Abstract
Background and objectives Remission in the majority of ANCA vasculitis patients is not sustained after a single course of rituximab, and risk of relapse warrants development of a successful strategy to ensure durable remission.
Design, setting, participants, & measurements A retrospective analysis of ANCA vasculitis patients who underwent maintenance therapy using rituximab-induced continuous B-cell depletion for up to 7 years was performed. Maintenance therapy with rituximab was initiated after achieving remission or converting from other prior maintenance therapy. Continuous B-cell depletion was achieved in all patients by scheduled rituximab administration every 4 months. Disease activity, serologic parameters, adverse events, and survival were examined.
Results In the study, 172 patients (mean age=60 years, 55% women, 57% myeloperoxidase–ANCA) treated from April of 2006 to March of 2013 underwent continuous B-cell depletion with rituximab. Median remission maintenance follow-up time was 2.1 years. Complete remission (Birmingham Vasculitis Activity Score [BVAS]=0) was achieved in all patients. Major relapse (BVAS≥3) occurred in 5% of patients and was associated with weaning of other immunosuppression drugs. Remission was reinduced in all patients. Survival mirrored survival of a general age-, sex-, and ethnicity-matched United States population.
Conclusion This analysis provides evidence for long-term disease control using continuous B-cell depletion. This treatment strategy in ANCA vasculitis patients also seems to result in survival rates comparable with rates in a matched reference population. These findings suggest that prospective remission maintenance treatment trials using continuous B-cell depletion are warranted.
Introduction
ANCA vasculitis is a systemic autoimmune disease characterized by small vessel inflammation caused by pathogenic autoantibodies directed against proteinase 3 (PR3) or myeloperoxidase (MPO) (1–3). Immunosuppressive therapy can result in remission; however, most patients relapse, which results in additional injury (4). Furthermore, chronic immunosuppression leads to additional toxicity.
Rituximab, a humanized murine monoclonal antibody directed against CD20 located on the surface of B lymphocytes (B cells), is effective in depleting B cells. The Rituximab in ANCA-Associated Vasculitis (RAVE) and Rituximab versus Cyclophosphamide in ANCA-Associated Renal Vasculitis (RITUXVAS) trials have shown efficacy of rituximab with steroids for induction of remission in ANCA vasculitis, similar to cyclophosphamide and steroids (5,6), and rituximab is now approved by the Food and Drug Administration and European Medicines Agency for this purpose. The use of anti–B cell therapy for early induction of remission in ANCA vasculitis is not surprising given that ANCA are pathogenic in vitro and in vivo (2,3). It is clear that remission in many patients is not sustained with a single induction course of rituximab (7–11). Relapses of ANCA vasculitis often occur after B-cell repopulation (9,10), suggesting that scheduled serial dosing of rituximab could result in sustained remission.
In April of 2006, our group began to give rituximab every 4 months to our most resistant cases. We subsequently reported that continuous B-cell depletion using rituximab was highly successful for early maintenance of remission in 39 patients with ANCA vasculitis (12). This maintenance strategy was the first to use a regimen of scheduled rituximab administration to prevent B-cell repopulation. With this regimen, we no longer waited for B-cell repopulation, ANCA titer, or signs and symptoms of relapse before giving the next dose. Here, we review our 7-year experience in 172 patients treated with rituximab-induced continuous B-cell depletion for maintenance of remission. Specific attention is directed to disease control, medication burden, adverse events, and survival.
Materials and Methods
Study Population
We performed a single center retrospective analysis of patients with ANCA vasculitis who underwent rituximab-induced continuous B-cell depletion for maintenance of remission. We included 172 consecutive patients treated between April of 2006 and March of 2013 at the Vasculitis and Glomerulonephritis Clinic in the Nephrology Division at the Massachusetts General Hospital. Patients were considered to have ANCA vasculitis if they had a positive test for PR3- or MPO-ANCA, which was detected by ELISA in the Massachusetts General Hospital ANCA Clinical Laboratory (12), together with a history of clinical and laboratory features consistent with granulomatosis with polyangiitis (GPA), microscopic polyangiitis (MPA), or related variant forms of vasculitis (13). New patients and patients with disease relapse (Birmingham Vasculitis Activity Score [BVAS]–Wegener's Granulomatosis [WG]≥3) were included in this retrospective analysis when they were transitioned to continuous maintenance rituximab after undergoing induction therapy and entering full remission (defined as BVAS-WG=0 while on prednisone<10 mg/d). Patients who were already in complete or partial remission (BVAS-WG≤2) were included in the study when continuous rituximab was added to their maintenance therapy.
We excluded patients who did not receive a subsequent scheduled rituximab maintenance dose. We also excluded patients whose disease was associated with levamisole-adulterated cocaine exposure and/or who showed active illicit drug abuse.
The Partners HealthCare System Human Research Committee approved this study, and this study was in adherence with the Declaration of Helsinki.
Treatment Protocol
Rituximab.
All patients were initiated on a strategy to maintain complete and continuous peripheral blood B cell depletion using rituximab. Patients who transitioned from induction therapy that included rituximab (Supplemental Figure 1) were continued on one dose every 4 months. Patients who transitioned from other maintenance therapy were started with two doses of rituximab separated by approximately 2 weeks followed by one dose every 4 months. Historically, B-cell return was seen by 6 months in several of our patients; therefore, we modified our dosing strategy early on to every 4 months. Most patients received 1000 mg rituximab at each infusion. Before each rituximab dose, patients received treatment with intravenous hydrocortisone (100 mg) one time, oral diphenhydramine, and oral acetaminophen. Rituximab administration frequency was shortened to every 3 months among patients who showed B-cell return at 4 months just before rituximab administration.
Other Immunosuppression Medications.
Other immunosuppressants were generally weaned at 2- to 4-month intervals as tolerated. For patients transitioned from other maintenance therapy, rituximab was added to the other therapy. Some of these patients had low levels of grumbling disease activity (BVAS-WG≤2) and persistent ANCA positivity at the time of rituximab initiation. For such patients, other immunosuppression was weaned slowly after BVAS-WG=0 was achieved and ANCA titers showed improvement. Weaning of prednisone was limited at times by adrenal insufficiency among patients who had been on prednisone continuously for several years. In patients on rituximab who relapsed, rituximab was continued along with some increase of other maintenance immunosuppression medications followed by a slower taper.
Ancillary Treatment Considerations.
Additional considerations included measurement of urine human chorionic gonadotropin before every rituximab infusion for women of childbearing age, trimethoprim-sulfamethoxazole or one of it equivalents for prevention of Pneumocystis jirovecii pneumonia, a proton pump inhibitor to prevent corticosteroid-induced gastric ulcers, and calcium and vitamin D to prevent corticosteroid-induced osteoporosis. A bisphosphonate was added for patients at highest risk for osteoporosis. Care was taken to give every patient an annual influenza vaccine as well as pneumococcal, tetanus, diphtheria, and acellular pertussis vaccines and test for Hepatitis B virus (HBV) before initiation of therapy. Lamivudine or entecavir was prescribed to any HBV core antibody-positive patient to prevent HBV reactivation, unless HBV surface antibodies were present in high titer. P. jirovecii pneumonia and gastric ulcer prophylaxis were typically discontinued on corticosteroid withdrawal.
Clinical Assessment, Definitions, Measurements, and Data Collection
Clinical assessments and laboratory measurements were maintained prospectively at each visit in a Filemaker Pro 9.0 database. Patients were evaluated at each rituximab infusion and typically, halfway between infusions. These flow sheets were reviewed retrospectively.
ANCA vasculitis disease activity was measured using the BVAS-WG (14). The BVAS-WG was adopted by our group at the time of the Wegener's Granulomatosis-Entretien trial, because it seemed more focused and specific to ANCA vasculitis than all vasculitis. The BVAS-WG was also used in the RAVE trial (5). BVAS-WG=2 means that a patient has two minor features of disease that are not organ-threatening. BVAS-WG=3 means one major feature of disease or three minor features. Patients were monitored by measurement of serum creatinine, ANCA, and peripheral B-cell counts every 4 months. Complete blood counts were checked every 2 months, and Ig levels were checked every 4–12 months. Complete disease remission was defined as BVAS-WG=0. A minor relapse was recorded if the patient developed BVAS-WG=2 after achieving BVAS-WG=0. A major relapse was recorded if the patient developed BVAS-WG≥3.
Adverse Events
All serious adverse events (AEs) were recorded at time of event and included any event requiring hospitalization and/or death. Causes of death were determined by chart review, discussion with outside clinicians, and/or discussion with first-degree relatives.
Statistical Analyses
Patient characteristics stratified by PR3- or MPO-ANCA are presented as mean±SD for age, median (interquartile range) for creatinine, and number (percent) for all categorical variables and compared using t tests for continuous variables and chi-squared tests where appropriate. ANCA titers were log-transformed because of non-normality for parametric testing.
Unadjusted overall time to relapse as well as mortality were described using Kaplan–Meier curves. A log-rank test was used to compare cumulative survival of the cohort with the survival of an age-, sex-, and ethnicity-matched general United States population based on data retrieved from National Vital Statistics Reports (15). Univariate Cox proportional hazards models were used to compare mortality by sex, serotype, and age.
Graphs were created using GraphPad Prism 5.0d. Heat maps were created using GENE-E Dev. Statistical analyses were conducted in SAS, version 9.2. Two-tailed P values<0.05 were considered significant.
Results
Baseline Characteristics
Table 1 illustrates the study population, which was comprised of 172 patients who underwent rituximab-induced continuous B cell depletion for maintenance of remission. We did not assign specific diagnoses of GPA or MPA given uncertainties involved in categorizing subsets of ANCA vasculitides (12). Furthermore, it is now becoming clear that ANCA autoantigen specificity may reflect phenotype more accurately than GPA or MPA (16). All patients were ANCA-positive at diagnosis: 98 (57%) patients had MPO-ANCA, and 74 (43%) patients had PR3-ANCA. There were no patients with both MPO-ANCA and PR3-ANCA. There were significantly more women in the MPO-ANCA group (62% versus 46%, P=0.04). There were more constitutional signs and symptoms (60% versus 38%, P=0.01) and ear, nose, and throat findings (76% versus 44%, P<0.001) measured by the modified BVAS-WG scoring system in the PR3-ANCA group.
Baseline characteristics of ANCA vasculitis patients undergoing continuous B-cell depletion
Induction of Remission and Weaning of Ancillary Immunosuppression
Patients transitioned from induction therapy were in remission at time of entry, and residual prednisone doses were weaned. All patients who transitioned from maintenance therapy had BVAS-WG scores of two or less. In each case, BVAS-WG became zero, and ancillary immunosuppression was weaned as shown in Figure 1.
Ancillary immunosuppression use in ANCA vasculitis patients undergoing continuous B-cell depletion. Heat maps depict use of prednisone, cyclophosphamide (CYC), azathioprine (AZA), mycophenolate mofetil (MMF), and methotrexate (MTX) throughout the study period. Each square represents maximum daily medication dose in milligrams for prednisone, CYC, AZA, and MMF and weekly medication dose in milligrams for MTX during each 90-day period. Squares from left to right represent each 90-day time period. Squares from top to bottom represent each individual patient. Gray squares within study period represent missing data, and each individual patient row becomes gray at the end of study period.
Continuous B-Cell Depletion and ANCA Titers
B-cell depletion was highly effective, because over 96% of tests showed no detectable B cells, defined as B-cell concentration<1 cell/mm3 (data not shown). Less than 1% of samples had >10 B cells/mm3 at any point during the entire study period. Rituximab dosing frequency was increased in these few instances. MPO- and PR3-ANCA titers were measured every 3–4 months in most patients (Figure 2). Despite continuous B-cell depletion, over 50% of MPO-ANCA patients and 25% of PR3-ANCA patients had positive titers throughout the study period.
ANCA titers in ANCA vasculitis patients undergoing continuous B cell depletion. (A) Myeloperoxidase (MPO)-ANCA and (B) proteinase 3 (PR3)-ANCA titers were measured in each patient throughout the study duration. The dashed line represents the lower limit of a positive test, which is 2.8 and 20 units for MPO- and PR3-ANCA, respectively. Error bars represent median and interquartile range. Titers were significantly lower at all time points compared with the 0- to 90-day time period for all time periods with greater than five samples (*P<0.05).
Disease Relapse
Median remission maintenance follow-up time was 2.1 years. All patients achieved complete remission (defined as BVAS-WG=0). Durable maintenance of remission was achieved in the majority of patients (Figure 3). Relapse (BVAS-WG≥2) occurred in 20% (n=35) of patients. Major relapse (BVAS-WG≥3) occurred in nine patients (5%, median BVAS-WG=4, range=4–5) and was associated with weaning of other immunosuppression. Remission was reinduced and maintained in all nine patients with the addition of prednisone and another immunosuppressive agent. Patients with minor relapse (BVAS-WG=2) responded to short courses of corticosteroid recycling and did not experience additional relapses. There were no statistically significant differences between patients who received induction therapy and were then transitioned to rituximab maintenance therapy and patients who were transitioned from other immunosuppressive maintenance therapy to rituximab maintenance therapy (Table 2).
Minor and major relapse-free survival of ANCA vasculitis patients undergoing continuous B-cell depletion. Kaplan–Meier curves comparing major (circles) and minor (triangles) relapse-free remission and survival.
Minor and major relapses among ANCA vasculitis patients undergoing continuous B cell depletion
AEs
Infusion Reactions.
All rituximab infusions were well tolerated and completed without acute anaphylactic reactions (Tables 3 and 4). Occasional minor infusion reactions occurred, and all were managed by immediately halting the infusion. Typical minor reactions included jaw tingling, throat irritation, and/or pruritus. In each case, the reaction dissipated, and the infusion was resumed at a slower rate. No major infusion reactions occurred.
Adverse events of ANCA vasculitis patients undergoing continuous B cell depletion
Patients with ANCA vasculitis undergoing continuous B-cell depletion who died
Delayed Infusion Reactions.
Delayed infusion-related symptoms were reported by several patients over subsequent days and included fever, hoarseness, and arthralgias. These symptoms were typically minor and inconsistently recorded and quantified. One patient was admitted to the hospital for 1 night for flu-like symptoms together with relative hypotension after an infusion.
Infections Requiring Hospitalization.
Twenty patients had at least 1 serious infection, with a total of 25 serious infections occurring overall, and 9 were pulmonary infections. Minor outpatient infections could not be quantitated.
Late-Onset Neutropenia.
Seventeen patients developed late-onset neutropenia (LON; defined as absolute neutrophil count<1000 cells/mm3). Four episodes resolved without intervention by the time of repeat blood work. However, 13 episodes were treated with granulocyte colony-stimulating factor (filgrastim). Four patients had associated fever and were hospitalized for intravenous antibiotics. In all patients, neutropenia subsequently resolved, and rituximab was not discontinued because of this effect in any patient.
Hypogammaglobulinemia.
Seventeen episodes of hypogammaglobulinemia to an IgG level of <400 mg/dl developed in 17 patients (10% of cohort), and rituximab was discontinued in 9 patients. None of these patients had a disease relapse, and six patients were hospitalized for infection; however, three infections were associated with fever in the setting of LON.
Malignancy.
Other than nonmelanoma skin cancers, there were two malignancies, which included one bladder cancer and one lung cancer.
Miscellaneous.
Three of five HBV core antibody-positive patients received lamivudine or entecavir along with rituximab treatment, and HBV reactivation did not occur. There were no cases of progressive multifocal leukoencephalopathy.
Pregnancy
While receiving rituximab, women were counseled extensively to avoid pregnancy or plan to conceive after stopping rituximab. Of note, 22 women under 40 years of age were treated with rituximab, and urine levels of human chorionic gonadotropin were measured and negative before each dose. When patients became pregnant, rituximab was discontinued, and maintenance immunosuppression was minimized to prednisone and/or azathioprine. High-risk obstetrical care at our center was initiated, and visit frequency in our clinic was increased to every 2 months. Five women in this age group achieved six pregnancies, resulting in five healthy offspring. One pregnancy resulted in fetal demise at 15 weeks. Only one patient (39 years) in the cohort who desired pregnancy was unable to conceive.
Survival
Survival is illustrated in Figure 4, and it mirrors survival of an age-, sex-, and ethnicity-matched United States population (P=0.94). There were 10 deaths, which are listed in Table 4. There was also no significant survival difference among ANCA patients based on sex or ANCA serotype. The only significant mortality risk factor in this group was age older than 80 years (hazard ratio, 9.02; 95% confidence interval, 2.41 to 33.79; P=0.001).
Survival of ANCA vasculitis patients undergoing continuous B-cell depletion compared with survival of the general population. Kaplan–Meier survival curves comparing cumulative survival of the cohort (diamonds) with the survival of an age-, sex-, and ethnicity-matched general United States population (dashes) based on data retrieved from National Vital Statistics Reports. There was no statistically significant difference in survival over time (P=0.94).
Discussion
This study describes the long-term results of 172 ANCA vasculitis patients who underwent rituximab-induced continuous B-cell depletion for maintenance of remission. Remission was maintained in the majority of patients, with a low rate of major relapse. This remission maintenance strategy also seems to result in survival rates comparable with a matched reference population, which may reflect a benefit of the treatment strategy described. Interestingly, most women in this study who wanted to achieve pregnancy were successful, even after perigestational rituximab exposure, perhaps related to improved quality of life. Only one patient has requested to be transitioned from rituximab to another primary maintenance immunosuppressant.
There have been 10 deaths to date in this cohort. The only significant predictor of death within this cohort was increased risk of death in patients who achieve remission at an age greater than 80 years, which is not surprising. Other AEs occurred, with pulmonary infections comprising the majority of serious AEs, which is expected based on prior reports of longitudinal examinations of ANCA vasculitis patients (17). Another main AE associated with rituximab use (also described with other rheumatic diseases) (18) was LON with concomitant risk of infection. Fortunately, LON resolves with granulocyte colony-stimulating factor use and typically does not recur, thereby allowing for rituximab continuation. Hypogammaglobulinemia was a late and modest complication without apparent major sequelae, which led to discontinuation of rituximab in several patients. These results differ from a previous report showing a marked decline in Ig levels, especially with rituximab administration after cyclophosphamide use (19). Explanations for this discrepancy may relate to small cohort size, cyclophosphamide dosing, and cumulative exposure to other immunosuppressants in the other cohort. Unfortunately, there does not seem to be any predictor for patients who develop LON or hypogammaglobulinemia.
Despite what seem to be impressive results, this study has several limitations. Success of this regimen may be less related to rituximab use and more related to the clinical care provided. Our patients are closely monitored through frequent clinic visits, telephone conversations, and email correspondence. Furthermore, many patients were exposed to other immunosuppressants because of grumbling disease activity (BVAS-WG=1–2), and these other drugs likely contributed to improved disease control. Interestingly, despite exposure to other immunosuppressants, overall toxicity was low, and survival was excellent. Success of this treatment regimen may also be, in part, related to the induction regimens used. A more complete induction regimen that includes cyclophosphamide, which was used in many of our patients, may be helpful to allow for maintenance therapy using rituximab alone. The results could be different compared with an induction regimen consisting of rituximab without a cytotoxic agent or high-dose steroids (5–8).
A fixed rituximab dosing schedule seems safe and effective; however, patients with ANCA vasculitis likely cannot remain on rituximab indefinitely. Redosing of rituximab based on B-cell return, rather than on a fixed schedule, may also be effective (9). A trial comparing continuous-scheduled rituximab with dosing based on B-cell return is now underway to address these questions (MAINtenance of remission using RITuximab in Systemic ANca associated vasculitis Trial, MAINRITSAN 2; NCT01731561). There must also be additional investigation into determining predictive markers for individuals who need fixed versus intermittent dosing. It seems that higher relative percentages of peripheral regulatory B cells at the time of B-cell repopulation after rituximab may be associated with more prolonged remission (20,21). This finding could imply an increased risk of relapse at the time of rituximab cessation. Regardless, these findings need to be validated in other centers, and predictors of sustained remission are needed.
Great strides have been made with regards to induction of remission in ANCA vasculitis with the introduction of rituximab-induced B-cell depletion (5,6); however, it is clear that long-term remission cannot be maintained without additional immunosuppression (10). We and others have shown successful maintenance of remission using continuous rituximab for 2 years (10,12); however, this study is the first long-term study showing success beyond the 2-year mark. Ideally, goals for durable maintenance of remission in ANCA vasculitis should include therapeutic agents that provide excellent protection from autoimmune insults with little to no toxicity, improved quality of life, and survival that matches survival of the general population. This study is the largest of its kind to date, which brings us closer to these goals. Our findings provide highly granular support for rituximab as an effective agent for durable maintenance of remission.
Disclosures
J.L.N. has served as a rituximab-specific advisory board member for Genentech, is currently participating in the Genentech-sponsored Rituximab in ANCA-Associated Vasculitis Registry (RAVER) study, and is receiving clinical research funding from Alexion Pharmaceuticals.
Acknowledgments
The authors thank the patients as well as the clinic nurses dedicated to their care in the Vasculitis and Glomerulonephritis Clinic, including Donna Hagstrom, Kate Cosgrove, Eleanor Coughlin, Laura Chambers White, Chelsea Barrett, Stefanie Navarro, and Luke Cogswell.
Part of this work was presented at the 16th International Vasculitis and ANCA Workshop in Paris, France (April 14–17, 2013).
Footnotes
Present address: William F. Pendergraft III, UNC Kidney Center, Division of Nephrology and Hypertension, Department of Medicine, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina.
Published online ahead of print. Publication date available at www.cjasn.org.
This article contains supplemental material online at http://cjasn.asnjournals.org/lookup/suppl/doi:10.2215/CJN.07340713/-/DCSupplemental.
See related editorial, “Should Rituximab Be Used to Prevent Relapse in Patients with ANCA-Associated Vasculitis?,” on pages 641–644.
- Received July 10, 2013.
- Accepted November 27, 2013.
- Copyright © 2014 by the American Society of Nephrology