Infections Complicating the Use of Monoclonal Antibodies for Malignancy
The most important infectious complication of rituximab therapy is reactivation of hepatitis B virus.
For decades the standard treatment of malignancies has been cytotoxic chemotherapeutic drugs. A by-product of this approach has been that normal cells, such as gastrointestinal mucosal cells or bone marrow cells, are also often affected, with significant constitutional and gastrointestinal side effects as well as infectious complications. In an effort to zero in on malignant cells and achieve better treatment success while causing less injury to normal cells, monoclonal antibody therapy and targeted drug therapy have been introduced into the oncologist's armamentarium. Monoclonal antibody therapy is designed to attack cell surface molecules on malignant cells. These antibodies may take the form of immunoglobulin G (IgG) antibodies alone, with the resulting complex identified by the immune system and cleared out, or take the form of "bispecific" IgG antibodies in complex, with one antibody directed against the malignant cell surface antigen and the other against a T-cell surface antigen, thereby engaging and activating T cells (bispecific T-cell engagers- "BITE"), or take the form of IgG conjugated with a toxin.1
Targeted drug therapy is designed to inhibit or activate molecules that may play a role in causation or control of the malignant disease. While these approaches are more directed than conventional chemotherapy, treatment complications have not been eliminated. This review focuses on infectious complications of monoclonal antibody treatments, which generally result because the therapy targets can in some cases be found on normal immune cells. This problem is most common in monoclonal antibody therapy to treat hematologic malignancies. It is important to note that the reason for predisposition to infection is often not clear in these patients, because in addition to monoclonal antibody the patients may be receiving or have recently received other targeted therapy, steroids, or conventional chemotherapy; often properly controlled trials are not available to best address this issue. Also, the malignancy itself may affect immune cells and predispose patients to infection, so studies including patients with different malignancies may be difficult to interpret. In this review, I will focus on the infectious complications from the use of monoclonal antibodies for malignancy.
Infectious Complications of Monoclonal Antibodies Used to Treat Lymphoma and CLL
Rituximab is a monoclonal antibody that targets CD20, a cell surface molecule found on B-cells. Rituximab causes rapid and profound loss of both pre-B cells and mature B-cells, and its effect can last as long as a year.2 It does not, however, appear to affect serum Ig levels produced by plasma cells (which do not express CD20), except with recurrent, prolonged use.2-4 It is approved for use in B-cell non-Hodgkin's lymphoma and chronic lymphocytic leukemia and some non-malignant diseases (rheumatoid arthritis, idiopathic thrombocytopenic purpura, renal transplant). Two other anti-CD20 antibodies, ofatumumab and obinutuzumab, have been approved for use but there is much more data and experience with rituximab.
The most important infectious complication of rituximab (and other anti-CD20) therapy is reactivation of the hepatitis B virus (HBV). Reactivation has been defined as increased blood HBV DNA and/or serum hepatitis B surface antigen reconversion (negative to positive), usually with at least a 3-fold increase in alanine aminotransferase. This complication of anti-CD20 therapy is not intuitive. It is not clear what exact role B-cells play in this process; it has been proposed that their role in antigen presentation to cytotoxic T-cells may be an important factor.5,6 Hepatitis B reactivation can occur in patients with chronic hepatitis B surface antigen (HBsAg+) or patients with "resolved" infection who are Hhepatitis B surface antigen negative (HBsAg−) but core antibody positive (HBcAb+), with negative or positive surface antibody status. Patients who are HBsAg−/HBcAb+ are considered to have a low level of hepatitis B virus in the liver.7,8
In HBsAg+ patients with lymphoma who receive conventional chemotherapy, the risk for reactivation is high. In a study that prospectively followed 78 HBsAg+ patients receiving conventional chemotherapy for any malignancy except hepatocellular carcinoma, hepatitis B reactivation occurred in 19%; however, of the 15 patients with lymphoma, 40% developed HBV reactivation.9 In another prospective study of HBsAg+ patients receiving conventional chemotherapy for various malignancies, the highest rate of HBV reactivation (58%) occurred in patients with non-Hodgkin's lymphoma.10 In 100 patients with non-Hodgkin's lymphoma receiving similar conventional chemotherapy, 48% of HBsAg+ patients developed HBV reactivation, compared with only 3% of HBsAg− patients (all of these HBsAg− patients were HBcAb+).11 HBV reactivation in patients with lymphoma who are HBsAg+ is therefore high and appears to be equally high in HBsAg+ patients treated with rituximab.12 The subtler issue is hepatitis B reactivation in patients who are HBsAg− but HBcAb+. The risk for hepatitis B reactivation is lower for HBsAg−/HBcAb+ patients than HBsAg+ patients, but the HBsAg−/HBcAb+ population is far larger than the chronic HBsAg+ population; up to 60% of people in an endemic area (such as China) may be HBsAg−/HBcAb+.7
Rituximab appears to increase the risk for hepatitis B reactivation when added to a conventional chemotherapy regimen in patients who are HBsAg−/HBcAb+. In a prospective study of 244 patients who received cytotoxic chemotherapy for lymphoma and were HBsAg−/HBcAb+, 7/88 patients (8%) who received rituximab developed hepatitis secondary to hepatitis B reactivation (defined as serum HBV DNA >105 copies/ml, HBsAg reconversion, and a >3-fold increase in serum alanine aminotransferase) as opposed to 1/156 patients who did not receive rituximab.13 Of note, 3 of the patients who had reactivation developed fulminant hepatitis and one of them died.13
In another prospective study of 150 HBsAg−/HBcAb+ non-Hodgkin's lymphoma patients receiving rituximab, cyclophosphamide, doxorubicin, vincristine, prednisone (R-CHOP) treatment, 17 (11%) developed hepatitis B reactivation (defined as a >10-fold increase in HBV DNA or >10,000 copies/ml if initially undetectable), 1 patient almost 12 months after chemotherapy was completed.14 While the likelihood of hepatitis B reactivation was lower in HBsAg−/HBcAb+ patients who were also hepatitis B surface antibody positive (HBsAb+), reactivation still occurred in some of these patients.14 This was also noted in a retrospective analysis of 178 HBsAg−/HBcAb+ patients from 6 Asian countries who received rituximab with conventional chemotherapy; the HBV reactivation rate (defined as HBsAg reconversion) was 9.6% and the median time to reactivation was 8.9 months. In addition, HBsAb− patients were significantly more likely to experience reactivation than HBsAb+ patients.15 Meta-analyses have corroborated the association between rituximab treatment and HBV reactivation in HBsAg−/HBcAb+ patients with non-Hodgkin's lymphoma.16,17
Based on this data, rituximab (as well as anthracycline chemotherapy and prolonged high dose corticosteroid use) is thought to place susceptible patients at high risk for HBV reactivation.12,18 Screening of patients is therefore important so that preventive therapy can be used. Lamivudine has been shown to be effective in preventing HBV reactivation in HBsAg+ patients receiving cytotoxic chemotherapy.19,20 In a meta-analysis including studies of lamivudine prophylaxis in HBsAg+ patients with lymphoma, most receiving cytotoxic chemotherapy only, HBV reactivation was demonstrated in 8.6% of patients receiving lamivudine compared with 50.6% who did not.21 Treatment-associated lamivudine resistance is known to occur, with the risk for developing the resistance mutation increasing with the duration of lamivudine therapy.19,22 This concept has a significant impact on HBsAg+ or HBsAg−/HbcAb+ patients with lymphoma receiving rituximab since the effects of rituximab therapy may last at least a year and maintenance rituximab therapy may be continued for months after the conventional chemotherapy course. In a retrospective analysis, 31% of 96 HBsAg+ patients with lymphoma on a rituximab-containing regimen and treated preventively with lamivudine developed HBV reactivation, vs no HBV reactivation in 31 HBsAg+ patients treated preventively with entecavir.15 In a prospective randomized study of 121 HBsAg+ patients with lymphoma on a rituximab-containing regimen, HBV reactivation occurred in 6.6% of patients who received preventive entecavir vs 30% of patients who received preventive lamivudine.23 In a prospective, randomized study of 80 HBsAg-/HBcAb+ lymphoma patients on a rituximab-containing regimen, HBV reactivation occurred in 1/41 patients who received preventive entecavir vs 7/39 who did not.24 Recommendations have emerged to support entecavir (or tenofovir by association, although there is no comparative rituximab treatment data) prophylaxis for HBsAg+ or HBsAg−/HBcAb+ patients who are treated with rituximab (or ofatumumab or obinutuzumab), with prophylaxis continuing at least 12 months after the completion of anti-CD20 therapy.6,25
Reports also suggest an increased incidence of other infections in patients receiving rituximab. One study noted a significant increase in hepatitis C viral load during chemotherapy compared with pre-chemotherapy in patients with hepatitis C-positive diffuse large B-cell lymphoma receiving rituximab, cyclophosphamide, doxorubicin hydrochloride, vincristine sulfate (oncovin), and prednisone (R-CHOP) compared with CHOP alone, as well as significantly increased hepatic toxicity in the R-CHOP group.26 In a Mayo Clinic experience over a period of 13 years, 30 patients receiving rituximab developed Pneumocystis jiroveci (PCP) pneumonia (29 had hematologic disorders); 3 of these patients did not receive concomitant chemotherapy or steroids.27 There have been other case reports of PCP and viral infections in patients receiving rituximab therapy, but in the absence of evidence from prospective comparative studies a causative role for rituximab has not been established.3,28 A number of cases of progressive multifocal leukoencephalopathy (PML) have been reported in patients receiving rituximab, and a black box warning to this effect appears in the US Food and Drug Administration (FDA) labeling.29 Most reports are in patients with lymphoproliferative disorders, which even prior to rituximab use had been associated with an increased incidence of PML. The increased incidence of PML in patients receiving rituximab is much lower than the well-documented increased incidence conferred by natalizumab.30