Exposure to hepatitis B virus (HBV) infection is common among HIV-infected patients because of shared routes of transmission,1 and serologic markers of prior or active HBV infection are identified in up to 68% of HIV patients.2 In the absence of HBV surface antigen (HBsAg), the presence of HBV core IgG antibody (anti-HBc) and HBV surface antibody (anti-HBs) usually indicates past resolved infection. The presence of isolated anti-HBc is often interpreted as evidence of remote HBV infection with subsequent loss of anti-HBs but may also represent a false-positive result. These patients are generally not considered to have detectable HBV DNA or to be at risk for complications from chronic HBV, including hepatic fibrosis, cirrhosis, and hepatocellular carcinoma.3
Sensitive molecular assays have detected HBV DNA in a subset of patients who have presumably recovered from acute HBV infection, however. This has resulted in the clinical entity known as occult HBV infection, defined as the presence of HBV DNA in the serum and/or liver tissue in the absence of HBsAg.4-6 Few studies have examined the prevalence of occult HBV in HIV, and no studies have evaluated risk factors for occult HBV among HIV-infected patients. In addition, the clinical impact of occult HBV on hepatic transaminitis and fibrosis remains unclear in HIV.
To address these questions, we first determined the prevalence of occult HBV among a random sample of HBsAg−/anti-HBc+ HIV-infected subjects followed in the Penn Center for AIDS Research (CFAR) prospective HIV cohort study. Second, we examined risk factors for occult HBV among these subjects. Finally, we determined whether occult HBV was associated with transaminitis and significant hepatic fibrosis as determined by the aspartate aminotransferase (AST)-to-platelet ratio index (APRI), a noninvasive measure of significant liver fibrosis in patients with chronic HBV and hepatitis C virus (HCV).7-10
We used 3 study designs to address our specific aims. First, to determine the prevalence of occult HBV in HIV, we performed a seroprevalence study among HBsAg−/anti-HBc+ HIV-infected subjects. Second, to determine risk factors for occult HBV, we performed a case-control study. Cases represented all subjects identified with occult HBV, and controls were all subjects without occult HBV. Using this study design, each hypothesized risk factor represented an exposure and occult HBV status represented the outcome. Finally, to examine if occult HBV was associated with transaminitis and significant hepatic fibrosis by APRI score, we collected liver transaminases and platelet values from the same date at which serum samples were evaluated for occult HBV. We then performed a cross-sectional study in which occult HBV represented the exposure and liver transaminases and APRI score were the study outcomes.
We examined subjects in the Penn CFAR Adult/Adolescent Database and Specimen Repository, initiated in November 1999 to track demographic, clinical, and laboratory data from HIV-infected patients cared for at University of Pennsylvania-affiliated hospitals (Hospital of the University of Pennsylvania [HUP], Penn Presbyterian Medical Center [PPMC], Philadelphia Veterans Affairs Medical Center [PVAMC], Pennsylvania Hospital, and the adolescent HIV clinic at the Children's Hospital of Philadelphia). Subjects in the CFAR Adult/Adolescent Database have laboratory-confirmed HIV, provide informed consent, and complete a standardized questionnaire that collects demographic, medical, psychosocial, and HIV data at enrollment. A serum sample is drawn from each enrolled subject and stored at −70°C in the CFAR Specimen Repository. Subjects complete a follow-up questionnaire and have a serum sample drawn every 6 months. Additional laboratory data, such as HBV and HCV serostatus, may be downloaded into the CFAR Adult/Adolescent Database from hospital laboratory computer systems.
All HBsAg−/anti-HBc+ subjects enrolled in the CFAR Adult/Adolescent Database at the HUP, PPMC, and PVAMC sites from November 1, 1999 through December 31, 2002 were eligible for inclusion. A simple random sample of 180 subjects was targeted to have 80% power to detect a 20% difference in the prevalence of occult HBV between those with and without each risk factor of interest. This sample size also provided 80% power to detect a 20% difference in transaminitis between subjects with and without occult HBV. We oversampled to account for subjects who might not have serum samples available in the CFAR Specimen Repository.
To avoid misclassification of study subjects by HBV and HCV serostatus, we repeated the HBsAg (Elecsys 2010; Roche Diagnostics, Indianapolis, IN), anti-HBc (HBV Core Antibody Assay; Diagnostic Products Corporation, Los Angeles, CA), and HCV antibody (anti-HCV; Abbott HCV EIA 2.0 or 3.0 enzyme immunoassay; Abbott Laboratories, Abbott Park, IL) tests on the same serum sample in which HBV DNA testing was performed. HBsAg and anti-HBc tests were performed on the serum of all subjects, whereas anti-HCV testing was repeated only for HIV-infected subjects who were previously recorded as anti-HCV−. All subjects with HCV coinfection had detectable HCV RNA.
Main Outcome Measures
The primary outcome for the determination of the prevalence and risk factors for occult HBV was the presence of HBV DNA as determined by the Gen-Probe HBV modified Ultrio transcription-mediated amplification (TMA) assay (Gen-Probe, San Diego, CA). This qualitative test, based on the same technology as the Procleix HIV-1/HCV (Gen-Probe) and Ultrio (Gen-Probe) assays,11 requires only 500 μL of serum for processing and has a lower limit of detection of 15 HBV copies/mL.12 The most recent sample of each eligible subject's banked serum (through December 31, 2002) was obtained from the CFAR Specimen Repository for HBV DNA testing.
The main outcomes for the determination of the clinical significance of occult HBV were (1) transaminitis, defined as alanine aminotransferase (ALT) >40 U/L or AST >30 U/L,13,14 and (2) significant liver fibrosis as determined by APRI score.
Demographic and clinical data, including age, gender, race/ethnicity, duration of HIV diagnosis, possible mode of HIV acquisition, self-reported alcohol use within the past 30 days, use of highly active antiretroviral therapy (HAART; defined as use of 3 antiretroviral agents), HAART regimen, CD4 T-lymphocyte count, HIV viral load (determined by Versant HIV-1 RNA 3.0 Assay; Bayer Diagnostics, Tarrytown, NY; lower limit of detection of 75 copies/mL), ALT, AST, and platelet count, were collected from the CFAR Adult/Adolescent Database. Antiretroviral use was classified into 3 categories: (1) receipt of HAART that included an anti-HBV antiretroviral, (2) receipt of HAART without an anti-HBV antiretroviral, and (3) no HAART.
Baseline cohort characteristics were described using proportions for categorical data and medians with interquartile ranges (IQRs) for continuous variables. Differences between subjects by occult HBV status were assessed using χ2 or Fisher exact tests for categorical data and Wilcoxon rank-sum tests for continuous data.
We used multivariable logistic regression analysis to identify risk factors for occult HBV. Hypothesized risk factors included CD4 count <200 cells/mm3, HIV RNA level >1000 copies/mL, chronic HCV, and lack of use of an antiretroviral with anti-HBV activity. Potential confounders included age, gender, race/ethnicity, HAART use, anti-HBs, and active alcohol use. Further model reduction was achieved by eliminating factors that proved to be potential risk factors only infrequently in 1000 bootstrap samples of the data.15 Robust 95% confidence intervals (CIs) that do not depend on large-sample assumptions were then estimated using bias-corrected bounds from another round of bootstrap resampling.16
We examined differences in ALT/AST levels and transaminitis by occult HBV status. Multivariable logistic regression was used to examine the association between occult HBV and transaminitis while controlling for potential confounders. The odds ratio (OR) and 95% CI of transaminitis between subjects with and without occult HBV were determined.
Finally, to assess the association between occult HBV and significant liver fibrosis, we calculated the APRI score: APRI = (AST [/ULN]*100)/platelet count [109/L]). The APRI has been validated to identify advanced hepatic fibrosis in HCV-monoinfected,7,9 HIV/HCV-coinfected,10 and HBV-monoinfected subjects8 with a high degree of accuracy. The APRI accurately represents liver fibrosis only when liver disease has reached a severely advanced stage. Significant fibrosis was defined by an APRI score >1.5, and no fibrosis was defined by an APRI score ≤0.5.
All data were analyzed using STATA 8.2 (Stata Corporation, College Station, TX). Statistical significance was declared with 2-sided probability values <0.05.
The study was approved by the Institutional Review Boards of the University of Pennsylvania and PVAMC.
Of 1193 HIV-infected patients enrolled in the CFAR Adult/Adolescent Database between November 1, 1999 and December 31, 2002, we identified 699 (59%) HBsAg−/anti-HBc+ subjects. A total of 222 subjects were randomly selected for the study. Thirty-three subjects did not have a serum sample in the CFAR Specimen Repository and so were excluded. On repeat HBsAg and anti-HBc testing, 3 subjects were HBsAg+ and 7 subjects were anti-HBc− and were also excluded. All subjects found to be anti-HCV− remained so on repeated testing. The final sample included 179 subjects.
Table 1 presents the baseline characteristics of these subjects. Subjects were primarily African American (75%) and male (88%), 24% self-reported a history of injection drug use, and 55% had chronic HCV. HAART was prescribed to 73% of subjects, and 59% were receiving an antiretroviral agent with anti-HBV activity.
After testing for HBV DNA, 17 (10%; 95% CI: 5% to 14%) subjects were identified with occult HBV. Subjects with and without occult HBV were similar with regard to age, gender, race, ethnicity, duration of HIV diagnosis, and use of HAART. Individuals with occult HBV less commonly had anti-HBs (7 [41%]) compared with those without (94 [58%]), but this difference was not statistically significant (P = 0.3).
On univariable analysis, subjects with occult HBV had lower median CD4 cell counts (273 [IQR: 123-386] cells/mm3 vs. 366 [IQR: 222-558] cells/mm3; P = 0.04) and higher median HIV RNA levels (23,205 [IQR: 75-53,676] copies/mL vs. 87 [IQR: 75-6199] copies/mL; P = 0.02), and were not receiving an antiretroviral with anti-HBV activity (5 [29%] vs. 99 [61%]; P = 0.01). Occult HBV was associated with an HIV RNA level >1000 copies/mL but not with CD4 cell counts <200 cells/mm3 or the presence of chronic HCV (see Table 1). A total of 7 (7%) of 99 (95% CI: 2% to 12%) HIV-positive/HCV+ subjects were identified with occult HBV compared with 10 (13%) of 80 (95% CI: 5% to 20%) HIV-positive/HCV− subjects (P = 0.1). Lack of receipt of an antiretroviral with anti-HBV activity was associated with occult HBV, and use of 2 such agents significantly decreased the likelihood of occult HBV.
Table 2 shows results of the multivariable analysis with bootstrap resampling examining risk factors for occult HBV. After controlling for all other hypothesized risk factors as well as age and use of HAART, only an HIV RNA level >1000 copies/mL (adjusted OR = 4.88, 95% CI: 1.01 to 30.26) and chronic HCV (adjusted OR = 0.26, 95% CI: 0.05 to 0.95) remained associated with occult HBV.
Median ALT levels were higher in the HBV DNA− group, but a greater proportion of these subjects were chronically infected with HCV (Table 3). Fewer subjects with occult HBV had transaminitis compared with those without occult HBV (4 [24%]) vs. 78 [48%]), but this difference was not statistically significant (P = 0.07). Univariable logistic regression demonstrated no association between occult HBV and transaminitis (OR = 0.33, 95% CI: 0.10 to 1.06). After controlling for chronic HCV, HAART, and active alcohol use, there was little change in the OR (adjusted OR = 0.42, 95% CI: 0.12 to 1.45). Occult HBV was also not associated with significant fibrosis as determined by APRI score (see Table 3).
In this study, we found the prevalence of occult HBV to be 10% among a randomly selected group of HBsAg−/anti-HBc+ HIV-infected patients. Almost half of the subjects with occult HBV had anti-HBs, suggesting that this antibody does not result in complete HBV elimination. In addition, anti-HBV antiretroviral use was evaluated among all subjects, and subjects with occult HBV were typically not receiving these medications. An HIV RNA level >1000 copies/mL and the absence of chronic HCV were identified as risk factors for occult HBV. Finally, occult HBV did not increase the risk of hepatic transaminitis among HIV-infected subjects.
In the absence of HIV, the prevalence of occult HBV has been reported to range from 7% to 60% among HBsAg−/anti-HBc+ patients.17 Previous studies examining the prevalence of occult HBV in HIV-infected patients have similarly reported a wide range of prevalences from 0% to 89.5% (Table 4).2,18-25 These observational studies used different study designs (cross-sectional in 6 studies and longitudinal in 3 studies) and employed a variety of methods with different sensitivities to identify occult HBV. They were often limited by their small sample sizes, did not always report the number of subjects on antiretrovirals with anti-HBV activity, and rarely examined the clinical significance of occult HBV. In contrast, our study is the largest to date to examine occult HBV in HIV-infected patients and evaluated clinically relevant outcomes. Furthermore, we repeated HBV and HCV serologic test results on the same serum sample in which HBV DNA testing was performed to avoid misclassification of hepatitis serostatus. Three subjects were subsequently identified with active HBV, presumably because of activities that put them at risk for HBV infection, and 7 were found to be anti-HBc−, suggesting that the initial anti-HBc results were false positive.
The mechanism for occult HBV remains unclear, but mutations in the S region of the HBV genome that could prevent production of HBsAg, host immune dysfunction allowing low-level HBV DNA levels, and inhibition of HBV replication by chronic HCV coinfection have been proposed.17 To help shed light on its mechanism, we examined potential risk factors for occult HBV. In particular, detectable HIV RNA was found to increase the risk of occult HBV, suggesting that immune dysfunction or dysregulation independent of CD4 cell count might impair elimination of HBV and allow HBV replication to persist at low levels. Chronic HCV was also found to be protective against occult HBV. Previous data among HIV-uninfected patients have suggested that chronic HCV might downregulate HBV expression and allow HBV replication at extremely low levels, promoting the development of occult HBV.17 Our results suggest that chronic HCV may act as a more dominant virus in the presence of HIV and downregulate HBV replication enough to suppress HBV replication completely, however, thereby reducing the risk of occult HBV in HIV/HCV-coinfected patients.
A major question about occult HBV is whether such small amounts of HBV DNA are associated with transaminitis and progressive liver damage.6,26 The clinical impact of occult HBV has been examined primarily among HIV-uninfected chronic HCV patients. Among these patients, occult HBV has been shown to promote hepatic inflammation27 and increase the risk of advanced fibrosis and cirrhosis.27-29 Occult HBV has also been found to increase the risk of hepatocellular carcinoma in these individuals, possibly by integration with the host genome and/or synthesis of pro-oncogenic proteins by free intrahepatic HBV genomes.4,17,30-34
The clinical significance of occult HBV in HIV has been unclear. Two prior studies among HIV-infected patients found that occult HBV increased the frequency of transaminitis and hepatic flares,19,25 but a third study found that occult HBV was associated with only minimal hepatocellular inflammation.23 In this study, occult HBV did not increase the risk of transaminitis, even after adjusting for HAART, chronic HCV, and alcohol use. We also did not find an association between occult HBV and significant liver fibrosis as determined by APRI score. Thus, the likely low levels of HBV DNA in occult HBV might not be sufficient to induce clinically significant inflammation.
The clinical implications of occult HBV in HIV require further examination, however. In particular, longitudinal studies are needed in HIV-infected patients to determine if occult HBV is associated with an increased incidence of transaminitis, more advanced hepatic fibrosis, and hepatocellular carcinoma. Furthermore, because occult HBV has been shown to reduce response rates to interferon-based treatment among HCV-monoinfected subjects,28,35-37 the effect of occult HBV on response to combination pegylated interferon and ribavirin therapy should be examined among HIV/HCV-coinfected patients. The results of these studies should provide data on the long-term effects of occult HBV in HIV and help to determine the need to test for occult HBV with sensitive HBV DNA assays and subsequently treat with anti-HBV agents.
Our study had several limitations. We identified only 17 subjects with occult HBV, and this small number limits our statistical power, particularly for determining risk factors for occult HBV. We used bootstrap resampling to obtain robust estimates of the association between hypothesized risk factors and occult HBV, however. Additionally, subjects came from study sites in a predominantly urban setting, potentially limiting the generalizability of our results.
In conclusion, we found a 10% prevalence of occult HBV infection among HBsAg−/anti-HBc+ HIV-infected subjects. The presence of anti-HBs did not rule out occult HBV. Detectable HIV RNA and the absence of chronic HCV were associated with occult HBV. Occult HBV did not increase the risk of transaminitis. Future studies should examine the long-term clinical implications of occult HBV in HIV-infected patients.
The authors acknowledge the invaluable assistance provided by Deborah Gudonis, RN (Penn CFAR Senior Research Coordinator), and Peggy Cummins (Coordinator of the Penn CFAR Specimen Repository) during this study. The authors also thank all Penn CFAR study subjects for their continued participation in the CFAR Adult/Adolescent Database.
1. Thio CL. Hepatitis B in the human immunodeficiency virus-infected patient: epidemiology, natural history, and treatment. Semin Liver Dis
2. Santos E, Yoshida CFT, Rolla VC, et al. Frequent occult hepatitis B virus infection in patients infected with human immunodeficiency virus type 1. Eur J Clin Microbiol Infect Dis
3. Lee WM. Hepatitis B virus infection. N Engl J Med
4. Hu KQ. Occult hepatitis B virus infection and its clinical implications. J Viral Hepat
5. Torbenson M, Thomas DL. Occult hepatitis B. Lancet Infect Dis
6. Conjeevaram HS, Lok AS. Occult hepatitis B virus infection: a hidden menace? Hepatology
7. Lackner C, Struber G, Liegl B, et al. Comparison and validation of simple noninvasive tests for prediction of fibrosis in chronic hepatitis C. Hepatology
8. Lebensztejn DM, Skiba E, Sobaniec-Lotowska M, et al. A simple noninvasive index (APRI) predicts advanced liver fibrosis in children with chronic hepatitis B. Hepatology
9. Wai CT, Greenson JK, Fontana RJ, et al. A simple noninvasive index can predict both significant fibrosis and cirrhosis in patients with chronic hepatitis C. Hepatology
10. Al-Mohri H, Cooper C, Murphy T, et al. Validation of a simple model for predicting liver fibrosis in HIV/hepatitis C virus-coinfected patients. HIV Med
11. McCormick MK, Dockter J. Evaluation of a new molecular assay for detection of human immunodeficiency virus type 1, hepatitis C virus RNA, and hepatitis B virus DNA. J Clin Virol
12. Linnen J, Phelps B. Implementation of Procleix Ultrio assay on the Procleix TIGRIS system: a fully automated Triplex NAT assay. Presented at: International Plasma Fractionation Association/Paul Ehrlich Institute 12th Nucleic Acid Testing Workshop on Surveillance and Screening of Bloodborne Pathogens; 2005; Bethesda.
13. Katkov WN, Friedman LS, Cody H, et al. Elevated serum alanine aminotransferase levels in blood donors: the contribution of hepatitis C virus. Ann Intern Med
14. Kratz A, Lewandrowski KB. Case records of the Massachusetts General Hospital. Weekly clinicopathological exercises. Normal reference laboratory values. N Engl J Med
15. Altman DG, Andersen PK. Bootstrap investigation of the stability of a Cox regression model. Stat Med
16. Efron B, Tibshirani RJ. An Introduction to the Bootstrap
. New York: Chapman and Hall; 1993.
17. Brechot C, Thiers V, Kremsdorf D, et al. Persistent hepatitis B virus infection in subjects without hepatitis B surface antigen: clinically significant or purely “occult”? Hepatology
18. 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
19. Hofer M, Joller-Jemelka HI, Grob PJ, et al. Frequent chronic hepatitis B virus infection in HIV-infected patients positive for antibody to hepatitis B core antigen only. Swiss HIV Cohort Study. Eur J Clin Microbiol Infect Dis
20. 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
21. Nunez M, Rios P, Perez-Olmeda M, et al. Lack of ‘occult’ hepatitis B virus infection in HIV-infected patients. AIDS
22. 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
23. Shire NJ, Rouster SD, Rajicic N, et al. Occult hepatitis B in HIV-infected patients. J Acquir Immune Defic Syndr
24. Wagner AA, Denis F, Weinbreck P, et al. Serological pattern ‘anti-hepatitis B core alone’ in HIV or hepatitis C virus-infected patients is not fully explained by hepatitis B surface antigen mutants. AIDS
25. Filippini P, Coppola N, Pisapia R, et al. Impact of occult hepatitis B virus infection in HIV patients naive for antiretroviral therapy. AIDS
26. Raimondo G, Pollicino T, Squadrito G. What is the clinical impact of occult hepatitis B virus infection? Lancet
27. Cacciola I, Pollicino T, Squadrito G, et al. Occult hepatitis B virus infection in patients with chronic hepatitis C liver disease. N Engl J Med
28. 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
29. Raimondo G. Occult hepatitis B virus infection and liver disease: fact or fiction? J Hepatol
30. Paterlini P, Driss F, Nalpas B, et al. Persistence of hepatitis B and hepatitis C viral genomes in primary liver cancers from HBsAg-negative patients: a study of a low-endemic area. Hepatology
31. Kubo S, Tamori A, Ohba K, et al. Previous or occult hepatitis B virus infection in hepatitis C virus-associated hepatocellular carcinoma without hepatic fibrosis. Dig Dis Sci
32. Brechot C, Jaffredo F, Lagorce D, et al. Impact of HBV, HCV and GBV-C/HGV on hepatocellular carcinomas in Europe: results of a European concerted action. J Hepatol
33. Shibata Y, Nakata K, Tsuruta S, et al. Detection of hepatitis B virus X-region DNA in liver tissue from patients with hepatitis C virus-associated cirrhosis who subsequently developed hepatocellular carcinoma. Int J Oncol
34. Pollicino T, Squadrito G, Cerenzia G, et al. Hepatitis B virus maintains its pro-oncogenic properties in the case of occult HBV infection. Gastroenterology
35. Fukuda R, Ishimura N, Hamamoto S, et al. Co-infection by serologically-silent hepatitis B virus may contribute to poor interferon response in patients with chronic hepatitis C by down-regulation of type-I interferon receptor gene expression in the liver. J Med Virol
36. Sagnelli E, Coppola N, Scolastico C, et al. Isolated anti-HBc in chronic hepatitis C predicts a poor response to interferon treatment. J Med Virol
37. Zignego AL, Fontana R, Puliti S, et al. Impaired response to alpha interferon in patients with an inapparent hepatitis B and hepatitis C virus coinfection. Arch Virol