Hepatitis B Virus and HIV Coinfection: Results of a Survey on Treatment Practices and Recommendations for Therapy
Clinical Infectious Diseases Sept 1, 2007
Paul J. Gaglio,1 Richard Sterling,2 Eric Daniels,3 Ellen Tedaldi,4 and the Terry Beirn Community Programs for Clinical Research on AIDS Hepatitis Working Groupa
1Department of Medicine, Columbia University College of Physicians and Surgeons, New York, New York; 2Divisions of Gastroenterology, Hepatology, and Infectious Diseases, Virginia Commonwealth University School of Medicine, Richmond; 3Social and Scientific Systems, Silver Springs, Maryland; and 4Department of Medicine, Temple University School of Medicine, Philadelphia, Pennsylvania
(See the article by Miailhes et al. on pages 624-32 and the editorial commentary by Tillman on pages 633-6)
Background. The management of hepatitis B virus (HBV) and human immunodeficiency virus (HIV) coinfection is challenged by the selection of patients for therapy, options for antiviral medications, and inconsistency in published treatment guidelines.
Methods. A survey was sent to 161 sites in a multicenter HIV clinical trials group to assess HBV screening, criteria for initiation of therapy, and treatment choices for patients coinfected with HBV and HIV.
Of 161 sites, 78 completed the survey (response rate, 48.4%). Of these sites, 98.7% screened for HBV infection, 86% vaccinated HIV-infected patients who were not immune to HBV infection, and 79% made treatment decisions without referral to a hepatologist or gastroenterologist. Treatment recommendations varied; 42% of the sites initiated therapy when patients' levels of alanine aminotransferase and aspartate aminotransferase were elevated and HBV DNA level was >105 copies/mL (100,000 c/ml), whereas 49% of the sites initiated therapy in the presence of any detectable HBV DNA level. Antiviral treatment choices for patients who were not concurrently receiving antiretroviral therapy were lamivudine plus tenofovir, adefovir, or interferon. Patients concurrently receiving antiretroviral therapy received lamivudine plus tenofovir preferentially, followed by tenofovir plus emtricitabine, adefovir, or interferon. Ninety-one percent of the sites screened for hepatocellular carcinoma.
Conclusions. The majority of HIV-infected patients were screened and vaccinated for HBV infection and underwent surveillance for hepatocellular carcinoma. Decisions regarding the performance of liver biopsy, threshold to initiate therapy, and criteria to discontinue therapy varied, reflecting inconsistencies in available treatment guidelines. Treatment decisions reflected concerns regarding future drug resistance in patients who are naive to antiretroviral therapy and the emergence of drug resistance in patients receiving antiretroviral therapy.
With the availability of therapy that can simultaneously treat hepatitis B virus (HBV) infection and HIV infection, treatment guidelines have been published and updated by various consensus panels [1-4]. These guidelines address screening for and diagnosis of HBV infection and recommend algorithms for therapy for patients who may or may not require concurrent treatment for HIV infection. An example is the complexity of management that is underscored by inconsistencies among the guidelines and the off-label use of tenofovir as an agent to treat HBV infection . Few studies have evaluated the use of and adherence to these guidelines by practicing HIV clinicians [6-8].
The aim of this survey was to ascertain how coinfection with HIV and HBV is managed "in the real world" by querying providers in the Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA), a federally funded, national network of community-based clinical treatment centers. These sites represent a geographically diverse collection of various facilities, such as community clinics, academic and federal hospital outpatient practices, and private physicians' offices. Specifically, we endeavored to address multiple issues regarding HBV-HIV coinfection, including screening for HBV infection and hepatocellular carcinoma, initiation or discontinuation of anti-HBV therapy, the requirement for liver biopsy, the threshold for HBV viremia to initiate therapy, and differences in treatment recommendations regarding e antigen-positive versus e antigen-negative HBV infection. In addition, we assessed treatment choices for HBV infection in patients receiving or not receiving concurrent antiretroviral therapy.
We sent a survey to 161 US sites participating in the CPCRA. This survey, using a multiple-choice format, was designed to assess how screening for HBV infection occurs, which HBV therapies are chosen for patients who are and are not receiving concurrent antiretroviral therapy, and criteria for initiating and discontinuing HBV therapy. Physicians, research nurses, and research assistants at the participating sites completed the survey in December 2004 (Before Entecavir was approved). No identifying data, including patients' names or initials and hospital numbers, were available. The data were reviewed in aggregate and then stratified by data source.
Rates of response to the survey and demographic characteristics of the population studied. One hundred sixty-one sites in the CPCRA received the survey, and results were obtained from 78 sites, representing a response rate of 48.4%; 52% of the sites provided information from a database, whereas 48% used estimated data. When response rates from sites with database-derived data were compared with those from sites with estimated data, there were no statistically significant differences.
The population surveyed included >52,000 HIV-infected patients (mean age, 36 years), 8% of whom were enrolled in 1 CPCRA-sponsored clinical study. Of the patients, 72% were male; 48% of the patients were black, 25% were white, 15% were Latino/Hispanic, 11% were of an unspecified race, and 1% were Asian. Thirty-eight percent of patients met the Centers for Disease Control and Prevention criteria for AIDS, and overall, 14% of patients were naive to previous antiretroviral therapy. There were 4074 patients characterized as having chronic HBV infection (defined as positivity for HBV surface antigen), reflecting 8% of the overall patient population. The demographic characteristics of the patients with chronic HBV infection were identical to those of the overall population of patients in the CPCRA. Thus, the prevalence of coinfection with HIV and HBV in this patient population correlates well with published data from similar patient populations [9, 10] (table 1).
Screening and vaccination for HBV infection and referral for treatment.
The vast majority of respondents to the survey (98.7%) indicated that all HIV-infected patients were screened for HBV infection by assessment of HBV total core antibody (at 92.3% of the sites), surface antigen (97.4%), and surface antibody (82.1%). HBV e antigen and/or antibody and HBV DNA level were rarely assessed during initial screening (at 3.8% and 1.3% of the sites, respectively). Assays were performed at either a local or commercial laboratory, depending on the preference of CPCRA personnel at each site. The majority of sites (86%) offered HBV vaccination to HIV-infected patients who did not have previous exposure to HBV infection, which was determined on the basis of serologic examination results indicating negativity for HBV core and/or surface antibody. Only 21% of the respondents indicated that when patients were identified as having chronic HBV infection, they were referred to a gastroenterologist or hepatologist for treatment, implying that the majority of treatment decisions for this patient population are being made by either an internist or an infectious diseases specialist who is treating the patient's HIV infection (table 2).
Table 2. Responses to a survey of Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA) sites regarding treatment practices for hepatitis B virus (HBV) and HIV coinfection.
98.7% said HIV+ patients are assessed for HBV exposure. 86% of HIV+ who did not have previous HBV exposure are vaccinated for HBV. 79% of docs said patients are NOT referred to gastroenterologist or hepatologist. 41% do not perform liver biopsy; 44% said they perform liver biopsy if patient has detectable HBV DNA and elevated transaminase levels; 10% perform liver biopsy if HBV DNA isdetectable regardless of transaminase levels.
WHEN IS HBV INFECTION TREATED IN COINFECTED PATIENTS?
Transaminase levels are elevated and HBV DNA >100,000 c/ml - 42%
HBV DNA >100,000 c/ml regardless of transaminase levels -- 49%
When is treatment started for antigen e-positive, HBV DNA>100k - 85%
When treatment started for e-antigen-negative, >10,000 c/ml - 76%
No therapy 8%
TREATMENT CHOICES for patients on ART
No therapy 7%
When is anti-HBV therapy discontinued in patients not receiving antiretroviral therapy?
--Therapy is continued indefinitely 46%
--Therapy is discontinued if a patient seroconverts from HBV e antigen to HBV e antibody 34%
--Therapy is discontinued only after persistent loss of HBV surface antigen 20%
When is anti-HBV therapy discontinued in patients receiving antiretroviral therapy?
--Therapy is continued indefinitely 71%
--Therapy is discontinued only after a persistent decrease in HBV surface antigen 23%
Indications for initiation of anti-HBV therapy.
A significant disparity regarding recommendations for liver biopsy was apparent from responses to the survey; 44% of the respondents recommended liver biopsy only for patients with evidence of active HBV replication and elevated serum transaminase levels, 41% did not perform liver biopsy despite the presence of HBV replication and elevated transaminase levels, and 10% recommended liver biopsy for any patient with HBV replication, regardless of elevated serum transaminase levels. In addition, treatment recommendations based on levels of viral replication varied considerably, with some respondents (42%) initiating therapy for patients with elevated aspartate and alanine aminotransferase levels and an HBV DNA level >105 copies/mL and others (49%) recommending therapy for all patients with an HBV DNA level >105 copies/mL, regardless of aspartate and alanine aminotransferase levels. The survey included an assessment of treatment strategies for e antigen-positive ("wild type") HBV infection versus e antigen-negative HBV infection on the basis of virus titer; 85% of the respondents recommended therapy for patients with e antigen-positive HBV infection, abnormal liver enzyme levels, and an HBV DNA level >105 copies/mL, whereas 76% recommended therapy for patients who were negative for HBV e antigen but had abnormal liver enzyme levels and an HBV DNA level >104 copies/mL.
The selection of anti-HBV therapy was stratified by whether the patient was also receiving concurrent antiretroviral therapy. For patients not receiving antiretroviral therapy, respondents selected several options for the treatment of HBV infection: lamivudine plus tenofovir (selected by 88% of the respondents), adefovir (41%), IFN (either standard or pegylated) (24%), no therapy (8%), or other nonspecified therapy (5%). The total percentage of treatment choices indicated by the respondents to the survey was >100%, because most sites used several treatment options. For patients receiving concurrent antiretroviral therapy, anti-HBV treatment included lamivudine plus tenofovir (selected by 88% of the respondents), tenofovir plus emtricitabine (75%), adefovir (28%), IFN (11%), or no therapy (7%). Again, because multiple treatment choices were provided by each site, the total percentage of responses is >100%.
A major concern in this patient population is access to specific drugs through the AIDS Drug Assistance Program, a federally funded program that provides medication for the treatment of HIV-infected patients. These programs are unique, because each state regulates which drugs are included in their formulary and establishes eligibility criteria for patient assistance . It could be hypothesized that treatment decisions initiated by specific sites participating in the CPCRA would reflect constraints on access to medication. We observed that the availability of anti-HBV therapy through the AIDS Drug Assistance Program did not influence treatment choices; 73% and 17% of respondents to the survey indicated that access to specific drugs was limited <10% of the time and 10%-25% of the time, respectively.
Decisions regarding duration and discontinuation of therapy. Among patients for whom anti-HBV therapy was initiated, 46% of the respondents recommended continuing anti-HBV therapy indefinitely for patients not receiving concomitant antiretroviral therapy. If patients developed a persistent loss of surface antigen while undergoing therapy, 20% of survey respondents recommended discontinuing HBV therapy; 34% recommended discontinuing therapy if patients converted from being e antigen positive to e antibody positive. For patients receiving concomitant anti-HBV and antiretroviral therapy, 71% of respondents recommended that anti-HBV therapy be continued indefinitely, whereas 23% would discontinue therapy only if patients developed a persistent loss of surface antigen during therapy.
Screening for hepatocellular carcinoma.
The majority of respondents (91%) indicated that screening for hepatocellular carcinoma was a standard practice for patients coinfected with HIV and HBV. This was accomplished by determination of alpha feto protein levels and by ultrasonograph.
Patients coinfected with HBV and HIV present multiple challenges; recommendations regarding the selection of patients for therapy, as well as the choices of therapy effective for both HIV infection and HBV infection, are evolving. Contributing to the complexity of these issues are inconsistencies in published treatment guidelines for coinfected patients and the increased availability of new antiviral therapies. The available literature regarding HIV-HBV coinfection suggest that the natural history of HBV infection in HIV-HBV-coinfected patients is accelerated, compared with that in HBV-monoinfected patients, because coinfected patients are at risk for rapidly progressing liver disease and hepatocellular carcinoma [12, 13], particularly in the context of a decreased CD4 cell count . In addition, HBV-associated hepatocellular injury may limit the ability to treat HIV infection with antiretroviral therapy because of concerns regarding hepatotoxicity [15, 16]. Thus, identifying appropriate candidates for therapy, as well as providing effective suppression of HBV replication, are important goals in the treatment of patients coinfected with HIV and HBV.
The aim of this survey was to query a group of experienced HIV providers to assess how HBV-HIV coinfection is managed in practice. This diverse group of experienced HIV providers demonstrates the variability of treatment decisions for coinfected patients (table 2).
The vast majority of respondents to the survey screened for HBV infection in the context of HIV infection, offered HBV vaccination to patients who did not have evidence of exposure to HBV infection, and surveyed for hepatocellular carcinoma, in accordance with published guidelines on HIV-HBV coinfection [1-4]. However, significant disparities were found among the sites with regard to the requirement for liver biopsy, as well as the threshold for HBV viremia, in relation to the initiation of HBV therapy. Liver biopsy was recommended by some respondents for patients with HBV replication and elevated serum transaminase levels, whereas others recommended biopsy in the context of active replication and normal transaminase levels. Other respondents stated that they did not perform biopsy despite active viremia and elevated transaminase levels. Treatment recommendations on the basis of levels of viral replication also varied significantly: elevated transaminase levels and an HBV DNA level >105 copies/mL was the threshold for initiation of therapy for some respondents, whereas others recommended therapy for any patient with an HBV DNA level >105 copies/mL, regardless of transaminase levels. The inconsistencies we observed in responses to the survey were not surprising; published treatment guidelines for coinfected patients vary significantly for these specific issues. Recently published results from a European consensus conference  indicated that liver biopsy is essential, whereas other published guidelines recommend treatment without biopsy in some cases  or performance of a biopsy based on the degree of hepatic injury and/or viral load . Inconsistencies also exist in published guidelines with regard to the recommended threshold of HBV DNA level to initiate therapy, either with a cutoff of 104 copies/mL [1, 3] or no specific threshold .
The vast majority of participants in the survey recommended therapy for patients with "wild type" (e antigen-positive) HBV infection, abnormal liver enzyme levels, and an HBV DNA level >105 copies/mL, as well as for HBV e antigen-negative patients with abnormal liver enzyme levels and an HBV DNA level >104 copies/mL, in accordance with recently published guidelines for HBV-monoinfected patients . Different thresholds for initiating therapy for HIV-infected patients coinfected with e antigen-positive HBV versus e antigen-negative HBV have also been published .
An important aspect of HBV-directed therapy is understanding when therapy can be safely discontinued. For coinfected patients for whom anti-HBV therapy was initiated, decisions regarding discontinuation of HBV-directed therapy varied considerably for patients receiving and not receiving concomitant HIV therapy. Although guidelines for the safe discontinuation of HBV therapy have been published for monoinfected patients , recommendations for patients coinfected with HBV and HIV are less clear; some authors suggest that HBV-directed therapy should be discontinued "with caution" [1, p. 624], whereas others make no specific recommendations [2, 3]. A major concern with regard to discontinuing anti-HBV therapy is the potential for significant "flares" in transaminase levels in the context of unopposed HBV replication. In light of recent reports documenting severe medical complications in patients discontinuing antiretroviral therapy as part of a treatment-sparing strategy, there is compelling evidence to maintain suppression of both HIV infection and HBV infection in appropriate patients [19, 20].
The respondents to the survey stratified HBV treatment selections depending on the requirement for concurrent antiretroviral therapy. For patients who were not concurrently receiving antiretroviral therapy, respondents recommended combination therapy with lamivudine plus tenofovir, adefovir monotherapy, and IFN monotherapy. The survey did not include entecavir or telbivudine as therapeutic options for HBV-infected patient, because the survey preceded approval of the drugs. Patients concurrently receiving antiretroviral therapy received lamivudine plus tenofovir therapy preferentially, followed by tenofovir plus emtricitabine therapy, adefovir monotherapy, and IFN monotherapy. These choices concur with published guidelines for patients coinfected with HIV and HBV and appropriately reflect concerns regarding future drug resistance in antiretroviral therapy-naive patients and the emergence of drug resistance in patients receiving antiretroviral therapy [1-4].
An interesting aspect identified after reviewing responses to the survey was the low rate of referral of coinfected patients to a gastroenterologist (21%), indicating that the vast majority of treatment decisions regarding HBV infection were made by infectious diseases specialists or internists. It could be postulated that the significant experience of the respondents in prescribing antiviral therapy in the context of unclear treatment guidelines contributed to the desire for them to treat coinfected patients rather than refer them to a gastroenterologist. Alternatively, a deficit in expertise, experience, or interest in treating patients coinfected with HIV and HBV among local gastroenterologists may have affected referral decisions. A potential drawback to the survey includes the possibility that the clinician filling out the responses to the survey may have answered the queries on the basis of personal preferences, rather than representing the practices of the group. In general, the CPCRA model encourages uniformity in practice patterns, and therefore, we are confident that responses to the survey appropriately reflected current treatment paradigms.
The results of this survey, completed by a group of clinicians who treat HIV-infected patients, identified multiple aspects of HBV-HIV coinfection that remain contentious. Although clear guidelines exist regarding surveillance for HBV infection and hepatocellular carcinoma, as well as vaccination policies for HIV-infected patients without previous exposure to HBV infection, published treatment guidelines are inconsistent with regard to important aspects of managing HBV infection in the context of HIV infection; inconsistencies include the requirement for liver biopsy prior to therapy, the threshold of HBV replication to initiate therapy, and the point (if any) when therapy can be safely discontinued. Therefore, it is not surprising that the respondents to the survey varied significantly in their recommendations related to these important issues.
On the basis of these inconsistencies, we believe that it is now appropriate for future consensus conferences and publications that address issues regarding coinfection with HBV and HIV to incorporate recommendations related to requirement for liver biopsy and the threshold of HIV DNA level to initiate therapy and to clearly define when HBV therapy can be discontinued for coinfected patients.
We recommend universal assessment of HBV infection for all HIV-infected patients and screening for hepatocellular carcinoma in coinfected patients, particularly in persons from areas where HBV is endemic.
HBV vaccination should be offered to all HIV-infected persons who have not had previous HBV exposure, although the efficacy of the vaccine is related to the degree of immunosuppression .
Because the available literature suggests a more rapid progression of the natural history of HBV infection in patients coinfected with HIV and control of HBV replication can be achieved with a low rate of resistance using appropriate antiviral therapy, it could be argued that any level of HBV DNA in a coinfected patient should be suppressed.
Liver biopsy may be valuable to assess histological injury in persons for whom treatment decisions may be less straightforward, specifically in coinfected patients with an HBV DNA level of 10,000 copies/mL (<2000 IU/mL) and normal serum transaminase levels.
Patients coinfected with HBV and HIV who are not receiving antiretroviral therapy and are not anticipated to require antiretroviral therapy in the near future should be treated with an antiviral therapy that does not target HIV infection or that is dosed in a manner that does not affect HIV replication (i.e., IFN, adefovir, or entecavir).
Coinfected patients receiving antiretroviral therapy should receive therapies that are effective against both viruses; lamivudine plus tenofovir or emtricitabine plus tenofovir are potential choices.
Finally, effective therapy against HBV infection should not be discontinued unless another drug with activity against HBV infection is substituted or the patient has achieved seroconversion from HBV surface antigen to surface antibody.