Intravenous immunoglobulin for juvenile idiopathic arthritis
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
The aim of this review is to determine the clinical effectiveness and safety of intravenous immunoglobulin for JIA.
This review will address two core questions:
1. What is the clinical efficacy of intravenous immunoglobulin for the treatment of JIA in terms of controlling the disease and reducing joint counts?
2. What are the risks (frequency and severity of adverse events) associated with immunoglobulin treatment in these patients?
Background
Juvenile idiopathic arthritis (JIA) represents a heterogeneous group of autoimmune diseases (Petty 1998). JIA arises before patients are 18 years old; diagnosis requires manifestation of symptoms lasting more than six weeks. It has an annual incidence of between 1.3 to 22.6 cases per 100,000 (Moe 1998; Prieur 1987). The disease tends to affect large and small joints and may interfere with growth and development. We can distinguish many arthritis sub‐types: (i) Systemic onset (Still's disease) occurs in about 20% of patients; (ii) Pauciarticular onset occurs in about 40% of patients; (iii) Polyarticular onset occurs in the remaining 40% of patients. Patients with polyarticular onset and a positive rheumatoid factor have a less favorable prognosis (Petty 1998; WHO Press 2005).
The treatment goals for JIA are straightforward: to achieve remission of disease with normal childhood growth and development (Agle 2003). However, there are no therapies that have been demonstrated to consistently achieve remission. For patients with mild disease, non‐steroidal anti‐inflammatory drugs (NSAIDs) or intra‐articular corticosteroid injections are the first‐line treatment choices. Naproxen and ibuprofen are frequently used because of equivalent efficacy, lower toxicity, and longer half‐life. However, in the absence of a response with these drugs, low doses of disease modifying anti‐rheumatic drugs (DMARDs) (such as methotrexate, sulfasalazine, and hydroxycholoroquine) are useful second‐line options for managing polyarticular disease. Except for treatment of severe systemic disease, systemic corticosteroids usually can be avoided (Goldenstein 1997).
The majority of children respond well to conventional treatment. However, a more aggressive disease course will be resistant to standard management. Biological agents (such as tumour necrosis factor antagonist and other cytokines inhibitors) and unique drugs (such as thalidomide) can obtain a satisfactory outcome; they provide new opportunities to suppress the inflammation in severe cases of JIA (CIGNA 2004; Hashkes 2005; McEvoy 2006).
Intravenous immunoglobulin (IVIG) is one such biologic agent. IVIG is a sterile freeze‐dried concentrate of immunoglobulin G (IgG), prepared from large pools of human plasma IVIG has a half‐life of approximately three to four weeks and provides a broad spectrum of IgG antibodies. Studies indicate several possible biological mechanisms for IVIG, including modulation of the expression and function of Fc receptors, interference with the cytokine network, and effects on T and B cells (Kazatchkine 2001).
The use of IVIGs in autoimmune diseases is not widely accepted, although the efficacy of intravenous immunoglobulin treatment has been demonstrated in several autoimmune diseases (i.e., Kawasaki disease, idiopathic thrombocytopenic purpura, dermatomyositis, and Guillan‐Barre syndrome) (Kazatchkine 2001; Sany 1994).
The use of IVIG in treating adult rheumatoid arthritis (RA) is controversial. Although some formal, randomised controlled trials suggest that IVIG is ineffective in treating rheumatoid arthritis (RA) (Kanik 1996; Pyne 2002), other trials have reported improvement in patients with RA and malignant RA (Maeda 2001; Pap 1998). Furthermore, anecdotal evidence and a few published articles suggest that IVIG treatments are effective in treating subsets of JIA; some studies show that IVIG has potential to decrease viral load, inflammation, and systemic manifestations such as fever (Lehmann 2004; Prieur 1990; Shaikov 1998; Silverman 1990; Silverman 1994; Uziel 1996).
Adverse events after administration of intravenous gamma‐globulin do occur but are seldom serious. There is a very small risk of anaphylaxis, almost always in patients with severe immunoglobulin A deficiency. Other reported side effects include risk of transmission of unknown infections, aseptic meningitis and headache, haemolytic anemia, infusion reactions, neurological involvement, skin reactions, neutropenia, and worsening of co‐morbid conditions (Bertorini 1996; Dalakas 2004; McCluskey 1990).
To the authors' knowledge, there is no previous systematic review on the use of IVIG in children. Thus, the role of IVIG in juvenile arthritis remains unclear. This review will revisit IVIG treatment as a viable alternative to first line JIA treatment failure.
Objectives
The aim of this review is to determine the clinical effectiveness and safety of intravenous immunoglobulin for JIA.
This review will address two core questions:
1. What is the clinical efficacy of intravenous immunoglobulin for the treatment of JIA in terms of controlling the disease and reducing joint counts?
2. What are the risks (frequency and severity of adverse events) associated with immunoglobulin treatment in these patients?
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled trials (RCTs) or controlled clinical trials (CCTs) comparing immunoglobulin to placebo and/or other therapies are eligible for inclusion. Studies with follow‐up less than one month will be excluded.
Types of participants
We will include studies on types of JIA fulfilling all criteria for diagnosis according to the established criteria from national and international organizations [such as International League of Associations for Rheumatology (ILAR), American College of Rheumatology (ACR), and European League Against Rheumatism (EULAR)], without any restrictions of disease duration, gender or concomitant treatment. This review will concern juvenile idiopathic arthritis [juvenile rheumatoid arthritis (JRA), juvenile chronic arthritis (JCA), juvenile idiopathic arthritis (JIA)] in patients under age 18, including all subgroups [oligo (pauci) articular JIA, rheumatoid factor (RF) negative and RF positive polyarthritis, systemic onset JIA, psoriasis arthritis, enthesitis‐related arthritis, and other arthritides] as diagnosed by a rheumatologist using established criteria from national and international organisations (ILAR, ACR, EULAR).
Participants with no clear diagnosis of JIA and those over 18 years of age will be excluded.
Types of interventions
Treatment trials with immunoglobulin versus placebo or immunoglobulin versus a DMARD are eligible for inclusion. Any doses of immunoglobulin will be eligible for inclusion. Patients can be on other DMARDS, non‐steroidal anti‐inflammatories or corticosteroids, provided that the doses are stable and randomly allocated to treatment with or without immunoglobulin.
Types of outcome measures
We will consider and include all the outcome measures recommended for use in clinical trials in the PRINTO‐core set (Giannini 1997). Moreover, side effects, total dropouts, and drug terminations will be included in the review. Articles will be excluded if there is lack of sufficient information in relation to the outcome measures of interest.
(i)Measures of disease activity, including active arthritis and limited range of motion joint counts, disability score, parent or patient's global assessment, and physician's global assessment, pain, and erythro‐sedimentation rate (ESR)
(ii)DAS (Disease Activity Score) (Prevoo 1995)
(iii)Remission rates
(iv)ACR improvement criteria (Giannini 1997)
(v)EULAR improvement criteria (van Gestel 1996)
(vi)Adverse events (i.e. anaphylaxis, transmission of unknown infections, aseptic meningitis and headache, haemolytic anemia, infusion reactions, neurological involvement, skin reactions, neutropia, etc.).
We will categorise under secondary outcome measures the following:
(vii)Health‐related quality of life
(viii)Functional ability (as measured on functional tests/questionnaires [e.g., JAFAS (Lovell 1989), CHAQ (Singh 1994), and JASI (Wright 1996)]
(ix)Joint damage as diagnosed radiographically (radiological damage)
(x)Drug terminations (Total dropouts, withdrawals due to inefficacy or side effects)
Search methods for identification of studies
Sources of published data will include electronic databases, handsearching of selected rheumatology journals, and handsearching of conference proceedings. We will not apply language restrictions.
Electronic databases: A number of electronic databases will be searched:
(i)Journal‐based databases: MEDLINE (via PubMed), EMBASE, CENTRAL, Literatura Latinoamericana en Ciencias de la Salud (LILACS), Database of Abstracts of Reviews of Effects and HTA Database
(ii)Clinical trials registries: ClinicalTrials.gov; National Research Register.
(iii)Regulatory agencies: Food and Drug Administration
Search strategies to identify studies will follow the Musculoskeletal Review Group recommendations.
Topic search terms (shown in PubMed format) are listed below and in Appendix 1. We will use similar searches for all other databases:
#1 Arthritis, Juvenile Rheumatoid[MeSH]
#2 Juvenile AND arthriti*[tw]
#3 Juvenile AND still* disease[tw]
#4 .#1 AND #2 AND #3
#5 gamma‐Globulins[MeSH]
#6 Immunoglobulins, Intravenous[MeSH]
#7 Intravenous immunoglobulin*[tw]
#8 IV Immunoglobulin*[tw]
#9 IVIG[tw]
#10 Alphaglobin[tw]
#11 Endobulin[tw]
#12 Gamimune[tw]
#13 Gammagard[tw]
#14 Gammonativ [tw]
#15 Intraglobin[tw]
#16 Intravenous IG[tw]
#17 Iveegam[tw]
#18 Sandoglobulin[tw]
#19 Venimmune[tw]
#20 Venoglobulin[tw]
#21 #5 AND #6 AND #7 AND #8 AND #9 AND #10 AND #11 AND #12 AND #13 AND #14 AND #15 AND #16 AND #17 AND #18 AND #19 AND #20
#22 #4 AND #21
Handsearching journals: Tables of contents of major rheumatology journals from 1980 to 2005 will be handsearched. Journals will include Arthritis and Rheumatism; Arthritis Care and Research; Journal of Rheumatology; Annals of Rheumatic Diseases; Rheumatology; Clinical and Experimental Rheumatology; and the Scandinavian Journal of Rheumatology.
Handsearching bibliographies: The bibliographies of all retrieved articles will also be handsearched for additional references not otherwise retrieved.
Conference Proceedings: We will review all abstracts presented at scientific meetings between 1998 and 2005 including the following sponsoring organisations: American College of Rheumatology (ACR) and the European League against Rheumatism (EULAR).
Grey literature: In order to obtain information on unpublished literature, we will contact the pharmaceutical companies that produce immunoglobulin and experts in the field.
We will also ask the manufacturers to declare the entire set of trials conducted on the drug.
Data collection and analysis
To determine each study's eligibility for inclusion, two separate investigators will independently review the title and abstracts of all the papers initially identified as potentially relevant. Disagreements will be resolved by consensus or by a third adjudicator if consensus cannot be reached. Once the relevant abstracts have been identified, the full publication will be reviewed by the two independent investigators to determine suitability for final inclusion. Inter‐observer agreement will be measured with kappa coefficients which are used as a means of classifying agreement between non‐numerical data. A kappa coefficient of 1 indicates a statistically perfect modeling; a kappa coefficient of 0 indicates that every model value was different from the actual value. A kappa statistic of 0.7 or higher is generally regarded as good statistic correlation‐the higher the value, the better the correlation (Artstein 2005).
Data handling
We will use the data extraction forms from the Cochrane Musculoskeletal Group and prior Cochrane systematic reviews pertaining to RA. Two independent investigators will extract the relevant data, including study characteristics, quality, results and conclusions. Both investigators will compare their data. Disagreements will be resolved by consensus and adjudication by other reviewers, if necessary. Where data are missing, further information will be sought from the authors. Data from studies with multiple publications will be extracted and reported as a single study.
Methodological quality
Methodological quality will be assessed of each study on the basis of the following components: (i) randomisation, (ii) adequate concealment of randomisation, (iii) blinding of provider, recipient and outcome assessor (iv) use of intention to treat analysis, and (v) description of dropouts and withdrawals. Each component will be assessed as met, not met, or unclear:
Allocation concealment will be assessed as per the Cochrane Handbook (Higgins 2005):
A. Adequate concealment ‐ Allocation of participants to different groups was not known until
the point of allocation (e.g. sequentially numbered, sealed, opaque envelopes; onsite computer
system with locked, unreadable files)
B. Unclear concealment (e.g. stating only that a list or table was used)
C. Inadequate concealment ‐ Transparent before allocation (e.g. alternation; case record
numbers; dates of birth or days of the week)
D. Not used ‐ clear that allocation concealment was not used
.
Grading of evidence
We will use the grading system described in Evidence‐based Rheumatology (Tugwell 2004) and recommended by the Musculoskeletal Group:
Platinum:
To achieve the platinum level of evidence, a published systematic review must have at least two individual controlled trials, each satisfying the following criteria:
-
Sample sizes of at least 50 per group‐If these do not find a statistically significant difference, then they are adequately powered for a 20% relative difference in the relevant outcome.
-
Blinding of patients and assessors for outcomes
-
Handling of withdrawals >80% follow up (imputations based on methods such as Last Observation Carried Forward (LOCF) are acceptable).
-
Concealment of treatment allocation
Gold:
The gold level of evidence requires at least one randomised clinical trial meeting all of the following criteria for the major outcome(s) as reported:
-
Sample sizes of at least 50 per group‐If these do not find a statistically significant difference, then they are adequately powered for a 20% relative difference in the relevant outcome.
-
Blinding of patients and assessors for outcomes
-
Handling of withdrawals > 80% follow up (imputations based on methods such as LOCF are acceptable)
-
Concealment of treatment allocation
Silver:
The silver level of evidence requires a randomised trial that does not meet the above criteria for gold or platinum ranking or evidence from at least one study of non‐randomised cohorts that did and did not receive the therapy, or evidence from at least one high quality case‐control study. A randomised trial with a "head‐to‐head" comparison of agents would be considered silver level ranking unless a reference were provided to a comparison of one of the agents to placebo showing at least a 20% relative difference.
Bronze:
The bronze ranking requires at least one high quality case series without controls (including simple before/after studies in which patients act as their own control) or a conclusion derived from expert opinion based on clinical experience without reference to any of the foregoing (for example, argument from physiology, bench research or first principles).
Data analysis
We will extract data (including study characteristics and outcome measures of efficacy and toxicity) and perform statistical analyses using RevMan 4.2.8 software. When possible, we will calculate the relative risk using a random effects model applied to each of the dichotomous outcomes. For the meta‐analysis of continuous outcomes, weighted mean differences (WMDs) between groups will be estimated and when different scales are used, standardized mean differences (SMDs) will be estimated. For standard measures of effect size, 0.80 will be considered large, 0.50 medium and 0.20 small. Toxicity will be analysed using relative risk (or Peto odds ratio for rare events) for total withdrawals and dropouts and withdrawals for specific reasons.
We will apply fixed effects models throughout, except when heterogeneity exists, in which cases a random effects model will be used to introduce less bias than excluding trials altogether. We will pool sufficiently homogeneous studies (e.g., similarities between participants, interventions, outcome assessment, etc.). We will include narrative summaries of heterogeneous outcomes.
Withdrawals and adverse events will be analysed. If the standard deviation (SD) for results is not available, we will input the baseline SD or an estimated value using the coefficient of variation (CV=SD/mean) from the other trials. If trials report means and ranges, the range will be divided by three to estimate the SD. We will try to be as conservative as possible; trials that report change from baseline scores with no SD will not be combined with trials that report end of trial results.
Clinical relevance tables
Clinical relevance tables will be presented under additional tables to improve the readability of the review. For dichotomous outcomes, the weighted absolute risk difference will be calculated using the risk difference (RD) statistic in RevMan. RR‐1 calculates the weighted relative per cent change. The number needed to treat (NNT) will be calculated from the control group event rate (unless the population event rate is known), and the relative risk will be calculated using the Visual Rx NNT calculator (Cates 2004).
Continuous outcome tables will also be presented under additional tables. For continuous variables, we will use the weighted mean difference (WMD) and standardised mean difference (SMD). Weighted absolute change will be calculated from the WMD statistic in RevMan when trials using the same scale are pooled. SMD will be used when studies use different scales to measure the same outcome. The SMD will be multiplied by the baseline standard deviation in the control group to obtain the weighted absolute change. Relative per cent change from baseline will be calculated as the absolute benefit divided by the baseline mean of the control group. NNT is calculated using the Wells calculator software available at the CMSG editorial office. The minimal clinically important difference (MCID) for each outcome will be determined for input into the calculator.
Heterogeneity
The results from the studies will be tested for heterogeneity using the chi‐square test when possible. Heterogeneity will be consider statically significant when a P value is less than 0.10 (Cochrane 2006; Higgins 2005).
Sensitivity analysis
We will perform sensitivity analyses by assessing different treatment combinations, patients with early or established RA, timing of drug initiation, trial design (step‐up, parallel or step‐down), adequate allocation concealment, blinding of patients and outcome assessors, and withdrawals and dropouts less than 15%. Trial results will be entered into RevMan 4.2.8 using the same plot direction to enable the pooling of results where the lowest value has improved and the highest value has worsened. Negative values in SMD will indicate a benefit of the active drug over the placebo.
