American Gastroenterological Association Institute Technical Review on Prevention and Treatment of Hepatitis B Virus Reactivation During Immunosuppressive Drug Therapy
2014; Elsevier BV; Volume: 148; Issue: 1 Linguagem: Inglês
10.1053/j.gastro.2014.10.038
ISSN1528-0012
AutoresRobert P. Perrillo, Robert G. Gish, Yngve Falck‐Ytter,
Tópico(s)Liver Disease Diagnosis and Treatment
ResumoHepatitis B virus reactivation (HBVr) is a potentially serious disorder that is frequently induced by long-term immunosuppressive drug therapy. HBVr is more often seen in hepatitis B surface antigen (HBsAg) and antibody to hepatitis B core antigen (anti-HBc)–positive patients (hereafter referred to as HBsAg-positive patients) but also occurs in patients with resolved infection as defined by negative HBsAg and positive anti-HBc.1Hoofnagle J.H. Reactivation of hepatitis B.Hepatology. 2009; 49: S156-S165Crossref PubMed Scopus (477) Google Scholar This clinical entity was first described nearly 50 years ago in the setting of cancer chemotherapy and kidney transplantation.2Galbraith R.M. Eddleston A.L. Williams R. et al.Fulminant hepatic failure in leukaemia and choriocarcinoma related to withdrawal of cytotoxic drug therapy.Lancet. 1975; 2: 528-530Abstract PubMed Scopus (218) Google Scholar, 3Nagington J. Reactivation of hepatitis B after transplantation operations.Lancet. 1977; 1: 558-560Abstract PubMed Scopus (134) Google Scholar Although specific cellular immunologic mechanisms have not been fully elucidated, the initial event is believed to be a disruption in the ability of the host immune system to control hepatitis B virus (HBV) replication.4Perrillo R.P. Acute flares in chronic hepatitis B: the natural and unnatural history of an immunologically mediated liver disease.Gastroenterology. 2001; 120: 1009-1022Abstract Full Text Full Text PDF PubMed Scopus (384) Google Scholar The best predictor of reactivation has been shown to be the level of HBV DNA at baseline.5Yeo W. Zee B. Zhong S. et al.Comprehensive analysis of risk factors associating with Hepatitis B virus (HBV) reactivation in cancer patients undergoing cytotoxic chemotherapy.Br J Cancer. 2004; 90: 1306-1311Crossref PubMed Scopus (273) Google Scholar Patients who are HBsAg positive are 5 to 8 times more likely to develop HBVr than those with resolved infection who are HBsAg negative but anti-HBc positive.6Shouval D. Shibolet O. Immunosuppression and HBV reactivation.Semin Liver Dis. 2013; 33: 167-177Crossref PubMed Scopus (117) Google Scholar The immunologic potency and complexity of the drug regimen are also likely to be important factors.5Yeo W. Zee B. Zhong S. et al.Comprehensive analysis of risk factors associating with Hepatitis B virus (HBV) reactivation in cancer patients undergoing cytotoxic chemotherapy.Br J Cancer. 2004; 90: 1306-1311Crossref PubMed Scopus (273) Google Scholar, 6Shouval D. Shibolet O. Immunosuppression and HBV reactivation.Semin Liver Dis. 2013; 33: 167-177Crossref PubMed Scopus (117) Google Scholar Reactivation is characterized by a sudden increase in serum HBV DNA level that is most often associated with a hepatitis flare several weeks later, as defined by an increasing alanine aminotransferase (ALT) level.1Hoofnagle J.H. Reactivation of hepatitis B.Hepatology. 2009; 49: S156-S165Crossref PubMed Scopus (477) Google Scholar, 4Perrillo R.P. Acute flares in chronic hepatitis B: the natural and unnatural history of an immunologically mediated liver disease.Gastroenterology. 2001; 120: 1009-1022Abstract Full Text Full Text PDF PubMed Scopus (384) Google Scholar Conventionally, reactivation is defined by the level of change in HBV DNA and ALT. In HBsAg carriers, this is often defined as either the de novo detection of HBV DNA or a ≥10-fold (1 log10) increase in HBV DNA level when compared with the baseline value before immunosuppressive drug therapy. A hepatitis flare is frequently determined to be present when there is at least a 2- to 3-fold elevation in ALT level above the patient’s baseline value or a predefined multiple of the upper limit of normal. In patients with resolved infection (HBsAg negative but anti-HBc positive), reactivation has been considered to occur with demonstration of reverse seroconversion to HBsAg-positive status. The clinical spectrum of medical interventions that induce HBVr has greatly expanded in the past 2 decades and now includes several types of biological agents that are used in the fields of rheumatology, pulmonology, dermatology, neurology, gastroenterology, hepatology, nephrology, and transplantation medicine.1Hoofnagle J.H. Reactivation of hepatitis B.Hepatology. 2009; 49: S156-S165Crossref PubMed Scopus (477) Google Scholar Tumor necrosis factor (TNF) inhibitors may represent the predominant class of noncancer drugs that are capable of inducing HBVr because they are immunologically potent, widely used across specialties, and generally given on a long-term basis. In a recent systematic review, nearly 40% of HBsAg carriers and 5% of anti-HBc–positive but HBsAg-negative patients developed HBVr during TNF inhibitor therapy, nearly half of whom were undergoing treatment for inflammatory bowel disease.7Perez-Alvarez R. Diaz-Lagares C. Garcia-Hernandez F. et al.Hepatitis B virus (HBV) reactivation in patients receiving tumor necrosis factor (TNF)-targeted therapy: analysis of 257 cases.Medicine (Baltimore). 2011; 90: 359-371Crossref PubMed Scopus (223) Google Scholar The fact that TNF inhibitors have been prescribed for more than 3 million people in the United States should raise concern, because this may result in several thousand cases of HBVr each year in the United States alone even when conservative prevalence rate estimates of HBsAg positivity of 0.4% and past infection of 3% are used. B cell–depleting agents have also become increasingly implicated in HBVr.8Dong H.J. Ni L.N. Sheng G.F. et al.Risk of hepatitis B virus (HBV) reactivation in non-Hodgkin lymphoma patients receiving rituximab-chemotherapy: a meta-analysis.J Clin Virol. 2013; 57: 209-214Abstract Full Text Full Text PDF PubMed Scopus (85) Google Scholar These drugs are used in patients with a variety of disorders, including leukemias, non-Hodgkin lymphoma, cryoglobulinemia, rheumatoid arthritis, and idiopathic thrombocytopenic purpura. Frequent reporting of rituximab-associated HBVr has led to a recent expanded box warning by the Food and Drug Administration (FDA) in which it is strongly recommended that all patients who are to undergo B cell–depletion therapy initially be screened for HBV and, if positive, referred to a specialist to evaluate the need for antiviral prophylaxis or close monitoring.9Rituxan package brochure. Available at: http://druginserts.com/lib/rx/meds/rituxan-1/. Accessed November 17, 2014.Google Scholar Several highly effective nucleos(t)ide analogues to treat hepatitis B have been licensed in the past 10 years. The newer agents, such as entecavir and tenofovir, have low resistance profiles, no significant drug-drug interactions, and an excellent safety record, which makes them suitable for long-term use. Real-world experience with both of these agents in more than 1200 patients has confirmed efficacy and safety features that are similar to those presented in registration trials.10Pol S. Lampertico P. First-line treatment of chronic hepatitis B with entecavir or tenofovir in 'real-life' settings: from clinical trials to clinical practice.J Viral Hepat. 2012; 19: 377-386Crossref PubMed Scopus (103) Google Scholar Nephrotoxicity is a potential concern with long-term use of tenofovir, but creatinine clearance rates in the above large study were shown to remain stable over 4 years; 2000 IU.12Lok A.S. McMahon B.J. Chronic hepatitis B: update 2009.Hepatology. 2009; 50: 661-662Crossref PubMed Scopus (2416) Google Scholar, 13EASL clinical practice guidelines: management of chronic hepatitis B virus infection.J Hepatol. 2012; 57: 167-185Abstract Full Text Full Text PDF PubMed Scopus (2710) Google Scholar, 14Liaw Y.F. Kao J.H. Piratvisuth T. et al.Asia-Pacific consensus statement on the management of chronic hepatitis B: a 2012 update.Hepatol Int. 2012; 6: 531-561Crossref PubMed Scopus (816) Google Scholar Unfortunately, because many patients who undergo immunosuppressive therapy are not screened for HBV and therefore not given prophylactic antiviral therapy, guidelines about how long to treat are often nonactionable recommendations. Several surveys have indicated that HBV screening is frequently overlooked in patients who undergo long-term immunosuppressive therapy. Physician surveys have indicated that routine screening is performed by 20% to 40% of oncologists, 40% of dermatologists, and 70% of rheumatologists.16Hwang J.P. Fisch M.J. Zhang H. et al.Low rates of hepatitis B virus screening at the onset of chemotherapy.J Oncol Pract. 2012; 8: e32-e39Crossref PubMed Scopus (72) Google Scholar, 17Stine J.G. Bass M. Ibrahim D. et al.Dermatologists' awareness of and screening practices for hepatitis B virus infection before initiating tumor necrosis factor-alpha inhibitor therapy.South Med J. 2011; 104: 781-788Crossref PubMed Scopus (18) Google Scholar, 18Stine J.G. Khokhar O.S. Charalambopoulos J. et al.Rheumatologists' awareness of and screening practices for hepatitis B virus infection prior to initiating immunomodulatory therapy.Arthritis Care Res (Hoboken). 2010; 62: 704-711Crossref PubMed Scopus (48) Google Scholar There are currently no reliable data with regard to the frequency of screening by gastroenterologists or other specialists who use immunosuppressive therapies. However, in the past decade, the treatment of inflammatory bowel disease has been marked by increasing use of TNF inhibitors such as infliximab and several other biological agents are newly available,19Cohen L.B. Nanau R.M. Delzor F. et al.Biologic therapies in inflammatory bowel disease.Transl Res. 2014; 163: 533-556Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar making this a relevant issue to address with gastroenterologists as well as other stakeholders. The primary purpose of the current report is to provide a technical review and critical update of the topic of HBVr due to immunosuppressive drug treatment. This review will focus on the results of clinical studies of the most frequently used oncological therapies and on biological agents that are either currently in use or soon will be used in various medical specialties. The review will not discuss other immunosuppressive states that can shape the natural history of hepatitis B, such as coinfection with human immunodeficiency virus, and it will not include descriptions of clinical studies in the areas of solid organ transplantation or hematopoietic stem cell transplantation. Due to the clinical relevance of this area to gastroenterologists, a specific set of guidelines to prevent and treat HBVr in patients undergoing immunosuppressive drug treatment has been given high priority by the American Gastroenterological Association Institute. This is a 2-step process in which the previously mentioned technical review will be used to facilitate final recommendations. The American Gastroenterological Association is using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology in this technical review to assess the extent to which immunosuppressive drug therapy is likely to cause HBVr and the degree to which intervention can be anticipated to improve clinical outcomes.20Sultan S. Falck-Ytter Y. Inadomi J.M. The AGA institute process for developing clinical practice guidelines part one: grading the evidence.Clin Gastroenterol Hepatol. 2013; 11: 329-332Abstract Full Text Full Text PDF PubMed Scopus (62) Google Scholar This was accomplished using a set of predetermined PICO (population, intervention, comparison, and outcome) questions, defining the importance of outcomes and rating the quality of evidence for those outcomes across studies. An accompanying report in this issue of Gastroenterology integrates the results of this technical review with the other GRADE criteria to produce a set of recommendations.21Reddy K.R. Beavers K.L. Hammond S.P. et al.American Gastroenterological Association Institute guideline on the prevention and treatment of hepatitis B virus reactivation during immunosuppressive drug therapy.Gastroenterology. 2015; 148: 215-219Abstract Full Text Full Text PDF PubMed Scopus (432) Google Scholar HBVr is defined in this document as either the de novo appearance of HBV DNA in a patient previously known to have nondetectable HBV DNA or a ≥10-fold increase when compared with a baseline value. This virological definition does not require concomitant elevation of ALT level. Permissible surrogates were new detection of HBsAg or hepatitis B e antigen (HBeAg). Flares of hepatitis due to HBVr are defined as elevated ALT level at least 3 times the baseline level that at a minimum is beyond the reference range. An information specialist developed a literature search with input from the authors. All search results were imported using bibliographic management software for de-duplication and title and abstract screening. The following bibliographic databases were searched through the Ovid interface: EBM Reviews, Cochrane Central Register of Controlled Trials (July 2013), Cochrane Database of Systematic Reviews (July 2005 to July 2013), Health Technology Assessment (3rd quarter), EMBASE (1980 to 2013 week 35), and Ovid MEDLINE. We applied a search filter for systematic reviews, meta-analyses, and health technology assessments for questions on the use of antiviral therapy. The primary search was performed from July to September 2013 and included all reports from 1998 to 2013 using the following search terms: hepatitis B, HBVr, anti-HBc, rituximab, immunosuppressive therapy, cancer chemotherapy, biological modifiers, antiviral prophylaxis, lamivudine, entecavir, telbivudine, and tenofovir (see Appendix 1 for search strategy). The initial search revealed 744 reports and their corresponding titles and abstracts. The authors discarded 606 reports by sequentially examining the titles and then the abstracts, and if applicable, the full text of the reports was retrieved. Reasons for exclusion were inappropriate content, such as relevance to solid organ transplantation or antiviral therapy in cohorts who were not treated with immunosuppressive drug therapy. We also excluded reports on bone marrow transplantation or hematopoietic stem cell transplantation due to the greater awareness of reactivation risk status and treatment policy in both HBsAg-positive and anti-HBc–positive patients. Case reports, abstracts, or conference proceedings were not preferred and were only used when there was a marked paucity of data. The remaining 98 references were sorted according to whether they would provide useful information to assess the individual PICO questions (see Appendix 2 for the trial flow diagram). Major databases such as MEDLINE and conference reports were also searched by the authors for studies that addressed the baseline risk of HBVr and outcomes of interest in defined populations. Prevalence studies were not included in the final data analysis if they did not provide reasonable evidence for consecutive case reporting, if baseline HBV DNA data were unavailable, or if the study lacked definable criteria by which reactivation could be diagnosed. Editorials and letters were deselected, as were all observational studies in which it was believed that the study design could lead to an unacceptable level of confounding either in the diagnosis of reactivation or in the assessment of outcomes due to antiviral therapy. Next, the authors systematically reviewed and partitioned the evidence for each outcome across studies, assessed the quality of evidence for each outcome, and then presented the evidence to answer each specific PICO question. The quality of the evidence was classified into 4 GRADE22Guyatt G.H. Oxman A.D. Vist G.E. et al.GRADE: an emerging consensus on rating quality of evidence and strength of recommendations.BMJ. 2008; 336: 924-926Crossref PubMed Google Scholar categories (high, moderate, low, and very low), and a summary of the evidence was documented in GRADE evidence profiles using GRADEpro software.23GRADEpro software. 2009. http://gradeworkinggroup.org/.Google Scholar According to GRADE criteria, evidence from randomized controlled trials (RCTs) started at high quality but was rated down if there was serious risk of bias, inconsistency or heterogeneity, indirectness, imprecision, and potential publication bias. Evidence from observational studies started at low quality but could be rated at a higher level if there was a large effect size. Observational studies were considered to be primarily helpful in determining the baseline risk of HBVr and providing additional information on patient outcomes. PICO questions were devised by the authors and approved for further study by the American Gastroenterological Association Governing Board in July 2013. Each PICO question asks if an intervention affects patient outcomes in a positive or negative way, and each independently required a careful and coordinated search of the medical literature as described in the preceding text (Table 1). The following clinical outcomes were considered critical or important for decision making: severity of hepatitis; disease morbidity; resource utilization, including the need for hospitalization; liver-related mortality; and interruption of cancer chemotherapy or other immunosuppressive drug treatment (Table 1).Table 1PICO QuestionsInformal questionPICO questionPopulationInterventionComparatorOutcome1.Is antiviral prophylaxis needed for HBsAg-positive patients who will undergo ISDT?All HBsAg-positive patients on ISDT associated with low, moderate, or high risk of HBVrAll oral antiviral drugs or any combination thereofaSearch included lamivudine, emtricitabine, entecavir, tenofovir, adefovir, telbivudine, or combination therapy.No antiviral prophylaxisHBVr, liver disease morbidity and mortality; treatment interruption; resource usebIncludes liver transplantation for patients without malignancy.2.Is antiviral prophylaxis needed for HBsAg-negative, anti-HBc–positive patients who will undergo ISDT?Patients with isolated anti-HBc who undergo ISDT associated with low, moderate, or high risk of HBVrAll oral antiviral drugs or any combination thereofNo antiviral prophylaxisHBVr, liver disease morbidity and mortality; treatment interruption; resource use3.Does the presence of anti-HBs in addition to anti-HBc in HBsAg-negative patients confer additional protection against HBVr?Patients with anti-HBc and anti-HBs who will undergo ISDT associated with low, moderate, or high risk of HBVrAll oral antiviral drugs or any combination thereofNo antiviral prophylaxisHBVr, liver disease morbiditymorbidity and mortality; treatment interruption; resource use4.Is prophylactic treatment with third-generation nucleos(t)ide analogues more effective than treatment with first- or second-generation nucleos(t)ide agents?Any patient undergoing antiviral prophylaxis for hepatitis BLamivudineAdefovirTelbivudineEntecavirTenofovirHBVr, liver disease morbidity and mortality; treatment interruption; resource use5.Is HBV DNA monitoring followed by on-demand antiviral therapy as effective as prophylactic antiviral therapy?Any patient treated with ISDT (malignant and nonmalignant disorders)Regular HBVDNA testing during ISDT followed by antiviral therapy if HBVr occursProphylactic treatmentHBVr, liver disease morbidity and mortality; treatment interruption; resource use6.Is treatment of established HBVr with third-generation nucleos(t)ide agents more effective than treatment with first- or second-generation drugs?Patients with delayed treatmentLamivudineAdefovirTelbivudineEntecavirTenofovirHBVr, liver disease morbidity and mortality; treatment interruption; resource use7.Should patients who will undergo long-term ISDT be screened for HBV before starting treatment?Patients who will undergo more than 4 wk of ISDT for malignant and nonmalignant conditionsTesting of HBsAg and anti-HBcNo testingHBVr, liver disease morbidity and mortality; discrimination; treatment interruption; resource useISDT, immunosuppressive drug therapy.a Search included lamivudine, emtricitabine, entecavir, tenofovir, adefovir, telbivudine, or combination therapy.b Includes liver transplantation for patients without malignancy. Open table in a new tab ISDT, immunosuppressive drug therapy. The numerator and denominator for each critical and important outcome were extracted from each study by using pretested data extraction sheets listing acceptable definitions for outcomes, such as HBVr, hepatitis, liver failure, liver-related mortality, and chemotherapy interruptions. When possible, the pooled relative risk (RR) was calculated for each outcome by using the Mantel-Haenszel random effect model in RevMan 5.2.24Review Manager (RevMan). Version 5.2. The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen2012Google Scholar Funnel plots were inspected for heterogeneity in addition to formal analysis of heterogeneity (χ2, P < .1) and residual heterogeneity that was not explained by chance (I2). The number of studies was insufficient to formally test for funnel plot asymmetry to detect possible publication bias. Because relative effects of interventions usually are stable across differing baseline risks, we initially pooled the results of all RCTs using antiviral regimens versus placebo from different populations and different antiviral regimens (Figure 1).25Hsu C. Hsiung C.A. Su I.J. et al.A revisit of prophylactic lamivudine for chemotherapy-associated hepatitis B reactivation in non-Hodgkin's lymphoma: a randomized trial.Hepatology. 2008; 47: 844-853Crossref PubMed Scopus (266) Google Scholar, 26Huang Y.H. Hsiao L.T. Hong Y.C. et al.Randomized controlled trial of entecavir prophylaxis for rituximab-associated hepatitis B virus reactivation in patients with lymphoma and resolved hepatitis B.J Clin Oncol. 2013; 31: 2765-2772Crossref PubMed Scopus (258) Google Scholar, 27Jang J.W. Choi J.Y. Bae S.H. et al.A randomized controlled study of preemptive lamivudine in patients receiving transarterial chemo-lipiodolization.Hepatology. 2006; 43: 233-240Crossref PubMed Scopus (170) Google Scholar, 28Lau G.K. Yiu H.H. Fong D.Y. et al.Early is superior to deferred preemptive lamivudine therapy for hepatitis B patients undergoing chemotherapy.Gastroenterology. 2003; 125: 1742-1749Abstract Full Text Full Text PDF PubMed Scopus (394) Google Scholar, 29Long M. Jia W. Li S. et al.A single-center, prospective and randomized controlled study: Can the prophylactic use of lamivudine prevent hepatitis B virus reactivation in hepatitis B s-antigen seropositive breast cancer patients during chemotherapy?.Breast Cancer Res Treat. 2011; 127: 705-712Crossref PubMed Scopus (48) Google Scholar Because relative effects appeared similar and little or no heterogeneity across studies was seen, the pooled relative effects were applied to typical baseline risks from different populations (those that were seen in the included RCTs but also from clinical settings in which baseline risks were not available directly from RCTs) to arrive at representative risk difference that would be most suitable to inform clinical guidance. Because well-performed cohort studies from well-defined populations (eg, cancer or rheumatic disease) may provide accurate estimates of baseline risk of HBVr and because the risk of reactivation is markedly different based on the patient’s baseline HBV serological tests, a comprehensive review of those prevalence rates, mostly from observational studies, was performed. When pooled estimates of baseline risk were obtained from untreated control arms of RCTs in addition to well-performed cohort studies that enrolled consecutive, untreated patients, baseline risk was transformed to natural log proportions and pooled using the fixed effects inverse variance method in OpenMeta[analyst].30OpenMeta[Analyst]. Version 3.17.14. 2014. http://www.cebm.brown.edu/open_meta.Google Scholar The baseline risk of HBVr during immunosuppressive drug therapy can be assumed to be that which exists in the absence of an intervention to prevent HBVr. The baseline risk was assessed by review of 5 RCTs and 25 observational studies by extracting the number of HBVr events from the untreated or deferred treatment arms of those studies. A positive HBeAg status and high baseline level of serum HBV DNA have been shown to be predictive of reactivation after immunosuppressive drug therapy, which is consistent with a state of poorer immunologic control over viral replication before immune suppression.1Hoofnagle J.H. Reactivation of hepatitis B.Hepatology. 2009; 49: S156-S165Crossref PubMed Scopus (477) Google Scholar, 4Perrillo R.P. Acute flares in chronic hepatitis B: the natural and unnatural history of an immunologically mediated liver disease.Gastroenterology. 2001; 120: 1009-1022Abstract Full Text Full Text PDF PubMed Scopus (384) Google Scholar, 5Yeo W. Zee B. Zhong S. et al.Comprehensive analysis of risk factors associating with Hepatitis B virus (HBV) reactivation in cancer patients undergoing cytotoxic chemotherapy.Br J Cancer. 2004; 90: 1306-1311Crossref PubMed Scopus (273) Google Scholar The baseline risk is also determined in part by the potency of the immunosuppressive drug regimen, and this will be addressed in the following section on drug risk. The level of uncertainty around the baseline risk estimates from the available literature was assessed and the confidence in the estimate judged as low, moderate, or high. Answering any of the testing or intervention PICOs requires knowledge of the extent to which an immunosuppressive drug can be anticipated to cause HBVr. The authors classified drugs as low, moderate, or high risk based on the knowledge imparted from a review of all studies used in the analysis (Table 2). Low risk was considered to be evident whenever immunosuppressive drugs were in use for decades (eg, azathioprine) yet published reports either did not describe use of the drug as the sole agent responsible for HBVr (eg, reactivation occurred mostly or exclusively in combination with corticosteroids) or there were not a significant number of post-marketing reports. Use of a low-risk drug was anticipated to result in HBVr in <1% of cases for all drugs in this category and substantially 1% of cases but 10% of cases. Drugs that are in the same class as other high-risk drugs but in which there were no defined instances of HBVr in published reports or the FDA Adverse Event Reporting System were considered by the authors to be provisional high risk.Table 2Risk GroupsRisk g
Referência(s)