To evaluate the safety and efficacy of infliximab in the treatment of juvenile rheumatoid arthritis (JRA).
To evaluate the safety and efficacy of infliximab in the treatment of juvenile rheumatoid arthritis (JRA).
This was an international, multicenter, randomized, placebo-controlled, double-blind study. One hundred twenty-two children with persistent polyarticular JRA despite prior methotrexate (MTX) therapy were randomized to receive infliximab or placebo for 14 weeks, after which all children received infliximab through week 44. Patients received MTX plus infliximab 3 mg/kg through week 44, or MTX plus placebo for 14 weeks followed by MTX plus infliximab 6 mg/kg through week 44.
Although a higher proportion of patients in the 3 mg/kg infliximab group than in the placebo group had achieved responses according to the American College of Rheumatology (ACR) Pediatric 30 (Pedi 30) criteria for improvement at week 14 (63.8% and 49.2%, respectively), the between-group difference in this primary efficacy end point was not statistically significant (P = 0.12). By week 16, after the crossover from placebo to infliximab 6 mg/kg when all patients were receiving infliximab, an ACR Pedi 30 response was achieved in 73.2% of all patients. By week 52, ACR Pedi 50 and ACR Pedi 70 responses had been reached in 69.6% and 51.8%, respectively, of patients. Infliximab was generally well tolerated, but the safety profile of infliximab 3 mg/kg appeared less favorable than that of infliximab 6 mg/kg, with more frequent occurrences of serious adverse events, infusion reactions, antibodies to infliximab, and newly induced antinuclear antibodies and antibodies to double-stranded DNA observed with the 3 mg/kg dose.
While infliximab at 3 mg/kg and 6 mg/kg showed durable efficacy at 1 year, achievement of the primary efficacy end point at 3 months did not differ significantly between infliximab-treated and placebo-treated patients. Safety data indicated that the 6-mg/kg dose may provide a more favorable risk/benefit profile. These results warrant further investigation in children with JRA.
Juvenile rheumatoid arthritis (JRA) is the most common chronic rheumatic disorder in children (1) and is also a major cause of acquired disability in childhood (2–4). Weekly methotrexate (MTX), at parenteral dosages of up to 15 mg/m2/week, has been established as an effective and safe therapy for polyarticular-course JRA (5–7). For children whose disease does not respond to MTX, anti–tumor necrosis factor (anti-TNF) therapy is a treatment option (8–11). One such anti-TNF agent, infliximab, is a chimeric monoclonal antibody that specifically and potently binds and neutralizes soluble TNFα and its membrane-bound precursor. Infliximab has been shown to be effective in combination with MTX for the treatment of adult RA (12, 13), and open-label studies of infliximab in JRA have been conducted (14, 15).
Infliximab has some potentially attractive features for treatment of JRA, including infrequent administration under the direct supervision of a health care professional and, in contrast to etanercept, lack of binding to lymphotoxin α, which is important in immune defense. The current trial was undertaken to investigate the efficacy and safety of infliximab in children with polyarticular-course JRA.
A total of 122 children with JRA from 34 sites (9 in North America, 3 in South America, and 22 in Europe) were enrolled between October 2001 and April 2004. An independent ethics committee at each center approved the protocol. Patients provided assent (if age appropriate and per local regulations), and parents/legal guardians provided written informed consent.
Criteria for study eligibility were as follows: age ≥4 years but <18 years, a diagnosis of JRA (16), suboptimal response to MTX after ≥3 months of treatment, ≥5 active joints, and no active systemic symptoms. Exclusion criteria included active uveitis, serious infection including tuberculosis, malignancy, or prior treatment with any TNF inhibitor. Disease-modifying drugs other than MTX and intraarticular corticosteroid injections were not permitted within 4 weeks prior to study entry and during the trial. Any patient with a positive purified protein derivative (PPD) skin test result had to receive adequate therapy for tuberculosis prior to receipt of study agent, according to the Centers for Disease Control and Prevention 2000 guidelines (17) or from the local tuberculosis control agency as appropriate. Concomitant treatment with low-dose corticosteroids (up to 0.2 mg/kg/day or 10 mg/day, whichever was less), 1 nonsteroidal antiinflammatory drug (NSAID), 1 analgesic that was not an NSAID, folic acid prophylaxis (required for all patients taking MTX), and narcotic or opioid analgesics was permitted during the study. If a mild or moderate reaction (e.g., headache, nausea, itching, urticaria) had been observed during a previous infusion, pretreatment with acetaminophen and an antihistamine was recommended for the patient prior to the start of all subsequent infusions.
This was a phase III, international, multicenter, randomized, double-blind, placebo-controlled study of infliximab therapy administered for 14 weeks, followed by a double-blind, all–active treatment extension through 44 weeks, in JRA patients receiving concomitant oral or parenteral MTX therapy (10–15 mg/m2/week). Patients were randomly assigned to 1 of 2 groups (Figure 1). Group 1 received placebo infusions plus MTX at weeks 0, 2, and 6, followed by an induction regimen of infliximab 6 mg/kg plus MTX at weeks 14, 16, and 20 and then every 8 weeks (termed “placebo/infliximab 6 mg/kg” or simply “infliximab 6 mg/kg”). Group 2 received infliximab 3 mg/kg plus MTX at weeks 0, 2, 6, and 14, placebo at week 16, and infliximab 3 mg/kg at week 20 and then every 8 weeks (termed “infliximab 3 mg/kg”). Patients completing treatment through week 44 who, in the judgment of the investigator, could benefit from continued treatment were eligible for an open-label extension beginning at week 52; this report contains data collected through week 52 only.
Infliximab was supplied as a lyophilized solid formulation containing 100 mg. Placebo was supplied as a lyophilized solid. Study medication was administered over a period of 40–120 minutes.
Serum samples were tested for antibodies to infliximab, using a bridging immunoassay (18), at weeks 0, 52, and 64. Serum infliximab concentrations were determined with an enzyme-linked immunoassay (18) at weeks 0, 2, 6, 14, 16, 20, 28, 36, 44, and 52. Clinical response evaluations, results of laboratory analyses (including testing for antinuclear antibodies [ANAs] and anti–double-stranded DNA [anti-dsDNA]), and adverse events were recorded throughout the study.
Response to therapy was defined based on the combination of the following 6 JRA core set parameters (19, 20): 1) physician global assessment of disease activity, on a 10-cm visual analog scale (VAS); 2) parent/patient global assessment of overall well-being, on a 10-cm VAS; 3) a measure of physical function, which for this trial was the cross-culturally adapted and validated version of a measure of functional ability in children with JRA, i.e., the Childhood Health Assessment Questionnaire (21, 22); 4) the number of joints with active arthritis, defined by the presence of swelling or, if no swelling was present, limitation of motion accompanied by pain, tenderness, or both (16); 5) the number of joints with limited range of motion; and 6) a laboratory measure of inflammation, which for this trial was the erythrocyte sedimentation rate. Whenever possible, each patient's joints were evaluated by the same assessor at each visit.
The primary end point of the trial was the proportion of patients meeting the American College of Rheumatology (ACR) Pediatric 30 (Pedi 30) criteria for improvement at week 14, defined as improvement of ≥30% in at least 3 of 6 core variables, with no more than 1 of the remaining variables worsened by >30% (19, 20). Patients were also evaluated using more stringent definitions of improvement, i.e., ≥50% or ≥70% improvement in at least 3 of 6 core variables, with no more than 1 of the remaining variables worsened by >30% (ACR Pedi 50 and ACR Pedi 70, respectively). At week 52, the number of patients who had 0 joints with active arthritis was determined.
This report complies with the recommendations of the CONSORT (Consolidated Standards of Reporting Trials group) statement (23). Statistical analyses were carried out according to the intent-to-treat principle, by 2 of the authors at Centocor, Inc. (Drs. Clark and Raychaudhuri). Members of the study steering committee (Drs. Ruperto, Lovell, Woo, Prieur, Petty, Martini, and Giannini) had full access to all statistical analysis reports.
For the primary efficacy end point, efficacy data were excluded due to withdrawal of consent for 1 patient in the placebo group (no efficacy data were available) and due to potential patient unblinding issues at a single site for 4 patients (2 in the placebo group and 2 in the infliximab 3 mg/kg group) (Figure 1). Thus, 117 patients were included in the primary efficacy analysis. In the analysis of the primary end point (the proportion of patients meeting the ACR Pedi 30 at week 14), patients who did not return for evaluation or patients with missing data were considered nonresponders. For efficacy analyses other than analysis of the primary end point, patients with available data at the given time point were included.
Statistical comparisons were made using the Cochran-Mantel-Haenszel chi-square test for categorical data and the van der Waerden test for continuous data. Ninety-five percent confidence intervals were calculated.
Patients were randomly assigned, at a 1:1 ratio, to 2 treatment groups. Patient allocation was performed using adaptive stratified randomization, with investigational site and age group (≥4 years but <8 years, ≥8 years but <12 years, and ≥12 years but <18 years) as the strata. Since age was considered as an independent variable that would affect probability of response, stratification according to age was implemented to assure proper balance among the treatment groups. It was determined that each treatment group would have to include at least 60 patients, to provide ∼79–97% power to detect a difference in the proportions of patients achieving the ACR Pedi 30 (α = 0.05, 2-sided) if the infliximab group had a frequency of response of at least 55–60% and the placebo group had a frequency of response of 25–30%.
One hundred twenty-two patients were randomized to 1 of 2 treatment groups: 62 to the placebo/infliximab 6 mg/kg group and 60 to the infliximab 3 mg/kg group. Prior to week 52, 13 (11%) of the 122 patients withdrew from the study (Figure 1). Baseline demographic and clinical characteristics were comparable between the 2 treatment groups (Table 1).
|Placebo/infliximab 6 mg/kg + MTX (n = 62)||Infliximab 3 mg/kg + MTX (n = 60)|
|Age at study entry, years||11.1 ± 4.0||11.3 ± 4.0|
|Disease duration, years||3.6 ± 3.4||4.2 ± 3.6|
|Age subgroup, no. (%)|
|≥4 years, <8 years||14 (22.6)||13 (21.7)|
|≥8 years, <12 years||18 (29.0)||14 (23.3)|
|≥12 years, <18 years||30 (48.4)||33 (55.0)|
|Female, no. (%)||49 (79.0)||53 (88.3)|
|White, no. (%)||53 (88.3)||50 (86.2)|
|JRA onset subtype, no. (%)|
|Systemic onset||8 (13.1)||11 (18.3)|
|Pauciarticular onset with polyarticular course||15 (24.6)||13 (21.7)|
|Polyarticular onset||38 (62.3)||36 (60)|
|Physician's global assessment of disease activity, 10-cm VAS||4.9 ± 1.9||5.2 ± 2.0|
|Parent's assessment of overall well-being, 10-cm VAS||4.1 ± 2.2||4.5 ± 2.2|
|C-HAQ score, 0–3 scale||1.2 ± 0.7||1.2 ± 0.7|
|No. of joints with active arthritis||18.5 ± 11.5||19.5 ± 12.3|
|No. of joints with limited range of motion||17.6 ± 12.0||18.4 ± 13.6|
|Erythrocyte sedimentation rate, mm/hour||32.0 ± 26.9||33.0 ± 25.0|
|Rheumatoid factor positive, no. (%)||14 (23.7)||13 (21.7)|
|Oral corticosteroid treatment, no. (%)||21 (34.4)||26 (43.3)|
|Prior DMARD treatment other than MTX, no. (%)||19 (31.1)||24 (40.0)|
|Weekly MTX dosage, mg/m2 (range 10–15)||12.2 ± 2.2||12.2 ± 2.4|
|Duration of MTX therapy, no. (%)|
|>6 months||37 (60.7)||35 (58.3)|
|≤6 months||8 (13.1)||9 (15.0)|
|Unknown†||16 (26.2)||16 (26.7)|
A rapid response was observed in the infliximab 3 mg/kg group, with 58.6% of the patients (34 of 58) exhibiting an ACR Pedi 30 response by week 6 (Figure 2). Although an ACR Pedi 30 response had been achieved in a higher proportion of patients in the infliximab 3 mg/kg group (37 of 58 [63.8%]) when compared with patients in the placebo group (29 of 59 [49.2%]) at week 14, the between-group difference in the proportion of patients achieving the primary efficacy end point was not statistically significant (P = 0.12). Also at week 14, higher proportions of patients in the infliximab 3 mg/kg group than in the placebo group met the ACR Pedi 50 improvement criteria (29 of 58 [50%] versus 20 of 59 [33.9%]; P = 0.078) and the ACR Pedi 70 improvement criteria (13 of 58 [22.4%] versus 7 of 59 [11.9%]; P = 0.130).
By week 16, following crossover from placebo to infliximab 6 mg/kg and at which time all patients were receiving infliximab, 73.2% of all patients (82 of 112) exhibited an ACR Pedi 30 response. By week 52, clinical responses meeting the ACR Pedi 50 and ACR Pedi 70 criteria were reached by 69.6% (78 of 112) and 51.8% (58 of 112) of the patients, respectively. There were no statistically significant differences between the infliximab dose groups when the proportions of patients with an ACR Pedi 30, ACR Pedi 50, or ACR Pedi 70 response at week 52 were compared. In addition, similar proportions of patients in the infliximab 3 mg/kg and infliximab 6 mg/kg groups had 0 joints with active arthritis at week 52 (26 of 59 [44.1%] and 25 of 58 [43.1%], respectively).
Findings for the individual components of the JRA core set are shown in Figure 3. By week 14, improvements in 4 of the 6 JRA core set components were greater in the infliximab 3 mg/kg group than in the placebo group. At week 14, the number of joints with active arthritis differed significantly between patients in the infliximab 3 mg/kg group and those in the placebo group (P = 0.016), whereas there were no significant differences for the other core set variables. By the end of the study, following crossover of placebo-treated patients to infliximab 6 mg/kg, improvement in the JRA core set components was comparable between the treatment groups.
Results of pharmacokinetic analyses demonstrated that, as expected, the infliximab 3 mg/kg group had lower median serum concentrations of infliximab early in the study compared with the infliximab 6 mg/kg group (at relative time frames). The majority of patients receiving infliximab 3 mg/kg had detectable serum infliximab concentrations up to week 36, after which more patients in this group had lower-than-quantifiable trough levels of infliximab in their sera. At these later sampling time points, more than half of the patients receiving infliximab 3 mg/kg (i.e., 51.1% at week 44 and 52.4% at week 52) did not maintain quantifiable infliximab concentrations by the end of the 8-week interval separating infusions (Figure 4). Overall, the median half-life of infliximab was 9.5 days with the 6-mg/kg dose, compared with 6.9 days with the 3-mg/kg dose. When the proportion of patients in whom an ACR Pedi 30 response was achieved was plotted against different categories of infliximab trough serum levels, a progressive increase in the proportion of patients achieving improvement was found to correspond to increasing serum infliximab concentrations (data not shown).
Overall, of the patients with evaluable samples, 25% (26 of 102) had antibodies to infliximab, with a higher incidence in the infliximab 3 mg/kg group (20 of 53 [38%]) compared with the infliximab 6 mg/kg group (6 of 49 [12%]). In addition, the antibody titers were notably higher among patients in the infliximab 3 mg/kg group, with 8 of 20 patients (40%) having a titer of 1:320 or higher (3 patients had titers of 1:20,480), compared with only 2 of 6 patients (33%) in the infliximab 6 mg/kg group.
Patients who did not test positive for antibodies to infliximab but following their last infusion had detectable serum concentrations of infliximab, which are known to interfere with interpretation of the analysis for antibodies to infliximab, were recorded as having inconclusive antibody status. This group of patients with inconclusive antibody status (determined 8 weeks after the last infliximab treatment) had a higher proportion of ACR Pedi 30 responders (46 of 49 [94%]) compared with the group of antibody-negative patients (19 of 24 [79%]) or the group of antibody-positive patients (16 of 24 [67%]). It should be noted that not all patients with appropriate samples for determination of antibodies to infliximab had ACR Pedi 30 data available for analysis.
Infusion-related reactions, defined as any adverse event that occurred during or within 1 hour following completion of an infusion, were observed in a higher proportion of patients in the infliximab 3 mg/kg group over 52 weeks (21 of 60 [35%]) than in either the infliximab 6 mg/kg group over 38 weeks (from week 14 to week 52) (10 of 57 [17.5%]) or the placebo group over 14 weeks (5 of 60 [8.3%]) (Table 2). Serious infusion reactions were more common in the infliximab 3 mg/kg group (4 patients) than in the infliximab 6 mg/kg group (2 patients). The most commonly reported events that were classified as infusion reactions were vomiting, fever, headache, and hypotension. In the placebo group, all infusion reactions were reported at week 14; those in the infliximab 3 mg/kg group were generally evenly distributed over weeks 14, 28, and 52, and those in the infliximab 6 mg/kg group were reported primarily at week 28. Three patients in the infliximab 3 mg/kg group (1 who continued in the study and 2 who discontinued study agent) and 1 in the infliximab 6 mg/kg group (who also discontinued study agent) had a possible anaphylactic reaction. No patient in the study had a delayed-type hypersensitivity reaction. In general, infusion-related reactions resolved upon slowing or stopping of the infusion.
|Placebo + MTX (0–14 weeks) (n = 60)||Infliximab 3 mg/kg + MTX (0–52 weeks) (n = 60)||Infliximab 6 mg/kg + MTX (14–52 weeks) (n = 57)|
|Mean ± SD weeks of followup||15.2 ± 7.2||50.8 ± 4.6||36.7 ± 4.5|
|Adverse events||49 (81.7)||58 (96.7)||54 (94.7)|
|Adverse events leading to discontinuation of study agent||1 (1.7)||2 (3.3)||5 (8.8)|
|Infusion reactions†||0 (0.0)||2 (3.3)||4 (7.0)|
|Circulatory failure||1 (1.7)||0 (0.0)||0 (0.0)|
|Depression||0 (0.0)||0 (0.0)||1 (1.8)|
|Serious adverse events‡||3 (5.0)||19 (31.7)||5 (8.8)|
|Infections||28 (46.7)||41 (68.3)||37 (64.9)|
|Serious infections||2 (3.3)||5 (8.3)||1 (1.8)|
|No. (%) of infusions with infusion reactions§||6/177 (3.4)||46/503 (9.1)||13/313 (4.2)|
|Patients with infusion reactions||5 (8.3)||21 (35.0)||10 (17.5)|
|Antinuclear antibodies (newly positive)||0/30 (0)||8/54 (14.8)||1/46 (2.2)|
|Anti–double-stranded DNA (newly positive)||0/30 (0)||7/54 (13.0)||0/46 (0)|
Among patients who had positive test results for antibodies to infliximab, the incidence of infusion reactions was 60% (12 of 20) in the infliximab 3 mg/kg group, compared with 50% (3 of 6) in the infliximab 6 mg/kg group. Across all treatment groups, there was a 3-fold higher incidence of infusion reactions among patients who tested positive for antibodies to infliximab (15 of 26 [58%]) than among patients who tested negative for antibodies to infliximab (5 of 26 [19%]) or who had inconclusive test results (6 of 50 [12%]). Similarly, serious infusion reactions occurred in 20% of patients (4 of 20) who were antibody positive, compared with none of the patients who had either negative or inconclusive test results.
The time frames during which adverse events were monitored, as shown in Table 2, differed among the treatment groups. The overall occurrence of adverse events was similar among the 3 treatment groups. More patients in the infliximab 6 mg/kg group (5 patients) discontinued study agent due to adverse events, compared with patients in the placebo and infliximab 3 mg/kg groups (1 patient and 2 patients, respectively) (Table 2). In addition, 2 patients receiving infliximab 3 mg/kg discontinued study agent as a result of abnormal laboratory test values, i.e., markedly abnormal alanine aminotransferase/aspartate aminotransferase levels that normalized after the MTX dosage was lowered (it should be noted that these laboratory abnormalities were not considered adverse events). After adjustment for the length of followup (i.e., incidence divided by length of followup), serious adverse events were approximately 2 times more frequent in the infliximab 3 mg/kg group (62.4%) than in the placebo group (32.9%) or the infliximab 6 mg/kg group (24.0%).
One patient died during the study, ∼10 days after the placebo infusion at week 2. After hospitalization for septic shock, cardiac function deteriorated, leading to death. A second patient, with systemic-onset JRA, died following study participation (infliximab 3 mg/kg group). This child experienced a severe JRA flare 3 months after the final infusion of infliximab and was hospitalized and treated. The patient withdrew from the study during participation in the open-label extension and 3 months later (6 months after the last infliximab infusion) died in the hospital, of cardiac arrest. The child had been placed on the stem cell transplant list because of JRA severity.
Infections occurred in 66.6% of patients treated with infliximab, compared with 46.7% of those receiving placebo (Table 2). The most commonly reported infection was upper respiratory tract infection. Three patients in the infliximab 3 mg/kg group had potential opportunistic infections: 1 with moniliasis presenting as vaginal thrush, 1 with moniliasis presenting as oral thrush, and 1 with herpes zoster. Serious infections occurred in 6 patients treated with infliximab and 2 patients receiving placebo. Among the serious infections reported, there were 4 cases of pneumonia in the infliximab 3 mg/kg group; 1 case followed a varicella zoster infection.
There was 1 report of asymptomatic pulmonary tuberculosis with negative mycobacterial cultures; this patient had a negative PPD skin test result at screening but a positive result during the routine screening that was performed prior to the start of the second year of the open-label extension. At that time, the patient was noted to have a PPD skin test result of 13 mm, as well as chest radiography and magnetic resonance imaging evidence of interstitial infiltration and nodularity. Results of polymerase chain reaction testing of a sputum sample from the patient were positive for Mycobacterium tuberculosis complex (Mycobacterium tuberculosis, Mycobacterium africanum, Mycobacterium bovis, and Mycobacterium microti), but all sputum cultures were read as negative. A subsequent lung scan revealed improvement of lung infiltration, and findings on repeat chest radiography were reported as being normal after 9 months of antituberculosis therapy. There were no reports of congestive heart failure or malignancy in this study.
Newly positive ANAs (titer ≥1:320) occurred in 8 of 54 patients (14.8%) and 1 of 46 patients (2.2%) during treatment with infliximab 3 mg/kg and infliximab 6 mg/kg, respectively, compared with none of the patients receiving placebo. Similarly, newly positive anti-dsDNA occurred in 7 of 54 patients (13.0%) and 0 of 46 patients (0%) during treatment with infliximab 3 mg/kg and infliximab 6 mg/kg, respectively, and in no patients during treatment with placebo. Overall, markedly abnormal changes in hematologic and clinical chemistry parameters were infrequent and were generally evenly distributed across treatment groups. A higher proportion of patients in the infliximab groups (1.7% and 8.8% in the 3 mg/kg and 6 mg/kg groups, respectively) as compared with placebo (0%) had at least 1 decrease in the absolute neutrophil count (<1.5 × 103 cells/μl and decrease ≥33%). Elevations in alanine aminotransferase levels to ≥90 IU/liter and elevations of ≥100% from baseline were observed in 3.3% of the patients treated with infliximab 3 mg/kg, 8.8% of those treated with infliximab 6 mg/kg, and no patients during placebo treatment.
This was a randomized placebo-controlled trial of a chimeric monoclonal anti-TNFα agent for the treatment of children with severe JRA that has been resistant to previous therapies. Patients treated with infliximab 3 mg/kg exhibited rapid improvement as early as week 2, based on the ACR Pedi 30 response. At week 14, an ACR Pedi 30 response had been achieved in a higher proportion of patients in the infliximab 3 mg/kg group compared with patients in the placebo group, but this difference in the primary efficacy end point was not statistically significant. At week 16, following crossover from placebo to the first dose of infliximab 6 mg/kg, 73% of all patients exhibited an ACR Pedi 30 response; this level of response was maintained for the remainder of the study. Similar trends were observed when higher levels of ACR Pedi response were assessed. Indeed, by week 52, an ACR Pedi 50 response had been achieved in just under 70% of the patients, and an ACR Pedi 70 response in just under 52%. Similar proportions of patients in the infliximab 3 mg/kg and infliximab 6 mg/kg groups had 0 joints with active disease at week 52.
These findings, despite the difference in the study design, are similar to those observed during the open-label phase of a trial assessing the use of etanercept in JRA (8), and demonstrate the value of infliximab in maintaining efficacy for up to 1 year. Of note, 78 of the 122 children enrolled in the study (63.9%) subsequently entered a 3-year open-label extension.
As noted above, the difference between the infliximab 3 mg/kg and placebo groups with regard to the percentage of patients meeting the ACR Pedi 30 criteria at week 14, the primary outcome measure for this trial, was not statistically significant. The ACR Pedi 30 criteria have been validated, and their reliability in discriminating between active agent and placebo has been established in studies of NSAIDs (24, 25), MTX (6), etanercept (8), and more recently, other biologic agents (26, 27). We believe the lack of statistical significance between treatment groups in the present study resulted from the combination of a small sample size (60 patients per treatment group) and a higher-than-anticipated rate of response (49.2%) in the placebo group (28), which was likely due to a greater placebo effect associated with agents administered by infusion and a placebo treatment phase that was too brief. Given that the study population, by design, consisted of patients with MTX-resistant disease, a placebo response of ∼20% was anticipated.
Overall, infliximab proved to be a beneficial treatment option, consistent with the risk/benefit profile in adult patients with RA (29). However, the safety profile for infliximab 3 mg/kg appeared to be less favorable in comparison with the 6-mg/kg dose, with an ∼2-fold higher incidence of infusion reactions (often severe), a higher incidence of antibodies to infliximab, and more frequent new formation of antinuclear and anti-dsDNA antibodies.
As implied by our pharmacokinetic findings and infusion reaction data, and as recently demonstrated in adults with RA (30), there is a correlation between maintenance of drug levels and avoidance of the development of antibodies to infliximab. In our study, children who received infliximab 6 mg/kg achieved better maintenance of drug levels and had a lower incidence of antibodies to infliximab. Indeed, if infliximab acts as an immunomodulator, the higher 6-mg/kg dose will be more likely than the 3-mg/kg dose to influence the incidence and titer of antibodies to infliximab. In addition, once development of infliximab antibodies is initiated, the neutralizing activity of the antibodies may lead to even more rapid elimination of infliximab at the lower dose. The suppressive effect of a higher infliximab dose on infliximab antibody formation may thus lead to a reduction in the incidence of adverse events associated with the presence of the antibodies.
In our study, 25% of the children developed antibodies to infliximab, with a higher incidence in the infliximab 3 mg/kg group (38%) than in the 6 mg/kg group (12%). These findings are consistent with trends previously observed in adult patients with Crohn's disease (18). However, since the increased incidence of antibody formation in JRA patients receiving the lower dose of infliximab in the present study led to comparatively lower median trough serum infliximab levels, we cannot rule out the possibility that if lower-dose treatment is given, a shorter interval between administrations might be necessary to maintain a more appropriate trough serum infliximab concentration, as has been recommended for the adult RA population (29), or that a higher dose of infliximab might be required at the beginning of the treatment regimen. Further evaluations in a larger group of patients will be needed to confirm or refute this unanticipated finding of a more favorable safety profile with infliximab 6 mg/kg than with infliximab 3 mg/kg.
In conclusion, while both infliximab 3 mg/kg and infliximab 6 mg/kg, in combination with MTX, produced an important, rapid, and durable clinical effect in children with JRA at 1 year, the primary efficacy end point of this study was not significantly different between infliximab treatment and placebo treatment at week 14. The 3-mg/kg dose of infliximab was associated with a substantially higher risk of serious adverse events, infusion reactions, and development of antibodies to infliximab, ANAs, and anti-dsDNA compared with the 6-mg/kg dose. Thus, the use of infliximab in children warrants further investigation.
Members of the study steering committee (Drs. Ruperto, Lovell, Woo, Prieur, Petty, Martini, and Giannini) had full access to all statistical analysis reports and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study design. Ruperto, Lovell, Travers, Keenan, Fasanmade, Martini, Giannini.
Acquisition of data. Ruperto, Lovell, Cuttica, Wilkinson, Woo, Espada, Wouters, Silverman, Balogh, Henrickson, Apaz, Baildam, Fasth, Gerloni, Lahdenne, Prieur, Ravelli, Saurenmann, Gamir, Wulffraat, Marodi, Petty, Joos, Zulian, McCurdy, Myones, Nagy, Reuman, Szer, Travers, Visvanathan, Fasanmade, Mendelsohn.
Analysis and interpretation of data. Ruperto, Lovell, Fasth, Ravelli, Nagy, Travers, Beutler, Keenan, Clark, Visvanathan, Fasanmade, Raychaudhuri, Mendelsohn, Martini, Giannini.
Manuscript preparation. Ruperto, Lovell, Cuttica, Wilkinson, Woo, Silverman, Henrickson, Baildam, Fasth, Lahdenne, Prieur, Ravelli, Saurenmann, Marodi, Zulian, Myones, Beutler, Keenan, Fasanmade, Raychaudhuri, Mendelsohn, Martini, Giannini.
Statistical analysis. Clark, Raychaudhuri, Mendelsohn.
Study logistics, study steering committee. Ruperto, Lovell.
Medical monitoring. Beutler.
Case report form design. Ruperto, Giannini.
The authors wish to acknowledge the writing support of Michelle Perate, MS and Mary Whitman, PhD. We also would like to thank Drs. Eric Veys (Ghent, Belgium), Richard Mouy (Paris, France), Ivan Foeldvari (Hamburg, Germany), Joachim Oppermann (Cottbus, Germany), Fernanda Falcini (Firenze, Italy), Wietse Kuis (Utrecht, The Netherlands), Stefan Hagelberg (Stockholm, Sweden), Polly Livermore (London, UK), Joyce Davidson (Liverpool, UK), and Lisa Imundo (New York, NY) for their participation in the trial.