Dr. Bingham has received consulting fees, speaking fees, and/or honoraria from Genentech and Roche (less than $10,000 each) and grant support from Genentech.
Rheumatoid Arthritis Clinical Studies
Immunization responses in rheumatoid arthritis patients treated with rituximab: Results from a controlled clinical trial†
Article first published online: 28 DEC 2009
Copyright © 2010 by the American College of Rheumatology
Arthritis & Rheumatism
Volume 62, Issue 1, pages 64–74, January 2010
How to Cite
Bingham, C. O., Looney, R. J., Deodhar, A., Halsey, N., Greenwald, M., Codding, C., Trzaskoma, B., Martin, F., Agarwal, S. and Kelman, A. (2010), Immunization responses in rheumatoid arthritis patients treated with rituximab: Results from a controlled clinical trial. Arthritis & Rheumatism, 62: 64–74. doi: 10.1002/art.25034
ClinicalTrials.gov identifier: NCT00282308.
- Issue published online: 28 DEC 2009
- Article first published online: 28 DEC 2009
- Manuscript Accepted: 8 SEP 2009
- Manuscript Received: 4 MAR 2009
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
To examine immunization responses in patients with rheumatoid arthritis (RA) treated with rituximab and to investigate the effects of rituximab-induced CD20+ B cell depletion on immune responses to tetanus toxoid (T cell–dependent antigen), pneumococcal polysaccharide (T cell–independent antigen), and keyhole limpet hemocyanin (KLH) (neoantigen) and on delayed-type hypersensitivity (DTH).
In a controlled trial, we enrolled 103 patients with active RA receiving a stable dose of methotrexate (MTX). Tetanus toxoid, pneumococcal polysaccharide, and KLH vaccines as well as a Candida albicans skin test were administered to 1 group of patients receiving rituximab plus MTX (called rituximab-treated patients) for 36 weeks and to 1 group of patients receiving MTX alone for 12 weeks. The primary end point was the proportion of patients with a ≥4-fold rise in antitetanus IgG levels. Antitetanus, antipneumococcal, and anti-KLH serum IgG levels were measured prior to and 4 weeks following vaccine administration. The DTH response to C albicans was measured 2–3 days following placement.
Responses to tetanus toxoid vaccine (≥4-fold rise) were similar in both groups (39.1% of rituximab-treated patients and 42.3% of patients treated with MTX alone). The ability to maintain a positive DTH response to the C albicans skin test was comparable in both groups (77.4% of rituximab-treated patients and 70% of patients treated with MTX alone), showing no effect of rituximab treatment. Rituximab-treated patients had decreased responses to pneumococcal polysaccharide vaccine (57% of patients had a 2-fold rise in titer in response to ≥1 serotype, compared with 82% of patients treated with MTX alone) and to KLH vaccine (47% of patients had detectable anti-KLH IgG, compared with 93% of patients treated with MTX alone).
Recall responses to the T cell–dependent protein antigen tetanus toxoid as well as DTH responses were preserved in rituximab-treated RA patients 24 weeks after treatment. Responses to neoantigen (KLH) and T cell–independent responses to pneumococcal vaccine were decreased, but many patients were able to mount responses. These data suggest that polysaccharide and primary immunizations should be administered prior to rituximab infusions to maximize responses.
Rituximab, a monoclonal antibody that targets CD20+ B cells, is effective for the treatment of rheumatoid arthritis (RA) (1–3). The CD20 antigen is restricted to B lymphocytes from the pre–B cell stage until terminal differentiation into plasma cells, and it is not present on mature plasma cells or stem cells (4). Peripheral B cell depletion lasts for an average of 6–9 months and in some instances longer (2, 5). Although rituximab results in near-to-complete CD20+ B cell depletion in peripheral blood, there is evidence that CD20+ cells in other tissues (e.g., spleen, lymph nodes, Peyer's patches, and synovium) may not be as susceptible (6, 7). In rituximab-treated RA patients, immune responses to influenza vaccine are not completely impaired (8), and this may be partly attributed to incomplete tissue B cell depletion as well as to the absence of CD20 on long-lived plasma cells. Rituximab treatment has been shown to differentially affect serum antibody levels, especially IgM (1, 9). Conversely, levels of IgG specific for infectious antigens, such as tetanus and pneumococcus, have remained stable over multiple treatment courses (1, 9).
Patients with RA are at an increased risk of infections, as a consequence both of immune dysregulation from the disease process and of immunomodulatory therapy (10, 11). As a result, vaccinations against infections are an important part of the management of patients with rheumatic diseases. However, treatment with immunosuppressive agents, such as methotrexate (MTX), has been associated with lower immunization response rates for antigens such as pneumococcal vaccines (12, 13). Given the effects of rituximab on B cells, an impact on immunization responses is postulated. Immunization responses in rituximab-treated patients have previously been assessed in small nonrandomized studies (14, 15). The objective of the current study was to characterize the pharmacodynamic effects of rituximab on humoral and cellular immunity in patients with RA by evaluating the effects of rituximab treatment on immunization responses to a variety of antigens in a randomized, controlled trial.
PATIENTS AND METHODS
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
The study enrolled 103 patients with RA, receiving a stable dosage of 10–25 mg/week MTX, at 26 centers in the US between January 2006 and December 2007. Patients provided signed written informed consent. Patients were age 18–65 years with active RA, defined as a swollen joint count ≥4 and a tender joint count ≥6.
Exclusion criteria included a history of pneumococcal vaccination within 3 years, tetanus vaccination within 5 years, hypersensitivity to immunizations, shellfish allergy, history of autoimmune disease or current inflammatory disease other than RA, history of cancer, current serious chronic infection, or other serious uncontrolled medical illness. Disease-modifying antirheumatic drugs (DMARDs) other than MTX were discontinued ≥4 weeks prior to randomization. Concomitant prednisone (≤10 mg/day or equivalent) was allowed and continued at a stable dose. We excluded patients with a neutrophil count <1.5 × 103/μl, elevated liver enzymes (aspartate aminotransferase or alanine aminotransferase ≥2.5 times the laboratory upper limit of normal), and levels of IgG <5.0 mg/ml and/or IgM <0.4 mg/ml. An upper age limit of 65 years was used because of known attenuation of vaccine response in older patients (16). Patients with a history of tetanus immunization (within 5 years) or pneumococcal immunization (within 3 years) were excluded because additional doses of these vaccines are not recommended at shorter intervals, and those patients may be less likely to mount additive immune responses to the respective vaccines (17, 18).
The Study Investigating the Effects of Rituximab on Immune Response in RA patients (SIERRA) was a randomized, controlled study. The study consisted of a screening period and a study period (36 weeks in the group receiving rituximab plus MTX [also called the rituximab-treated group] and 12 weeks in the group receiving MTX alone).
The study period for the group receiving rituximab plus MTX was temporally longer to enable immunization at times when patients were depleted of peripheral blood B cells but might have B cell repletion in lymphoid organs. This was thought to be the optimum time for a humoral immune response as well as a convenient time for RA patients to be immunized, given that most would receive retreatment with rituximab shortly after this time.
Patients were stratified by age (18–50 years and 51–65 years) and site and randomized 2:1 to receive either open-label rituximab (2 × 1,000 mg given 2 weeks apart) plus a stable dose of MTX or no rituximab (MTX alone). Methylprednisolone (100 mg administered intravenously [IV]) was administered before each rituximab infusion only in the group receiving rituximab plus MTX. Patients could also receive antihistamine and/or acetaminophen. Patients randomized to the group receiving MTX alone continued to receive 10–25 mg/week MTX.
Tetanus toxoid adsorbed vaccine (Aventis Pasteur, Swiftwater, PA), 23-valent pneumococcal polysaccharide vaccine (Pneumovax; Merck, Rahway, NJ) (19, 20), keyhole limpet hemocyanin (KLH; Intracel, Frederick, MD), and a Candida albicans skin test (Allermed, San Diego, CA) were administered according to the protocol schedule (Figure 1A) in both groups. Tetanus and pneumococcal vaccines were administered as intramuscular injections in the deltoid muscle. KLH was administered subcutaneously, and C albicans was administered as an intradermal skin test. Antitetanus, antipneumococcal, and anti-KLH serum IgG levels were measured prior to and 4 weeks following each respective vaccine administration. The delayed-type hypersensitivity (DTH) skin test response was measured 2–3 days following placement.
Response to tetanus vaccine was tested to assess the integrity of a T cell–dependent anamnestic humoral response. The 23-valent pneumococcal polysaccharide vaccine was chosen to assess a mostly T cell–independent or “pure B cell” humoral response. KLH, a mollusk-derived metalloprotein, an antigen unknown to recipients, was chosen to assess neoantigen response (21). The C albicans skin test was chosen to assess (T cell–mediated) DTH responses.
End points and assays of immunization response.
The primary end point was the proportion of patients with positive responses to tetanus toxoid vaccine. A positive response to tetanus was a ≥4-fold increase from prevaccination baseline in antitetanus IgG titer for patients with prevaccination tetanus antibody titers ≥0.1 IU/ml, or a titer ≥0.2 IU/ml for patients with prevaccination titers <0.1 IU/ml (22).
Secondary end points included additional measures of tetanus response and responses to pneumococcal vaccine, KLH, and the Calbicans skin test. For response to tetanus vaccine, we measured the proportion of patients with a ≥2-fold increase from prevaccination baseline in antitetanus IgG titer for patients with prevaccination tetanus antibody titers ≥0.1 IU/ml, or the proportion of patients with a titer ≥0.2 IU/ml for patients with prevaccination titers <0.1 IU/ml.
For pneumococcal vaccinations, end points included the proportion of patients with positive antibody responses against an individual pneumococcal serotype for 12 serotypes and the proportion of patients with a positive response against at least 1, 2, 3, 4, 5, or 6 of the 12 serotypes measured. A positive response against a serotype was defined as a 2-fold increase or an increase of >1 μg/ml from prevaccination levels (23, 24). For KLH response, anti-KLH antibody levels 4 weeks postimmunization were assessed. Antitetanus and anti-KLH antibodies were measured by standard enzyme-linked immunosorbent assay, and antipneumococcal antibodies were measured by fluoroimmunoassay.
The proportion of patients who maintained a positive DTH response to C albicans was assessed from day 1 to week 24 in the rituximab-treated group and from day 1 to week 12 in the group treated with MTX alone. A positive response to the C albicans skin test was defined as at least 5 mm in the mean of the longest and midpoint orthogonal diameters of induration at 48–72 hours (25).
All adverse events (AEs) and serious AEs were recorded. Serious AEs were those that required hospitalization or IV antibiotics, and as previously defined (2). Infusion reactions were defined as any AE reported during or within 24 hours of an infusion. Serious infusion reactions were defined as infusion reactions that met criteria for serious AEs.
Assessments included immunoglobulin levels (IgM, total IgG and IgG subsets [for a subset of patients], and IgA) and fluorescence-activated cell sorting analyses for T cell (CD3+, CD4+, CD8+) and B cell (total CD19+, CD19+CD27+, CD19+CD27–) subsets. Sera were collected from both groups.
Differences in immune responses were assessed using descriptive statistics. Missing data were not imputed, and all analyses were based on available data.
Covariates at baseline that may influence vaccination responses were examined using logistic regression models. The regression model was adjusted for covariates, including continuous variables (age and MTX dose) and categorical variables (treatment, sex, background corticosteroid use, diabetes mellitus history, and tobacco use history). The complete model was fit for each of the 3 vaccination responses and treatment effects, and P values were examined. No selection was performed for these models. In addition, logistic regression analyses were performed using a forward selection methodology to detect any potential relationships between vaccination responses and the following continuous laboratory variables: CD19+ counts, IgA, IgM, IgG, IgG1, IgG2, IgG3, and IgG4. These statistical models were fit separately for baseline and postbaseline laboratory values to allow for both static and dynamic testing of effects. In addition, because IgG subclasses were only available for a subset of patients and postbaseline levels were only available for rituximab-treated patients, models were fit separately to examine effects of immunoglobulin levels. After all testing was completed, another forward selection was employed to integrate the results of the models, and a final multivariable model was determined.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Baseline characteristics and disposition.
Of 103 patients (69 randomized to the group receiving rituximab plus MTX and 34 to the group receiving MTX alone), 68 in the rituximab plus MTX group and 32 in the MTX alone group received at least 1 vaccine and were included in the safety-evaluable population. Baseline patient and disease characteristics are summarized in Table 1. Patient characteristics were generally well balanced between groups, although the proportion of white patients in the rituximab plus MTX group was lower (71% versus 84%), and a higher proportion of patients in the rituximab plus MTX group were receiving background corticosteroids (41% versus 25%) and demonstrated anergy to the C albicans skin test (<5 mm orthogonal induration) at pretreatment baseline (52%). In the MTX alone group, 28 of 34 patients (82%) completed the 12-week study period, and in the rituximab plus MTX group, 65 of 69 patients (94%) completed the 36-week treatment period (Figure 1B). The analysis populations for immunization and skin test responses consisted of patients who had pre- and postimmunization titers and received the given vaccine, and those who had received both skin tests. Therefore, sample sizes differed slightly for the various immune response analyses.
|Characteristic||MTX alone (n = 32)||Rituximab + MTX (n = 68)|
|Age, mean ± SD years (range)||49.7 ± 10.5 (30–64)||49.7 ± 9.6 (20–65)|
|Age ≥50 years||17 (53)||38 (56)|
|Female||25 (78)||51 (75)|
|White||27 (84)||48 (71)|
|Black||3 (9)||7 (10)|
|Hispanic||2 (6)||10 (15)|
|Other||0 (0)||3 (4)|
|Never||17 (53.1)||34 (50.0)|
|Current||4 (12.5)||14 (20.6)|
|Previous||11 (34.4)||20 (29.4)|
|RA disease duration, mean ± SD years||8.4 ± 7.7||8.5 ± 7.9|
|Previous DMARDs, mean ± SD†||2.7 ± 1.9||3.1 ± 2.3|
|Previous use of a TNF inhibitor||18 (56.3)||38 (55.9)|
|Baseline concomitant corticosteroid use‡||8 (25.0)||28 (41.2)|
|Corticosteroid dose, mean ± SD mg/day§||8.5 ± 4.2||6.4 ± 3.0|
|Baseline MTX dose, mean ± SD mg/week||16.4 ± 4.3||17.2 ± 4.2|
|RF positive||21 (65.6)||44 (64.7)|
|Anti-CCP positive||21 (65.6)||43 (63.2)|
|Baseline Candida albicans anergy¶||8 (29)||33 (52)|
|CRP level, mean ± SD mg/dl||NC||1.2 ± 1.6|
|DAS28-ESR, mean ± SD||NC||6.2 ± 1.1|
The majority of rituximab-treated patients had peripheral CD19+ B cell depletion during immunization. At week 24 (tetanus immunization and repeated skin test), 92.3% of patients had peripheral CD19+ B cells below the lower limit of normal (LLN) (<80 cells/μl). At week 28 (administration of 23-valent pneumococcal polysaccharide vaccine), 88.9% of patients had CD19+ B cell counts below the LLN, and at week 36 (4 weeks after KLH immunization), 76.4% of patients had peripheral CD19+ B cell counts below the LLN. Peripheral naive B cells (CD19+CD27–) began to increase by week 28, while peripheral memory B cells (CD19+CD27+) remained depleted (Figure 2A). T cell subset counts (CD3+, CD4+, and CD8+) were stable.
Immunoglobulin levels over time are shown in Figure 2B. Rituximab-treated patients had a decrease in median IgM levels; median levels remained above the LLN at all time points. There were no notable changes in median IgA or IgG levels. At baseline, prior to receiving rituximab, the proportion of rituximab-treated patients with IgM and IgG levels below the laboratory LLN (0.5 mg/ml and 6.7 mg/ml, respectively) was 6% and 9%, respectively. At week 36 following rituximab treatment, the corresponding percentages of patients with IgM and IgG levels below the LLN were 14% and 7%, respectively. In the MTX alone group, 1 patient (3.7%) had IgM or IgG below the LLN at both baseline and week 12.
Of rituximab-treated patients who had IgG subsets analyzed (n = 44 for rituximab-treated patients and n = 14 for MTX alone–treated patients), pretreatment baseline levels were above the LLN for all but 1 patient for IgG1 and IgG2 and for 4 patients for IgG4 (all in the rituximab-treated group) (Figure 2C). No overall changes were observed in IgG1 or IgG2 levels over time. Small decreases were seen in both IgG3 and IgG4 in rituximab-treated patients by the first postbaseline visit, after which levels stabilized. Median levels remained above the LLN. The rituximab-treated patients also received premedication with corticosteroids, which may affect IgG subclass levels (26).
Responses to tetanus toxoid vaccine.
Responses to tetanus toxoid vaccine were similar in both groups (Table 2). A total of 25 rituximab-treated patients (39.1%) and 11 MTX alone–treated patients (42.3%) demonstrated a ≥4-fold rise in antitetanus IgG titer, yielding a difference between groups of –3.2% (95% confidence interval [95% CI] –25.7%, 19.2%). A total of 35 rituximab-treated patients (54.7%) and 16 MTX alone–treated patients (61.5%) demonstrated a ≥2-fold rise in antitetanus IgG titer (difference between groups –6.8% [95% CI –29.2%, 15.5%]).
|MTX alone (n = 26)||Rituximab + MTX (n = 64)||% difference (95% CI)|
|Patients with ≥4-fold titer increase 4 weeks after vaccine, no. (%)†‡||11 (42.3)||25 (39.1)||−3.2 (−25.7, 19.2)|
|Patients with ≥2-fold titer increase 4 weeks after vaccine, no. (%)‡||16 (61.5)||35 (54.7)||−6.8 (−29.2, 15.5)|
|No. of patients with nonprotective titer (≤0.1 IU/ml) prior to vaccine||3||3||–|
|No. of patients with nonprotective titer (≤0.1 IU/ml) before vaccine and with protective titer 4 weeks after vaccine||0||2||–|
|GMT prior to vaccine (95% CI)||1.0 (0.49, 2.14)||1.2 (0.88, 1.69)||–|
|GMT 4 weeks after vaccine (95% CI)||5.2 (2.25, 12.00)||3.9 (2.72, 5.74)||–|
Geometric mean titers (GMTs) of antitetanus IgG before and after immunization were similar between treatment groups, with similar point estimates. Only 3 patients in each treatment group had nonprotective (≤0.1 IU/ml) antitetanus titers before immunization, and 2 patients in the rituximab-treated group mounted protective titers after vaccination.
Responses to skin test.
A total of 31 rituximab-treated patients (48.4%) and 20 MTX alone–treated patients (71.4%) demonstrated positive DTH responses of at least 5 mm induration at baseline prior to receiving rituximab (Table 3). Of those, 77.4% of the rituximab-treated patients and 70% of the MTX alone–treated patients showed the same level of response on repeat testing (difference between groups 7.4% [95% CI –17.5%, 32.3%]). The ability to maintain a positive response to the C albicans skin test was comparable in both groups. A sensitivity analysis was performed with a lower threshold definition of DTH response of ≥2 mm induration. DTH response maintenance rates were consistent with the 5 mm threshold, with comparable rates of response maintenance of 80% and 85% in rituximab-treated patients and MTX alone–treated patients, respectively.
|MTX alone, no. (%)||Rituximab + MTX, no. (%)||% difference (95% CI)|
|Five mm induration|
|Patients positive at baseline||20 (71.4)||31 (48.4)||–|
|Patients maintaining a positive response†||14 (70.0)||24 (77.4)||7.4 (−17.5, 32.3)|
|Two mm induration|
|Patients positive at baseline||21 (75)||41 (64)||–|
|Patients maintaining a positive response†||18 (85)||33 (80)||5.0 (−14.9, 24.9)|
Responses to 23-valent pneumococcal polysaccharide vaccine.
The proportion of patients with responses to each of the 12 pneumococcal serotypes tested was decreased in rituximab-treated patients compared with MTX alone–treated patients (Figure 3A). The proportion of rituximab-treated patients with a positive response to at least 1, 2, 3, 4, 5, and 6 serotypes was decreased compared with MTX alone–treated patients (Figure 3B). Postimmunization GMTs of antipneumococcal IgG were lower overall in rituximab-treated patients (data not shown).
Responses to KLH.
The proportion of patients who mounted a quantifiable anti-KLH IgG response 4 weeks postimmunization was lower in the rituximab-treated patients (30 of 64 patients [46.9%]) than in the MTX alone–treated patients (25 of 27 patients [92.6%]). The GMT of anti-KLH IgG 4 weeks postimmunization for MTX alone–treated patients (1,585.5 titer units [95% CI 1,065.15, 2,360.17]) was ∼3-fold greater than the corresponding GMT for rituximab-treated patients (539.5 titer units [95% CI 461.54, 630.61]).
Predictors of immunization response.
For all 3 vaccines, the only consistently significant predictor was IgG2 level at time of immunization. The odds ratios for IgG2 were 1.01 (95% CI 1.003, 1.016) (P = 0.0028), 1.008 (95% CI 1.002, 1.015) (P = 0.015), and 1.009 (95% CI 1.003, 1.015) (P = 0.0039) for tetanus, pneumococcal, and KLH vaccines, respectively. The net result of each mg/dl increase in IgG2 at the time of vaccination was an ∼1% increase in the odds of a vaccination response. Age, MTX dose, concomitant corticosteroid use, diabetes mellitus, smoking, baseline anergy to C albicans skin test, IgM, IgA, total IgG, IgG1, IgG3, IgG4, and peripheral CD19+ B cell counts were tested and found not to be significant predictors of immunization response.
The safety evaluable population was defined as all patients who received any amount of rituximab or vaccine. Because of the open-label nature of the study and different duration of the study periods of the 2 treatment groups, safety between the groups could not be directly compared. There were no withdrawals due to AEs in the group treated with rituximab plus MTX, and there were 2 withdrawals due to AEs in the group treated with MTX alone (RA flare and essential tremor prior to receiving vaccine). In the MTX alone–treated patients, 2 (6.1%) experienced serious AEs (RA flare and ovarian cyst). In the rituximab-treated patients, 3 (4.4%) experienced serious AEs (hip fracture, chest pain, and coronary artery disease). No deaths occurred during the study periods.
No cases of serious infections were reported in either group during the study periods. Infusion reactions in the group treated with rituximab plus MTX occurred in 20 patients (29%) for the first infusion and in 11 patients (16%) for the second infusion. For the first infusion, the most common AEs were throat irritation, rash, and pruritus. For the second infusion, the most common AE was headache. There was 1 serious AE within 24 hours of an infusion (a hip fracture) but no serious infusion reactions.
AEs attributed to immunizations or skin tests occurred in 15 rituximab-treated patients (22%) and in 8 MTX alone–treated patients (24.2%). These included itching, rash, soreness at the injection site, and malaise. Of the 3 administered vaccines, the most common immunization to which an AE was attributed was KLH (9 patients).
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
We report the first randomized, controlled study of immunization responses in patients with RA treated with rituximab. Four types of immune responses were assessed when the majority of rituximab-treated patients were depleted of peripheral blood B cells: T cell–dependent recall responses to the protein antigen tetanus toxoid vaccine, T cell–independent responses to pneumococcal polysaccharide vaccine, T cell–dependent primary responses to the protein KLH, and maintenance of (T cell–mediated) DTH responses to the C albicans skin test. Responses to tetanus toxoid vaccine were comparable in both groups of patients. In addition, rituximab did not impair DTH responses, since maintenance of DTH response was comparable between groups. Responses to neoantigen (KLH) and T cell–independent responses to pneumococcal polysaccharide vaccine were decreased in a higher proportion of rituximab-treated patients. However, more than half of the patients who received rituximab were able to mount antipneumococcal responses to ≥1 serotype.
Smaller, nonrandomized studies of immunization response in patients with lymphoma and lupus treated with rituximab showed decreased responses to tetanus toxoid vaccine (14, 27). Because these studies were not randomized and examined different patient populations and the immunizations occurred at earlier time points following rituximab treatment, it is difficult to compare their results with those of the present trial. However, responses to tetanus toxoid vaccine were also decreased in a study of rhesus monkeys treated with rituximab (28). Immunizations in that study were administered earlier in the course of rituximab treatment than in the present study, and it is possible that immune responses to tetanus toxoid vaccine are more robust later in the time course when the tissue compartment has been repopulated with CD20+ B cells, but the peripheral blood has not (29).
Taken together, all studies of antitetanus immune response suggest that memory responses are decreased shortly after rituximab treatment, but not at later time points when tissue B cell repletion is likely advanced. Mechanistically, the antitetanus B cell memory compartment is partially depleted after rituximab treatment, resulting in decreased responses if immunization is administered in the first few months after B cell depletion. However, during B cell repletion it is possible that the residual tetanus-specific B cell memory compartment expands and results in good responses when the tetanus boost is administered 6 months after B cell depletion. Whether the antitetanus memory resides in CD20– plasma cells, in a partially depleted CD20+ memory cell compartment, or both is unknown and could be the focus of further research.
Our study confirms findings of previous smaller clinical studies and preclinical studies which suggested that immune responses to neoantigens and polysaccharides are decreased following rituximab treatment. Bearden et al showed that patients with pretransplant chronic renal failure treated with rituximab had decreased responses to the neoantigen PhiX174, and van der Kolk et al showed decreased responses to the neoantigens KLH and hepatitis A vaccine (14, 30). Animal studies have shown decreased primary and memory responses to DNP-KLH in baboons and mice treated with antibodies to CD20 (7, 31). In a study of 15 patients with systemic lupus erythematosus treated with rituximab, Albert et al showed that responses to the pneumococcal polysaccharide vaccine were decreased (27).
Because neoantigen and polysaccharide responses are B cell dependent, decreased responses to KLH and 23-valent pneumococcal polysaccharide vaccine are consistent with rituximab's mechanism of action. Despite peripheral B cell depletion, however, responses to the KLH and pneumococcal vaccine were not completely abrogated in the present study. A total of 47% of patients had a quantifiable anti-KLH response, and 57% and 43% of rituximab-treated patients mounted responses to at least 1 and 2 pneumococcal serotypes, respectively. The 23-valent pneumococcal polysaccharide vaccine and the KLH were administered 7 and 8 months, respectively, after rituximab infusions. Similar to the results with tetanus toxoid vaccine, these data suggest that the ability to mount a response to these antigens increases with increasing time after rituximab infusion, potentially at a time when B cells have begun to repopulate tissue.
This study is the first to examine the effects of rituximab on DTH responses. Given that DTH responses are T cell dependent but B cell independent, it is not surprising that rituximab treatment did not have an effect. Because RA patients have relatively high incidences of cutaneous anergy (32) and baseline anergy was detected in up to 20% of patients in the present study, the presence of a group receiving MTX alone was important in showing that DTH responses were not affected by rituximab treatment.
Predictors of immunization were also assessed. IgG2 levels were predictive of immune response to vaccines, although levels did not change significantly and IgG2 subclass deficiency was rare, with only 1 patient having IgG2 below the LLN at baseline. Previous data have shown that in normal healthy adults, IgG2 levels correlate with responses to polysaccharide antigens (33). In the current study, there were no notable changes in IgG. This is in contrast to other studies of rituximab for RA in which small decreases in IgG have been observed (2). In previous studies, patients were older and treated with higher doses of corticosteroids, both of which are associated with decreased IgG levels (16, 26). Other studies of anti-CD20 therapies in which steroids were used minimally or not at all also showed minimal to no effect on IgG (34, 35). IgM levels do decrease with rituximab treatment; however, IgM level was not an immune response predictor. Interestingly, and in contrast to other studies (13, 36), smoking, age, and diabetes were also not predictive of immune response in the present study, likely due to exclusion of patients age >65 years and the small numbers of patients who had diabetes or were current smokers.
The results of this study give us additional insight into the mechanistic effects of rituximab treatment and inform immunization practices. The results are also informative for understanding immune responses in RA patients, since patients treated with MTX alone had low immune responses to tetanus toxoid vaccine, pneumococcal vaccine, and C albicans skin tests compared with historical normal controls (17, 24, 37, 38). An earlier study in RA patients has shown reduction in responsiveness to pneumococcal vaccines compared with healthy control subjects, with one-third of patients responding either to none or to only 1 of 7 serotypes (10). Thus, factors in RA disease pathogenesis and immunosuppressive medication use may contribute to decreased production of antipneumococcal antibody. Similarly, responses to pneumococcal vaccine in healthy individuals also show a wide variability (24).
Randomized, controlled data are limited concerning immunization responses in patients with RA and other autoimmune diseases treated with other biologic therapies (24, 25). In a study in patients with psoriatic arthritis treated with the tumor necrosis factor α (TNFα) inhibitor etanercept, patients receiving MTX were less likely to respond to pneumococcal antigen challenge than those who were not, and older patients (age ≥47 years) were less likely to respond than younger patients (36). In another study, responses to pneumococcal vaccine in RA patients receiving the TNFα inhibitor adalimumab were similar to those in patients receiving placebo ∼1 month after starting therapy (13). RA patients treated with anakinra with or without MTX and RA patients treated with placebo had similar responses to tetanus toxoid vaccine (39). In abatacept-treated patients with psoriasis, decreased responses to the neoantigens KLH and PhiX172 were observed (40).
Immunosuppressive drugs used in RA treatment, including MTX and corticosteroids, may increase the risk of infection and respiratory disease and, subsequently, mortality. Therefore, immunization is essential to afford maximal protection. The results of the present study suggest that to maximize response, immunizations should be administered either prior to rituximab infusion or with as great a time interval as possible after an infusion and before the next infusion. This is not always practical, especially with seasonal vaccines like influenza or vaccines needed for travel. Although responses to pneumococcal polysaccharide vaccine and neoantigens were decreased, many patients did mount responses, suggesting that patients who require immunizations should receive them, even if they have been treated with rituximab.
We could not assess influenza vaccine, a protein vaccine, in this trial due to the length of the study and seasonal variability of the vaccine. Oren et al reported that vaccination against influenza generated a humoral response for 2 of 3 antigens tested in RA patients treated with rituximab (8). The proportion of responders to 1 of the antigens only (California) was significantly lower in the patients treated with rituximab, while the rate of response to the other 2 antigens was similar in rituximab- and nonbiologic DMARD–treated patients. We did not study live vaccines, since they could pose safety risks to immunosuppressed individuals. We also did not study patients age >65 years, who are known to have attenuated vaccine responses, and these results may therefore not be generalizable to older patients. The safety profile in this smaller study was consistent with observations from other trials of rituximab in patients with RA (1–3, 5). Immunization responses do not necessarily correlate with infection risk (40), and there were no serious infections during the study period. In addition, immunizations were well tolerated.
In summary, treatment with rituximab was well tolerated and did not appear to diminish antibody response to tetanus toxoid immunization and DTH skin testing, although responses to pneumococcal and neoantigen vaccines were reduced. These results suggest that patients with RA treated with rituximab can be effectively and safely vaccinated; however, to maximize response, polysaccharide and primary immunizations should be administered prior to rituximab infusions.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Bingham had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Bingham, Halsey, Martin, Agarwal, Kelman.
Acquisition of data. Bingham, Looney, Deodhar, Greenwald, Codding, Agarwal, Kelman.
Analysis and interpretation of data. Bingham, Halsey, Greenwald, Trzaskoma, Agarwal, Kelman.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
We acknowledge the contribution of the following SIERRA investigators: Jeff Alloway, MD, Eugene Boling, MD, Marcy Bolster, MD, Alan Brodsky, MD, Jeffrey Butler, MD, Saima Chohan, MD, Christine Codding, MD, John Donohue, MD, Justus Fiechtner, MD, Maria Greenwald, MD, Joseph Huffstutter, MD, Philip Kempf, MD, Alan Kivitz, MD, Eric Lee, MD, George Liang, MD, James Loveless, MD, Alan Nussbaum, MD, Robert Roschmann, MD, Philippe Saxe, MD, Joy Schechtman, MD, Steve Stern, MD, James Taborn, MD, Robert G. Trapp, MD, and Michael Weitz, MD. Keith Del Villar, PhD, at Genentech provided assistance in the preparation of the manuscript.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
- 2Rituximab for rheumatoid arthritis refractory to anti–tumor necrosis factor therapy: results of a multicenter, randomized, double-blind, placebo-controlled, phase III trial evaluating primary efficacy and safety at twenty-four weeks. Arthritis Rheum 2006; 54: 2793–806., , , , , , et al.
- 3The efficacy and safety of rituximab in patients with active rheumatoid arthritis despite methotrexate treatment: results of a phase IIb randomized, double-blind, placebo-controlled, dose-ranging trial. Arthritis Rheum 2006; 54: 1390–400., , , , , , et al.
- 9Long-term safety of patients receiving rituximab in rheumatoid arthritis clinical trials. J Rheumatol. In press., , , , , , et al.
- 17Centers for Disease Control and Prevention. Prevention of pneumococcal disease: recommendations of the advisory committee on immunization practices (ACIP). URL: www.cdc.gov/mmwr/preview/mmwrhtml/00047135.htm.
- 18Centers for Disease Control and Prevention. Recommended adult immunization schedule — United States, 2009. URL: www.cdc.gov/mmwr/PDF/wk/mm5753-Immunization.pdf.
- 20Pneumovax package insert. Whitehouse Station (NJ): Merck; 2008.
- 22The Immunological Basis for Immunization Series Module 3: Tetanus. Geneva: World Health Organization. URL: http://www.who.int/vaccines-documents/DocsPDF-IBI-e/mod3_e.pdf. 1993 (WHO/EPI/Gen 93.13)..
- 25Candin Candida Albicans Skin Test Antigen for Cellular Hypersensitivity package insert. Allermed Laboratories, Inc. April 2000.
- 35Ocrelizumab, a humanized anti-CD20 monoclonal antibody, in the treatment of patients with rheumatoid arthritis: a phase I/II randomized, blinded, placebo-controlled, dose-ranging study. Arthritis Rheum 2008; 58: 2652–61., , , , , , et al.
- 39Effects of Anakinra (KINERET) on vaccine response in patients with rheumatoid arthritis [abstract]. Arthritis Rheum 2004; 50 Suppl: S564., , , , , , et al.