MAJOR ARTICLE. Downloaded from by guest on 22 December 2018

Transcription

1 MAJOR ARTICLE 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 Rajesh T. Gandhi, 1 Alysse Wurcel, 1 Hang Lee, 1 Barbara McGovern, 2 Janet Shopis, 1 Meghan Geary, 1 Rohini Sivamurthy, 1 Paul E. Sax, 3 and Chinweike Ukomadu 3 1 Massachusetts General Hospital, 2 Lemuel Shattuck Hospital and New England Medical Center, and 3 Brigham and Women s Hospital, Boston Background. Whether human immunodeficiency virus type 1 (HIV-1) positive subjects who test positive for isolated antibody to hepatitis B core antigen (anti-hbc) should be vaccinated with hepatitis B vaccine is not certain. Development of an anamnestic response after vaccination would suggest previous hepatitis B virus (HBV) infection, in which case vaccination is not necessary. Methods. Sixty-nine HIV-1 positive subjects who tested negative for hepatitis B surface antigen (HBsAg) and antibody to HBsAg (anti-hbs) received vaccination with standard hepatitis B vaccine. Twenty-nine subjects (42%) tested positive for anti-hbc, and 40 (58%) tested negative for anti-hbc. An anamnestic response was defined as an anti-hbs titer of 10 IU/L within 4 weeks of the first vaccination. Results. The overall anamnestic response rate was 16% and was not significantly different between subjects who tested positive for anti-hbc (24%) and those who tested negative for anti-hbc (10%) before vaccination ( P p.18). Approximately 50% of subjects who tested positive for anti-hbc also tested positive for antibody to hepatitis Be antigen (anti-hbe). The anamnestic response rate was higher in subjects who tested positive for both anti-hbc and anti-hbe (43%) than in subjects who tested positive for anti-hbc but negative for anti-hbe (7%) ( P p.035). After a complete series of vaccinations, HIV-1/hepatitis C virus (HCV) coinfected subjects were less likely to achieve high anti-hbs titers than were subjects infected with HIV-1 alone. Conclusions. After hepatitis B vaccination, the anamnestic response rate in HIV-1 positive subjects who tested positive for isolated anti-hbc but negative for anti-hbe was low and was comparable to that in subjects who tested negative for anti-hbc. This finding suggests that testing for anti-hbc alone may not be a reliable assessment of protection from HBV infection. HIV-1/HCV coinfection may be associated with impaired responses to hepatitis B vaccine, and evaluation of strategies to improve immunogenicity of the vaccine in such individuals is warranted. The Advisory Committee on Immunization Practices recommends that adults at risk for hepatitis B undergo testing for either antibody to hepatitis B surface antigen (anti-hbs) or antibody to hepatitis B core antigen (anti- HBc) to evaluate susceptibility to hepatitis B virus (HBV) Received 7 October 2004; accepted 30 November 2004; electronically published 28 March Presented in part: 42nd annual meeting of the Infectious Diseases Society of America, Boston, 30 September 3 October 2004 (abstract 871). Financial support: Harvard Medical School Center for AIDS Research. Reprints or correspondence: Dr. Rajesh T. Gandhi, Infectious Diseases Unit, Massachusetts General Hospital, GRJ 504, 55 Fruit St., Boston, MA The Journal of Infectious Diseases 2005; 191: by the Infectious Diseases Society of America. All rights reserved /2005/ $15.00 infection [1]. A positive test result for either anti-hbs or anti-hbc for subjects who have not been immunized against hepatitis B is generally interpreted as being indicative of previous infection, in which case hepatitis B vaccination is usually not offered. This recommendation is based on the most common serologic pattern of previously cleared HBV infection, in which individuals test negative for hepatitis B surface antigen (HBsAg) and positive for both anti-hbs and anti-hbc. However, some individuals who test negative for both HBsAg and anti-hbs will test positive for IgG anti- HBc. This finding, known as isolated anti-hbc, is particularly common in individuals who are infected with HIV-1 or hepatitis C virus (HCV) [2 6]. In subjects who are coinfected with HIV-1 and HCV, more than Hepatitis B Vaccine and Isolated Anti-HBc JID 2005:191 (1 May) 1435

2 three-fourths of those who test negative for both HBsAg and anti-hbs will test positive for anti-hbc [7]. Despite its frequent occurrence, the implication of this finding is not certain. Isolated IgG anti-hbc may represent either (1) resolved HBV infection with loss of anti-hbs, (2) occult chronic HBV infection with undetectable HBsAg, or (3) a false-positive test result. If a positive test result for anti-hbc is interpreted as evidence of resolved HBV infection, as has been recommended, then patients who test positive for isolated anti-hbc will not be offered hepatitis B vaccination. Whether individuals who test positive for isolated anti-hbc should be vaccinated with hepatitis B vaccine is not certain. The presence of other markers of previous HBV infection, such as antibody to hepatitis Be antigen (anti-hbe), may suggest resolved infection [8]; vaccination would not be indicated in this situation. If subjects who test positive for isolated anti- HBc develop more rapid seroprotective antibody responses after exposure to hepatitis B antigen, this finding would suggest that, in terms of susceptibility to infection, subjects who test positive for isolated anti-hbc differ from subjects who test negative for anti-hbc. That is, the development of an anamnestic response after hepatitis B vaccination in individuals who test positive for isolated anti-hbc would suggest that they have been previously infected with hepatitis B virus, in which case vaccination would not be necessary. In previous studies, most individuals who tested positive for isolated anti-hbc mounted a slow or primary, rather than a rapid or anamnestic, response after hepatitis B vaccination [9 11], suggesting that isolated anti-hbc may represent a falsepositive antibody response. However, no studies of anamnestic response rates after hepatitis B vaccination have been reported for HIV-1 positive subjects who tested positive for isolated anti-hbc. HIV-1 positive individuals have a particularly high probability of previous HBV infection, because of shared routes of transmission. If the presence of isolated anti-hbc is indicative of previous HBV infection, then groups with a high pretest probability of previous infection may be more likely to have an anamnestic response after hepatitis B vaccination. Therefore, we hypothesized that, if the presence of isolated anti-hbc reflects previous infection, then HIV-1 positive subjects who test positive for isolated anti-hbc would have a higher anamnestic response rate after hepatitis B vaccination than would those who test negative for anti-hbc. Thus, we conducted a study to assess the anamnestic response rate in HIV-1 positive subjects who test positive for isolated anti-hbc. SUBJECTS, MATERIALS, AND METHODS Study population. Ninety-seven HIV-1 positive subjects were recruited from the infectious diseases clinics of Massachusetts General Hospital, Brigham and Women s Hospital, and Lemuel Shattuck Hospital (all in Boston, MA). Individuals with no history of hepatitis B vaccination who tested negative for both HBsAg and anti-hbs were eligible for this study. Information regarding HCV antibody and RNA status, most recent CD4 cell count, nadir CD4 cell count, most recent HIV-1 RNA level, current antiretroviral therapy, and most recent alanine aminotransferase (ALT) level were abstracted from the chart. The institutional review boards of Massachusetts General Hospital, Brigham and Women s Hospital, and Lemuel Shattuck Hospital approved this study. Study procedures. Subjects enrolled in the study underwent baseline testing for HBsAg and qualitative and quantitative testing for anti-hbs and IgG anti-hbc (Abbott Laboratories, Abbott Park, IL). The absorbance ratio for anti-hbc was calculated as the absorbance for a particular subject s sample divided by the control cutoff absorbance (an absorbance less than the threshold is indicative of a positive test result). Samples with absorbance values within 30% of the control cutoff absorbance were retested in triplicate, to confirm the original results. Subjects with a baseline CD4 cell count of 1200 cells/mm 3 and without current use of immunomodulatory agents (e.g., chemotherapy or treatment with corticosteroids or interferon) received a series of 3 vaccinations with standard hepatitis B vaccine (either 10 mg of Recombivax [Merck] or 20 mg of Engerix B [GlaxoSmithKline]), at 0, 4, and 24 weeks. We chose this CD4 cell count threshold so that we could examine the response to vaccination in a relatively immunocompetent group of subjects. Blood samples were obtained for qualitative and quantitative testing for anti-hbs, at 2 and 4 weeks after the first vaccination and 4 weeks after the second and third vaccinations (weeks 8 and 28, respectively). All samples in which the quantitative anti-hbs titer was!10 IU/L but for which the qualitative assay result was positive underwent repeat testing in triplicate, and the mean titer was used to assign the final response category (see Study definitions). Subjects from whom frozen serum samples collected before vaccination were available underwent additional testing for HBe antigen (HBeAg) and anti-hbe (DiaSorin). Study definitions. For anti-hbs, a seroprotective titer was defined as 10 IU/L. As in previous studies, an anamnestic response was defined as an anti-hbs titer of 10 IU/L at either 2 or 4 weeks after the first vaccination. A subject was classified as being HCV positive if he or she had either a positive HCV antibody or a positive HCV RNA test result. A subject was classified as receiving a hepatitis B active drug if he or she was taking lamivudine, tenofovir, or adefovir at study entry. Statistical analysis. The primary end point of the study was the anamnestic response rate in subjects receiving vaccination. We examined the effect of the following baseline characteristics on this end point: age, sex, race, current CD4 cell count, nadir CD4 cell count, HIV-1 RNA level, use of highly active antiretroviral therapy (HAART), use of a hepatitis B active drug, HCV status, and anti-hbc and anti-hbe status. We 1436 JID 2005:191 (1 May) Gandhi et al.

3 used Fisher s exact test and the x 2 test for comparison of proportions between the 2 groups and the Wilcoxon rank sum test for comparison of numerically measured outcomes between the 2 groups. All comparisons were performed by use of 2-sided tests, and resulting P values!.05 were considered to be statistically significant. We performed all analyses with SAS (version 8.2; SAS Institute). RESULTS Baseline Results Prevalence of and risk factors for isolated anti-hbc. Of 97 HIV-1 positive subjects who tested negative for both HBsAg and anti-hbs, 44 tested positive for anti-hbc (prevalence of isolated anti-hbc, 45%). Thirty-six subjects (37%) had a history of injection drug use, 3 (3%) had a history of hemophilia, and the remaining subjects (60%) were thought to have acquired HIV-1 by sexual transmission. In univariable analysis, male sex, white race, HCV positivity, and elevated ALT level were associated with isolated anti-hbc positivity (table 1). Anti- HBc positivity was not associated with current CD4 cell count, nadir CD4 cell count, use of HAART, use of a hepatitis B active drug, or HIV-1 RNA level. In multivariable analysis that included sex, race, HCV status, and ALT level, HCV positivity and male sex remained independently associated with anti-hbc positivity. After adjustment for HCV status, a higher ALT level was no longer associated with anti-hbc positivity, suggesting that the difference in ALT level was probably due to a higher rate of HCV infection in the group of subjects who tested positive for isolated anti-hbc. Prevalence of anti-hbe positivity. Of the 44 HIV-1 positive subjects who tested positive for isolated anti-hbc, 41 underwent testing for anti-hbe and HBeAg. Twenty (49%) of the Table 1. Univariable analysis of risk variables for antibody to hepatitis B core antigen (anti-hbc) in HIV-1 positive subjects who tested negative for hepatitis B surface antigen (HBsAg) and antibody to HBsAg. Characteristic 41 subjects tested positive for anti-hbe. In contrast, only 2 (5%) of the 40 subjects who tested negative for HBsAg, anti- HBs, and anti-hbc tested positive for anti-hbe. No subject tested positive for HBeAg. In univariable analysis, HCV positivity and higher ALT level were associated with anti-hbe positivity, whereas age, sex, race, current CD4 cell count, nadir CD4 cell count, HIV-1 RNA level, use of HAART, and use of a hepatitis B active drug were not. In multivariable analysis, only HCV positivity ( P p.02) remained independently associated with anti-hbe positivity. Strength of positive anti-hbc antibody response. Since false-positive results may be more likely for subjects with weaker anti-hbc responses, we estimated the strength of this response by examining the absorbance ratio in subjects who tested positive for anti-hbc (see Study procedures). Of 14 subjects who tested positive for both anti-hbc and anti-hbe, all had an absorbance ratio of 0.10, indicating a strongly positive response (an absorbance less than the threshold is indicative of a positive test result). In contrast, of 15 subjects who tested positive for anti-hbc but negative for anti-hbe, 9 (60%) had an absorbance ratio of 10.10, suggesting a weaker response. Thus, subjects who tested positive for both anti-hbc and anti- HBe were more likely to have a strongly positive anti-hbc response than were subjects who tested positive for anti-hbc but negative for anti-hbe ( P!.001). Vaccination Results Vaccination results in the entire cohort. The subjects with CD4 cell counts of 1200 cells/mm 3 at the time of testing received vaccination with hepatitis B vaccine. The baseline characteristics of the 69 vaccinees are shown in table 2. The mean CD4 cell count at the time of vaccination was 518 cells/mm 3, Anti-HBc negative (n p 53) Anti-HBc positive (n p 44) P Age, years 41 (42) 8 46 (45) 7 Male sex a 25/53 (47) 36/44 (82)!.001 White race a 27/51 (53) 32/43 (74).04 HCV positive a 5/52 (10) 30/44 (68)!.0001 Current CD4 cell count, b cells/mm (402) (367) 318 NS Nadir CD4 cell count, b cells/mm (192) (179) 198 NS HIV-1 RNA level, b copies/ml 9002 (83) 27,970 13,999 (91) 37,886 NS Use of HAART a 37/53 (70) 31/44 (70) NS Use of a hepatitis B active drug a 34/53 (64) 25/44 (57) NS ALT level b 23 (17) (49) 33!.0001 NOTE. Data are mean (median) SD or no. of subjects with characteristic/no. (%) of subjects tested, unless otherwise noted. ALT, alanine aminotransferase; HAART, highly active antiretroviral therapy; HCV, hepatitis C virus; NS, not significant. a Fisher s exact test was used for the group comparison. b Wilcoxon rank sum test was used for the group comparison. Hepatitis B Vaccine and Isolated Anti-HBc JID 2005:191 (1 May) 1437

4 Table 2. Baseline characteristics of HIV-1 positive subjects who received hepatitis B vaccination. Characteristic Subjects (n p 69) Age, years 43 (44) 8 Male sex 43 (62) Race African American 28 (41) White 38 (55) Other 3 (4) HCV positive 23 (33) Current CD4 cell count, cells/mm (457) 276 Nadir CD4 cell count, cells/mm (203) 195 HIV-1 RNA level, copies/ml 6719 (38) 24,072 Use of HAART 50 (72) Use of a hepatitis B active drug 41 (59) Anti-HBc positive 29 (42) NOTE. Data are mean (median) SD or no. (%) of subjects, unless otherwise noted. Anti-HBc, antibody to hepatitis B core antigen; HAART, highly active antiretroviral therapy; HCV, hepatitis C virus. and the mean nadir CD4 cell count was 232 cells/mm 3. At baseline, 23 subjects (33%) tested positive for HCV, and 29 (42%) tested positive for anti-hbc. All of the 69 subjects who initiated vaccination returned for evaluation at either week 2 or week 4 after the first vaccination and had anti-hbs titers that could be examined for development of an anamnestic response, the primary end point of the study. Twenty-two subjects (32%) did not have anti-hbs titers measured after the third vaccination. With respect to age, sex, race, HCV status, CD4 cell count at the time of vaccination, nadir CD4 cell count, and baseline ALT level, subjects who did not have anti-hbs titers measured after the third vaccination did not differ from subjects who did. The development of a seroprotective anti-hbs titer occurred in 16%, 22%, and 62% of subjects after the first, second, and third vaccinations, respectively. For the entire group, the anamnestic response rate was 16%. The anamnestic and final response rates were not affected by age, sex, race, current CD4 cell count, nadir CD4 cell count, HIV-1 RNA level, use of HAART, use of a hepatitis B active drug, or baseline ALT level. Subjects who had anamnestic responses developed high anti- HBs titers: the mean titer was 1604 IU/L (range, 13 13,008 IU/ L), the median titer was 150 IU/L, and the geometric mean titer was 270 IU/L within 4 weeks of the first vaccination, in the 11 subjects who had anamnestic responses. Effect of baseline anti-hbc status on response to vaccine. We examined whether the rate of response to hepatitis B vaccine differed by baseline anti-hbc status. We found that the anamnestic response rate was not significantly different between subjects who tested positive for anti-hbc and subjects who tested negative for anti-hbc. Only 7 (24%) of 29 subjects who tested positive for anti-hbc had an early response after the first vaccination, which was not statistically significantly different from the rate of 4 (10%) of 40 subjects who tested negative for anti-hbc ( P p.18). Thus, the majority of HIV-1 positive subjects who tested positive for isolated anti-hbc did not develop an anamnestic response after receiving hepatitis B vaccine. Since the presence of anti-hbe at baseline may identify subjects with a higher probability of previous HBV infection, we examined whether subjects who tested positive for both anti- HBc and anti-hbe at baseline had a different anamnestic response rate than did subjects who tested positive for anti-hbc but negative for anti-hbe and subjects who tested negative for anti-hbc. Approximately 50% of subjects who tested positive for anti-hbc also tested positive for anti-hbe. The anamnestic response rate in those who tested positive for anti-hbc but negative for anti-hbe was not different from the rate in those who tested negative for anti-hbc: 1 (7%) of 15 and 4 (10%) of 40, respectively. However, subjects who tested positive for both anti-hbc and anti-hbe had a higher anamnestic response rate (6/14 [43%]) than did those who tested positive for anti- HBc but negative for anti-hbe (1/15 [7%]; P p.035) and those who tested negative for anti-hbc (4/40 [10%]; P p.01). Most of the anamnestic responses in subjects who tested positive for anti-hbc occurred in those who also tested positive for both anti-hbc and anti-hbe (6 of the 7 anamnestic responses). We also assessed whether the likelihood of an anamnestic response after vaccination was associated with the strength of the anti-hbc response in subjects who tested positive for this marker at baseline. Of the 7 subjects who tested positive for anti-hbc and who developed an anamnestic response, all had baseline anti- HBc absorbance ratios of 0.1 (indicating a strongly positive response). None of the 9 subjects with anti-hbc absorbance ratios of 10.1 had an anamnestic response after vaccination. There was a trend for an association between a baseline anti-hbc absorbance ratio of 0.1 and development of an anamnestic response ( P p.066). Subjects who tested positive for anti-hbc at baseline were more likely to develop low nonseroprotective anti-hbs titers within 2 4 weeks of the first vaccination than were subjects who tested negative for anti-hbc at baseline (table 3). Of the 40 subjects who tested negative for anti-hbc at baseline, none developed an anti-hbs titer of 1 9 IU/L after the first vaccination. In contrast, 6 (21%) of 29 subjects who tested positive for anti-hbc at baseline developed an anti-hbs titer of 1 9 IU/ L after the first vaccination ( P p.004); all of these subjects tested positive for anti-hbc but negative for anti-hbe. After the third vaccination, 61% of subjects who tested negative for anti-hbc and 63% of subjects who tested positive for anti-hbc developed a seroprotective anti-hbs response; these 2 response rates were not significantly different. In subjects who developed a seroprotective response, the geometric mean anti- HBs titer after the third vaccination was lower in subjects who tested positive for anti-hbc than in subjects who tested negative 1438 JID 2005:191 (1 May) Gandhi et al.

5 Table 3. Titers of antibody to hepatitis B surface antigen (anti-hbs) 2 4 weeks after the first vaccination. Baseline status (no. of subjects) Anti-HBs titer, IU/L Anti-HBc negative (40) 0 (0) a 0 (0) 1 (3) 3 (8) Anti-HBc positive (29) 6/29 (21) a 3 (10) 2 (7) 2 (7) Anti-HBc positive and anti-hbe negative (15) 6 (40) 0 (0) 0 (0) 1 (7) Anti-HBc positive and anti-hbe positive (14) 0 (0) 3 (21) 2 (14) 1 (7) NOTE. Data are no. (%) of subjects. Anti-HBc, antibody to hepatitis B core antigen; anti-hbe, antibody to hepatitis Be antigen. a P p.004. for anti-hbc 174 versus 586 IU/L, respectively but this difference was not statistically significant. Effect of baseline HCV status on response to vaccine. We also compared the response rate to hepatitis B vaccine in HCVpositive and HCV-negative subjects. The anamnestic response rate was the same in HCV-negative subjects and in HCV-positive subjects: 16% and 17%, respectively. There was a trend toward a higher seroprotective response rate after the third vaccination in HCV-negative subjects, compared with that in HCVpositive subjects (71% vs. 42%; P p.09). The quantitative anti- HBs titer after the third vaccination was significantly higher in HCV-negative subjects than in HCV-positive subjects: the mean and geometric mean anti-hbs titers were 6850 (interquartile range [IQR], ) and 200 IU/L, respectively, whereas, in HCV-positive subjects, the mean and geometric mean titers were 49 (IQR, 0 47) and 46 IU/L ( P p.02). In addition, after the third vaccination, a lower proportion of HCV-positive subjects achieved a high anti-hbs titer, compared with HCV-negative subjects (0% vs. 33% for anti-hbs titer of 1000 IU/L; P p.04, Fisher s exact test) (figure 1). DISCUSSION Isolated anti-hbc is frequently detected in HIV-1 positive patients who test negative for both HBsAg and anti-hbs; in the present study, the prevalence of this finding was 45% and was significantly associated with HCV positivity. Approximately 50% of HIV-1 positive subjects who tested positive for isolated anti- HBc also tested positive for anti-hbe, suggesting previous HBV infection. The association between anti-hbe positivity and HCV positivity supports the idea that these subjects had previously been infected with both HBV and HCV, which share routes of transmission. In these anti-hbc/anti-hbe positive subjects, the anamnestic response rate after hepatitis B vaccination was higher than that in HIV-1 positive subjects who tested negative for anti-hbc. Subjects who tested positive for anti-hbc but negative for anti-hbe had an anamnestic response rate after vaccination that was comparable to that in subjects who tested negative for anti-hbc. Although difficult to prove, it is possible that, for HIV-1 positive individuals who test negative for anti- HBe, the presence of anti-hbc represents a false-positive serologic test result. The finding that many of these individuals had a weaker anti-hbc response and that anamnestic responses in subjects who tested positive for anti-hbc occurred only in those with stronger anti-hbc responses supports this possibility. False-positive serologic test results are more frequent in subjects infected with either HIV-1 or HCV [12 14], perhaps because of nonspecific immune activation in these conditions. Although the anamnestic response rate was not significantly different between subjects who had tested positive or negative for anti-hbc, subjects who tested positive for anti-hbc (and, in particular, positive for anti-hbc but negative for anti-hbe) were more likely to develop low anti-hbs titers after the first vaccination than were subjects who tested negative for anti-hbc (table 3). The reason for these low-level responses after exposure to HBsAg is not clear. It is possible that some of these individuals had occult HBV infection and a diminished ability to mount an anti-hbs response. However, in a previous study, we tested this cohort for HBV DNA by use of an ultrasensitive assay and found the prevalence of ongoing hepatitis B viremia to be only 2.4% [15]. Although this suggests that ongoing occult infection is un- Figure 1. Lower proportion of hepatitis C virus (HCV) positive subjects with high titers of antibody to hepatitis B surface antigen (anti-hbs) after 3 vaccinations with hepatitis B vaccine, compared with HCV-negative subjects. The anti-hbs titer was measured 4 weeks after the third hepatitis B vaccination. Hepatitis B Vaccine and Isolated Anti-HBc JID 2005:191 (1 May) 1439

6 common, we cannot rule out intermittent hepatitis B viremia in these subjects, since we did not test sequential samples for HBV DNA. In addition, the seroprotective response rate after a complete series of vaccinations was comparable between subjects who tested positive for anti-hbc and subjects who tested negative for anti-hbc ( 60%), indicating that the subjects who tested positive for anti-hbc were not completely anergic. An unexpected finding in the present study was that HIV- 1/HCV coinfected subjects were less likely to develop high anti- HBs titers after the third vaccination. HCV infection has been proposed to result in dendritic cell dysfunction [16 18], which may impair antigen presentation. HCV-positive individuals have been found to have slower response rates and lower antibody titers than HCV-negative individuals, after receiving other immunogens, such as hepatitis A vaccine [19]. Previous studies of the effect of HCV status on response to hepatitis B vaccine in HIV-1 negative individuals have yielded conflicting results [19 24], and it has been proposed that variable schedules and doses of hepatitis B vaccine may be important in explaining these conflicting results [20]. To our knowledge, the present study is the first to evaluate the effect of HCV status on response to hepatitis B vaccine in an HIV-1 infected population. On the basis of our findings, further study of response to hepatitis B vaccine in larger groups of HIV-1/HCV coinfected subjects is warranted. Although the HIV-1 positive subjects in the present study had relatively high CD4 cell counts (mean, 518 cells/mm 3 ), the overall response rate after a complete series of vaccinations was only 62%. This response rate is much lower than that in healthy individuals, who generally have a seroconversion rate of 190% after the third vaccination. Other studies have also shown that HIV-1 positive individuals have lower response rates to hepatitis B vaccine than do HIV-1 negative individuals [25 28]. Given that HIV-1 positive individuals are at a high risk for HBV infection and may have a worse clinical outcome after HBV infection, compared with HIV-1 negative individuals [29], a more effective HBV vaccine would be beneficial in this population of patients. There are several limitations to the present study. First, given the small sample size in the present study, we had limited power to detect small differences in the effect of baseline anti-hbc, HCV, or anti-hbe on rates of response to hepatitis B vaccination. However, the observed anamnestic response rate of 7 (24%) of 29 anti-hbc positive subjects makes it very unlikely that, even with a larger sample size, the true response rate would be 150% ( P p.004). Second, data on anti-hbs titer after the third vaccination were not available for a relatively high proportion of subjects. With respect to known factors that could potentially affect vaccine response, subjects who did not have titers measured after the third vaccination did not differ significantly from subjects who did. In addition, all of the subjects in the present study had data available for assessment of anamnestic response rate, the primary end point of the study. Third, HIV-1 positive subjects may have limited ability to generate anamnestic responses after hepatitis B vaccination. We studied relatively immunocompetent HIV-1 positive subjects, and subjects who had a rapid response developed high anti-hbs titers, suggesting that the HIV-1 positive subjects were able to generate anamnestic responses. In addition, our findings on HIV-1 positive subjects are comparable to those on HIV-1 negative subjects in other studies [9, 10]. Fourth, defining subjects as HCV positive on the basis of a positive antibody test may result in inclusion of some individuals who have cleared HCV RNA. However, most HIV-1 positive subjects who test positive for HCV antibody will have persistent HCV infection [30]. Finally, whether protection from hepatitis B can be predicted on the basis of an anamnestic response is not certain. However, since a large trial to definitively examine vaccine efficacy in subjects who test positive for anti-hbc is not feasible, the use of an anamnestic response as indirect evidence of previous infection still provides useful information. In conclusion, the rate of anamnestic response to vaccination in HIV-1 positive subjects who tested positive for isolated anti- HBc and negative for anti-hbe was low and was comparable to the rate in subjects who tested negative for anti-hbc. If our findings are confirmed in larger studies, the results of the present study suggest that some HIV-1 positive subjects who test positive for isolated anti-hbc may have a false-positive test result and should be offered a complete series of vaccinations. Although some HIV-1 positive subjects who test positive for isolated anti-hbc have had previous HBV infection, it is difficult to distinguish such individuals from those with a falsepositive anti-hbc test result. In the future, there may be methods to identify subjects who test positive for isolated anti-hbc who are more likely to have an anamnestic response, such as those who test positive for anti-hbe or have a strong baseline anti-hbc response. In such patients, it may be reasonable to measure the anti-hbs titer 2 4 weeks after the first hepatitis B vaccination. If subjects have an anamnestic response (i.e., an anti-hbs titer of 10 IU/L), further vaccination may not be necessary. If they do not, then they should receive a complete series of vaccinations. To be adopted, this approach should be validated in a larger cohort; we are in the process of conducting such a validation study. Acknowledgments We thank Audrey Byrne, Nicole Burgett, Sarah Toner, Melinda Boczanowski, Roslyn Gerwin, Colleen P. Corcoran, Rosemary Delabre, Zbigniew Szczepiorkowski, Nan Midgley, Shean Marley, Kevin Murphy, William Theisen, Lynn Dumas, Tauheed Zaman, and the providers and patients who participated in this study JID 2005:191 (1 May) Gandhi et al.

7 References 1. Hepatitis B virus: a comprehensive strategy for eliminating transmission in the United States through universal childhood vaccination. Recommendations of the Immunization Practices Advisory Committee (ACIP). MMWR Recomm Rep 1991; 40(RR-13): Shire NJ, Rouster SD, Rajicic N, Sherman KE. Occult hepatitis B in HIV-infected patients. J Acquir Immune Defic Syndr 2004; 36: Grob P, Jilg W, Bornhak H, et al. Serological pattern anti-hbc alone : report on a workshop. J Med Virol 2000; 62: 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: Davaro RE, Cheeseman SH, Keroack MA, Ellison RT. The significance of isolated antibody to hepatitis B core antigen seropositivity in patients infected with human immunodeficiency virus. Clin Infect Dis 1996; 23: Berger A, Doerr HW, Rabenau HF, Weber B. High frequency of HCV infection in individuals with isolated antibody to hepatitis B core antigen. Intervirology 2000; 43: Gandhi RT, Wurcel A, Lee H, et al. Isolated antibody to hepatitis B core antigen in human immunodeficiency virus type-1 infected individuals. Clin Infect Dis 2003; 36: Ganem D, Prince AM. Hepatitis B virus infection natural history and clinical consequences. N Engl J Med 2004; 350: Lok AS, Lai CL, Wu PC. Prevalence of isolated antibody to hepatitis B core antigen in an area endemic for hepatitis B virus infection: implications in hepatitis B vaccination programs. Hepatology 1988; 8: McMahon BJ, Parkinson AJ, Helminiak C, et al. Response to hepatitis B vaccine of persons positive for antibody to hepatitis B core antigen. Gastroenterology 1992; 103: Silva AE, McMahon BJ, Parkinson AJ, Sjogren MH, Hoofnagle JH, Di Bisceglie AM. Hepatitis B virus DNA in persons with isolated antibody to hepatitis B core antigen who subsequently received hepatitis B vaccine. Clin Infect Dis 1998; 26: Rompalo AM, Cannon RO, Quinn TC, Hook EW 3rd. Association of biologic false-positive reactions for syphilis with human immunodeficiency virus infection. J Infect Dis 1992; 165: Thomas DL, Rompalo AM, Zenilman J, Hoover D, Hook EW 3rd, Quinn TC. Association of hepatitis C virus infection with false-positive tests for syphilis. J Infect Dis 1994; 170: Clifford BD, Donahue D, Smith L, et al. High prevalence of serological markers of autoimmunity in patients with chronic hepatitis C. Hepatology 1995; 21: Gandhi RT, Wurcel A, McGovern B, et al. Low prevalence of ongoing hepatitis B viremia in HIV-positive individuals with isolated antibody to hepatitis B core antigen. J Acquir Immune Defic Syndr 2003; 34: Anthony DD, Yonkers NL, Post AB, et al. Selective impairments in dendritic cell associated function distinguish hepatitis C virus and HIV infection. J Immunol 2004; 172: Sarobe P, Lasarte JJ, Casares N, et al. Abnormal priming of CD4 + T cells by dendritic cells expressing hepatitis C virus core and E1 proteins. J Virol 2002; 76: Auffermann-Gretzinger S, Keeffe EB, Levy S. Impaired dendritic cell maturation in patients with chronic, but not resolved, hepatitis C virus infection. Blood 2001; 97: Keeffe EB, Iwarson S, McMahon BJ, et al. Safety and immunogenicity of hepatitis A vaccine in patients with chronic liver disease. Hepatology 1998; 27: Wiedmann M, Liebert UG, Oesen U, et al. Decreased immunogenicity of recombinant hepatitis B vaccine in chronic hepatitis C. Hepatology 2000; 31: Navarro JF, Teruel JL, Mateos ML, Marcen R, Ortuno J. Antibody level after hepatitis B vaccination in hemodialysis patients: influence of hepatitis C virus infection. Am J Nephrol 1996; 16: Leroy V, Bourliere M, Durand M, et al. The antibody response to hepatitis B virus vaccination is negatively influenced by the hepatitis C virus viral load in patients with chronic hepatitis C: a case-control study. Eur J Gastroenterol Hepatol 2002; 14: Lee SD, Chan CY, Yu MI, Lu RH, Chang FY, Lo KJ. Hepatitis B vaccination in patients with chronic hepatitis C. J Med Virol 1999; 59: Awofeso N, Levy M, Harper S, et al. Response to HBV vaccine in relation to vaccine dose and anti-hcv positivity: a New South Wales correctional facilities study. Vaccine 2001; 19: Wong EK, Bodsworth NJ, Slade MA, Mulhall BP, Donovan B. Response to hepatitis B vaccination in a primary care setting: influence of HIV infection, CD4 + lymphocyte count and vaccination schedule. Int J STD AIDS 1996; 7: Loke RH, Murray-Lyon IM, Coleman JC, Evans BA, Zuckerman AJ. Diminished response to recombinant hepatitis B vaccine in homosexual men with HIV antibody: an indicator of poor prognosis. J Med Virol 1990; 31: Collier AC, Corey L, Murphy VL, Handsfield HH. Antibody to human immunodeficiency virus (HIV) and suboptimal response to hepatitis B vaccination. Ann Intern Med 1988; 109: Carne CA, Weller IV, Waite J, et al. Impaired responsiveness of homosexual men with HIV antibodies to plasma derived hepatitis B vaccine. Br Med J (Clin Res Ed) 1987; 294: Thio CL, Seaberg EC, Skolasky R Jr, et al. HIV-1, hepatitis B virus, and risk of liver-related mortality in the Multicenter Cohort Study (MACS). Lancet 2002; 360: Thomas DL, Astemborski J, Rai RM, et al. The natural history of hepatitis C virus infection: host, viral, and environmental factors. JAMA 2000; 284: Hepatitis B Vaccine and Isolated Anti-HBc JID 2005:191 (1 May) 1441