The prognosis for children and adults with systemic lupus erythematosus (SLE) has considerably improved over the past 30 years because of advances in diagnosis and treatment.1 Corticosteroids remain the first line of treatment for SLE patients.2 Cyclophosphamide is a cytotoxic agent with significant side effects that modifies the course of the disease.2 Azathioprine and mycophenolate mofetil are commonly used to control the disease, and they allow corticosteroids to be weaned.2 However, in some SLE patients, the disease course is aggressive and is not controlled by these treatments.3 An important advance is B cell targeted therapy with rituximab, which is a chimeric monoclonal antibody specific for human CD20 that depletes circulating B cells and has been shown to be beneficial in childhood and adulthood SLE case series.3^–6 The incidence of adverse effects after treatment with rituximab varies in paediatric SLE studies.5 6 The aim of this study is to report the safety and efficacy of rituximab in paediatric SLE patients treated in one centre.

PATIENTS AND METHODS

Medical records of rituximab-treated SLE children seen at the paediatric nephrology and rheumatology units at the Great Ormond Street Hospital for Children were reviewed. All patients fulfilled four or more of the 1982 revised American College of Rheumatology criteria for the diagnosis of SLE. Patient data including current and previous immunosuppressive therapy were collected. Seven patients from the present study have been included in a previous report.6 Indications for rituximab were failure to respond sufficiently to standard immunosuppressive therapy and acute life- or organ-threatening disease (table 1).

Treatment protocol

All patients received two rituximab infusions, usually 14 days apart, at a dose of 750 mg/m²/infusion. The second rituximab infusion was delayed in two patients owing to an upper respiratory tract infection in one and to leucopenia in another. Two patients required re-treatment with a further rituximab infusion at a dose of 750 mg/m² (administered without cyclophosphamide). All patients received premedication with chlorphenamine and paracetamol 1 hour prior to the rituximab infusion and an intravenous dose of 100 mg methylprednisolone just prior to the rituximab infusion. Most patients received cyclophosphamide infusions the day after the rituximab infusions. The cyclophosphamide dose of 250–750 (median 375) mg/m² was adjusted according to the degree of renal dysfunction. A dose of oral prednisolone was given for 3 days after the rituximab infusion (30 mg on the first day, 20 mg on the second and then 10 mg on the third) before continuing with the patients’ previous maintenance prednisolone dose.

Evaluation

The disease activity of all patients was assessed using the British Isles Lupus Assessment Group (BILAG) index and by laboratory measures (haemoglobin, white blood cells, neutrophils, lymphocytes, platelets, erythrocyte sedimentation rate, complement C3 and C4, antinuclear antibody (ANA), anti-double-stranded DNA antibodies, serum immunoglobulins (IgG, IgM), plasma creatinine, serum albumin and urine albumin/creatinine ratio). Normal ranges of values, depending on the age and sex of the patient, used at our laboratory are the following: haemoglobin 11–16 g/dl, white blood cells 4.0–12.0×10⁹/l, lymphocytes 1.2–5.2×10⁹/l, platelets 150–450×10⁹/l, erythrocyte sedimentation rate 0–10 mm/h, C3 0.75–1.65 g/l, C4 0.14–0.54 g/l, antinuclear antibody negative, anti-ds DNA antibodies negative, IgG 5.4–16.1 g/l, IgM 0.5–1.9 g/l, plasma creatinine 30–100 μmol/l, serum albumin 35–50 g/l and urine albumin/creatinine ratio <25 mg/mmol. B cell counts (CD19) were used to monitor the effects of the rituximab treatment. BILAG is a validated tool for SLE disease activity assessment. It is divided into eight organ systems (general, mucocutaneous, neurological, musculoskeletal, cardiorespiratory, vasculitis, renal and haematological parameters) and has been validated in childhood disease.7 8 BILAG scores were assessed prior to rituximab therapy and at follow-up after therapy after 1–2 months, 6–8 months and 11–14 months and at the last follow-up visit. Estimated glomerular filtration rate (eGFR) was calculated using the Schwartz formula (eGFR  =  height×40/serum creatinine).9 The biochemical, immunological and haematological parameters at 1, 3, 6, 12, 18, 24, 30 and 36±1 months were compared with those prior to the treatment. Analysis was performed separately both for the whole group of children and for the children who had an initial abnormal value.

Ethical approval for this study was obtained from the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust Research Ethics Committee. The study was conducted in accordance with the Declaration of Helsinki with verbal informed consent obtained in all cases.

All analyses were performed using one-tailed Wilcoxon matched-pairs signed-rank test in Excel (USA). p Values of 0.05 or less were considered statistically significant.

RESULTS

Patient data

Nineteen children (89% (17/19) female) with a median age of 14 (range 6–16) years received a total of 21 rituximab treatment episodes and were followed up for a median of 20 (range 6–38) months with 100% patient survival. Re-treated patients were analysed for both initial and subsequent treatment episodes.

Guidelines for the prescribing of rituximab for patients with lupus resistant to standard treatments were created at our hospital and approved by the Drug and Therapeutics Committee. The selection criteria for patients were severe active lupus disease, previously treated with standard lupus treatment including cyclophosphamide infusions for 6 months (if no obvious contra-indication to that treatment) and no known severe reaction to humanised chimeric antibodies. Chronic active infection, recent severe infection, pregnancy or planned pregnancy led to the exclusion of the patient for the therapy with rituximab.

Almost all of our patients received cyclophosphamide before the treatment with rituximab, as was required according to the guidelines. As one can see in table 1, the patients (number 17, 18 and 19) who were treated recently did not receive cyclophosphamide beforehand. We, with time, widened our indication for rituximab as we got more experienced with the usage of the drug and had the explicit consent of patients and patients’ parents. The cumulative cyclophosphamide dose pre-treatment was 3750 (range 250–800) mg/m² in the 15 cases where this could be calculated. Three patients had received cyclophosphamide before starting medical care in our hospital.

The children were treated for a variety of disease manifestations: generalised symptoms (48%), renal impairment (48%), central nervous system (19%) and skin (29%) involvement, joint involvement (10%), haematological changes (5%) and growth failure (5%). Each indication was assessed individually according to the aforementioned guidelines and according to our clinical impression. Full recovery, improvement and some improvement of the predominant symptoms were achieved according to the clinical impression after treatment (table 1). Full recovery (52%) is defined as remission of the disease, with a post-treatment asymptomatic patient, partial improvement (43%) is defined as partial remission with some remaining symptoms after treatment, and some improvement (5%) is defined as minor symptomatic benefit after treatment.

All patients were treated with individual maintenance therapy after the treatment with rituximab, which consisted of corticosteroids, azathioprine, mycophenolate mofetil and hydroxychloroquine. The daily dosage of prednisolone could be decreased after the treatment with rituximab from the initial median dose of 0.35 (range 0.06–1.95) mg/kg to 0.14 (range 0.05–0.39) mg/kg within 6 months (p<0.001) after the treatment and was kept at the same level of median 0.13 (range 0.05–0.25) mg/kg at the follow-up at 12 months (p = 0.0014).

The British Isles Lupus Assessment Group

Reduction of SLE disease activity after the treatment with rituximab was represented by a rapid decrease in BILAG scores from a median of 14 (range 3–43) to 6 (range 1–19) within 1 month (p<0.001). A significant decrease in the BILAG score was also seen at the follow-up at 6 and 12 months. At the last follow-up of 10–37 (median 25) months, a stable reduction of BILAG scores (p = 0.004) compared with those prior to treatment was found (fig 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 BILAG and renal involvement. Comparison of the data following treatment with the data prior to the treatment: *p<0.05; **p<0.005. The number of cases evaluated at each timepoint are included in the figure. (A) BILAG (all cases) development in primarily 21 cases. (B) eGFR (<90 ml/min/1.73 m² prior to treatment) development in primarily 10 cases. (C) Serum albumin (<35 g/l prior to treatment) development in primarily 12 cases. (D) UAUC (>25 mg/mmol prior to treatment) development in primarily 13 cases.

Renal involvement

Renal function was impaired in nine out of 19 children (eGFR lower than 90 ml/min/1.73 m²) before rituximab therapy with eGFR of 14–85 ml/min/1.73 m² (median 54) in those nine patients. Three patients had progressed to renal failure, and two of them were on renal replacement therapy (RRT) prior to receiving rituximab therapy. Significant improvement in renal function was achieved within 3 months (p = 0.046) of rituximab treatment and was maintained for 18 months (p = 0.04) with the discontinuation of RRT in both patients (fig 1). Hypoalbuminaemia of 20–34 (median 29) g/l and albuminuria with elevated urine albumin/creatinine ratio (UAUC) 42–5871 (median 254) mg/mmol improved within the first month and achieved statistical significance 3 months after the treatment both for serum albumin and UAUC (p = 0.0016 and p = 0.002, respectively). The improvement was maintained at 18 months for UAUC (p = 0.014) and at 24 months for serum albumin (p = 0.034) (fig 1). Increasing serum albumin levels and declining UAUC levels (p<0.001 and p = 0.003 after 3 months, respectively) were observed after all treatment episodes, even in cases with normal values prior to the treatment.

To classify and specify renal involvement as one of the disease symptoms, 15 of the 19 patients underwent a renal biopsy. Four patients had the WHO class II of lupus nephritis (LN), two had the WHO class III LN and nine of them had the WHO class IV LN. We did not perform routine biopsies for disease-monitoring purposes.

Immunology

Abnormal immunological parameters responded quickly to rituximab. Within 1 month, C3 (median 0.46, range 0.15–0.53 g/l, p = 0.02) and C4 values (median 0.08, range 0.03–0.13 g/l, p = 0.05) increased, and anti-dsDNA antibodies values (median 106, range 20–541 IU/ml, p = 0.008) decreased. Improvements in anti-dsDNA, C3 and C4 persisted for 12, 24 and 24 months after treatment respectively (fig 2). ANA levels fell (p = 0.048) in the 12 cases where ANA levels had been measured prior to treatment (fig 2). Two patients were ANA-negative prior to rituximab treatment, and six patients did not have their ANA status recorded as they were positive for dsDNA. Seven cases had baseline IgG and IgM values measured, as well as regular follow-up values. Two patients had low IgM levels prior to treatment, and in three patients their IgM had dropped to below the normal range (with three not returning to normal by the time of the last follow-up at 7, 19 and 28 months). A fall in IgG was seen in three of the cases, including two with low baseline IgG. Levels of IgG returned to normal values by 2 to 18 months after the treatment.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Immunology. Comparison of the data after treatment with the data prior to the treatment: *p<0.05; **p<0.005. The numbers of cases evaluated at each time point are included in the figure. (A) ANA (positive prior to treatment) development in primarily 12 cases. (B) C3 (<0.75 g/l prior to treatment) development in primarily 10 cases. (C) C4 (<0.14 g/l prior to treatment) development in primarily 15 cases. (D) Anti-dsDNA (>9.9 IU/ml prior to treatment) development in primarily 16 cases.

B cell values were evaluated for 17 treatment episodes, and 16 children experienced depletion of their CD19⁺ B cells. Nine recovered their B cell levels during the observation period after 2–9 (median 6) months, whereas the other eight still had depleted levels at 6–24 (median 9) months (fig 3).

• Download figure
• Open in new tab
• Download powerpoint

Figure 3 B cell depletion. The numbers of the examined patients are included in the figure.

Haematological involvement

Rapid improvement of haematological parameters was observed for erythrocyte sedimentation rates (ESR) from a median of 75 (range 7–150) mm/h to 24 (range 4–89) mm/h in all 21 treatment episodes. Haemoglobin and platelet counts increased significantly in all 21 episodes, especially in those who had anaemia and thrombocytopenia prior to rituximab. Statistical significance was achieved within 1 (p = 0.038), 1 (p = 0.013) and 12 months (p = 0.027) after the treatment for haemoglobin, platelets and ESR, respectively, with low ESR persisting after 36 months. Haemoglobin and platelet counts continued to be significantly increased at 24 (p = 0.04) and 18 months (p = 0.034), respectively. Baseline lymphopenia (median 0.49×10⁹/l, range 0.10–1.11×10⁹/l) improved significantly six months (p = 0.046) after the treatment, with an improvement sustained at 24 months (p = 0.022) (fig 4). A trend towards increasing white blood cell values was observed after rituximab treatment. No significant changes were found in neutrophil levels during the longer term follow-up of 36 months.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4 Haematology. Comparison of the data after treatment with data prior to the treatment: *p<0.05; **p<0.005. The numbers of the cases evaluated at each time point are included in the figure. (A) ESR (all cases) development in primarily 21 cases. (B) Haemoglobin (<11.5 g/dl prior to treatment) development in primarily 13 cases. (C) Lymphocytes (<1.2×10⁹/l prior to treatment) development in primarily 15 cases. (D) Platelets (<150×10⁹/l prior to treatment) development in primarily nine cases.

Adverse effects

No acute adverse effects were observed in the 21 patients who received rituximab treatment. One patient suffered nausea and vomiting after one of the cyclophosphamide infusions. Five patients experienced herpes zoster episodes 6 (0.5–15) months after the treatment, which were uncomplicated and controlled using acyclovir. One of the patients had had his first herpes zoster viral infection before rituximab treatment.

DISCUSSION

Improvements in BILAG scores, renal, immunological and haematological parameters were observed in our study population of 19 children with refractory SLE who were treated with rituximab. Recovery of abnormal values was achieved within 1 month for several parameters, including BILAG, complement C3 and C4, anti-dsDNA antibodies, haemoglobin and platelet counts. Serum albumin, eGFR, UAUC, ANA and ESR responded after a longer time period of 3 to 12 months. Different parameters maintained their improvement for different time intervals within the observation period. Loss of significance can be explained by the reduction in the number of cases during the observation period: only three cases had a follow-up of 30 months.

Nine studies of differing doses of rituximab in adults (seven studies) and children (two studies) with refractory SLE have recently been undertaken.4^–6 10^–17 All studies showed improvement of SLE symptoms and laboratory parameters. B cell depletion was observed in 65 to 100% of examined patients.4 13 B cells were depleted in 94% of examined children in our study. Different disease activity scores (Systemic Lupus Activity Measure score, BILAG, MEX-SLE Disease Activity Index score and SLE Disease Activity Index score) were used to define the primary outcome of clinical efficacy of rituximab, but all improved significantly in the reported studies.6 10 13 14 16 18 19 Improvement in anti-dsDNA antibodies and C3 and C4 levels differed from study to study; significant changes were observed in two studies.14 15 17 Significant improvements in anti-dsDNA antibodies, and C3 and C4 levels were seen in this cohort of patients.

Early response of refractory SLE symptoms and laboratory markers was observed in the first month after rituximab therapy in 19 patients with 21 treatment episodes in our study. Two other studies have shown a rapid response, especially in patients with neurological involvement, within days and up to one month.10 19 Some longer intervals (2–4 months) between the treatment with rituximab and response were seen in other studies.4 5 13 14 16^–18

The commonest side effect seen in this series was herpes zoster infection in 26% of the patients. All these patients had extensive immunosuppressive therapy (including corticosteroids, azathioprine, hydroxychloroquine, mycophenolate mofetil and/or cyclophosphamide) both before and after the treatment with rituximab. Increased incidence of herpes zoster was also reported in three other papers about rituximab in SLE patients.5 10 18 In comparison, Willems et al reported “severe side effects” (septicaemia, transient severe haematological toxicity with thrombocytopenia and neutropenia, transient lymphopenia) in 45% of their patients.5 No severe haematological toxicity was found in this group of patients, although haematological parameters were not routinely tested in the first two weeks post-rituximab infusion. There were no severe bacterial infections.5 10 18 Development of human anti-chimeric antibodies has been reported after treatment with rituximab but was not routinely looked for in this study.3 18

A safety concern regarding rituximab was published by the Food and Drug Administration in the United States in December 2006 regarding the development of progressive multifocal leukoencephalopathy (PML), a fatal polyoma viral infection of the central nervous system that had been observed in two patients treated for SLE with rituximab.22 Several cases of PML after rituximab (used as chemotherapy or as a part of stem-cell transplantation) have been reported in patients with haematologic malignancies as early as 2002.23 24 It is, however, important to notice that several SLE patients have suffered from PML without being treated with rituximab.22

The precise mechanism of B cell involvement still needs to be fully evaluated, but the loss of B cell tolerance is believed to play a key part in the development and continuation of childhood-onset SLE.3 4 CD20 is a membrane protein expressed during the maturation of B cells and is not expressed on plasma cells or pro B-cells.2 Rituximab is believed to induce the elimination of B cells through three different pathways: stimulation of apoptosis, complement-dependent cytotoxicity and antibody-dependent cellular cytotoxicity.25

High doses of rituximab for treatment of refractory SLE in paediatric patients were found to be effective and safe in this study. Further randomised controlled trials (such as comparing rituximab and cyclophosphamide) are required to establish the role of rituximab in the treatment of childhood SLE.