A systematic literature search was performed from 1962 to ...



Updated Consensus Statement on the Use of Rituximab in Patients with Rheumatoid Arthritis

Supplementary Material

Page

Table of Contents

Systematic Literature Review 3

Search terms and limits 3

Outcome measures and data extraction 4

Data analysis 4

Mechanism of Action of Rituximab in RA 5

Recommendation 6

Indication – serological status 6

Screening before initiating rituximab 6

Hepatitis B 6

Vaccination 7

Treatment Dosage & Co-medication 7

Treatment dosage 7

Evaluation & management of response & non-response 8

Consideration for repeated treatment 8

Post TNF-inhibitor failure & biological DMARD therapy post rituximab 9

Post TNF-inhibitor failure 9

Role of rituximab in other autoimmune diseases 9

Contraindication and serious adverse events 9

Serious Infections 9

Progressive Multifocal Leukoencephalopathy (PML) 9

Immunoglobulin Levels & Infection Risk 10

Additional aspects to be considered and research agenda 10

Figures 13

References 24

Systematic Literature Review

The scope of the literature search was outlined by the expert group at the first meeting A systematic literature search was performed from 1962 to February 2009 using MEDLINE, EMBASE and COCHRANE databases with all studies evaluating rituximab (RTX) for rheumatoid arthritis (RA) included (by MHB). Abstracts were also obtained from the 2008-2009 American College of Rheumatology (ACR) and 2008-2010 EULAR conferences. Where full papers of these abstracts were published online during this period, the latter were obtained and used for data extraction.

Search terms and limits

1. arthritis, rheumatoid/ or caplan's syndrome/ or felty's syndrome/ or rheumatoid nodule/

2. rheumatoid arthritis.mp.

3. (early adj2 arthritis).mp.

4. inflammatory arthritis.mp

5. Biological Therapy/

6. biologic$.mp.

7. rituximab.mp.

8. mabthera.mp.

9. rituxan

10. 7 or 8 or 9

11. 4 or 1 or 3 or 2

12. 10 and 11

13. Limit 12 to english language

14. Limit 13 to yr-’1962-2010’

The above search was run through each database, Medline (255 references), Embase (1065 references) and Cochrane (35 references). After de-duplication, 1225 references were available for examination.

From these, main reasons for exclusion were (i) articles not related to the search topic; (ii) review articles and case reports; (iii) non-randomised, efficacy studies +/- less than 6 months +/- less than 50 patients; (iv) duplicate articles obtained from the databases searched. Existing meta-analyses and SLR were obtained.

The criteria for study selection for efficacy analysis were: (1) double blind randomised controlled trials (RCTs); (2) patients with RA; (3) studies evaluating rituximab; (4) trials of ≥ 6 months duration; (5) studies with ≥50 patients; (6) publications in English.

In addition hand-searched references and abstracts were also included. Despite different diseases and often associated co-morbidities, data on the toxicity of rituximab from the oncology literature were also considered where appropriate to provide insights in areas where there remains a paucity of data in RA cohorts.

Outcome measures and data extraction

Where relevant, demographic characteristics, disease duration, treatment allocation (dose and duration), concomitant treatments, follow-up duration and outcomes were extracted for each trial. Outcome measures for signs and symptoms - the ACR response criteria (20/50/70), and EULAR-response criteria were recorded. To evaluate structural damage, baseline and change in Sharp scores (including modifications) and Larsen scores, and numbers of patients achieving radiographic non-progression (defined in individual studies) were collected.

Safety aspects were evaluated using data from large cohorts with long-term follow up data including registries and observational studies and pooled data from RCTs.

Queries were raised and resolved by discussion at the expert meeting. For efficacy analysis, studies were in particular identified based on prior treatment (where relevant): Disease modifying anti-rheumatic drug (DMARD) naïve, methotrexate (MTX) naïve, MTX inadequate response (MTX IR) and tumour necrosis factor inhibitor inadequate response (TNFi IR).

Data analysis

Efficacy data were analysed using Review Manager (RevMan Version 5.0. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2008) software and expressed as relative risks (RR) or standardised mean differences (SMD). Toxicity data were not subject to formal statistical analysis. Levels of evidence were assigned according to the Oxford Centre for Evidence-based Medicine levels of evidence (index.aspx?o=1025).

Mechanism of Action of Rituximab in RA

B-cells have a number of important functions including precursor of plasma cells and (auto-)antibody production, antigen presentation and cytokine production 1. The clinical efficacy of rituximab relies on B-cell depletion, induced by antibody dependent cell-mediated cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC) and apoptosis, although the actual mechanism by which this leads to improvement is unclear. B-cell depletion affects other cell types as well, such as monocytes/macrophages and T-cells 2, 3; the extent to which this contributes to the efficacy of rituximab requires further investigation. Recently, it was shown that a sub-group of T-cells express functionally active CD20 and are depleted with rituximab therapy 4; the biological role in vivo and value of this cell population, however, remains uncertain.

To date, B-cell depletion of synovial tissue has not been proven to be of clinical utility. Rituximab has been shown to reduce CD68-macrophages in synovial tissue biopsies, an effect which was associated with clinical response, consistent with results from other anti-rheumatic therapies 5-7. In addition, rituximab does not completely abolish the generation of plasmablasts likely generated in mucosal tissues 8.

Currently, there are no data on the effects on immunglobulin (Ig) levels with long-term treatment. Compared to global Ig levels, rituximab leads to a more pronounced decrease of circulating autoantibodies, including rheumatoid factor (RF) and anti-citrullinated peptide antibodies (ACPA), particularly after the first cycle, and is more effective in autoantibody positive patients. However, the decline in autoantibodies does not correlate with response and the reason why autoantibody positive patients respond better is not fully clarified. Since B-cells, autoantibody production and susbsequent immune complex formation may contribute to the pathogenetic pathway of autoantibody positive disease, B-cell depletion could be thought to interfere with this pathway, thereby representing a rational target in this group of patients.

B-cells are also effective antigen presenting cells, particularly those that express RF 9. Murine data suggest that antigen presentation by B-cells independent of antibody production is important in induction and maintenance of autoimmunity 10, 11. Additionally, B-cells are important for the secretion of a variety of cytokines, TNF-alpha, interleukin (IL)-6 and IL-1, which play a major role in RA 12. In other autoimmune disease models, such as experimental autoimmune encephalomyelitis (EAE), IL-10 producing B-cells have been reported to play a key role in controlling autoimmunity 13 with more recent data in EAE suggesting that these cells are important in the early phase but not in established disease 14; a role for certain B-cell subsets in arthritis has not been reported and the clinical value of IL-10 producing B-cells with regulatory functions currently remains unclear 15, 16.

Recommendation

Indication – serological status

Association of serological status and response rates to RTX have been further addressed upon combining MIRROR and SERENE populations (approximately 550 seropositive and 116 seronegative patients) 17, 18; the seropositive patients showed a greater fall in disease activity score (DAS28), plus a higher odds ratio (OR) of achieving American College of Rheumatology (ACR) 20, 50 and 70 responses and DAS28 low disease activity (LDA) and remission rates. Apart from DAS28 remission, all differences were statistically significant at 24 and 48 weeks (DAS28 LDA only at 48 weeks). The OR for a seropositive patient achieving a particular response was between 2.2 and 3.4 times the likelihood for a seronegative patient 19. Another study first identified the best four biomarkers from the REFLEX study (IgA RF, IgG anti-CCP3 antibodies, C-reactive protein and soluble CD25)20; these together with IgM and IgG isotypes of RF were applied to the SERENE cohort 20. Seropositivity for any RF isotype or IgG ACPA was associated with a higher ACR 50 response rate compared to placebo. The placebo-controlled OR for the biomarker positive group (positive IgA RF and raised inflammatory markers) achieving ACR 50 response at week 24 was 10.4 (compared to 2.7 in the biomarker negative group). These data are in line with those of the DANCER study where ~70% of the patients were MTX-inadequate responders 21. An IMAGE sub-study that evaluated poor prognosis baseline characteristics also suggested that the seropositive sub-group was likely to benefit more 22.

In addition to such sub-analyses from RCTs, registry data also supports the association between seropositivity and favourable response to rituximab. The European collaboration (CERRERA database) recently reported on 794 patients 23; a high DAS28 at baseline and positive RF strongly predicted a good/moderate EULAR response. When correcting for baseline DAS28, RF remained predictive of response (OR 2.0, 95% CI 1.1-3.7). The German RABBIT registry concluded similar results [OR 2.0 (95%CI: 1.2 - 3.4)] 24. Recently, it was shown that a baseline serum IgG concentration above normal synergised with the presence of RF and/or ACPA for predicting response to rituximab 25.

Screening before initiating rituximab

Hepatitis B

Currently, aside from 1 case report of HBV reactivation in a patient with RA treated with rituximab 26 (with another in ANCA-associated vacsulitis) 27 there remains a general lack of experience in this area. The oncology and hepatology literature however offers important insights; although these include several reports of patients with HBV undergoing rituximab treatment, cases of reactivated HBV have been widely described with rituximab. Successful treatment with rituximab in patients with HBV with the use of anti-viral (usually lamivudine) prophylaxis has been reported 28 (category IV) with 2 RCTs 29, 30 and a recent meta-analysis also confirming the benefits of anti-viral prophylactic therapy 31 (category Ia).

A recent review paper identified 12 individual case reports of HBV reactivation in association with rituximab therapy with a mortality rate of 83% 32; 5 cases occurred in patients that were hepatitis B surface antigen (HBsAg)-negative; all these patients came from the lymphoma population in which patients also received concomitant chemotherapy. Additional cases in oncology patients have also been reported 33-38.

Different categories of HBV infection status exist, determined by testing for HBV infection serological markers. These include acute, chronic, occult or resolved HBV infection. Chronic HBV infection may be associated with either active carrier status [HBsAg+, anti-HB core antibody (HBc) +, HBV DNA+ with often liver damage] or indeed inactive carrier status (HBsAg+, anti-HBc+, HBV DNA- with no liver damage). Occult or resolved infection is associated with HBsAg- and anti-HBc+ with anti-HBs+/- status. The management of patients should clearly be undertaken with an expert hepatologist. The nature of the antiviral agent including the specific treatment used and duration of therapy would be influenced by several factors (not least whether the HBV disease is active or not), with each case being individually considered.

Two oncology patients with HCV were successfully treated with rituximab without prophylaxis 39 (category IV), whilst others have suggested rituximab may be implicated in cases of hepatic injury in patients with HCV 40 (category IV). Indeed, rituximab is being increasingly used for HCV-associated cryoglobulinaemic vasculitis 41-44 and in several studies no evidence of increased HCV replication- reactivation or acute liver injury has been seen. Nevertheless, more data are needed generally regarding HCV and rituximab treatment.

Vaccination

The study by van Assen et al 45 showed impaired responses to influenza vaccination which had also been demonstrated in two previous smaller studies 46, 47. Moreover, in the study by Bingham et al 48 primary responses against pneumococcus and keyhole limpet Hemocyanin (KLH) were impaired, whereas secondary responses to tetanus toxoid were maintained or preserved. Importantly, impairment of response was more marked and prevalent in the ‘early rituximab’ group that received vaccination 4-8 weeks after rituximab administration compared to the group that were vaccinated 6-8 months after receiving rituximab (late group) (category Ib).

Treatment Dosage & Co-medication

Treatment dosage

The main consensus document summarises the clinical efficacy and radiographic benefit reported in the REFLEX study of TNF inhibitor failure patients 49, 50. Also mentioned is the SERENE study of MTX-inadequate responder patients 18, in which 2x500mg and 2x1000mg rituximab dose regimens were compared to placebo. Cycle 2 was administered if a DAS28 ≥ 2.6 from 6 months was observed. 89.5%/90.5% patients in all the randomised groups (rituximab 500mg and 1000mg) were re-treated. In this study there was no statistically significant difference between the two active cohorts for any of the outcomes which were numerically similar. However, the study was powered to compare both doses to placebo, and both demonstrated superior outcomes to placebo (table 4, main document).

The MIRROR study (n=346) evaluated MTX-inadequate responder patients randomised to one of three rituximab dosage regimens, with 2 courses of rituximab given 24 weeks apart 17; the randomised doses were either two courses of 2 x 500mg, one course of 2 x 500mg followed by one course of 2 x 1000mg (dose escalation) and two courses of rituximab 2 x 1000mg. Equivalent efficacy of outcomes was observed.

Figure 1 (a-e) illustrates Forrest plots with the risk ratios of a pooled analysis of the 6 month ACR 20/50/70 and EULAR response rates and proportion of patients achieving low disease activity at 6 months for the 2 x 500mg versus 2 x 1000mg doses in these 2 studies. Placebo patients in the SERENE study that were re-treated after 6 months with rituximab 2x 500mg have not been included in this pooled analysis. Figure 2(a-d) similarly shows risk ratios of 12 month response rates comparing 2 courses of 2x500mg versus 2x500mg followed by 2x1000mg dosage regimens (dose escalation).

The IMAGE study evaluated 2x500mg and 2x1000mg rituximab doses to placebo in MTX-naïve patients with RA 51. Figure 3 (a-f) shows the risk ratios of response rates at 12 months for both the rituximab doses compared to placebo. Figure 4 summarises the radiographic data.

Evaluation & management of response & non-response

Consideration for repeated treatment

In a retrospective comparison of MTX-inadequate responder patients from phase II and phase III studies, patients that were treated at flare received less than half the dose of rituximab than treatment-to-target patients 52. Treatment to target patients were treated, on average, every 25 weeks whereas treatment as needed patients (i.e. at flare), every 60 weeks. However, disease in treatment as needed patients was less well controlled (with smaller fall in DAS28 and higher HAQ scores over 2 years, as well as more flares of disease activity; see figure 5). It is worth noting that the treatment as needed, upon flare, patient group had on average longer disease duration and greater DMARD exposure. Any differences in associated risk of toxicity and serious infection between the two groups cannot be ascertained from this study with only short-term toxicity evaluated. In the SUNRISE trial (TNF-inhibitor inadequate responders) all patients initially received open label rituximab (2 x 1000mg) 53. A significant improvement was observed in patients who received a second rituximab cycle versus placebo, but only in 1st cycle responders. There was no difference in serious adverse events (7%) or serious infection events (SIE) (2%) between the two arms of the study. Nevertheless, of patients responding at 24 weeks, a significant number maintained a response to 48 weeks without further rituximab treatment (e.g. 29% with ACR 70 response). Preliminary data from the German registry also suggests re-treatment after 6 months in complete responders and early re-treatment of non-responders is associated with improved longer-term benefits 54.

Post TNF-inhibitor failure & biological DMARD therapy post rituximab

Post TNF-inhibitor failure

The Swiss registry suggested a superior response to rituximab in patients failing their first TNF-inhibitor compared with a second TNF-inhibitor, but the groups were heterogeneous, requiring propensity scoring 55. Sub-group analysis suggested RTX’s effectiveness was greatest when used in the context of previous TNFi inefficacy (as opposed to intolerance) 56. The Swedish registry demonstrated the usefulness of either rituximab or an alternative TNF-inhibitor but in contrast to the Swiss report, switching to RTX following TNFi intolerance seemed to lead to better results 57, 58. In the REFLEX study, rituximab was assessed in an RCT after TNF-inhibitor failure49 (see above); data from RCTs on an alternative TNF-inhibitor 59 and other biological therapies 60, 61 following TNF-inhibitor failure demonstrated similar efficacy, but radiographic outcomes were not investigated in those trials contrasting with REFLEX.

Role of rituximab in other autoimmune diseases

Although most studies of RTX in other autoimmune disease were open label 62-69, there have been a couple of RCTs of interest. The ‘EXPLORER’ study 70 in patients with SLE failed to confirm the positive findings reported in non-controlled studies (category 1b) although other factors related to study design and disease assessment may have been relevant. The French AIR registry reported positive experiences of RTX in 136 patients with SLE 71. A very recent RCT in ANCA associated vasculitides (the ‘RAVE’ study) demonstrated the benefit of RTX in this patient cohort 72 although a smaller study did not reproduce such findings 73. RTX has also been evaluated in other CTDs including systemic sclerosis 74-76 and Sjogren’s syndrome 77-79.

Contraindication and serious adverse events

Serious Infections

There has been one confirmed case of pneumocystis jiroveci with the rate of these infections similar to control data.

Progressive Multifocal Leukoencephalopathy (PML)

PML is an often fatal demyelinating disorder caused by the JC polyoma virus, a common, usually asymptomatic virus (residing in the kidney). Historically, PML has been associated with haematological disorders and emerged as a major complication of HIV. The mechanism of PML development in relation to monoclonal antibody treatment remains unclear (with recent review by Carson K, et al80). Inhibition of T-lymphocyte trafficking into the CNS may partly account for cases associated with natalizumab and efalizumab. With rituximab treatment the B-lymphocyte depletion seen has been postulated to be followed by repletion of the mature B-lymphocyte with expansion of pre-B-lymphocytes harbouring latent JC virus. Whilst the development of PML in patients with RA has been associated with other established risk factors and possibly rituximab therapy, in the case of SLE, data suggest this disease may be an independent risk factor for PML - PML has been reported previously in 30 patients with SLE who had not received rituximab.

Immunoglobulin Levels & Infection Risk

A review of immunoglobulin (Ig) levels in a variety of pathologies and the risk of infection was recently reported 81. The results confirmed that in conditions such as primary immunodeficiency syndromes associated with sustained, very low levels of IgA, IgG, or IgM, a significantly increased risk of infection is observed, primarily respiratory tract infections of bacterial origin. The risk seems to be particularly at IgG levels ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download