley RT, Britton HA. Loss of hepatitis B antibody in human immunodeficiency virus-positive hemophilia patients. Pediatr Infect Dis J 1987; 6:1051 4. 25. Nunez M, Ramos B, Diaz-Pollan B, et al. Virological outcome of chronic hepatitis B virus infection in HIV-coinfected patients receiving anti-hbv active antiretroviral therapy. AIDS Res Hum Retroviruses 2006; 22:842 8. 26. Cacciola I, Pollicino T, Squadrito G, Cerenzia G, Orlando M, Raimondo G. Occult hepatitis B virus infection in patients with chronic hepatitis C liver disease. N Engl J Med 1999; 341:22 6. 27. De Maria N, Colantoni A, Friedlander L, et al. The impact of previous HBV infection on the course of chronic hepatitis C. Am J Gastroenterol 2000; 95:3529 36. 28. Neau D, Winnock M, Jouvencel AC, et al. Occult hepatitis B virus infection in HIV-infected patients with isolated antibodies to hepatitis B core antigen: Aquitaine cohort, 2002 2003. Clin Infect Dis 2005; 40: 750 3. 29. Gandhi RT, Wurcel A, Lee H, et al. Response to hepatitis B vaccine in HIV-1 positive subjects who test positive for isolated antibody to hepatitis B core antigen: implications for hepatitis B vaccine strategies. J Infect Dis 2005; 191:1435 41. HIV/AIDS CID 2007:45 (1 November) 1229