The effects of rituximab on immunocompetency in patients with autoimmune disease


  • R. John Looney,

    Corresponding author
    1. University of Rochester, Rochester, New York
    • Department of Medicine, Division of Allergy, Immunology and Rheumatology, University of Rochester, Box 695, 601 Elmwood Avenue, Rochester, NY 14642
    Search for more papers by this author
    • Dr. Looney has received consulting fees (less than $10,000 each) from Genentech, IDEC/Biogen, Roche, Amgen, Wyeth, Trubion, Coley, and MedImmune; he also participates in multicenter clinical trials sponsored by some of these companies. Dr. Calabrese has received consulting and speaking fees (more than $10,000) from Genentech.

  • Renganathan Srinivasan,

    1. University of Rochester, Rochester, New York
    Search for more papers by this author
  • Leonard H. Calabrese

    1. Cleveland Clinic Foundation, and College of Medicine of Case Western Reserve University, Cleveland, Ohio
    Search for more papers by this author
    • Dr. Looney has received consulting fees (less than $10,000 each) from Genentech, IDEC/Biogen, Roche, Amgen, Wyeth, Trubion, Coley, and MedImmune; he also participates in multicenter clinical trials sponsored by some of these companies. Dr. Calabrese has received consulting and speaking fees (more than $10,000) from Genentech.


Advances made in the last several years have contributed to the rapid development of B cell–targeting therapies for autoimmune diseases. Rituximab, the first B cell–targeting therapeutic agent approved for use in humans, has played a major role in this progress, with studies, initially in oncology and hematology and more recently in rheumatology, demonstrating that B cell depletion is reasonably well tolerated. There have also been double-blind, placebo-controlled trials in which B cell depletion has been shown to be an effective treatment for at least one autoimmune disease, rheumatoid arthritis (RA) (1–5). In addition, recent double-blind, placebo-controlled trials have shown the efficacy of rituximab in relapsing-remitting multiple sclerosis (MS) and IgM paraprotein-associated neuropathy, and additional controlled trials are under way in antineutrophil cytoplasmic antibody–associated vasculitis (AAV), systemic lupus erythematosus (SLE), MS, dermatomyositis/polymyositis, and type 1 diabetes mellitus. Furthermore, a large number of new B cell–targeting therapeutic agents are in clinical development.

At this point, it seems clear that the use of therapies targeting B cells in autoimmune diseases will become an important part of the practice of rheumatology. Better understanding of the consequences of B cell–targeting therapies should therefore be a high priority. The effect of B cell–targeting therapies on immunocompetency is of particular importance, and is the present focus of this review.

Rituximab is a chimeric monoclonal antibody (human constant regions and mouse variable regions) that recognizes human CD20, a cell-surface glycoprotein expressed on B cells from early development in the bone marrow until terminal differentiation into plasma cells (Figure 1). After a single course of rituximab, peripheral blood routinely remains depleted of B cells for 6–12 months. In addition, depletion of B cells occurs in the tissue, but may not be as dramatic. It is important to note that although treatment with rituximab depletes mature B cells from blood and tissue, it does not eliminate long-lived plasma cells, the major source of protective antibodies (1, 2).

Figure 1.

Targets of rituximab. Shown are the B cell subsets and expression of surface markers, with evidence that long-lived plasma cells are spared. Note that CD19 and CD20 (diamonds) are expressed on all B cells, from immature precursors in the bone marrow to plasmablasts. In contrast, mature plasma cells do not express CD20, and CD19 expression on mature plasma cells is variable. After treatment with rituximab, binding of most anti-CD20 monoclonal antibodies, used in flow cytometry, is blocked; therefore, for accurate enumeration of residual B cells, many investigators have assayed CD19 expression. Plasmablasts and short-lived plasma cells are the major source of anti–double-stranded DNA (anti-dsDNA), anti–rheumatoid factor (anti-RF), and anti–cyclic citrullinated peptide (anti-CCP) antibodies, while long-lived plasma cells produce most of the antimicrobial and anti–RNA binding protein (anti-RBP) antibodies.

In the present review, the concept of immunocompetency is understood to be the ability of integrated immune defenses to defend the host from infection. Guidelines for the evaluation of immunocompetency in patients receiving immunosuppressive agents for autoimmune disease have recently been developed by the Autoimmunity Centers of Excellence (6).

Effect of rituximab on rates of infection

The gold standard for evaluating immunocompetency is assessment of the rates of infection by various organisms. However, determining the effect of rituximab on the rates of infection is problematic. Patients treated with rituximab are often susceptible to infections, which can be attributed to the underlying disease and/or the use of other immunosuppressive drugs. Therefore, rates of infection in the general population cannot be compared with rates of infection in patients treated with rituximab. Some of the most informative findings come from data on adverse reactions during controlled trials. Based on the combined data from phase IIb and phase III placebo-controlled clinical trials of methotrexate with or without rituximab in patients with RA, the rate of serious infection in the group receiving rituximab was 5.0 per 100 patient-years as compared with 3.4 per 100 patient-years in the group receiving placebo (P = 0.49) (7, 8). Thus, there were more serious infections with rituximab treatment, but this was not a statistically significant difference.

Unfortunately, with an infection rate of ∼3 per 100 patient-years, more than 800 subjects would need to be in each group to have sufficient power to detect a doubling of the rate of infection (6). Therefore, the current data, derived from 200–300 RA patients per group, are insufficient to detect a clinically important increase in the rate of serious infections. Rituximab was first approved for use in cancer patients, and thus considerably more data are available from the oncology literature. In patients undergoing treatment with rituximab for B cell lymphomas, the incidence of serious infections (grades III or IV, including sepsis) is ∼2% (see the Genentech and IDEC/Biogen package inserts and prescribing information for rituximab [Rituxan]). In most trials in oncology, there has not been an increase in infection in patients receiving rituximab therapy added to chemotherapy (9). However, an increase in the infection rate has been found after treatment with rituximab in a subset of patients with lymphoma, i.e., patients with human immunodeficiency virus (HIV) infection and lymphoma. In this group of patients, a regimen of rituximab plus CHOP combination treatment (comprising cyclophosphamide, doxorubicin, vincristine, and prednisolone) increased the infection-related mortality rate from 2% to 14% (P = 0.035) (10). This effect was most pronounced in patients with a CD4 T cell count lower than 50/μl. There was also a trend toward increased neutropenia among patients in the rituximab treatment arm (P = 0.11) (10).

One reassuring finding from many studies is that the serum IgG levels generally are maintained after treatment with rituximab (11, 12). However, there have been reports of hypogammaglobulinemia occurring in rituximab-treated patients, especially in those receiving rituximab in combination with aggressive chemotherapy or bone marrow transplantation (13–16). In a prolonged followup study from the University College London, patients with RA treated with multiple courses of rituximab showed declines in the levels of several Ig isotypes, as well as extremely low levels of IgM. Rates of infection were not increased in the patients with low Ig levels, but the number of Ig-deficient subjects was small (17). In industry-sponsored studies of patients with RA undergoing long-term treatment with rituximab, the levels of IgM dropped below the lower limit of normal in ∼20% of patients, and IgG levels also declined below normal in ∼5% of patients. In these long-term followup studies of a relatively large number of patients (n = 1,039), the rate of serious infections was 4.7 per 100 patient-years in patients with normal IgM and IgG levels, 5.6 per 100 patient-years in patients with low IgM levels, and 4.8 per 100 patient-years in patients with low IgG levels (18). Thus, the relatively modest decrease in the expression of Ig found in these studies was not associated with an increased rate of serious infections.

These studies on infection rates and Ig levels after treatment with rituximab raise the question, “When should replacement with intravenous immunoglobulin (IVIG) be started?” Unfortunately, there is no easy answer to this question when considering an appropriate treatment strategy for patients with secondary causes of immunodeficiency (19, 20). Criteria supporting the use of IVIG include 1) low levels of protective antibodies to encapsulated organisms, viruses, or bacterial toxins, as well as a poor or absent antibody response to immunization or infection; 2) severe hypogammaglobulinemia (serum IgG levels <200 mg/dl); and 3) an increased rate, severity, and duration of infections, especially infections of the respiratory system or infections caused by encapsulated organisms such as pneumococcus, Haemophilus influenzae, and meningococcus (19, 20).

In our experience with patients treated with rituximab, we use the following approach. When patients develop severe or frequent infections, especially involving encapsulated organisms, we evaluate the total IgG, IgM, and IgA levels, and also measure specific IgG to pneumococcal polysaccharides and tetanus toxoid. Total IgG levels below 600 mg/dl raise concerns about immunodeficiency, but can be seen in the absence of true immunodeficiency. Levels of specific IgG are better predictors of infection and the need for IVIG.

Patients who have had problems with infections but have good levels of protective antibodies should be evaluated for defects in innate immunity, e.g., expression of classical complement and levels of mannose-binding lectin. Patients with intermediate levels of IgG, i.e., between 300 and 600 mg/dl, who are not having problems with infections and who have protective levels of antipneumococcal polysaccharide IgG and antitetanus IgG (or who develop protective IgG levels after immunization) may have less severe immunodeficiency, and therefore such patients can be followed up closely without initiating therapy with IVIG. Similar recommendations have been made for patients who receive long-term steroid treatment for asthma and whose IgG levels are between 300 and 400 mg/dl (21). Prophylactic IVIG should definitely be started in any patient whose IgG level is below 200 mg/dl, and should be strongly considered for patients with IgG levels lower than 300 mg/dl. Children treated with rituximab will not have the normal increase in Ig levels seen during development (22). For this reason, replacement IVIG should be administered in all infants and very young children who are treated with rituximab (23).

Rituximab and immunizations

An expected complication of rituximab therapy is the failure of the humoral immune system to respond to vaccines. This defect was originally described in nonhuman primates treated with rituximab, and subsequently confirmed in patients with lymphoma, those undergoing dialysis, and those with lupus (24–27). In a study sponsored by the Autoimmunity Centers of Excellence, 9 of 15 patients with SLE failed to show a response to either the tetanus toxoid or the pneumococcal vaccine at 7 months post–rituximab treatment (28). In that study, humoral responses to the vaccine tended to be more likely in patients with earlier reconstitution of B cells. In a study of patients with RA who were immunized with the influenza vaccine at various times after treatment with rituximab, a humoral response to the vaccine was observed, but responded less well than control patients with RA (29).

Immune responses to the polysaccharide antigens of Streptococcus pneumoniae and other encapsulated organisms (TI-2 antigens) present a unique challenge to the immune system (30). TI-2 antigen responses are defective in many different patient populations, including infants younger than age 2 years, patients with asplenia or those who have undergone splenectomy, and patients after high-dose chemotherapy or bone marrow transplantation (31). Splenic marginal zone B cells play a critical role in the responses to TI-2 antigens. Peripheral blood IgM+,CD27+ B cells (nonswitched memory B cells) appear to represent a pool of cells that are related to marginal zone B cells. Peripheral blood IgM+,CD27+ B cells, representing approximately one-third of the circulating B cells in normal adults, are absent in infants, patients with congenital asplenia, and patients who have undergone splenectomy (32). Treatment with rituximab is associated with depletion of both switched and nonswitched memory B cells, and IgM+,CD27+ B cell depletion appears to be especially prolonged. The prolonged defect in nonswitched memory B cells after rituximab therapy suggests that TI-2 responses may remain defective for years. Thus, evaluation of patients after immunization with polysaccharide vaccines in conjunction with assessment of B cell subsets in these patients will be needed to verify this prediction.

Inhibition by rituximab of humoral responses to vaccines raises the possibility that the humoral immune response to infections will also be defective in rituximab-treated patients. Defective humoral responses pose a special problem in treating patients who have previously been exposed to a variety of organisms, including varicella zoster virus (VZV). Indeed, postexposure prophylaxis should be considered in VZV-seronegative patients treated with rituximab. Parvovirus B19 infection and infections producing toxins represent other clinical situations in which the humoral responses may be important, and for which treatment with IVIG may be crucial (33).

Additional studies are needed to determine at what point the humoral response to vaccines and infections might return, and whether immune biomarkers, such as B cell numbers and subsets, can be used to predict a return of the responses. In the meantime, it would be best to immunize with pneumococcal polysaccharide vaccine before administering rituximab. Despite the possibility of poor humoral responses, routine immunizations, including yearly influenza vaccination, should still be given to patients receiving rituximab. However, it should be remembered that some patients will not have an adequate response to influenza vaccine, and that the Centers for Disease Control and Prevention recommendations for prophylactic antiviral therapy should be considered.

Effects of rituximab on innate immunity

Innate immunity can also be affected by treatment with rituximab. Similar to the effects after treatment with polyclonal IVIG, cytopenias, including neutropenia, have been reported to occur immediately after rituximab infusion (20, 34). These early cytopenias are usually transient. Surprisingly, late-onset neutropenia (LON) can also occur following treatment with rituximab. There are a number of reports of LON associated with rituximab in patients with lymphoma (35). In a recent report, LON was investigated in 130 patients with B cell lymphoma who were receiving chemotherapy with or without rituximab (36). LON was detected in 6 of the 76 patients receiving rituximab and in none of the 54 patients not receiving rituximab (P = 0.04). The median nadir neutrophil count in the patients who developed LON was 200; the lowest counts were 23 and 32. Onset occurred at an average of 175 days (range 77–204 days) posttreatment and lasted a median of 14 days (range 11–16 days). One patient developed a buccal cellulitis; the other 5 patients were asymptomatic. The neutropenia responded to treatment with growth factors.

Preliminary data from that study suggest an association between B cell recovery and granulocyte decline (36). These changes in B cell recovery and granulocyte decline were also correlated with changes in the level of stromal cell–derived factor 1, a chemokine affecting the homeostasis of both B cells and granulocytes. Since blood cell counts were checked only every 3 months, the authors estimated that they probably missed more cases of neutropenia than they found. Whether LON will be identified in patients with nonmalignant diseases who are treated with rituximab is unclear, but until we have more information, neutrophil counts in these patients should be monitored.

In addition to the induction of neutropenia, rituximab may also have effects on phagocytes. In a recent review, it was hypothesized that many of the effects of rituximab may be related to interactions with Fcγ receptors (FcγR), and the authors predicted that FcγR functions would be blocked (37). Although this blockade might be beneficial for some types of autoimmunity, it could also block host defenses. Additional studies will be needed to determine the frequency and consequences of abnormal phagocyte function in the presence of normal cell counts.

Effects of rituximab on surveillance of opportunistic and viral infections

As with all immunosuppressive drugs, there is concern about the reactivation of latent or chronic infections. Reactivation of tuberculosis or other intracellular facultative organisms, which have been a problem associated with anti–tumor necrosis factor therapies, has so far not been found with rituximab. The effect of rituximab on chronic or latent viral infections is a more complex issue.

Rituximab has been used extensively in individual cases and small open trials to treat the extrahepatic complications of cryoglobulinemia in patients with chronic hepatitis C virus (HCV) infection, with no compelling evidence of treatment-related adverse effects on liver enzyme levels or clinically significant increases in viral load (38, 39). The effects of rituximab on hepatitis B virus (HBV) are more pernicious, but similarly complicated. In the oncology literature, there are numerous case reports of serious and even fatal HBV reactivation, but most of these cases have been in the context of concomitant chemotherapy (40). There is one report of HBV reactivation in a patient with lymphoma who was receiving rituximab alone (41). HBV reactivation is a potentially serious breach on immunocompetency, and therefore should be included in the screening of all patients being considered for rituximab therapy, regardless of cotherapy.

There are many cases of reactivation of Epstein-Barr virus (EBV) or cytomegalovirus (CMV) in patients receiving rituximab. Nearly all of these cases occurred in patients receiving concomitant chemotherapy, the latter of which may itself lead to reactivation. Reactivation of EBV and CMV has not been reported in patients receiving rituximab alone. Moreover, EBV- and CMV-specific cellular immunity, as assessed using tetramers to detect virus-specific CD8 T cells, appears to persist in these patients (42). Whether the function of these virus-specific CD8 T cells, as well as CD4 T cells, is normal remains to be determined. In patients receiving rituximab for the treatment of malignancies, unusually severe cases of herpes zoster virus have been reported; the role of rituximab in these unusual cases is unclear (16, 43).

The US Food and Drug Administration (FDA) recently issued a warning about a possible association between progressive multifocal leukoencephalopathy (PML) and rituximab therapy (44). Although cases of PML have occurred predominantly in patients with HIV and hematologic malignancies, the FDA warning was sparked by 2 recent case reports of PML in patients with SLE who were receiving rituximab.

PML is a demyelinating disease caused by the JC virus, a human polyoma virus. Approximately 80% of the population is seropositive for the JC virus, which is acquired in childhood but persists in a latent form. In healthy individuals, the JC virus is periodically shed in the urine without any clinical sequelae. In settings of profound compromise of host defense, the JC virus can be found not only in the urine but also in the blood. According to prevailing theory, such viremia may lead to seeding of the central nervous system with infection of myelin-producing oligodendrocytes, resulting in clinical disease.

In recent years, PML has been reported in a variety of conditions associated with immunosuppression or immunodeficiency, including in recipients of organ transplantation, patients undergoing chemotherapy, patients with HIV infection, and, most recently, patients with MS or Crohn's disease treated with natalizumab (45–47). PML has been reported on at least 23 occasions in patients with lymphoma undergoing treatment with rituximab, with the majority having received rituximab in combination with chemotherapy or stem cell transplantation (48–51). The 2 patients with SLE who developed PML after treatment with rituximab had received prior treatment with immunosuppressive medications, including cyclophosphamide, and the PML developed after 6 infusions of rituximab (4 in 2004 and 2 in 2005) in the first patient and after 3 courses of rituximab (from 2002 to 2005) in the second patient. PML infection was the cause of death in both patients.

Since patients with SLE who are receiving rituximab for off-label use generally exhibit resistance to conventional treatment and often have severe immunosuppression, it is not clear whether the rate of PML in such patients is higher than expected. Investigators found that the trial of rituximab for relapsing-remitting MS had fulfilled its primary end point, a reduction in enhancing lesions on magnetic resonance imaging (52), which suggests that rituximab can affect immune responses in the central nervous system. In MS, rituximab decreases both B cells and T cells in the cerebrospinal fluid (53). Whereas most treatments for MS, such as interferon-β and glatiramer, are not associated with PML, natalizumab, an antibody against the α4 integrin, is associated with a strikingly increased rate of PML (1 in 1,500 patient-years) (46). A recent systematic review of PML in patients with rheumatic diseases has shed additional light on this issue but has not clarified the risks (54).

Calabrese and colleagues identified 37 cases of PML in patients with rheumatic diseases, as reported in the MEDLINE database of peer-reviewed literature (54). Interestingly, nearly two-thirds of these cases occurred in patients with SLE, and only a single, well-documented case was reported in a patient with RA. Intriguingly, ∼40% of the SLE patients with PML were not heavily immunosuppressed, having only received glucocorticoids, plaquenil, or no treatment in the 6 months preceding the symptoms of PML. Thus, SLE may in itself represent a poorly understood risk factor for PML, making the assignment of risk associated with rituximab in these isolated cases difficult to assess. Long-term pharmacovigilance will be needed to further sort out this complex issue. The concerns about PML coinciding with rituximab therapy need to be taken seriously. Patients need to be informed about these concerns and educated about the presentation of PML.

Risks and benefits of combination therapy

In patients with severe SLE, those with AAV, or those with myositis, the ability to get the disease under control quickly in order to prevent damage and, at the same time, minimize the risk of infection is a major therapeutic challenge. Using rituximab in combination with other agents seems likely to play an important role in achieving these goals, since it has played a similar role in the treatment of lymphomas. Thus, a number of small, uncontrolled trials have studied the use of rituximab in combination with a variety of immunosuppressive agents.

For SLE, the use of bolus cyclophosphamide and steroids in combination with rituximab, as was originally proposed by Edwards and colleagues at the University College London for patients with RA, has been the focus of most studies in Europe, in which rituximab has been administered in 1–4 doses and intravenous cyclophosphamide in 1 or 2 doses (55–60). Because the disease in these patients was refractory to prior therapies, including, in most cases, cyclophosphamide, the results from those case series are impressive. Nevertheless, there are also case series comprising patients with SLE who have received rituximab as monotherapy added to background therapy or in combination with a course of steroids (61–64). On the whole, it would seem that patients undergoing treatment with the combination of rituximab and cyclophosphamide have presented with more severe disease and also have had more impressive responses after receiving this combination therapy as compared with patients receiving rituximab alone. Since both rituximab and cyclophosphamide are active against B cells, it is reasonable to expect that synergy occurs with these agents.

The current evidence supporting the use of rituximab in AAV is remarkably impressive. The 2 case series from the Mayo Clinic have yielded striking results, e.g., nearly all of the patients had previously failed treatment with cyclophosphamide, and yet the combination of high-dose steroids and rituximab induced complete responses in all patients (65, 66). However, there has also been a study in which rituximab and a single bolus of cyclophosphamide induced equally impressive results without the use of high-dose steroids (58).

For any given non-RA autoimmune disease, it would certainly be a mistake, in the absence of controlled trials, to assume too much about the benefits of rituximab combination therapy. Data on rituximab combination therapy should be forthcoming from the ongoing trials of rituximab in SLE (combined with high-dose steroids or high-dose steroids and mycophenolate) and in AAV (combined with high-dose steroids). Additional controlled studies with rituximab combined with cyclophosphamide would be of particular interest. While waiting for the results of these studies, we continue to recommend rituximab for off-label use to treat patients with a variety of unusually severe or refractory autoimmune diseases. For these patients whose condition is very difficult to manage, we have provided bolus steroids and cyclophosphamide at the time of the rituximab infusions for the initial induction treatment, and then have switched to immunosuppressive drugs such as mycophenolate, azathioprine, or methotrexate for the maintenance phase. We are not providing rituximab for off-label use in patients with less severe autoimmune diseases, except as part of clinical trials.


The current state of our understanding regarding the effects of rituximab on immunocompetency can be summarized as follows.

Rates of infection.

Rates of infection are the gold standard for the assessment of immunocompetency, while all other measures are surrogates. Unfortunately, infection rates are notoriously difficult to determine outside of clinical trials. The rates of serious infections have not been found to be significantly increased in patients receiving rituximab as compared with those receiving placebo in controlled trials in RA; however, with the available data, a doubling of serious infections could easily be missed.

Increased infection-related mortality in HIV patients with lymphoma has been reported after patients underwent treatment with rituximab added to CHOP, and patients with very low CD4 T cell counts were at highest risk. Therefore, by analogy, there should be concern that patients with autoimmune diseases similarly treated with aggressive chemotherapy and rituximab might have increased infection-related mortality.

Concerns have also been raised regarding the possibility of reactivation of HBV, which has been observed in patients with cancer undergoing chemotherapy in conjunction with rituximab therapy. In addition, 2 cases of PML developed in patients with SLE who were treated with rituximab and other immunosuppressive agents. Thus far, reactivation of tuberculosis does not appear to be a problem.

Response to immunization.

It is very clear that rituximab blocks humoral responses to immunization for many months. The duration of this effect and the immunologic correlates of recovery have not been determined. Similarly, the relationship between B cell subsets and the response to immunization will be of particular interest, especially in terms of possible correlations between the recovery of nonswitched memory B cells (the circulating equivalent of marginal zone B cells) and the recovery of the response to pneumococcal polysaccharide and other TI-2 antigens. Data on cellular immune responses and investigations of whether or not B cell depletion shifts the CD4 T cell response more toward Th1 would also be of interest.

Surveillance of latent infections.

There is not much evidence of increased rates or greater severity of herpes virus reactivation after treatment with rituximab. There is a potential risk of reactivation of HBV, based on the observations in oncology, but nearly all of these cases have been in patients receiving chemotherapy in addition to rituximab. The therapy does not seem to exacerbate active HCV infections, according to the results from trials of rituximab for cryoglobulinemia.

Recent reports have raised concerns about PML occurring in association with rituximab therapy. Although the FDA issued a warning based on 2 cases of PML in patients with SLE, the occurrence of PML in these patients, who had only mild immunosuppression while taking conventional medications, makes it difficult to assess the risk attributable to rituximab.

Analysis of cellular immunity.

Investigations of the effects of rituximab on cellular immunity have been limited. In general, these studies have been restricted to enumeration of T cell subsets rather than to functional studies (64, 67, 68).

Analysis of innate immunity.

Rituximab can cause neutropenia, and this can be associated with infections. The effects of rituximab on the function of neutrophils and other phagocyte functions have not been evaluated.

Genetic polymorphisms.

FcγR polymorphisms have been linked to the clinical efficacy of rituximab in lymphoma and to B cell depletion in SLE (69–71). The influence of other genetic polymorphisms on B cell depletion or recovery, as well as the role of genetic polymorphisms on immunocompetency have not been evaluated.


Our current knowledge regarding the effects of rituximab on immunocompetency, as summarized above, is consistent with the experience in oncology and autoimmune diseases, i.e., rituximab is generally well tolerated. Nevertheless, precautions should be taken to minimize the risk of infections. Precautions recommended by the authors are presented in Table 1. These precautions are a response to the problems with immunocompetency that could develop, as has been discussed in this review.

Table 1. Precautions for the use of rituximab in autoimmune diseases*
  • *

    HBV = hepatitis B virus; HIV = human immunodeficiency virus; TB = tuberculosis; HBsAg = hepatitis B surface antigen; VZV = varicella zoster virus; RA = rheumatoid arthritis; IVIG = intravenous immunoglobulin; G-CSF = granulocyte colony-stimulating factor; PML = progressive multifocal leukoencephalopathy.

1. Patients need to be informed about the potential effects of rituximab on immunocompetency
 Decreased response to immunization
 Possible increased rate of infection
 Possible neutropenia and hypogammaglobulinemia
 Possible reactivation of infections, including HBV and JC virus
2. Screening prior to use of rituximab
 History of severe, unusual, or frequent infections
 History of high risk for HIV or TB
 Routine laboratory tests: complete blood cell count with differential, metabolic survey, and urinalysis
 Serology for hepatitis B (HBsAg, anti-HBs, anti-HBc)
 Serology for VZV for patients who have not had chickenpox or been immunized should be considered in patients who are heavily immunosuppressed; this should not be necessary in RA patients
 IgM, IgG, IgA
3. Immunization
 Avoid live virus vaccines
 If possible, complete immunizations 1 month prior to starting rituximab
 After rituximab, continue to administer recommended vaccines, including yearly influenza vaccine, but try to give as late as possible after the last dose
4. At the time of any serious or unusual infection, evaluate neutrophil count and IgG level
 For IgG levels below normal, check IgG antibody titers (antitetanus, antipneumococcal polysaccharide); if IgG antibodies are low, response to immunization should be evaluated
 When to start IVIG and/or prophylactic antibiotics will depend on severity of hypogammaglobulinemia, IgG antibody titers, frequency and type of infections, comorbid conditions, concomitant medications, and patient age
 Neutropenia secondary to rituximab responds well to administration of G-CSF
5. Prophylaxis
 Influenza: for patients who are unlikely to respond to immunization, consider antiviral chemotherapy during outbreak
 HBV: although the risk of reactivation after treatment with rituximab alone is not well established, risk is clearly dependent on the status of the infection before treatment, comorbidities, concomitant medications, and the primary diagnosis; for patients at high risk for reactivation, prophylaxis with lamovudine should be considered
 VZV: postexposure prophylaxis should be considered in seronegative patients who have additional factors putting them at high risk
6. Early recognition and treatment
 Patients and caregivers need to be educated about the importance of early recognition of infections, and systems need to be in place to ensure rapid response by health professionals
 In addition to a general admonition to call their health care provider if they have questions or if they are ill, patients should be specifically instructed to call if they have shaking chills, fever >101°F, respiratory illness other than a mild cold, symptoms of a urinary infection, or if they think they may be developing the flu
 Patients should be specifically informed about the symptoms of hepatitis, zoster, and PML (confusion, lethargy, dizziness, difficulty talking or walking, and vision problems)

It is important to note that studies of B cells and B cell subsets were not included in our recommendations. Studies of B cells to assess immunocompetency would be of great interest, but at this point their clinical relevance is not clear. This may change as the correlation between B cells and immunocompetency is better established. The recent European League Against Rheumatism consensus statement on the use of rituximab in patients with RA, as well as the recent reviews by Calabrese and colleagues on reactivation of viral hepatitis and the JC virus in rheumatic diseases provide additional discussions on these infections as they pertain to the use of rituximab (40, 54, 72, 73).


The authors thank Drs. Anolik and Sanz for their helpful discussion, and Dr. Sanz for providing Figure 1.