Reactivation of hepatitis B virus after rituximab-containing treatment in patients with CD20-positive B-cell lymphoma

Authors


Abstract

BACKGROUND:

Reactivation of hepatitis B virus (HBV) after rituximab-containing chemotherapy in patients with B-cell lymphoma has been recognized as a potentially serious complication in HBV immune patients.

METHODS:

To determine the HBV reactivation in patients treated with rituximab, a retrospective study of HBV-related markers was performed before and after rituximab-containing treatment in 261 consecutive patients with CD20-positive B-cell lymphoma.

RESULTS:

Of the 261 patients, 230 patients were tested for both hepatitis B surface antigen (HBsAg) and antibody to hepatitis B core antigen (anti-HBc) before treatment. Fifty-six (24.3%) of 230 patients were anti-HBc positive, and the remaining 174 (75.6%) patients were anti-HBc negative. Among the 56 anti-HBc–positive patients, 5 (8.9%) became HBsAg positive (HBV reactivation), whereas none of the 174 anti-HBc–negative patients became HBsAg positive with a median follow-up of 24 months (P = .001). Among the 5 patients with HBV reactivation, 4 were negative for antibody to HBsAg (anti-HBs), and 1 patient was positive for anti-HBs. All 5 of these patients were treated successfully with entecavir on detection of HBsAg, although 4 of the 5 patients exhibited mild to moderate elevation of alanine aminotransferase. Among 56 anti-HBc–positive patients, those negative for anti-HBs had a higher probability of developing HBV reactivation compared with those positive for anti-HBs (4 of 19; 21.1% vs 1 of 37; 2.7%, P = .014).

CONCLUSIONS:

Patients with isolated anti-HBc are at high risk of HBV reactivation and should be monitored closely for HBsAg, anti-HBs, HBV-DNA, and transaminase levels during and after rituximab-containing treatment. Although preemptive use of entecavir enabled successful management of HBV reactivation, mild to moderate hepatic flare was still observed. These approaches should be further evaluated in a prospective study with regard to clinical usefulness, safety, and cost-effectiveness. Cancer 2010. © 2010 American Cancer Society.

Hepatitis B virus (HBV) infection is a well-known complication in patients receiving cytotoxic chemotherapy or intensive immunosuppression, and preventive use of antiviral agents, such as lamivudine, is recommended in such cases.1 Although HBV infection has a wide clinical spectrum, ranging from asymptomatic self-limiting hepatitis to severe potentially fatal progressive hepatic failure, loss of hepatitis B surface antigen (HBsAg) and development of antibody to hepatitis B surface antigen (anti-HBs) and antibody to hepatitis B core antigen (anti-HBc) after acute HBV infection are thought to represent clearance of HBV. However, studies have indicated that a low level of virus replication persists in the liver and peripheral blood mononuclear cells in these patients.2, 3

Rituximab is a chimeric mouse human monoclonal antibody and has been used effectively alone or in combination with cytotoxic agents for treating both low- and high-grade B-cell lymphoma. Several recent reports have suggested that use of rituximab is related to the development of fatal HBV reactivation in HBsAg-negative patients with prior HBV exposure.4-7 HBV reactivation was a rare complication in these patients before the increased use of rituximab. Recently, Yeo et al8 reported a high incidence of HBV reactivation in HBsAg-negative and anti-HBc–positive patients receiving rituximab-containing chemotherapy in Hong Kong. Umemura et al9 reported a high mortality rate of patients with malignant lymphoma associated with HBV reactivation after use of rituximab in Japan. As HBV infection is endemic in Asia, these observations are extremely important with regard to therapy of B-cell lymphoma patients in this area. Here, we performed a retrospective investigation of the occurrence of HBV reactivation after rituximab-containing chemotherapy in 252 B-cell lymphoma patients admitted to Kameda General Hospital, Kamogawa-shi, Japan, over the past 5 years.

MATERIALS AND METHODS

Study Population

Consecutive patients with B-cell lymphoma admitted to Kameda General Hospital between April 2004 and April 2009, who received rituximab-containing chemotherapy and were followed up for at least 2 months after the start of treatment, were included in this study (Table 1). After admission, all of the patients were given a physical examination, blood examination, and serum biochemistry analysis. All of the patients received chest and abdominal computed tomography and ultrasonography of the abdomen at the time of admission as an initial evaluation.

Table 1. Characteristics of 252 Hepatitis B Surface Antigen-Negative Patients Treated With Rituximab-Containing Chemotherapy
CharacteristicsNo. of Patients Checked for Anti-HBc, n=230Anti-HBc Was Not Checked
Anti-HBc NegativeAnti-HBc Positive
  1. Anti-HBc indicates antibody to hepatitis B core antigen; M, male; F, female; anti-HBs, antibody to hepatitis B surface antigen; NA, not assessed; DLBCL, diffuse large B-cell lymphoma; FL, follicular lymphoma; BL, Burkitt lymphoma; MCL, mantle cell lymphoma; CLL, chronic lymphocytic lymphoma; auto-SCT; autologous stem cell transplantation, allo-SCT; allogeneic stem cell transplantation.

No. of Patients174 (75.6%)56 (24.0%)22
Sex (M/F)97/7737/197/15
Age, median (range)66 (15-92)71 (43-90)74 (40-93)
Anti-HBs positive5 (2.9%)37 (66.1%)NA
Anti-HBs negative168 (97.1%)19 (33.9%)NA
Diagnosis   
 DLBCL1014016
 FL4761
 BL600
 MCL623
 CLL420
 Others1062
No. of rituximab infusions, median (range)6 (1-14)6 (1-12)6 (1-8)
Observation period, median mo (range)21 (2-55)26 (2-53)28 (2-56)
No. of patients receiving auto-SCT731
No. of patients receiving allo-SCT410
No. of patients who became HBsAg positive050

The study population consisted of 261 patients (147 male, 57%; 114 female, 43%). The median age of the patients was 70 years, with a range of 15 to 93 years. Diagnosis of lymphoma subtypes included diffuse large B-cell lymphoma (DLBCL; n = 163, 62.5%), follicular lymphoma (n = 57, 21.8%), mantle cell lymphoma (n = 11, 4.2%), Burkitt lymphoma (n = 6, 2.3%), chronic lymphocytic leukemia (n = 6, 2.3%), and other B-cell lymphomas (n = 18, 6.9%). The median number of rituximab infusions (375 mg/m2) was 6, with a range of 1 to 14, and the median observation period from the start of treatment was 24 months, with a range of 2 to 56 months. All patients received varied courses of glucocorticoid-containing chemotherapy, such as cyclophosphamide, doxorubicin, vincristine, and prednisone10; etoposide, methylprednisolone, high-dose cytarabine, and cisplatin11; and fractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone,12 in combination with rituximab.

Because of the retrospective nature of this study, individual informed consent was not obtained from all of the patients. Instead, this study protocol including the use of frozen stored plasma was approved by the institutional review board of Kameda General Hospital.

HBV Serology and HBV-DNA

All patients were screened for HBsAg at the time of admission. However, before September 2006, anti-HBs and anti-HBc tests were performed at the discretion of the treating physician. Therefore, in patients whose anti-HBs or anti-HBc were not checked, those antibodies were checked using frozen sera obtained within 1 month of admission. There were 13 patients whose anti-HBc and anti-HBs were not checked their who were found to be negative for both anti-HBc and anti-HBs after completion of chemotherapy and were considered to be HBV-naive. After October 2006, anti-HBs and anti-HBc tests were performed for all patients. The sera from 31 of 55 anti-HBc–positive patients were tested retrospectively for the presence or absence of immunoglobulin (Ig)M class anti-HBc. When a patient was found to have developed hepatitis or became HBsAg positive during the course of treatment, quantitative determination of HBV-DNA was performed by transcription-mediated amplification13 or by the Taqman polymerase chain reaction method.14

HBsAg, anti-HBs, anti-HBc, and anti-HBc IgM were detected using commercially available enzyme immunoassay kits (Abbot Japan, Tokyo, Japan). The results of anti-HBs and anti-HBc are expressed as percentage inhibition (negative control optic density [OD] − positive control OD), with >50% defined as a positive result.

Definition of HBV Hepatitis and HBV Reactivation

Hepatitis was defined as a serum alanine aminotransferase (ALT) level >3-fold higher than the normal upper limit of 2 consecutive determinations >5 days apart in the absence of clinical and laboratory features of acute infection of hepatitis A, hepatitis C, or other systemic infections. Although there are no clear diagnostic criteria for HBV reactivation, we defined HBV reactivation as seroconversion from HBsAg negative to HBsAg positive with or without an increase in HBV-DNA level from baseline (>2.6 log copies/mL).

Statistical Analysis

The rates of HBV reactivation in anti-HBc–negative, anti-HBc–positive, anti-HBc-positive/anti-HBs-positive, and anti-HBc-positive/anti-HBs-negative patients were determined and compared using Fisher exact test.

RESULTS

Patient Characteristics (Table 1)

Nine of the 261 patients were found to be HBsAg positive, and were excluded from the analysis. All 9 patients were treated successfully with rituximab-containing regimens with use of antiviral agents, such as lamivudine or entecavir, and showed no development of severe hepatitis.

The background characteristics of the rest of the 252 HBsAg-negative patients are shown in Table 1. Twenty-two patients were not tested for anti-HBc and anti-HBs before use of rituximab because of unavailability of serum. Therefore, a total of 230 patients were checked for HBV serology; 174 (75.6%) patients were found to be anti-HBc negative and 56 (24.3%) patients were anti-HBc positive. Among the 174 patients negative for anti-HBc, 4 were positive for anti-HBs. Thirty-three of the 55 anti-HBc–positive patients were further tested for anti-HBc IgM, and all of them were found to be negative for anti-HBc IgM, indicating the absence of recent HBV infection in these patients.

There was a predominance of male patients in both the anti-HBc–negative and anti-HBc–positive groups. However, patients who were not checked for anti-HBc were predominantly female. Five (2.9%) of the 174 anti-HBc–negative patients were positive for anti-HBs. Among the 55 anti-HBc–positive patients, 36 (65.4%) were also positive for anti-HBs, and 19 (34.5%) were negative for anti-HBs. DLBCL was the predominant lymphoma subtype among the 3 groups. However, the incidences of follicular lymphoma were lower in the anti-HBc–positive and anti-HBc–unchecked groups. The median number of rituximab infusions and observation period after the start of rituximab treatment were comparable among the 3 groups.

Autologous peripheral blood stem cell transplantation (PBSCT) was performed in 7 and 3 patients in the anti-HBc–negative and anti-HBc–positive groups, respectively. Allogeneic stem cell transplantation was performed in 4 patients in the anti-HBc–negative group and in 1 patient in the anti-HBc–positive group.

Consequences of HBV Serology After Rituximab-Containing Treatment (Fig. 1)

Among the 252 HBsAg-negative patients, 5 (2.0%) became HBsAg-positive (HBV reactivation) after rituximab-containing treatment. The details of the patients who became HBsAg positive are shown in Table 2. None of the anti-HBc–negative or anti-HBc–unchecked patients became HBsAg positive after median observation periods of 21 and 28 months, respectively. In contrast, 5 of the 56 anti-HBc–positive patients (8.9%) became HBsAg positive, and 4 of these patients were negative for anti-HBs before treatment. The incidence of HBsAg reappearance in anti-HBc–positive patients was significantly higher (P = .001) compared with anti-HBc–negative patients. Among the anti-HBc–positive patients, 4 of the 19 patients negative for anti-HBs became HBsAg positive, whereas only 1 of the 37 anti-HBs–positive patients became HBsAg positive after treatment. Positivity for anti-HBc only showed a significant association with the reappearance of HBsAg after rituximab-containing therapy (P = .014).

Figure 1.

Longitudinal representation of hepatitis B virus (HBV) reactivation after rituximab-containing chemotherapy in patients with CD20-positive B-cell lymphoma is shown. HBsAg indicates hepatitis B surface antigen; anti-HBc, antibody to hepatitis B core antigen; anti-HBs, antibody to hepatitis B surface antigen.

Table 2. Details and Outcome of the 5 Patients Who Became Hepatitis B Surface Antigen Positive After Rituximab-Containing Treatment
DetailsCase 1Case 2Case 3Case 4Case 5
  1. M indicates male; F, female; DLBCL, diffuse large B-cell lymphoma; IVL, intravascular lymphomatosis; FL, follicular lymphoma; CLL, chronic lymphocytic lymphoma; BM, basement membrane; HBV, hepatitis B virus; HBsAg, hepatitis B surface antigen; HBAb, hepatitis B antibody; HBcAb, antibody to hepatitis B core antigen; R-CHOP, cyclophosphamide, doxorubicin, vincristine, and prednisone with rituximab; R-ESHAP, etoposide, cytarabine, cisplatinum, and methylprednisolone with rituximab; HyperCVAD, hyperfractionated cyclophosphamide, vincristine, doxorubicin, and dexamethasone; allo-BMT, allogeneic bone marrow transplantation; ALT, alanine aminotransferase.

Age/sex67/M46/F63/F72/F56/M
DiagnosisDLBCLFLIVLCLLDLBCL
Stage434NA3
BM invasionNoNoYesYesNo
HBV status (HBsAg/HBAb/HBcAb)     
 Before rituximab−/−/+−/−/+−/−/+−/−/+0
 After treatment of reactivation−/−/+−/−/+−/−/++/−/+−/−/+
 Treatment before reactivationR-CHOP ×6, R-ESHAP ×3R-CHOP ×6R-CHOP ×5, R-ESHAP ×2R-CHOP ×1, HyperCVAD ×2R-CHOP ×8, R-ESHAP ×1, allo-BMT
HBV reactivation after last rituximab, mo77111132.2
 Peak ALT (IU/L)22911177978191
 Peak total bilirubin (mg/dl)5.50.80.51.21.5
Peak HBV-DNA level (log copies/μL)>7.68.54.67.63.7
HBV-DNA after treatment<2.5<3.5<2.5<3.7<2.6
Antiviral treatmentEntecavirEntecavirEntecavirEntecavirEntecavir
OutcomeAlive wellAlive wellAlive wellDied of CLLAlive well

One patient (Case 5) who was initially positive for anti-HBs became HBsAg positive 21 months after completion of rituximab treatment. He received PBSCT from his human leukocyte antigen-matched brother after failure of 8 cycles of cyclophosphamide, doxorubicin, vincristine, and prednisone with rituximab15 chemotherapy using high-dose cyclophosphamide and total-body irradiation. Graft-versus-host disease (GVHD) prophylaxis included standard short-term methotrexate and cyclosporine A. Although his clinical course was complicated with development of generalized chronic GVHD, which was treated with cyclosporine A and prednisone, he obtained clinical remission of lymphoma. He was anti-HBs positive before and after rituximab-containing treatment, but its titer gradually decreased and disappeared by 7 months, and HBsAg seroconversion was confirmed 17 months after transplantation (21 months after completion of rituximab therapy). However, HBV-DNA remained below the level of detection (2.6 log copies/mL), and hepatitis B e antigen was negative. He received entecavir preemptively based on detection of HBsAg and showed only mild liver function abnormality. Four other patients received varied courses of rituximab-containing chemotherapy. One patient (Case 2) developed hepatitis with the appearance of HBsAg. Serum levels of ALT and aspartate aminotransferase (AST) were elevated to 1117 IU/L and 602 IU/L, respectively, with no increase in total bilirubin level. HBV-DNA was increased to 8.5 log copies/mL. The patient was treated successfully with entecavir (0.5 mg/d), and her HBV-DNA became undetectable (3.5 log copies/mL) by 5 months after the start of entecavir, although she remained positive for HBsAg until 16 months after the start of entecavir and became negative thereafter.

Three other patients did not develop clinical hepatic flare despite the appearance of HBsAg and detection of HBV-DNA in the serum, although slight elevation of ALT was observed. All patients received entecavir on detection of HBsAg, and HBV-DNA decreased promptly in all of these cases.

DISCUSSION

HBV reactivation is a common complication in HBsAg-positive patients receiving hematopoietic transplantation or immunosuppressive anticancer therapy. Recently, the risk of developing hepatitis B virus reactivation after chemoimmunotherapy or organ transplantation in HBs-Ag–negative but anti-HBs– and/or anti-HBc–positive patients has been described and characterized by measuring HBV-associated markers.16-18 However, the precise frequency and risk factors for developing hepatic flare may vary among ethnic groups or geographic regions. In addition, differences in anticancer therapy, use of stem cell transplantation, definition of reactivation, lack of prospective study, and possibility of missing seroconversion because of the absence of symptoms could be responsible for these differences.

Rituximab has significant efficacy in the treatment of a wide range of B-cell lymphomas. Addition of rituximab to anthracycline-based chemotherapy, such as cyclophosphamide, doxorubicin, vincristine, and prednisone, is considered a standard front-line treatment for CD20-positive B-cell lymphomas.19, 20 In addition, clinical benefits of rituximab maintenance have also been observed after different treatment protocols.21 However, rituximab induces profound and prolonged B-cell depletion and has been reported to be associated with varicella-zoster virus infection, chronic parvovirus infection, and cytomegalovirus infection.22-24 Rituximab binds to CD20 antigen, which is expressed not only on neoplastic B cells but also on normal B lymphocytes. As B lymphocytes act as antigen-presenting cells for CD8-postive cytotoxic T cells25 that limit the expansion of HBV infection in hepatocytes, depletion of B lymphocytes by rituximab may result in defective T-cell responses to HBV, thereby allowing replication of the virus. Use of corticosteroids may also play a role in expansion of HBV, because HBV-DNA contains a glucocorticoid response element that facilitates HBV replication.26 Use of lamivudine as prophylaxis in HBsAg-positive patients scheduled to receive cytotoxic chemotherapy is advised in HBV endemic areas, including Japan. However, reactivation of HBV in HBsAg-negative patients after cytotoxic chemotherapy has been considered a rare event except in patients undergoing stem cell transplantation.

In the present study, we investigated the occurrence of HBV reactivation after rituximab-containing chemotherapy in 261 consecutive patients with CD20-positive B-cell lymphoma. This could be the largest study on the incidence of HBV reactivation associated with rituximab use, compared with the previous reported series. In our study, performed in Kamogawa, Chiba, in the central part of Japan, 5 (8.9%) of 56 HBsAg-negative, anti-HBc–positive patients developed HBV reactivation, and the prevalence rate of occult HBV infection was 24.3%. Four (21.1%) of 19 patients positive for anti-HBc alone developed HBV reactivation within 7 to 11 months after the use of rituximab, whereas among 37 patients who were both anti-HBc- and ant-HBs positive, only 1 (0.4%) patient who received allogeneic stem cell transplantation developed HBV reactivation. Similar to our study, Yeo et al8 described 80 patients in Hong Kong with HBsAg-negative DLBCL who received cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy with or without rituximab. In 46 patients with anti-HBc, 5 (23.8%) of the 21 patients treated with cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy with rituximab developed HBV reactivation, whereas none of 25 patients who received cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy without rituximab developed reactivation. The high incidence of HBV reactivation in anti-HBc–positive and HBsAg-negative patients was attributed to the incorporation of rituximab into cyclophosphamide, doxorubicin, vincristine, and prednisone chemotherapy. In addition, our current observation, together with those of Yeo et al,8 suggests that anti-HBc–positive patients without anti-HBs have a greater likelihood of developing HBV reactivation after rituximab-containing treatment compared with patients positive for anti-HBs. However, Koo et al27 in Singapore and Targhetta et al28 in Cagliari, Sardinia, Italy, observed HBV reactivation in only 1 (2.2%) of 46 and 2 (2.7%) of 74 patients with anti-HBc, respectively. Fukushima et al29 in Saga, in the southern part of Japan, reported that 2 (6.3%) of 32 patients with occult HBV infection developed HBV reactivation after rituximab. Although the risk of HBV reactivation in patients with occult HBV infection after rituximab-containing therapy seems real, the incidence of reactivation varies substantially among studies. The reason for the difference in the incidence of HBV reactivation among the studies remains to be elucidated. However, the intensity of treatment, patient characteristics, and geographic differences in HBV may account for these differences. Remarkable genotype variation has been reported among the regions in which HBV is endemic. Studies on the clinical significance of HBV genotype suggest that the genotype may be correlated with clinical events,30 although the relationship between HBV genotype and reactivation of HBV after chemotherapy has not been determined.

The results of the present study suggest that HBsAg-negative B-cell lymphoma patients who are anti-HBc–positive but anti-HBs–negative are at high risk of HBV reactivation after rituximab-containing treatment. Thus, complete HBV serology, including HBsAg, anti-HBs, and anti-HBc, should be determined before the start of chemotherapy or intensive immunosuppression to identify patients at high risk of HBV reactivation. Although prophylactic use of antiviral agents in HBsAg-positive patients receiving cytotoxic chemotherapy is the standard strategy,1 the effect of prophylactic use of such agents in patients with occult HBV infection remains uncertain. However, when it occurs, the mortality rate in this group of patients is high.9 In addition, a proportion of patients exhibited increases in ALT and HBV-DNA simultaneously, and some patients showed an increase in HBV-DNA level before the increase in ALT.31 It is also possible that HBV-DNA could decline when liver enzyme levels begin to rise. In the present study, close monitoring of HBV-related markers and preemptive use of entecavir after the detection of HBsAg or HBV-DNA successfully controlled HBV replication and prevented the development of fulminant hepatic failure, although 1 of 5 patients showed AST elevated to 1171 U/mL. Thus, our current approach may not be appropriate for prevention of severe hepatic injury because of HBV reactivation. In addition, mutation of HBsAg may occur under immunological pressure, which could result in false-negative results for detection of HBsAg.32 Intense monitoring of HBV-DNA with more sensitive assays may be a rational approach. However, the sampling schedule would be arbitrary, and HBV-DNA methodology is both expensive and not widely available. Furthermore, we could not completely eliminate hepatic flare because of HBV reactivation with this approach.

Alternatively, use of prophylactic lamivudine until 6 months after completion of anticancer therapy, as recommended by Yeo et al,8 might also be feasible, as use of lamivudine before the commencement of chemotherapy successfully reduced the incidence of HBV reactivation in HBsAg-positive cancer patients.1 Lamivudine is a relatively safe agent and is cost-effective. However, long-term use of lamivudine may induce mutation in the tyrosine-methionine-aspartate-aspartate motif of the HBV DNA polymerase gene (YMDD mutation), which can cause hepatic flare.33 Entecavir is currently the most potent anti-HBV agent and showed 5 to 6 log suppression of viral load in HBsAg-positive patients without development of resistance. However, it is expensive compared with lamivudine, and long-term prophylactic use of entecavir in patients with HBV reactivation has not been examined to date. Furthermore, the optimal duration of antiviral therapy in these patients has not been studied in depth. Although 6 to 12 months of lamivudine treatment after discontinuation of chemotherapy has been recommended, extended prophylactic treatment for >12 months may be advisable in patients with high HBV-DNA levels.34 These approaches should be evaluated in the context of clinical usefulness, safety, and cost-effectiveness for prevention of HBV reactivation-related morbidity and mortality.

Combined use of rituximab and intensive chemotherapy is associated with the increased incidence of HBV reactivation in HBsAg-negative but anti-HBc–positive patients with B-cell lymphoma, especially in endemic areas. The present study indicated that the risk of HBV reactivation is significantly elevated in patients who are HBsAg negative but anti-HBc positive. Among anti-HBc–positive patients, anti-HBs–negative patients appear to have higher likelihood of developing HBV reactivation compared with anti-HBs–positive patients. Although further studies are needed with regard to prevention of HBV reactivation, close monitoring of HBV-related markers including HBV-DNA, and prophylactic or preemptive use of antiviral agents should be considered for high-risk patients scheduled to receive rituximab-containing chemotherapy.

Acknowledgements

We thank the staffs of the blood transfusion section and central laboratory of Kameda General Hospital for their cooperation in this study.

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

Ancillary