Journal of Viral Hepatitis, 2015
doi:10.1111/jvh.12402
Hepatitis B reactivation in HBsAg-negative/HBcAb-positive patients receiving rituximab for lymphoma: a meta-analysis L. Mozessohn,1 K. K. W. Chan,2 J. J. Feld3 and L. K. Hicks1
1
Division of Hematology/Oncology, St. Michael’s
2
Hospital, Toronto, ON, Canada; Division of Hematology/Oncology, Sunnybrook Odette Cancer Centre, Toronto, ON, Canada; and 3Toronto Centre for Liver Disease, Toronto Western Hospital Liver Centre, Toronto, ON, Canada Received December 2014; accepted for publication February 2015
SUMMARY. Patients with chronic hepatitis B (HBsAg-posi-
tive) are at risk of viral reactivation if rituximab is administered without antiviral treatment, a potentially fatal complication of treatment. Patients with so-called ‘resolved hepatitis B virus infection’ (HBsAg-negative/cAb-positive) may also be at risk. We performed a systematic review of the English and Chinese language literature to estimate the risk of hepatitis B virus (HBV) reactivation in HBsAg-negative/cAb-positive patients receiving rituximab for lymphoma. A pooled risk estimate was calculated for HBV reactivation. The impact of HBsAb status and study design on reactivation rates was explored. Data from 578 patients in 15 studies were included. ‘Clinical HBV reactivation’, (ALT >3 9 normal and either an increase in HBV DNA from baseline or HBsAg seroreversion), was estimated at 6.3% (I2 = 63%, P = 0.006). Significant heterogeneity was
Rituximab is a chimeric anti-CD20 antibody widely used in the treatment of B-cell non-Hodgkin’s lymphoma (NHL). The addition of rituximab to conventional chemotherapy results in a significant survival advantage in both indolent and aggressive NHL and has become the standard of care for these diseases [1–3]. Rituximab is also used to treat conditions such as rheumatoid arthritis and other autoimmune conditions [4,5]. Rituximab relatively spares neutrophils, but does cause prolonged B-cell depletion and thus is uniquely immunosuppressive [6]. In particular, when administered to an at-risk host, rituximab can trigger reactivation of hepatitis B virus (HBV), which is a well-recognized and potentially fatal complication of lymphoma treatment. Abbreviations: AASLD, American Association for the Study of Liver Diseases; CDC, Centers for Disease Control; CTCAE, Common Terminology Criteria for Adverse Events; FDA, Food and Drug Administration; HBV, hepatitis B virus; NCCN, National Comprehensive Cancer Network; NHL, non-Hodgkin’s lymphoma. Correspondence: Lee Mozessohn, Division of Hematology/Oncology, St. Michael’s Hospital, 30 Bond St, Room 2-084 Donnelly Wing, Toronto, ON M5B 1W8, Canada. E-mail: [email protected]
© 2015 John Wiley & Sons Ltd
detected. Reactivation rates were higher in prospective vs retrospective studies (14.2% vs 3.8%; OR = 4.39, 95% CI 0.83–23.28). Exploratory analyses found no effect of HBsAb status on reactivation risk (OR = 0.083; P = 0.151). Our meta-analysis confirms a measurable and potentially substantial risk of HBV reactivation in HBsAgnegative/cAb-positive patients exposed to rituximab. However, heterogeneity in the existing literature limits the generalizability of our findings. Large, prospective studies, with uniform definitions of HBV reactivation, are needed to clarify the risk of HBV reactivation in HBsAg-negative/ cAb-positive patients. Keywords: hepatitis B, hepatitis B reactivation, non-Hodgkin’s lymphoma, rituximab.
Chronic HBV (HBsAg-positive) affects 350 million people worldwide, many of whom are asymptomatic and unaware that they are infected [7]. The prevalence of patients with so-called ‘resolved HBV infection’ (HBsAg-negative/cAbpositive) is much higher – up to 60% in endemic areas [8– 11]. These individuals have been exposed to the virus and have mounted a successful immune response to control viral replication. They have an excellent long-term prognosis; however, they do harbour trace amounts of replication-competent HBV DNA in their liver, likely lifelong. With potent immunosuppression, immune control may be lost, leading to HBV reactivation with or without reappearance of HBsAg (seroreversion) [12–14]. The risk of HBV reactivation among patients with lymphoma and chronic, HBsAg-positive-HBV is approximately 50% with conventional chemotherapy [15]. Accurate risk estimates in patients receiving rituximab-based chemotherapy are unknown, largely due to an unwillingness to expose HBsAg-positive patients to chemoimmunotherapy without viral prophylaxis; it is likely that the risk of HBV reactivation in this setting is greater than 50%. Pre-emptive use of nucleoside analogues such as lamivudine can largely prevent HBV reactivation in patients with
2
L. Mozessohn et al.
chronic HBV [16–19]. As a result, the Food and Drug Administration (FDA), the Centers for Disease Control (CDC), the National Comprehensive Cancer Network (NCCN) and the American Association for the Study of Liver Diseases (AASLD) all recommend screening for HBV prior to the administration of anti-CD20 therapy [20–23]. Moreover, a recent cost-effectiveness analysis suggests that screening for HBV prior to lymphoma treatment is not only clinically effective, but is also cost saving [15]. Nonetheless, although many groups advocate screening for HBcAb prior to rituximab-based treatment, the risk of reactivation in HBsAg-negative/cAb-positive patients is unknown, and management strategies in this more prevalent population have not been fully agreed upon by various organizations. Prior to the widespread use of rituximab, reactivation in patients with resolved HBV (HBsAg-negative, HBcAb-positive) was a relatively rare event and was largely limited to those undergoing hematopoietic stem cell transplantation [24–27]. However, with the increased use of rituximabcontaining regimens, there have been a growing number of reports describing HBV reactivation in patients with resolved HBV – some of which have been fatal [28–30]. We completed a systematic review of the English and Chinese literature to estimate the risk of HBV reactivation in HBsAg-negative/cAb-positive patients receiving rituximab-based therapy for lymphoma. We also explored factors that may increase or mitigate risk, and the extent to which study design may have influenced risk estimates.
MATERIALS AND METHODS Literature search strategy Ovid MEDLINE (1996 to June 2013 [week 3]) and EMBASE (1996 to June 2013 [week 26]) were searched using the medical subject headings ‘lymphoma’ and ‘hepatitis B’. Search results were limited to English and Chinese language publications.
Study selection and data extraction Eligible studies were limited to those describing HBsAg-negative/cAb-positive patients receiving rituximab for lymphoma. We excluded reviews, case reports and case series with less than five patients. A single author (LM) reviewed the titles and abstracts of all identified studies. Two authors (LM and LH) independently reviewed the full text of all studies identified as potentially eligible during title and abstract screening. Agreement was reached through consensus. A third author (KC) reviewed the titles of Chinese language studies and provided translation of those identified as potentially eligible. Reference lists of included studies were reviewed for additional studies meeting eligibility criteria.
Data were independently abstracted from source papers by LM and LH. Differences were resolved through consensus. Corresponding authors were contacted when data were unavailable in published papers.
Outcome measures The primary outcome for this review was the rate of HBV reactivation in HBsAg-negative/cAb-positive patients receiving rituximab plus chemotherapy for lymphoma. For this meta-analysis, two definitions of HBV reactivation were used. ‘Clinical HBV reactivation’ was defined as an increase in ALT to greater than three times the upper limit of normal and either a documented increase in HBV DNA from baseline or HBsAg seroreversion. ‘Preclinical HBV reactivation’ was defined as an increase in HBV DNA from baseline (including a change from undetectable to detectable) and/or HBsAg seroreversion with or without an abnormal ALT level. Studies were excluded from the ‘preclinical’ definition if they did not explicitly state that HBV DNA was part of routine post-rituximab surveillance to avoid a biased risk estimate. The effect of HBsAb status on hepatitis B reactivation risk was explored.
Statistical analysis Probabilities of reactivation were pooled using a random effects model using the method of DerSimonian and Laird [31]. Meta-analyses were performed using OpenMetaAnalyst (http://www.cebm.brown.edu/open_meta) [32]. The effect of HBsAb on reactivation rates was examined with mixed effect logistic regression with a random effects model for studies where HBsAb data were available. This analysis was conducted using SAS version 9.3 (Cary, NC, USA).
Assessment of heterogeneity The Cochrane chi-square for heterogeneity was used to determine whether significant statistical heterogeneity existed in the pooled estimates. Statistical heterogeneity was considered significant when the test produced a P value 3 9 ULN and either increase in HBV DNA from baseline or HBsAg seroreversion. †Viral expansion (increased HBV DNA from baseline including change from undetectable to detectable) and/or HBsAg seroreversion +/ increase in ALT. ‡Unable to calculate rate because did not systematically monitor HBV DNA levels in all patients at all visits. §Unable to calculate rate because treatment trial (if DNA increased from baseline, started on antiviral irrespective of transaminase level). ¶Unable to calculate rate because of missing data from study.
Fig. 2 Pooled risk of ‘clinical’ HBV reactivation among HBsAg /cAb+ patients receiving rituximab for lymphoma.
HBV reactivation rate of 6.1% (14 reactivations of 260 patients) vs 9.0% (4 reactivations of 81 patients) in European studies. There was only 1 North American study and as such, it was not included in the geographic analysis. Significant heterogeneity was again observed suggesting that the variability observed is unlikely to be due to variation in viral genotype alone. We explored whether or not the presence or absence of antibodies to hepatitis B surface antigen (HBsAb status)
was predictive of HBV reactivation risk. Using the clinical definition of reactivation, 3 studies (145 patients) had sufficient data to be included in this analysis. There was a nonstatistically significant trend to a lower risk of HBV reactivation in patients with measurable antibodies to HBsAg (OR = 0.083; P = 0.151). In addition, mortality data were available in only 6 studies, and thus, it was not possible to compare mortality in ‘clinical HBV reactivation’ vs ‘preclinical HBV reactivation’. © 2015 John Wiley & Sons Ltd
4
L. Mozessohn et al.
Table 1 Studies with HBsAg /HBcAb+ patients receiving rituximab for lymphoma
Study
Study design
Persico, 2006 [42] Yeo, 2009 [44] Francisci, 2010 [33] Niitsu, 2010 [40] Koo, 2011 [37] Huang, 2013 [11] Targhetta, 2008 [43] Hanbali, 2008 [35] Fukushima, 2009 [34]
Prospective Prospective Prospective Prospective Prospective Prospective Retrospective Retrospective Retrospective and prospective Retrospective Retrospective Retrospective Retrospective Retrospective Retrospective
Ji, 2010 [36] Mendez-Navarro, 2010 [39] Matsue, 2010 [38] Watanabe, 2011 [46] Zhu, 2012 [45] Oh, 2013 [41]
Country of origin
Total HBcAb+/ HBsAg patients
Reactivation rate with ‘clinical’ definition*
Reactivation rate with ‘preclinical’ definition†
Italy Hong Kong Italy Japan Singapore Taiwan Italy USA Japan
7 21 38 43 62 39 74 26 32
2/7 (28.9%) 5/21 (23.8%) N/A§ N/A§ 2/62 (3.2%) N/A§ 2/74 (2.7%) N/A¶ 1/11 (9.1%)
2/7 (28.9%) N/A‡ 3/38 (7.9%) 6/43 (14.0%) N/A‡ 7/39 (17.9%) N/A‡ N/A¶ 1/21 (4.8%)
China USA Japan Japan China Korea
43 25 56 20 25 67
1/43 0/25 3/56 N/A§ N/A§ 2/67
N/A‡ N/A‡ N/A‡ 5/20 (25.0%) N/A‡ N/A‡
(2.3%) (0%) (5.4%)
(3.0%)
*Increase in ALT >3 9 ULN and either increase in HBV DNA from baseline or HBsAg seroreversion. †Viral expansion (increased HBV DNA from baseline including change from undetectable to detectable) and/or HBsAg seroreversion +/ increase in ALT. ‡Unable to calculate rate because did not systematically monitor HBV DNA levels in all patients at all visits. §Unable to calculate rate because treatment trial (if DNA increased from baseline, started on antiviral irrespective of transaminase level). ¶Unable to calculate rate because of missing data from study.
Fig. 2 Pooled risk of ‘clinical’ HBV reactivation among HBsAg /cAb+ patients receiving rituximab for lymphoma.
HBV reactivation rate of 6.1% (14 reactivations of 260 patients) vs 9.0% (4 reactivations of 81 patients) in European studies. There was only 1 North American study and as such, it was not included in the geographic analysis. Significant heterogeneity was again observed suggesting that the variability observed is unlikely to be due to variation in viral genotype alone. We explored whether or not the presence or absence of antibodies to hepatitis B surface antigen (HBsAb status)
was predictive of HBV reactivation risk. Using the clinical definition of reactivation, 3 studies (145 patients) had sufficient data to be included in this analysis. There was a nonstatistically significant trend to a lower risk of HBV reactivation in patients with measurable antibodies to HBsAg (OR = 0.083; P = 0.151). In addition, mortality data were available in only 6 studies, and thus, it was not possible to compare mortality in ‘clinical HBV reactivation’ vs ‘preclinical HBV reactivation’. © 2015 John Wiley & Sons Ltd
6
L. Mozessohn et al.
types are found in China, while Japan and Korea have almost exclusively genotype B and C isolates, respectively. In Europe, genotypes A and D predominate. Although there are differences in clinical outcome between the different genotypes including the risk of cirrhosis and hepatocellular carcinoma, whether this would affect the probability of reactivation during immunosuppression is unknown and overall it seems unlikely that genotype is a major factor in the variable HBV reactivation rates reported in different studies. Ultimately, the most important source of heterogeneity in this study was likely study design. As described above, we attempted to standardize the definitions of ‘clinical’ and ‘pre-clinical’ of HBV reactivation – however, because this was not always consistent with the definition of HBV reactivation used by individual studies, the reliability of data may have been affected. As well, studies varied with regard to how closely they monitored patients for HBV reactivation, which may have influenced the reported rates of reactivation. Finally, how long patients were monitored for HBV reactivation varied between studies. Recent data suggest that a risk of HBV reactivation persists for months after rituximab therapy is concluded [11] – thus studies that followed patients for longer may be more likely to recognize HBV reactivation. We observed a difference in the rate of HBV reactivation reported by prospective studies, as compared to retrospective studies. Larger, prospective studies tended to report higher rates of HBV reactivation. For example, in a prospective study comparing the rate of HBV reactivation in lymphoma patients receiving rituximab-based chemoimmunotherapy vs chemotherapy alone, there were significantly more reactivations in the rituximab group (23.8% vs 0%, respectively) [44]. This study defined reactivation as a threefold or greater increase in serum ALT levels with evidence of HBsAg seroreversion and an increase in HBV DNA levels when compared to baseline. The higher rate of HBV reactivation observed in prospective studies likely reflects closer, systematic surveillance for HBV reactivation among the prospective studies. As a result, it is likely that the available prospective studies provide an estimate of the risk of HBV reactivation that is closer to the true risk. Recently, Dong et al. [51] estimated the HBV reactivation risk in lymphoma patients receiving rituximab. Their analysis focused on the HBsAg-positive population, but a small subset of studies included in the paper described reactivation rates in HBsAg-negative, HBcAb-positive patients. In this group, they estimated the relative risk of HBV reactivation as 5.52 (for rituximab vs no rituximab
exposure) combining retrospective and prospective studies. However, they defined HBV reactivation as a >10-fold rise in serum HBV DNA levels with an increase in serum ALT compared to baseline, a somewhat less stringent criterion than our definition for ‘clinical reactivation’. Our analysis also includes additional trials not included in their estimate, including the only randomized controlled trial of viral prophylaxis in this population [11]. Nonetheless, our results are consistent with the Dong paper in both of us report an important increased risk of HBV reactivation among HBsAg-negative, HBcAb-positive patients exposed to rituximab. At present, it is not possible to precisely estimate the risk of HBV reactivation in patients with so-called ‘resolved HBV’ who are receiving rituximab due to marked heterogeneity in existing studies; nonetheless, it is evident that reactivation does occur, and that the risk is likely clinically significant. The best evidence for recommendations regarding surveillance is from the recent randomized controlled trial examining antiviral prophylaxis [11]. In this paper, they reported superior outcomes with antiviral prophylaxis compared to therapeutic antiviral treatment at the time of reactivation. However, the therapeutic arm appeared safe. In that trial, serum markers (liver function, HBV viral load and HBsAg) were monitored during each cycle of chemoimmunotherapy (every 3 weeks). Upon completion of treatment, serum markers were monitored every month during follow-up for 12 months. Pending further trials, it seems reasonable to adopt this as a minimum surveillance strategy in patients not given antiviral prophylaxis at the start of rituximabbased treatment. An evaluation of this strategy to postpone antiviral prophylaxis until a detected increase in viral load and/or HBsAg seroreversion occurs should be confirmed in future prospective studies. However, given the substantial heterogeneity between studies, the field would benefit from the development of consensus diagnostic criteria for ‘preclinical’ and ‘clinical HBV reactivation’ by leading organizations. Using these standardized definitions, prospective studies should then be conducted with the goal of more accurately defining the risk of reactivation and the duration of time during which reactivations can occur.
DISCLOSURES KC, JF and LH have received research funding from Gilead. No other relevant disclosures.
REFERENCES 1 Coiffier B, Thieblemont C, Van Den Neste E et al. Long-term outcome of patients in the LNH-98.5 trial, the
first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a
study by the Groupe d’Etudes des Lymphomes de l’Adulte. Blood 2010; 116(12): 2040–2045.
© 2015 John Wiley & Sons Ltd
HBV reactivation 2 Marcus R, Imrie K, Solal-Celigny P et al. Phase III study of R-CVP compared with cyclophosphamide, vincristine, and prednisone alone in patients with previously untreated advanced follicular lymphoma. J Clin Oncol 2008; 26(28): 4579–4586. 3 Schulz H, Bohlius JF, Trelle S et al. Immunochemotherapy with rituximab and overall survival in patients with indolent or mantle cell lymphoma: a systematic review and meta-analysis. J Natl Cancer Inst 2007; 99(9): 706–714. 4 Edwards JC, Szczepanski L, Szechinski J et al. Efficacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med 2004; 350(25): 2572–2581. 5 Cohen SB, Emery P, Greenwald MW et al. Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: results of a multicenter, randomized, doubleblind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks. Arthritis Rheum 2006; 54(9): 2793–2806. 6 Maloney DG. Anti-CD20 antibody therapy for B-cell lymphomas. N Engl J Med 2012; 366(21): 2008–2016. 7 Lavanchy D. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J Viral Hepat 2004; 11(2): 97–107. 8 Liu CJ, Chen DS, Chen PJ. Epidemiology of HBV infection in Asian blood donors: emphasis on occult HBV infection and the role of NAT. J Clin Virol 2006; 36(Suppl. 1): S33–S44. 9 Uhm JE, Kim K, Lim TK et al. Changes in serologic markers of hepatitis B following autologous hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 2007; 13(4): 463–468. 10 Hsu C, Tsou HH, Lin SJ et al. Chemotherapy-induced hepatitis B reactivation in lymphoma patients with resolved HBV infection: a prospective study. Hepatology 2014; 59(6): 2092–2100. 11 Huang YH, Hsiao LT, Hong YC et al. Randomized controlled trial of entecavir prophylaxis for rituximabassociated hepatitis B virus reactivation in patients with lymphoma and
© 2015 John Wiley & Sons Ltd
12
13
14
15
16
17
18
19
20
21
resolved hepatitis B. J Clin Oncol 2013; 31(22): 2765–2772. Li L, Chen PJ, Chen MH, Chak KF, Lin KS, Tsai SJ. A pilot study for screening blood donors in Taiwan by nucleic acid amplification technology: detecting occult hepatitis B virus infections and closing the serologic window period for hepatitis C virus. Transfusion 2008; 48(6): 1198–1206. Hui CK, Sun J, Au WY et al. Occult hepatitis B virus infection in hematopoietic stem cell donors in a hepatitis B virus endemic area. J Hepatol 2005; 42(6): 813–819. Zhu Y, Yamamoto T, Cullen J et al. Kinetics of hepadnavirus loss from the liver during inhibition of viral DNA synthesis. J Virol 2001; 75(1): 311–322. Zurawska U, Hicks LK, Woo G et al. Hepatitis B virus screening before chemotherapy for lymphoma: a cost-effectiveness analysis. J Clin Oncol 2012; 30(26): 3167–3173. Lau GK, Yiu HH, Fong DY et al. Early is superior to deferred preemptive lamivudine therapy for hepatitis B patients undergoing chemotherapy. Gastroenterology 2003; 125(6): 1742–1749. Hsu C, Hsiung CA, Su IJ et al. A revisit of prophylactic lamivudine for chemotherapy-associated hepatitis B reactivation in non-Hodgkin’s lymphoma: a randomized trial. Hepatology 2008; 47(3): 844–853. Loomba R, Rowley A, Wesley R et al. Systematic review: the effect of preventive lamivudine on hepatitis B reactivation during chemotherapy. Ann Intern Med 2008; 148(7): 519–528. Coffin CS, Fung SK, Ma MM. Canadian Association for the Study of the L. Management of chronic hepatitis B: Canadian Association for the Study of the Liver consensus guidelines. Can J Gastroenterol 2012; 26(12): 917–938. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Hodgkin’s Lymphomas Version 2.2013. Available at: http:// www.nccn.org/professionals/physi cian_gls/pdf/nhl.pdf (accessed 15 December 2013). Lok ASF, McMahon BJ. Chronic Hepatitis B: Update 2009. American
22
23
24
25
26
27
28
29
7
Association for the Study of Liver Diseases. Available at: http:// www.aasld.org (accessed 15 December 2013). “Arzerra (ofatumumab) and Rituxan (rituximab): Drug Safety Communication - New Boxed Warning, Recommendations to Decrease Risk of Hepatitis B Reactivation”. Available at: http://www.fda.gov (accessed 15 October 2013). “Recommendations for Routine Testing and Follow-up for Chronic Hepatitis B Virus (HBV) Infection”. Available at: http://www.cdc.gov (accessed 10 January 2014). Lalazar G, Rund D, Shouval D. Screening, prevention and treatment of viral hepatitis B reactivation in patients with haematological malignancies. Br J Haematol 2007; 136(5): 699–712. Liang R. How I treat and monitor viral hepatitis B infection in patients receiving intensive immunosuppressive therapies or undergoing hematopoietic stem cell transplantation. Blood 2009; 113(14): 3147–3153. Chen PM, Fan S, Liu CJ et al. Changing of hepatitis B virus markers in patients with bone marrow transplantation. Transplantation 1990; 49(4): 708–713. Onozawa M, Hashino S, Izumiyama K et al. Progressive disappearance of anti-hepatitis B surface antigen antibody and reverse seroconversion after allogeneic hematopoietic stem cell transplantation in patients with previous hepatitis B virus infection. Transplantation 2005; 79(5): 616–619. Wu JM, Huang YH, Lee PC, Lin HC, Lee SD. Fatal reactivation of hepatitis B virus in a patient who was hepatitis B surface antigen negative and core antibody positive before receiving chemotherapy for nonHodgkin lymphoma. J Clin Gastroenterol 2009; 43(5): 496–498. Law JK, Ho JK, Hoskins PJ, Erb SR, Steinbrecher UP, Yoshida EM. Fatal reactivation of hepatitis B post-chemotherapy for lymphoma in a hepatitis B surface antigen-negative, hepatitis B core antibody-positive patient: potential implications for future prophylaxis recommendations. Leuk Lymphoma 2005; 46(7): 1085–1089.
6
L. Mozessohn et al.
types are found in China, while Japan and Korea have almost exclusively genotype B and C isolates, respectively. In Europe, genotypes A and D predominate. Although there are differences in clinical outcome between the different genotypes including the risk of cirrhosis and hepatocellular carcinoma, whether this would affect the probability of reactivation during immunosuppression is unknown and overall it seems unlikely that genotype is a major factor in the variable HBV reactivation rates reported in different studies. Ultimately, the most important source of heterogeneity in this study was likely study design. As described above, we attempted to standardize the definitions of ‘clinical’ and ‘pre-clinical’ of HBV reactivation – however, because this was not always consistent with the definition of HBV reactivation used by individual studies, the reliability of data may have been affected. As well, studies varied with regard to how closely they monitored patients for HBV reactivation, which may have influenced the reported rates of reactivation. Finally, how long patients were monitored for HBV reactivation varied between studies. Recent data suggest that a risk of HBV reactivation persists for months after rituximab therapy is concluded [11] – thus studies that followed patients for longer may be more likely to recognize HBV reactivation. We observed a difference in the rate of HBV reactivation reported by prospective studies, as compared to retrospective studies. Larger, prospective studies tended to report higher rates of HBV reactivation. For example, in a prospective study comparing the rate of HBV reactivation in lymphoma patients receiving rituximab-based chemoimmunotherapy vs chemotherapy alone, there were significantly more reactivations in the rituximab group (23.8% vs 0%, respectively) [44]. This study defined reactivation as a threefold or greater increase in serum ALT levels with evidence of HBsAg seroreversion and an increase in HBV DNA levels when compared to baseline. The higher rate of HBV reactivation observed in prospective studies likely reflects closer, systematic surveillance for HBV reactivation among the prospective studies. As a result, it is likely that the available prospective studies provide an estimate of the risk of HBV reactivation that is closer to the true risk. Recently, Dong et al. [51] estimated the HBV reactivation risk in lymphoma patients receiving rituximab. Their analysis focused on the HBsAg-positive population, but a small subset of studies included in the paper described reactivation rates in HBsAg-negative, HBcAb-positive patients. In this group, they estimated the relative risk of HBV reactivation as 5.52 (for rituximab vs no rituximab
exposure) combining retrospective and prospective studies. However, they defined HBV reactivation as a >10-fold rise in serum HBV DNA levels with an increase in serum ALT compared to baseline, a somewhat less stringent criterion than our definition for ‘clinical reactivation’. Our analysis also includes additional trials not included in their estimate, including the only randomized controlled trial of viral prophylaxis in this population [11]. Nonetheless, our results are consistent with the Dong paper in both of us report an important increased risk of HBV reactivation among HBsAg-negative, HBcAb-positive patients exposed to rituximab. At present, it is not possible to precisely estimate the risk of HBV reactivation in patients with so-called ‘resolved HBV’ who are receiving rituximab due to marked heterogeneity in existing studies; nonetheless, it is evident that reactivation does occur, and that the risk is likely clinically significant. The best evidence for recommendations regarding surveillance is from the recent randomized controlled trial examining antiviral prophylaxis [11]. In this paper, they reported superior outcomes with antiviral prophylaxis compared to therapeutic antiviral treatment at the time of reactivation. However, the therapeutic arm appeared safe. In that trial, serum markers (liver function, HBV viral load and HBsAg) were monitored during each cycle of chemoimmunotherapy (every 3 weeks). Upon completion of treatment, serum markers were monitored every month during follow-up for 12 months. Pending further trials, it seems reasonable to adopt this as a minimum surveillance strategy in patients not given antiviral prophylaxis at the start of rituximabbased treatment. An evaluation of this strategy to postpone antiviral prophylaxis until a detected increase in viral load and/or HBsAg seroreversion occurs should be confirmed in future prospective studies. However, given the substantial heterogeneity between studies, the field would benefit from the development of consensus diagnostic criteria for ‘preclinical’ and ‘clinical HBV reactivation’ by leading organizations. Using these standardized definitions, prospective studies should then be conducted with the goal of more accurately defining the risk of reactivation and the duration of time during which reactivations can occur.
DISCLOSURES KC, JF and LH have received research funding from Gilead. No other relevant disclosures.
REFERENCES 1 Coiffier B, Thieblemont C, Van Den Neste E et al. Long-term outcome of patients in the LNH-98.5 trial, the
first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a
study by the Groupe d’Etudes des Lymphomes de l’Adulte. Blood 2010; 116(12): 2040–2045.
© 2015 John Wiley & Sons Ltd
doi:10.1111/jvh.12402
Hepatitis B reactivation in HBsAg-negative/HBcAb-positive patients receiving rituximab for lymphoma: a meta-analysis L. Mozessohn,1 K. K. W. Chan,2 J. J. Feld3 and L. K. Hicks1
1
Division of Hematology/Oncology, St. Michael’s
2
Hospital, Toronto, ON, Canada; Division of Hematology/Oncology, Sunnybrook Odette Cancer Centre, Toronto, ON, Canada; and 3Toronto Centre for Liver Disease, Toronto Western Hospital Liver Centre, Toronto, ON, Canada Received December 2014; accepted for publication February 2015
SUMMARY. Patients with chronic hepatitis B (HBsAg-posi-
tive) are at risk of viral reactivation if rituximab is administered without antiviral treatment, a potentially fatal complication of treatment. Patients with so-called ‘resolved hepatitis B virus infection’ (HBsAg-negative/cAb-positive) may also be at risk. We performed a systematic review of the English and Chinese language literature to estimate the risk of hepatitis B virus (HBV) reactivation in HBsAg-negative/cAb-positive patients receiving rituximab for lymphoma. A pooled risk estimate was calculated for HBV reactivation. The impact of HBsAb status and study design on reactivation rates was explored. Data from 578 patients in 15 studies were included. ‘Clinical HBV reactivation’, (ALT >3 9 normal and either an increase in HBV DNA from baseline or HBsAg seroreversion), was estimated at 6.3% (I2 = 63%, P = 0.006). Significant heterogeneity was
Rituximab is a chimeric anti-CD20 antibody widely used in the treatment of B-cell non-Hodgkin’s lymphoma (NHL). The addition of rituximab to conventional chemotherapy results in a significant survival advantage in both indolent and aggressive NHL and has become the standard of care for these diseases [1–3]. Rituximab is also used to treat conditions such as rheumatoid arthritis and other autoimmune conditions [4,5]. Rituximab relatively spares neutrophils, but does cause prolonged B-cell depletion and thus is uniquely immunosuppressive [6]. In particular, when administered to an at-risk host, rituximab can trigger reactivation of hepatitis B virus (HBV), which is a well-recognized and potentially fatal complication of lymphoma treatment. Abbreviations: AASLD, American Association for the Study of Liver Diseases; CDC, Centers for Disease Control; CTCAE, Common Terminology Criteria for Adverse Events; FDA, Food and Drug Administration; HBV, hepatitis B virus; NCCN, National Comprehensive Cancer Network; NHL, non-Hodgkin’s lymphoma. Correspondence: Lee Mozessohn, Division of Hematology/Oncology, St. Michael’s Hospital, 30 Bond St, Room 2-084 Donnelly Wing, Toronto, ON M5B 1W8, Canada. E-mail: [email protected]
© 2015 John Wiley & Sons Ltd
detected. Reactivation rates were higher in prospective vs retrospective studies (14.2% vs 3.8%; OR = 4.39, 95% CI 0.83–23.28). Exploratory analyses found no effect of HBsAb status on reactivation risk (OR = 0.083; P = 0.151). Our meta-analysis confirms a measurable and potentially substantial risk of HBV reactivation in HBsAgnegative/cAb-positive patients exposed to rituximab. However, heterogeneity in the existing literature limits the generalizability of our findings. Large, prospective studies, with uniform definitions of HBV reactivation, are needed to clarify the risk of HBV reactivation in HBsAg-negative/ cAb-positive patients. Keywords: hepatitis B, hepatitis B reactivation, non-Hodgkin’s lymphoma, rituximab.
Chronic HBV (HBsAg-positive) affects 350 million people worldwide, many of whom are asymptomatic and unaware that they are infected [7]. The prevalence of patients with so-called ‘resolved HBV infection’ (HBsAg-negative/cAbpositive) is much higher – up to 60% in endemic areas [8– 11]. These individuals have been exposed to the virus and have mounted a successful immune response to control viral replication. They have an excellent long-term prognosis; however, they do harbour trace amounts of replication-competent HBV DNA in their liver, likely lifelong. With potent immunosuppression, immune control may be lost, leading to HBV reactivation with or without reappearance of HBsAg (seroreversion) [12–14]. The risk of HBV reactivation among patients with lymphoma and chronic, HBsAg-positive-HBV is approximately 50% with conventional chemotherapy [15]. Accurate risk estimates in patients receiving rituximab-based chemotherapy are unknown, largely due to an unwillingness to expose HBsAg-positive patients to chemoimmunotherapy without viral prophylaxis; it is likely that the risk of HBV reactivation in this setting is greater than 50%. Pre-emptive use of nucleoside analogues such as lamivudine can largely prevent HBV reactivation in patients with
2
L. Mozessohn et al.
chronic HBV [16–19]. As a result, the Food and Drug Administration (FDA), the Centers for Disease Control (CDC), the National Comprehensive Cancer Network (NCCN) and the American Association for the Study of Liver Diseases (AASLD) all recommend screening for HBV prior to the administration of anti-CD20 therapy [20–23]. Moreover, a recent cost-effectiveness analysis suggests that screening for HBV prior to lymphoma treatment is not only clinically effective, but is also cost saving [15]. Nonetheless, although many groups advocate screening for HBcAb prior to rituximab-based treatment, the risk of reactivation in HBsAg-negative/cAb-positive patients is unknown, and management strategies in this more prevalent population have not been fully agreed upon by various organizations. Prior to the widespread use of rituximab, reactivation in patients with resolved HBV (HBsAg-negative, HBcAb-positive) was a relatively rare event and was largely limited to those undergoing hematopoietic stem cell transplantation [24–27]. However, with the increased use of rituximabcontaining regimens, there have been a growing number of reports describing HBV reactivation in patients with resolved HBV – some of which have been fatal [28–30]. We completed a systematic review of the English and Chinese literature to estimate the risk of HBV reactivation in HBsAg-negative/cAb-positive patients receiving rituximab-based therapy for lymphoma. We also explored factors that may increase or mitigate risk, and the extent to which study design may have influenced risk estimates.
MATERIALS AND METHODS Literature search strategy Ovid MEDLINE (1996 to June 2013 [week 3]) and EMBASE (1996 to June 2013 [week 26]) were searched using the medical subject headings ‘lymphoma’ and ‘hepatitis B’. Search results were limited to English and Chinese language publications.
Study selection and data extraction Eligible studies were limited to those describing HBsAg-negative/cAb-positive patients receiving rituximab for lymphoma. We excluded reviews, case reports and case series with less than five patients. A single author (LM) reviewed the titles and abstracts of all identified studies. Two authors (LM and LH) independently reviewed the full text of all studies identified as potentially eligible during title and abstract screening. Agreement was reached through consensus. A third author (KC) reviewed the titles of Chinese language studies and provided translation of those identified as potentially eligible. Reference lists of included studies were reviewed for additional studies meeting eligibility criteria.
Data were independently abstracted from source papers by LM and LH. Differences were resolved through consensus. Corresponding authors were contacted when data were unavailable in published papers.
Outcome measures The primary outcome for this review was the rate of HBV reactivation in HBsAg-negative/cAb-positive patients receiving rituximab plus chemotherapy for lymphoma. For this meta-analysis, two definitions of HBV reactivation were used. ‘Clinical HBV reactivation’ was defined as an increase in ALT to greater than three times the upper limit of normal and either a documented increase in HBV DNA from baseline or HBsAg seroreversion. ‘Preclinical HBV reactivation’ was defined as an increase in HBV DNA from baseline (including a change from undetectable to detectable) and/or HBsAg seroreversion with or without an abnormal ALT level. Studies were excluded from the ‘preclinical’ definition if they did not explicitly state that HBV DNA was part of routine post-rituximab surveillance to avoid a biased risk estimate. The effect of HBsAb status on hepatitis B reactivation risk was explored.
Statistical analysis Probabilities of reactivation were pooled using a random effects model using the method of DerSimonian and Laird [31]. Meta-analyses were performed using OpenMetaAnalyst (http://www.cebm.brown.edu/open_meta) [32]. The effect of HBsAb on reactivation rates was examined with mixed effect logistic regression with a random effects model for studies where HBsAb data were available. This analysis was conducted using SAS version 9.3 (Cary, NC, USA).
Assessment of heterogeneity The Cochrane chi-square for heterogeneity was used to determine whether significant statistical heterogeneity existed in the pooled estimates. Statistical heterogeneity was considered significant when the test produced a P value 3 9 ULN and either increase in HBV DNA from baseline or HBsAg seroreversion. †Viral expansion (increased HBV DNA from baseline including change from undetectable to detectable) and/or HBsAg seroreversion +/ increase in ALT. ‡Unable to calculate rate because did not systematically monitor HBV DNA levels in all patients at all visits. §Unable to calculate rate because treatment trial (if DNA increased from baseline, started on antiviral irrespective of transaminase level). ¶Unable to calculate rate because of missing data from study.
Fig. 2 Pooled risk of ‘clinical’ HBV reactivation among HBsAg /cAb+ patients receiving rituximab for lymphoma.
HBV reactivation rate of 6.1% (14 reactivations of 260 patients) vs 9.0% (4 reactivations of 81 patients) in European studies. There was only 1 North American study and as such, it was not included in the geographic analysis. Significant heterogeneity was again observed suggesting that the variability observed is unlikely to be due to variation in viral genotype alone. We explored whether or not the presence or absence of antibodies to hepatitis B surface antigen (HBsAb status)
was predictive of HBV reactivation risk. Using the clinical definition of reactivation, 3 studies (145 patients) had sufficient data to be included in this analysis. There was a nonstatistically significant trend to a lower risk of HBV reactivation in patients with measurable antibodies to HBsAg (OR = 0.083; P = 0.151). In addition, mortality data were available in only 6 studies, and thus, it was not possible to compare mortality in ‘clinical HBV reactivation’ vs ‘preclinical HBV reactivation’. © 2015 John Wiley & Sons Ltd
4
L. Mozessohn et al.
Table 1 Studies with HBsAg /HBcAb+ patients receiving rituximab for lymphoma
Study
Study design
Persico, 2006 [42] Yeo, 2009 [44] Francisci, 2010 [33] Niitsu, 2010 [40] Koo, 2011 [37] Huang, 2013 [11] Targhetta, 2008 [43] Hanbali, 2008 [35] Fukushima, 2009 [34]
Prospective Prospective Prospective Prospective Prospective Prospective Retrospective Retrospective Retrospective and prospective Retrospective Retrospective Retrospective Retrospective Retrospective Retrospective
Ji, 2010 [36] Mendez-Navarro, 2010 [39] Matsue, 2010 [38] Watanabe, 2011 [46] Zhu, 2012 [45] Oh, 2013 [41]
Country of origin
Total HBcAb+/ HBsAg patients
Reactivation rate with ‘clinical’ definition*
Reactivation rate with ‘preclinical’ definition†
Italy Hong Kong Italy Japan Singapore Taiwan Italy USA Japan
7 21 38 43 62 39 74 26 32
2/7 (28.9%) 5/21 (23.8%) N/A§ N/A§ 2/62 (3.2%) N/A§ 2/74 (2.7%) N/A¶ 1/11 (9.1%)
2/7 (28.9%) N/A‡ 3/38 (7.9%) 6/43 (14.0%) N/A‡ 7/39 (17.9%) N/A‡ N/A¶ 1/21 (4.8%)
China USA Japan Japan China Korea
43 25 56 20 25 67
1/43 0/25 3/56 N/A§ N/A§ 2/67
N/A‡ N/A‡ N/A‡ 5/20 (25.0%) N/A‡ N/A‡
(2.3%) (0%) (5.4%)
(3.0%)
*Increase in ALT >3 9 ULN and either increase in HBV DNA from baseline or HBsAg seroreversion. †Viral expansion (increased HBV DNA from baseline including change from undetectable to detectable) and/or HBsAg seroreversion +/ increase in ALT. ‡Unable to calculate rate because did not systematically monitor HBV DNA levels in all patients at all visits. §Unable to calculate rate because treatment trial (if DNA increased from baseline, started on antiviral irrespective of transaminase level). ¶Unable to calculate rate because of missing data from study.
Fig. 2 Pooled risk of ‘clinical’ HBV reactivation among HBsAg /cAb+ patients receiving rituximab for lymphoma.
HBV reactivation rate of 6.1% (14 reactivations of 260 patients) vs 9.0% (4 reactivations of 81 patients) in European studies. There was only 1 North American study and as such, it was not included in the geographic analysis. Significant heterogeneity was again observed suggesting that the variability observed is unlikely to be due to variation in viral genotype alone. We explored whether or not the presence or absence of antibodies to hepatitis B surface antigen (HBsAb status)
was predictive of HBV reactivation risk. Using the clinical definition of reactivation, 3 studies (145 patients) had sufficient data to be included in this analysis. There was a nonstatistically significant trend to a lower risk of HBV reactivation in patients with measurable antibodies to HBsAg (OR = 0.083; P = 0.151). In addition, mortality data were available in only 6 studies, and thus, it was not possible to compare mortality in ‘clinical HBV reactivation’ vs ‘preclinical HBV reactivation’. © 2015 John Wiley & Sons Ltd
6
L. Mozessohn et al.
types are found in China, while Japan and Korea have almost exclusively genotype B and C isolates, respectively. In Europe, genotypes A and D predominate. Although there are differences in clinical outcome between the different genotypes including the risk of cirrhosis and hepatocellular carcinoma, whether this would affect the probability of reactivation during immunosuppression is unknown and overall it seems unlikely that genotype is a major factor in the variable HBV reactivation rates reported in different studies. Ultimately, the most important source of heterogeneity in this study was likely study design. As described above, we attempted to standardize the definitions of ‘clinical’ and ‘pre-clinical’ of HBV reactivation – however, because this was not always consistent with the definition of HBV reactivation used by individual studies, the reliability of data may have been affected. As well, studies varied with regard to how closely they monitored patients for HBV reactivation, which may have influenced the reported rates of reactivation. Finally, how long patients were monitored for HBV reactivation varied between studies. Recent data suggest that a risk of HBV reactivation persists for months after rituximab therapy is concluded [11] – thus studies that followed patients for longer may be more likely to recognize HBV reactivation. We observed a difference in the rate of HBV reactivation reported by prospective studies, as compared to retrospective studies. Larger, prospective studies tended to report higher rates of HBV reactivation. For example, in a prospective study comparing the rate of HBV reactivation in lymphoma patients receiving rituximab-based chemoimmunotherapy vs chemotherapy alone, there were significantly more reactivations in the rituximab group (23.8% vs 0%, respectively) [44]. This study defined reactivation as a threefold or greater increase in serum ALT levels with evidence of HBsAg seroreversion and an increase in HBV DNA levels when compared to baseline. The higher rate of HBV reactivation observed in prospective studies likely reflects closer, systematic surveillance for HBV reactivation among the prospective studies. As a result, it is likely that the available prospective studies provide an estimate of the risk of HBV reactivation that is closer to the true risk. Recently, Dong et al. [51] estimated the HBV reactivation risk in lymphoma patients receiving rituximab. Their analysis focused on the HBsAg-positive population, but a small subset of studies included in the paper described reactivation rates in HBsAg-negative, HBcAb-positive patients. In this group, they estimated the relative risk of HBV reactivation as 5.52 (for rituximab vs no rituximab
exposure) combining retrospective and prospective studies. However, they defined HBV reactivation as a >10-fold rise in serum HBV DNA levels with an increase in serum ALT compared to baseline, a somewhat less stringent criterion than our definition for ‘clinical reactivation’. Our analysis also includes additional trials not included in their estimate, including the only randomized controlled trial of viral prophylaxis in this population [11]. Nonetheless, our results are consistent with the Dong paper in both of us report an important increased risk of HBV reactivation among HBsAg-negative, HBcAb-positive patients exposed to rituximab. At present, it is not possible to precisely estimate the risk of HBV reactivation in patients with so-called ‘resolved HBV’ who are receiving rituximab due to marked heterogeneity in existing studies; nonetheless, it is evident that reactivation does occur, and that the risk is likely clinically significant. The best evidence for recommendations regarding surveillance is from the recent randomized controlled trial examining antiviral prophylaxis [11]. In this paper, they reported superior outcomes with antiviral prophylaxis compared to therapeutic antiviral treatment at the time of reactivation. However, the therapeutic arm appeared safe. In that trial, serum markers (liver function, HBV viral load and HBsAg) were monitored during each cycle of chemoimmunotherapy (every 3 weeks). Upon completion of treatment, serum markers were monitored every month during follow-up for 12 months. Pending further trials, it seems reasonable to adopt this as a minimum surveillance strategy in patients not given antiviral prophylaxis at the start of rituximabbased treatment. An evaluation of this strategy to postpone antiviral prophylaxis until a detected increase in viral load and/or HBsAg seroreversion occurs should be confirmed in future prospective studies. However, given the substantial heterogeneity between studies, the field would benefit from the development of consensus diagnostic criteria for ‘preclinical’ and ‘clinical HBV reactivation’ by leading organizations. Using these standardized definitions, prospective studies should then be conducted with the goal of more accurately defining the risk of reactivation and the duration of time during which reactivations can occur.
DISCLOSURES KC, JF and LH have received research funding from Gilead. No other relevant disclosures.
REFERENCES 1 Coiffier B, Thieblemont C, Van Den Neste E et al. Long-term outcome of patients in the LNH-98.5 trial, the
first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a
study by the Groupe d’Etudes des Lymphomes de l’Adulte. Blood 2010; 116(12): 2040–2045.
© 2015 John Wiley & Sons Ltd
HBV reactivation 2 Marcus R, Imrie K, Solal-Celigny P et al. Phase III study of R-CVP compared with cyclophosphamide, vincristine, and prednisone alone in patients with previously untreated advanced follicular lymphoma. J Clin Oncol 2008; 26(28): 4579–4586. 3 Schulz H, Bohlius JF, Trelle S et al. Immunochemotherapy with rituximab and overall survival in patients with indolent or mantle cell lymphoma: a systematic review and meta-analysis. J Natl Cancer Inst 2007; 99(9): 706–714. 4 Edwards JC, Szczepanski L, Szechinski J et al. Efficacy of B-cell-targeted therapy with rituximab in patients with rheumatoid arthritis. N Engl J Med 2004; 350(25): 2572–2581. 5 Cohen SB, Emery P, Greenwald MW et al. Rituximab for rheumatoid arthritis refractory to anti-tumor necrosis factor therapy: results of a multicenter, randomized, doubleblind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks. Arthritis Rheum 2006; 54(9): 2793–2806. 6 Maloney DG. Anti-CD20 antibody therapy for B-cell lymphomas. N Engl J Med 2012; 366(21): 2008–2016. 7 Lavanchy D. Hepatitis B virus epidemiology, disease burden, treatment, and current and emerging prevention and control measures. J Viral Hepat 2004; 11(2): 97–107. 8 Liu CJ, Chen DS, Chen PJ. Epidemiology of HBV infection in Asian blood donors: emphasis on occult HBV infection and the role of NAT. J Clin Virol 2006; 36(Suppl. 1): S33–S44. 9 Uhm JE, Kim K, Lim TK et al. Changes in serologic markers of hepatitis B following autologous hematopoietic stem cell transplantation. Biol Blood Marrow Transplant 2007; 13(4): 463–468. 10 Hsu C, Tsou HH, Lin SJ et al. Chemotherapy-induced hepatitis B reactivation in lymphoma patients with resolved HBV infection: a prospective study. Hepatology 2014; 59(6): 2092–2100. 11 Huang YH, Hsiao LT, Hong YC et al. Randomized controlled trial of entecavir prophylaxis for rituximabassociated hepatitis B virus reactivation in patients with lymphoma and
© 2015 John Wiley & Sons Ltd
12
13
14
15
16
17
18
19
20
21
resolved hepatitis B. J Clin Oncol 2013; 31(22): 2765–2772. Li L, Chen PJ, Chen MH, Chak KF, Lin KS, Tsai SJ. A pilot study for screening blood donors in Taiwan by nucleic acid amplification technology: detecting occult hepatitis B virus infections and closing the serologic window period for hepatitis C virus. Transfusion 2008; 48(6): 1198–1206. Hui CK, Sun J, Au WY et al. Occult hepatitis B virus infection in hematopoietic stem cell donors in a hepatitis B virus endemic area. J Hepatol 2005; 42(6): 813–819. Zhu Y, Yamamoto T, Cullen J et al. Kinetics of hepadnavirus loss from the liver during inhibition of viral DNA synthesis. J Virol 2001; 75(1): 311–322. Zurawska U, Hicks LK, Woo G et al. Hepatitis B virus screening before chemotherapy for lymphoma: a cost-effectiveness analysis. J Clin Oncol 2012; 30(26): 3167–3173. Lau GK, Yiu HH, Fong DY et al. Early is superior to deferred preemptive lamivudine therapy for hepatitis B patients undergoing chemotherapy. Gastroenterology 2003; 125(6): 1742–1749. Hsu C, Hsiung CA, Su IJ et al. A revisit of prophylactic lamivudine for chemotherapy-associated hepatitis B reactivation in non-Hodgkin’s lymphoma: a randomized trial. Hepatology 2008; 47(3): 844–853. Loomba R, Rowley A, Wesley R et al. Systematic review: the effect of preventive lamivudine on hepatitis B reactivation during chemotherapy. Ann Intern Med 2008; 148(7): 519–528. Coffin CS, Fung SK, Ma MM. Canadian Association for the Study of the L. Management of chronic hepatitis B: Canadian Association for the Study of the Liver consensus guidelines. Can J Gastroenterol 2012; 26(12): 917–938. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines): Non-Hodgkin’s Lymphomas Version 2.2013. Available at: http:// www.nccn.org/professionals/physi cian_gls/pdf/nhl.pdf (accessed 15 December 2013). Lok ASF, McMahon BJ. Chronic Hepatitis B: Update 2009. American
22
23
24
25
26
27
28
29
7
Association for the Study of Liver Diseases. Available at: http:// www.aasld.org (accessed 15 December 2013). “Arzerra (ofatumumab) and Rituxan (rituximab): Drug Safety Communication - New Boxed Warning, Recommendations to Decrease Risk of Hepatitis B Reactivation”. Available at: http://www.fda.gov (accessed 15 October 2013). “Recommendations for Routine Testing and Follow-up for Chronic Hepatitis B Virus (HBV) Infection”. Available at: http://www.cdc.gov (accessed 10 January 2014). Lalazar G, Rund D, Shouval D. Screening, prevention and treatment of viral hepatitis B reactivation in patients with haematological malignancies. Br J Haematol 2007; 136(5): 699–712. Liang R. How I treat and monitor viral hepatitis B infection in patients receiving intensive immunosuppressive therapies or undergoing hematopoietic stem cell transplantation. Blood 2009; 113(14): 3147–3153. Chen PM, Fan S, Liu CJ et al. Changing of hepatitis B virus markers in patients with bone marrow transplantation. Transplantation 1990; 49(4): 708–713. Onozawa M, Hashino S, Izumiyama K et al. Progressive disappearance of anti-hepatitis B surface antigen antibody and reverse seroconversion after allogeneic hematopoietic stem cell transplantation in patients with previous hepatitis B virus infection. Transplantation 2005; 79(5): 616–619. Wu JM, Huang YH, Lee PC, Lin HC, Lee SD. Fatal reactivation of hepatitis B virus in a patient who was hepatitis B surface antigen negative and core antibody positive before receiving chemotherapy for nonHodgkin lymphoma. J Clin Gastroenterol 2009; 43(5): 496–498. Law JK, Ho JK, Hoskins PJ, Erb SR, Steinbrecher UP, Yoshida EM. Fatal reactivation of hepatitis B post-chemotherapy for lymphoma in a hepatitis B surface antigen-negative, hepatitis B core antibody-positive patient: potential implications for future prophylaxis recommendations. Leuk Lymphoma 2005; 46(7): 1085–1089.
6
L. Mozessohn et al.
types are found in China, while Japan and Korea have almost exclusively genotype B and C isolates, respectively. In Europe, genotypes A and D predominate. Although there are differences in clinical outcome between the different genotypes including the risk of cirrhosis and hepatocellular carcinoma, whether this would affect the probability of reactivation during immunosuppression is unknown and overall it seems unlikely that genotype is a major factor in the variable HBV reactivation rates reported in different studies. Ultimately, the most important source of heterogeneity in this study was likely study design. As described above, we attempted to standardize the definitions of ‘clinical’ and ‘pre-clinical’ of HBV reactivation – however, because this was not always consistent with the definition of HBV reactivation used by individual studies, the reliability of data may have been affected. As well, studies varied with regard to how closely they monitored patients for HBV reactivation, which may have influenced the reported rates of reactivation. Finally, how long patients were monitored for HBV reactivation varied between studies. Recent data suggest that a risk of HBV reactivation persists for months after rituximab therapy is concluded [11] – thus studies that followed patients for longer may be more likely to recognize HBV reactivation. We observed a difference in the rate of HBV reactivation reported by prospective studies, as compared to retrospective studies. Larger, prospective studies tended to report higher rates of HBV reactivation. For example, in a prospective study comparing the rate of HBV reactivation in lymphoma patients receiving rituximab-based chemoimmunotherapy vs chemotherapy alone, there were significantly more reactivations in the rituximab group (23.8% vs 0%, respectively) [44]. This study defined reactivation as a threefold or greater increase in serum ALT levels with evidence of HBsAg seroreversion and an increase in HBV DNA levels when compared to baseline. The higher rate of HBV reactivation observed in prospective studies likely reflects closer, systematic surveillance for HBV reactivation among the prospective studies. As a result, it is likely that the available prospective studies provide an estimate of the risk of HBV reactivation that is closer to the true risk. Recently, Dong et al. [51] estimated the HBV reactivation risk in lymphoma patients receiving rituximab. Their analysis focused on the HBsAg-positive population, but a small subset of studies included in the paper described reactivation rates in HBsAg-negative, HBcAb-positive patients. In this group, they estimated the relative risk of HBV reactivation as 5.52 (for rituximab vs no rituximab
exposure) combining retrospective and prospective studies. However, they defined HBV reactivation as a >10-fold rise in serum HBV DNA levels with an increase in serum ALT compared to baseline, a somewhat less stringent criterion than our definition for ‘clinical reactivation’. Our analysis also includes additional trials not included in their estimate, including the only randomized controlled trial of viral prophylaxis in this population [11]. Nonetheless, our results are consistent with the Dong paper in both of us report an important increased risk of HBV reactivation among HBsAg-negative, HBcAb-positive patients exposed to rituximab. At present, it is not possible to precisely estimate the risk of HBV reactivation in patients with so-called ‘resolved HBV’ who are receiving rituximab due to marked heterogeneity in existing studies; nonetheless, it is evident that reactivation does occur, and that the risk is likely clinically significant. The best evidence for recommendations regarding surveillance is from the recent randomized controlled trial examining antiviral prophylaxis [11]. In this paper, they reported superior outcomes with antiviral prophylaxis compared to therapeutic antiviral treatment at the time of reactivation. However, the therapeutic arm appeared safe. In that trial, serum markers (liver function, HBV viral load and HBsAg) were monitored during each cycle of chemoimmunotherapy (every 3 weeks). Upon completion of treatment, serum markers were monitored every month during follow-up for 12 months. Pending further trials, it seems reasonable to adopt this as a minimum surveillance strategy in patients not given antiviral prophylaxis at the start of rituximabbased treatment. An evaluation of this strategy to postpone antiviral prophylaxis until a detected increase in viral load and/or HBsAg seroreversion occurs should be confirmed in future prospective studies. However, given the substantial heterogeneity between studies, the field would benefit from the development of consensus diagnostic criteria for ‘preclinical’ and ‘clinical HBV reactivation’ by leading organizations. Using these standardized definitions, prospective studies should then be conducted with the goal of more accurately defining the risk of reactivation and the duration of time during which reactivations can occur.
DISCLOSURES KC, JF and LH have received research funding from Gilead. No other relevant disclosures.
REFERENCES 1 Coiffier B, Thieblemont C, Van Den Neste E et al. Long-term outcome of patients in the LNH-98.5 trial, the
first randomized study comparing rituximab-CHOP to standard CHOP chemotherapy in DLBCL patients: a
study by the Groupe d’Etudes des Lymphomes de l’Adulte. Blood 2010; 116(12): 2040–2045.
© 2015 John Wiley & Sons Ltd
