Rituximab as maintenance therapy following remission induction in relapsing or refractory systemic lupus erythematosus

Abstract

Objective

To investigate the efficacy and safety of rituximab (RTX) maintenance therapy compared with traditional immunosuppressive agent (ISA) maintenance therapy in patients with relapsing or refractory SLE.

Methods

It is a prospective observational non-randomized cohort study. The study enrolled SLE patients in four centres who had received at least one course of RTX induction treatment. Patients with a clinical response to RTX were divided into two groups based on their maintenance therapy in the first 12 months: the RTX group and the ISA group. The relapse-free survival times were compared between the two groups. Univariate and multivariate analyses were conducted to identify predictive factors for disease relapse.

Results

Among the 82 patients included in the cohort, 67 (81.7%) patients had a clinical response at 6 months. RTX maintenance therapy was applied in 34 (50.7%) patients and ISA maintenance therapy was applied in the remaining 33 (49.3%) patients. After a median follow-up of 24 months, a total of 13 (19.4%) patients had experienced disease relapse, comprising three in the RTX group and 10 in the ISA group. Patients in the RTX group had a higher relapse-free survival rate than patients in the ISA group. Multivariate analysis identified hydroxychloroquine use, RTX maintenance therapy and haematological system involvement as independent predictors for sustained remission.

Conclusion

This multicentre prospective cohort study demonstrated that long-term RTX maintenance therapy has high efficacy and acceptable safety in relapsing or refractory SLE patients who had a clinical response to RTX induction therapy.

Introduction

SLE is a chronic systemic autoimmune disease characterized by the emergence of large numbers of autoantibodies and multiple system involvement. Current therapies are designed to manage inflammation, stave off disease relapses and reduce clinical symptoms, with the goal of preventing permanent organ injury. The standard treatments for SLE include corticosteroids, antimalarials and immunosuppressive agents (ISAs). Among the ISAs, cyclophosphamide is used most widely as an induction treatment for severe lupus with important organ involvement. Other effective ISAs include methotrexate, MMF, AZA, CSA and tacrolimus (TAC). However, these interventions are not always effective and are sometimes associated with serious side effects. Therefore, multiple clinical trials have incorporated biologics into treatment regimens for SLE patients with poor responses or side effects to standard therapies.

Although the pathogenesis of SLE is not fully understood, growing evidence suggests that B lymphocytes play a central role. Use of B-cell targeted therapies in SLE treatment has generated great interest, based on strong evidence provided by genetic, immunological and clinical observations [1]. Rituximab (RTX), a chimeric monoclonal antibody that selectively targets CD20⁺ B cells, can dramatically improve abnormalities in B-cell homeostasis, and appears reasonable for induction of long-term remission in SLE [2]. However, the LUNAR and EXPLORER trials for lupus nephritis or extrarenal lupus, respectively, failed to reveal additive effects for RTX use on top of background immunosuppressants [3, 4]. These findings may have arisen through the selection of the study subjects and the design of the study protocols. Research on B-cell depletion therapy has remained very attractive. In studies without background immunosuppression beyond corticosteroid administration, numerous retrospective and prospective nonrandomized studies have indicated that RTX is effective for induction in moderate or severe SLE, especially in patients with refractory lupus nephritis, autoimmune haemocytopenia and neuropsychiatric lupus [5–10]. With several guideline recommendations [11, 12], RTX is frequently used as an induction regimen in relapsing and refractory SLE patients.

Other than initial remission induction, sustained remission is an important therapeutic goal in SLE management. This goal has not been well-met by traditional therapeutic regimens. Currently, there is no standard maintenance protocol to prevent relapse in RTX-treated patients. However, RTX has been proven to act as a maintenance agent in other autoimmune diseases. The MAINRITSAN study showed that RTX was effective for maintenance therapy in ANCA-associated vasculitis, with a lower recurrence rate than AZA [13]. However, evidence is still lacking on whether RTX can be used as a maintenance regimen in SLE. A recent multicentre retrospective study showed that RTX maintenance therapy is a potential treatment option for patients with refractory SLE [14]. However, no prospective studies on RTX as long-term maintenance therapy in SLE have been reported to date.

The present prospective study was designed to compare the efficacy and safety for remission maintenance between RTX and traditional ISAs in relapsing or refractory SLE patients who had achieved remission by RTX induction treatment. Predictors for disease relapse were also explored in the study cohort.

Materials and methods

This prospective observational cohort study was conducted at four centres: Department of Rheumatology in the Second Affiliated Hospital ZheJiang University School of Medicine, Department of Nephrology in the First Affiliated Hospital ZheJiang University School of Medicine, Department of Rheumatology in Lishui Central Hospital and Department of Rheumatology in Ningbo First Hospital. The study complies with the Declaration of Helsinki, and was approved by the local ethics committee in each centre, including the Ethics Committee of the Second Affiliated Hospital of Zhe Jiang University School of Medicine, the First Affiliated Hospital ZheJiang University School of Medicine, Lishui Central Hospital and Ningbo First Hospital. Written informed consent has been obtained from the subjects before participation in the study.

Patients and treatment protocol

The inclusion criteria for the cohort were: history of meeting the 1997 American College of Rheumatology criteria or 2012 Systemic Lupus International Collaborating Clinics criteria for SLE; moderate or severe active disease involving at least one important organ in the renal, haematological or neuropsychiatric systems, defined as ≥1 domain with British Isles Lupus Assessment Group (BILAG) B score; presence of relapsing (treatment with at least one ISA and relapse after achieving remission) or refractory (no remission after 3–6 months under traditional treatment) disease; and acceptance of RTX induction treatment. The exclusion criteria were: <1 domain with BILAG B score prior to first administration of RTX; non-completion of first course of RTX because of adverse effects such as allergy or infection; RTX initiation in other centres without sufficient clinical data; and completion of regular follow-up examinations for <6 months.

All included patients received at least one course of RTX. In the first course as the induction regimen, RTX was provided with different patterns: two doses of 375 mg/m² infused at 2-week intervals, one single dose of 375 mg/m², two doses of 200 mg infused at 2-week intervals, or four doses of 100 mg infused at 1-week intervals. The first pattern was defined as standard-dose RTX and the other patterns with a total dose below 2 × 375 mg/m² were defined as low-dose RTX. Each patient received 5 mg dexamethasone intravenous and 12.5 mg promethazine intramuscular before each RTX infusion to avoid infusion reaction. Effective B-cell depletion was defined as CD19⁺ B-cell count <1% of the total peripheral blood lymphocytes or absolute cell count <5 cells/μL.

Among the patients who showed a clinical response within 6 months, some received RTX maintenance therapy with an interval of 6–12 months and others received traditional ISA maintenance therapy. The treatment options were determined by the individual physicians in consultation with their patients. For the maintenance regimen, RTX was provided at 200–500 mg in a single dose. All patients were treated with corticosteroid at the same time, and the dose of corticosteroid was gradually tapered according to the disease activity. HCQ was used in patients without contraindications. In cases with disease recurrence or adverse drug reactions (ADRs), the treatment regimen could be adjusted at any time during the study.

Clinical assessment and data collection

The revised BILAG-2004 and SLEDAI-2000 were used to assess SLE disease activity [15, 16]. The BILAG index grades ranged from A to E, with A as severely active, B as moderately active, C as stable mild disease, D as once active and E as never active, and were converted into a numerical score for statistical analysis (A = 9, B = 3, C = 1, D = 0 and E = 0). Major clinical response (MCR) was defined as achieving BILAG C score or better in all organs at 6 months without experiencing a relapse within the first 6 months. Partial clinical response (PCR) was defined as achieving ≤1 organ with BILAG B score and at least one BILAG A or BILAG B score less than baseline at 6 months. No clinical response (NCR) was defined as failure to meet the definition of MCR or PCR at 6 months. Relapse was defined as emergence of at least one new BILAG A or BILAG B score.

All patients were prospectively followed up every 1–3 months during the active disease stage and every 3–6 months during remission. Demographic data (age, sex, disease duration), antinuclear antibody profiles and previous immunosuppressive therapies were recorded at the baseline visit before the first infusion of RTX. The following data were collected at each visit: clinical data (lupus manifestations), biochemical data (anti-dsDNA antibody positivity, complement levels, B-cell count) and treatment-related data (prednisone dose, ISA type). Disease activity was also assessed. The observation end point was death, disease relapse or study termination.

Statistical analysis

The percentage of patients who had a clinical response (MCR and PCR) was evaluated at 6 months. The percentage of patients who experienced a relapse was evaluated at the observation end point. Other evaluation parameters included changes in disease activity scores (BILAG and SLEDAI), reduction in corticosteroid dosage and severe adverse events (SAEs).

Statistical analyses were performed using SPSS software V22.0 (IBM, Armonk, NY, USA). In descriptive analyses, the baseline characteristics were compared between the two groups. The data for these descriptive analyses were presented as median [interquartile range (IQR)]. The Mann–Whitney U test was used for comparisons of quantitative variables. Comparisons of categorical variables were performed with the χ² test or Fisher exact test as appropriate. A multivariate logistic regression model was used to identify predictive factors for disease relapse. Kaplan–Meier plots were created to determine the relapse-free survival times in the two maintenance groups. Two-sided P-values of <0.05 were considered statistically significant.

Results

Baseline characteristics

A total of 82 patients were included in the cohort from November 2018 until October 2021. At baseline, the median (IQR) age was 32.0 (25.0–43.3) years and the median (IQR) disease duration was 5.5 (3.0–11.0) years. Of the 82 patients, 73 (89.0%) were female. With regard to system involvement, 49 (59.8%) patients had active lupus nephritis, 49 (59.8%) had haematological system involvement and 18 (22.0%) had neuropsychiatric lupus. All patients were under corticosteroid treatment, with 29 (35.4%) patients receiving pulses of intravenous methylprednisolone therapy. During the first course of induction treatment, 38 (46.3%) patients received standard-dose RTX and 44 (53.7%) received low-dose RTX.

Remission induction

Within 1–3 months after the first infusion of RTX, 68 of 82 (82.9%) patients achieved B-cell depletion. At 6 months, 67 (81.7%) patients had a clinical response (MCR 57.3%, PCR 24.4%), while 15 (18.3%) experienced treatment failure (NCR). The clinical response rates did not differ significantly between the patients treated with standard-dose or low-dose RTX at 3 months (P = 0.852) and 6 months (P = 0.438). The changes in corticosteroid dosage, BILAG score and SLEDAI score in the first 24 months are shown in Fig. 1. For the 67 patients who had a clinical response, the median (IQR) daily prednisone dosage reduced from 45 (25–75) mg to 7.5 (6–10) mg, the BILAG score decreased from 9 (6–12)–1 (0–1), and the SLEDAI score decreased from 9 (6–14)–2 (0–4) at 12 months. The daily prednisone dosage was reduced to ≤7.5 mg in 21 (31.3%) patients at 6 months and 37 (55.2%) patients at 12 months.

Maintenance therapy

The 67 patients who had a clinical response to RTX were divided into two groups according to their maintenance therapy in the first 12 months. Fig. 2 shows the numbers of courses of RTX maintenance therapy and the choice of ISA. RTX maintenance therapy was applied in 34 (50.7%) patients, including three patients who did not receive a second course of RTX within 12 months because of a persistent B-cell depletion state. The median number of RTX courses was 3 (range 1–6) and the median cumulative RTX dose was 800 (range 400–1000) mg per year. The remaining 33 (49.3%) patients received ISA maintenance therapy. The ISAs selected were MMF (n = 13), TAC (n = 7), AZA (n = 4), CSA (n = 7) and MMF combined with TAC (n = 2). The baseline characteristics of these patients are summarized in Table 1. No significant differences in the demographic characteristics, organ involvement, BILAG score and SLEDAI score were observed between the two groups (P > 0.05). The median (IQR) baseline corticosteroid dosage was 50 (35–100) mg in the RTX group and 30 (25–50) mg in the ISA group, with a significant difference (P = 0.044). However, no significant difference was found in daily prednisone dosage during the maintenance phase between the two groups.

The median (IQR) overall follow-up time was 24 (18–36) months. Twenty-four (70.6%) patients received sustained RTX maintenance therapy without disease relapse until 24 months. Overall, 13 (19.4%) patients experienced disease relapse during follow-up, comprising three in the RTX group and 10 in the ISA group (Table 2). The median (IQR) relapse time was 18 (12–23) months. Among these 13 patients, four patients (three in the ISA group, one in the RTX group) continued RTX treatment and obtained disease remission again. Four patients on RTX maintenance therapy transferred to ISA maintenance therapy (three between 12 months and 24 months, one between 24 months and 36 months), and all four achieved sustained remission. Patients who received standard-dose RTX had a lower recurrence rate than patients who received low-dose RTX (9.1% vs 29.4%, P = 0.035). The relapse-free survival times of patients in the different maintenance therapy groups are illustrated in Fig. 3. Patients who received RTX maintenance therapy had a higher relapse-free survival rate than patients who received ISA maintenance therapy either grouped by maintenance therapy within 12 months (P = 0.029) or 24 months (P = 0.015).

Predictors for disease relapse

Univariate and multivariate analyses were conducted to examine the associations between clinical characteristics and the risk of recurrence. The univariate analyses were conducted on 10 factors: age, sex, disease duration, RTX dose, maintenance therapy, HCQ use, MCR at 6 months and active disease in the renal, haematological or neuropsychiatric systems (Table 2). Variables identified with P <0.1 in the univariate analyses were applied to a multivariate logistic regression analysis. The multivariate model contained four parameters (Table 3). This analysis showed that HCQ use [OR 0.067 (95% CI 0.010, 0.463), P = 0.006], RTX maintenance therapy [OR 0.088 (95% CI 0.012, 0.636), P = 0.016] and haematological system involvement [OR 0.124 (95% CI 0.021, 0.744), P = 0.022] were identified as independent predictors for sustained remission.

Safety

During a median follow-up of 24 months in the total cohort, 48 (58.5%) patients had no reported ADRs, eight (9.8%) had minor ADRs and 26 (31.7%) had SAEs. In the 26 patients with SAEs, 43 SAEs occurred with a rate of 26 SAEs per 100 patient-years (Supplementary Table S1, available at Rheumatology online). The majority of the SAEs (33/43, 76.7%) were infections, including respiratory tract infection, urinary tract infection, central nervous system infection, gastroenteritis, biliary tract infection and herpes zoster. Non-infection-related SAEs included infusion reaction, fever, deep venous thrombosis and acute kidney injury. Twenty-four (55.8%) SAEs occurred during the induction phase within the first 6 months. RTX was discontinued due to SAEs in six patients (four infection, two allergy). During the maintenance therapy phase, 19 SAEs occurred in 16 patients, with no significant difference between the RTX and ISA groups (16.2 vs 12.1 SAEs per 100 patient-years). No malignant neoplasms was observed in the cohort during the follow-up period.

Discussion

Despite a lack of evidence in randomized controlled trials, RTX is widely used as an induction regimen in relapsing and refractory SLE patients. In the present cohort, all patients with relapsing or refractory SLE had active disease in at least one important organ in the renal, haematological or neuropsychiatric systems at baseline. Due to death or loss to follow-up, there were some patients who received one course of RTX therapy but were not included in the cohort. The clinical response rate to the first course of RTX was 81.7% at 6 months; because of the above reason, it may be a little higher than the actual level. Simultaneously, RTX administration significantly decreased the glucocorticoid dosage, and this effect was more obvious in the patients on RTX maintenance therapy. This finding is of major importance, because a higher cumulative glucocorticoid dose increases the risk of organ damage [17].

In addition to achieving remission in the active phase of SLE, sustained remission in the maintenance phase is another important therapeutic goal in SLE management [18]. Prevention of disease flares is a key aim in SLE treatment because flares are associated with both increased SLE-related damage and increased glucocorticoid use [19]. Despite multiple therapies examined to date, this goal has not been well met. The crude flare rate was >30 flares per 100 person-years in a prospective cohort of 1460 patients [20]. The present study prospectively explored the efficacy and safety of RTX as a maintenance therapy after remission induction in SLE patients.

Previous research showed that the duration of B-cell depletion caused by RTX induction therapy was ∼6–12 months in the majority of patients and that longer duration of B-cell depletion was associated with better disease outcomes [21]. Therefore, a second course of RTX therapy was generally provided during this period. In our study, the relapse rates were similar between the RTX and ISA groups in the first 12 months after induction. After 12 months, the relapse rates were consistently increased in the ISA group and significantly higher than those in the RTX group, which may be interpreted by the gradual recovery of B-cell activity. A large multicentre retrospective study found that RTX maintenance therapy can assist with glucocorticoid and immunosuppressive drug sparing while stabilizing the overall disease activity. However, disease flares were still common after RTX discontinuation [14]. In our multivariate analysis, RTX maintenance therapy was identified as one of the independent predictors for sustained remission.

Other independent predictors for sustained remission were HCQ use and haematological system involvement. HCQ is a cornerstone of SLE treatment. There is evidence for multiple beneficial effects of HCQ in SLE [22]. However, application of HCQ does not suit all SLE patients. HCQ is sometimes reduced or discontinued over a long course of SLE to limit the cumulative exposure and avoid HCQ-related toxicity. Recently, a large prospective cohort study demonstrated that the SLE flare risk was increased after HCQ tapering or discontinuation [20]. Similarly, our findings suggested that continuous maintenance with HCQ was helpful, even in patients treated with RTX. The haematological manifestations in SLE patients mainly included thrombocytopenia and autoimmune haemolytic anaemia. Based on the results in some retrospective studies [8, 23, 24], RTX has been recommended as a second-line induction regimen in refractory haematological diseases [12]. Our data showed that patients with haematological system involvement had a higher possibility of achieving sustained remission with RTX treatment than patients with renal and neurological system involvement.

The most common SAEs in the present study were infections, including bacterial, viral and fungal infections, and these were the main causes of RTX withdrawal. In a retrospective study on 115 SLE patients who received a mean of 1.95 RTX cycles, severe infections were found in 7.0% of patients [25]. In another retrospective study, 70 patients with different autoimmune diseases received long-term RTX therapy for a median of 54 months, and severe infections were observed in 25.7% of patients [26]. In the present study, the safety profile of RTX maintenance therapy was acceptable with no treatment-related deaths and no significant differences in SAE occurrence compared with ISA maintenance therapy. More than half of the SAEs occurred during the induction phase, which may be partly due to the high disease activity and simultaneous use of high-dose glucocorticoid. Allergy was reported to be a common ADR for RTX, especially during the first administration. In the present cohort, infusion reaction only occurred in two patients. Transient fever and rash were the most common manifestations of allergy, and were usually mild to moderate. These findings may be due to pre-treatment with anti-allergy drugs (dexamethasone and promethazine) before RTX administration in each patient.

RTX was administered with an interval of 6–12 months in the maintenance phase. Most patients accepted long-term maintenance RTX therapy because of the convenient usage. Overall, our results are consistent with the findings in previous observational studies that long-term RTX therapy was well tolerated.

The key strength of the present study lies in the multicentre design and the number of patients recruited, with the study having one of the largest cohorts of patients with relapsing or refractory SLE treated with RTX. Furthermore, the study examined the first prospective cohort of SLE patients on RTX maintenance therapy in real-world practice. The major limitation of the study was its non-randomized controlled design. Potential selection bias may have existed between the ISA and RTX groups. In addition, the dosage and interval of RTX were determined by individual physicians, rather than being fixed. Finally, the follow-up time differed among the patients in the cohort. Therefore, the recurrence end point may not have been observed in some patients due to a short follow-up period.

Conclusions

This multicentre prospective cohort study demonstrated that long-term RTX maintenance therapy has high efficacy and acceptable safety in relapsing or refractory SLE patients who had a clinical response to RTX induction therapy. However, the details of RTX usage requires further investigation.

Acknowledgements

The authors thank Alison Sherwin, PhD, from Liwen Bianji (Edanz) (www.liwenbianji.cn/) for editing the English text of a draft of this manuscript. The authors would like to thank all physicians and patients who participated in this study.

Funding: This work is supported by Zhejiang Provincial Natural Science Foundation (Grant number: LY22H100004).

Disclosure statement: The authors have declared no conflicts of interest.

Data availability statement

The data underlying this article are available in the article and in its online supplementary material.

Supplementary data

Supplementary data are available at Rheumatology online.

References