Roche trials database

Protocol number:

WA19977

Title of Study:

A 24 week randomized, double-blind, placebo-controlled withdrawal trial with a 16 week open-label lead-in phase, and 64 week open-label follow-up, to evaluate the effect on clinical response and the safety of tocilizumab in patients with active polyarticular-course juvenile idiopathic arthritis.

Sponsor:

Hoffmann-La Roche

Company division:

Pharmaceutical

Product name:

RoActemra/Actemra

Generic name:

tocilizumab [RoActemra/Actemra]

Therapeutic area:

• Juvenile Idiopathic Arthritis

Clinical study summary:

This was a multicenter, 3-part study: Part I consisted of a 16-week active RoActemra/Actemra treatment lead-in period followed by Part II, a maximum of 24 weeks randomized double-blind placebo-controlled withdrawal period, followed by Part III, a 64-week open-label period beginning at Week 40 to examine the long term use of RoActemra/Actemra on safety and efficacy. In Part II, patients were randomized in a 1:1 ratio to either receive placebo or continue on active RoActemra/Actemra treatment at the same dose received in Part I. Each patient continued in Part II of the study until Week 40 or until they satisfied JIA ACR30 flare criteria (relative to Week 16) and qualified for escape.

Study center(s):

58 centers in Argentina, Australia, Belgium, Brazil, Canada, France, Germany, Italy, Mexico, Peru, Poland, Russian Federation, Spain, United Kingdom, U.S.A.

Phase of development:

III

Objectives:

The primary objective (from Part II) was to compare the proportion of patients on RoActemra/Actemra (tocilizumab) versus placebo who developed a juvenile idiopathic arthritis (JIA) American College of Rheumatology 30 (ACR30) flare (relative to Week 16) by Week 40

The secondary objectives were as follows:

• To evaluate the efficacy of open-label RoActemra/Actemra therapy (Part I)
• To evaluate the long-term effect of RoActemra/Actemra on the maintenance of clinical response and safety in patients with polyarticular JIA (pJIA) (Parts II and III)
• To evaluate the efficacy and safety of RoActemra/Actemra 8 mg/kg vs 10 mg/kg in patients < 30 kg (Parts I, II and III)

Methodology:

Part I: 16-week active treatment lead-in period: A blinded joint assessor performed all the joint examinations in Part I and Part II. No reductions or changes of concomitant NSAIDs, corticosteroids or methotrexate dosing were permitted during Part I, except for documented safety reasons.

Part II: 24-week double-blind withdrawal period: No reductions or changes of study drug or concomitant medications were permitted during Part II of the study, except for documented safety reasons. Patients who entered escape in Part II could receive standard of care

Part III: A 64-week open-label long term extension: If a patient had successfully completed Part II of the study (Week 40) or entered escape, they received IV tocilizumab at the dose given at baseline. In Part III only, patients receiving tocilizumab 10 mg/kg, that increased their weight to ≥ 30 kg and at least 5 kg over baseline weight for 3 consecutive visits, had their tocilizumab dose decreased to 8 mg/kg (at investigator discretion) at the third consecutive visit.

Throughout the study, patients were assessed for clinical efficacy and safety. Patients who received prohibited therapy were withdrawn from study drug. Reductions in methotrexate or NSAIDs were permitted at any time for safety issues. All patients that discontinued the study for any reason returned for follow-up safety assessments.

Number of patients (planned/analyzed):

188 patients enrolled: 166 patients analyzed

Diagnosis and main criteria for inclusion:

Patients aged 2 years up to and including age 17 years with at least 6 months active polyarticular-course juvenile idiopathic arthritis (pJIA) with an inadequate response to methotrexate, due to lack of efficacy or toxicity, who were receiving standard of care, either with or without non-steroidal anti-inflammatory drugs (NSAIDs), either with or without or low-dose corticosteroids, and either with or without concomitant methotrexate therapy.

Test product, dose and mode of administration or test procedure:

Part I: Active treatment lead-in period, every 4 weeks for 4 doses.

• < 30 kg randomized 1:1 to either tocilizumab 8 mg/kg or 10 mg/kg intravenous (IV) infusion
• ≥ 30 kg tocilizumab 8 mg/kg IV infusion

Part II: Double-blind withdrawal period.

• tocilizumab (at the same dose as Part I) or placebo

Part III: Open-Label (Part I dose resumed)

Duration of treatment:

40 weeks

Reference therapy, dose and mode of administration or reference procedure:

Part II: Double-blind withdrawal period.

• Placebo to tocilizumab

Criteria for evaluation (efficacy, safety):

Efficacy

Primary endpoint: proportion of patients with JIA ACR30 flare (in Part II)

Secondary endpoints: JIA ACR30/50/70/90 responses, JIA ACR core components (parent/patient global assessment of overall well-being, physician global assessment of disease activity, number of joints with active arthritis, number of joints with limitation of movement, Erythrocyte Sedimentation Rate (ESR), functional ability determined by the Childhood Health Assessment Questionnaire Disability Index [CHAQ-DI]), pain visual analog scale, C-reactive protein (CRP), Juvenile Arthritis Damage Index, (ADI), inactive disease status, and Juvenile Arthritis Disease Activity Score (JADAS)

Pharmacokinetics and Pharmacodynamics

Tocilizumab serum concentration, interleukin-6, interleukin-6-receptor

Safety

Adverse events, laboratory results, physical examination, vital signs, electrocardiogram, tuberculosis test

Statistical methods:

The primary analysis was based on an analysis using the Cochran-Mantel-Haenszel (CMH) test adjusted for stratification factors. All statistical hypotheses for the primary and secondary endpoints were tested at the 5% significance level using a 2-sided test. To control for the Type I error rate the secondary endpoints were tested in a hierarchical fixed sequence.

Nonlinear mixed effects modeling (NONMEM) was used to analyze the serum concentration-time tocilizumab data collected in this study. Systemic pharmacokinetic exposure parameters were computed using the population pharmacokinetic model for patients who had received all 4 scheduled tocilizumab infusions for Part I and 6 scheduled tocilizumab infusions for Part II. The model computed pharmacokinetic parameters and observed exposure parameters for each of the tocilizumab dosing regimens were compared and the geometric mean ratio together with the associated 90% confidence interval (CI) were presented for each pairwise comparison.

Values of the pharmacodynamic parameters were summarized descriptively.

Safety analyses included examination of overall summaries and characteristics or types of AEs/serious adverse events (SAEs).

Summary (efficacy, safety, other results):

Study population

Demographic and disease characteristics at baseline were as expected for this study population and well balanced between treatment groups. The majority of the patients were females (77%), white (80%) and of non-Hispanic ethnicity (66%).

The average age of the patients in the study was 11 years and the majority of patients weighed ≥ 30 kg (119 [63%]) at baseline. Differences in the demographic characteristics (age, height, weight and body surface area) were reflected in the weight groups of < 30 kg and ≥ 30 kg.

The mean disease duration at baseline was 4.2 years with patients ≥ 30 kg having a longer duration of disease than patients < 30 kg as would be expected based on the average ages of patients in these categories. The majority of patients were RF-negative (126 [67%]) at baseline. Demographic and disease characteristics were similar for the patient populations in Part I and Part II.

There were no meaningful differences in the values of the JIA ACR core set components at baseline between treatment groups of RoActemra/Actemra 8 mg/kg (< 30 kg), RoActemra/Actemra 10 mg/kg (< 30 kg) and RoActemra/Actemra 8 mg/kg (≥ 30 kg). Patients had severe disease at baseline with a mean (standard deviation [SD]) number of active joints of 20.3 (14.3), mean (SD) patient/parent and physician global assessment VAS scores of 52.9 (25.0) and 61.4 (20.7), respectively and mean (SD) CHAQ-DI score of 1.39 (0.74).

The randomization stratification variables were similar between treatment groups. 46% of patients were receiving concomitant oral corticosteroids at baseline with a mean dose of 0.13 mg/kg/day. 79% of patients were receiving concomitant methotrexate at baseline with a median dose of 11.8 mg/m²/week.

Efficacy

The study met its primary endpoint at Week 40 with a significantly higher proportion of placebo patients experiencing a JIA ACR30 flare between Week 16 and Week 40 relative to Week 16 compared to the all RoActemra/Actemra patients (48.1% vs 25.6% [p = 0.0024]). All of the supportive analyses of the primary endpoint (alternative missing data imputation, PP population and alternative statistical techniques) similarly detected a statistically significant treatment difference.

The hierarchical chain of assessment for secondary endpoints for the study was broken at the assessment of number of joints with limitation of movement and hence treatment significance was not assessed below that point in the chain of assessments. The main secondary efficacy findings were as follows:

• After 16 weeks of open-label treatment, 168 out of 188 patients (89.4%) achieved at least a JIA ACR30 response and were eligible to continue to Part II of the study. The JIA ACR30 response rate in the open-label treatment period (Part I) was slightly lower in the RoActemra/Actemra 8 mg/kg (< 30 kg) patients (26/34 patients, 76.5%) compared to the RoActemra/Actemra 10 mg/kg patients (31/35 patients, 88.6%). The response rate in RoActemra/Actemra 8 mg/kg (≥ 30 kg) patients (111/119 patients, 93.3%) was comparable to that of RoActemra/Actemra 10 mg/kg (< 30 kg) patients. This pattern of response was also evident in relation to JIA ACR50/70/90 endpoints.
• There was a rapid onset of response to open-label treatment with 50% of patients having a JIA ACR30 response by Week 2.
• Improvement in all of the JIA ACR core components was observed during Part I.
• There was a statistically significant difference in the rate of JIA ACR30/50/70 response at Week 40 with a higher response rate for patients randomized to RoActemra/Actemra.
• Efficacy in the primary endpoint was driven by improvements in all JIA ACR core components. The effect of the dose difference between RoActemra/Actemra 10 mg/kg (< 30 kg) and RoActemra/Actemra 8 mg/kg (< 30 kg) was seen in all components except the patient/parent global assessment VAS.
• Patients who received RoActemra/Actemra after Week 16 showed a statistically significant decrease compared to placebo in the pain VAS
• Twice as many patients in the all RoActemra/Actemra group compared to the all placebo group had inactive disease at Week 40.
• Patients treated with concurrent methotrexate showed enhanced JIA ACR responses and reduced flare rate at Week 40 in both the RoActemra/Actemra and placebo groups.
• Biologic-naïve patients showed enhanced JIA ACR responses and reduced flare rate compared to patients previously treated with biologics in both the RoActemra/Actemra and placebo groups.
• JADAS-27 improvement occurred rapidly and continued over time to Week 40 for patients receiving RoActemra/Actemra.

Pharmacodynamic results:

In both Part I and Part II of the study, following RoActemra/Actemra administration to Week 16 or Week 40, pharmacodynamic profiles for CRP, ESR and sIL-6R concentrations were similar between RoActemra/Actemra 10 mg/kg (< 30 kg) and 8 mg/kg (≥ 30 kg) groups, whereas pharmacodynamic profiles for RoActemra/Actemra 8 mg/kg (< 30 kg) consistently demonstrated inferior response compared to the pharmacodynamic profiles from the other 2 groups.

Pharmacokinetic results:

Pharmacokinetic exposure parameters were more comparable for the RoActemra/Actemra 10 mg/kg dose in pJIA patients weighing < 30 kg and the RoActemra/Actemra 8 mg/kg dose in pJIA patients weighing ≥ 30 kg. Pharmacodynamic data results (CRP, ESR, IL-6 and sIL-6R) showed that the RoActemra/Actemra 10 mg/kg dose in pJIA patients weighing <30 kg and the RoActemra/Actemra 8 mg/kg dose in patients weighing ≥30 kg achieved comparable pharmacodynamics responses. The comparison of efficacy outcome (JIA ACR30/50/70/90) by treatment group confirmed that efficacy achieved for the RoActemra/Actemra 10 mg/kg dose in children with a bodyweight < 30 kg was comparable to that for the RoActemra/Actemra 8 mg/kg dose in children weighing≥ 30 kg. Analysis of JIA ACR responses by pharmacokinetic exposure quartiles further illustrated that lower pharmacokinetic exposure resulted in lower clinical efficacy.

In addition, there was no consistent trend for the rates of AEs per 100 patient-years across the 3 treatment groups and 4 pharmacokinetic exposure quartiles.

Safety results:

The safety results have been presented using the all exposure population; this includes all the patients until their last visit before 04 November 2011. During Part III of the study (after the double-blind period) patients who at baseline weighed < 30 kg and were randomized to RoActemra/Actemra 10 mg/kg were allowed to switch dose to RoActemra/Actemra 8 mg/kg if they had sufficient sustained weight gain.

• The rate of AEs in the all exposure population was 479.8 per 100 patient-years. AE rates per 100 patient-years were similar in patients ≥ 30 kg (501.9 [RoActemra/Actemra 8 mg/kg]) and patients < 30 kg (445.6 [RoActemra/Actemra 10 mg/kg] and 471.9 [RoActemra/Actemra 8 mg/kg]). Most AEs were mild (493 AEs) or moderate (185 AEs). AEs occurring at least once in ≥ 5% of patients were juvenile arthritis (which were underlying disease flares reported as AEs), nasopharyngitis, headache, upper respiratory tract infection, cough, pharyngitis, nausea, diarrhea, rhinitis, vomiting, abdominal pain, oropharyngeal pain, and rash.
• There were no deaths during the study. The rate of SAEs in the all exposure population was 12.5 per 100 patient-years. Given the low number of SAEs there were no determinable differences in the rates of SAEs between treatment groups. The SOC that had the most patients reporting at least 1 SAE in the all exposure population was infections and infestations (9 patients [4.8%]), followed by injury, poisoning and procedural complications (3 patients [1.6%]). The majority of SAEs were individual cases occurring in individual patients with the exceptions of pneumonia, bronchitis and cellulitis. Pneumonia was reported in 4 patients (3 patients receiving RoActemra/Actemra 8 mg/kg [≥ 30 kg] and 1 patient receiving RoActemra/Actemra 10 mg/kg [< 30 kg]). Bronchitis was reported in 2 patients receiving RoActemra/Actemra 10 mg/kg (< 30 kg) and cellulitis was reported in 2 patients receiving RoActemra/Actemra 8 mg/kg (≥ 30 kg).
• The rate of infections in the RoActemra/Actemra all exposure population was 163.7 per 100 patient years. The most common events observed were nasopharyngitis and upper respiratory tract infections.  The rate of serious infections was numerically higher in patients weighing <30 kg treated with 10 mg/kg RoActemra/Actemra (12.2 per 100 patient years) compared to patients weighing ≥30 kg, treated with 8 mg/kg RoActemra/Actemra (4.0 per 100 patient years). The incidence of infections leading to dose interruptions was also numerically higher in patients weighing <30 kg treated with 10 mg/kg RoActemra/Actemra (21.4%) compared to patients weighing ≥30 kg, treated with 8 mg/kg RoActemra/Actemra (7.6%).
• Seven patients were withdrawn from the study due to an AE. This included 1 patient who was withdrawn following an event reported as juvenile arthritis, which indicated that the reason to discontinue treatment in this patient was insufficient efficacy. Since the calculation of the rates of AEs per 100 patient-years only included events reported during RoActemra/Actemra treatment, 1 patient who was withdrawn from treatment due to gastroenteritis while receiving placebo in Part II was excluded from the rate of AEs leading to withdrawal (3.3 events per 100 patient-years [6 events in 6 patients]). The rate of AEs leading to withdrawal was low and similar across all treatment groups.
• Patients with baseline oral corticosteroid use had a similar rate of AEs to those not receiving corticosteroids (440.4 and 510.9 per 100 patient-years, respectively).
• The majority of patients were taking methotrexate at baseline (148/188 patients), so comparison of AE rates across baseline methotrexate use is of limited value due to the low number of patients not taking methotrexate at baseline (40/188 patients). However the observed data suggested a similar AE incidence rate across the 2 groups of methotrexate use (473.0 AEs per 100 patient-years for methotrexate users, compared to 505.7 AEs per 100 patient-years for patients not taking methotrexate at baseline). A similar trend was observed across the most frequently reported SOCs.
• There were no anaphylactic reactions during the study. Infusion-related AEs were defined as those reported during infusion or within 24 hours following infusion, and occurred at a rate of 23.3 per 100 patient-years, with the most frequently reported events being nausea, dizziness, headache, and hypotension.
• During study treatment, mean neutrophil counts decreased in most subgroups of patients. Six patients (3.2%) experienced CTC Grade 3 decreases in neutrophils and 1 patient (0.5%) experienced a CTC Grade 4 decrease in neutrophils. No AEs of neutropenia were reported.
• The majority of patients maintained a reduced but normal platelet count (172 patients [91.5%]) throughout study treatment. CTC Grades 2, 3 and 4 decreases in platelets were observed in 1 patient (0.5%) each. There was 1 AE of thrombocytopenia and 1 event of platelet count decreased reported; both were non-serious and no concurrent bleeding events were reported. There were no serious bleeding AEs reported.
• ALT and AST values remained within the normal range in the majority of patients throughout study treatment. CTC Grades 2 and 3 ALT elevations were observed in 4.3% and 0.5% of patients, respectively. CTC Grade 2 AST elevations were observed in 2 patients. In addition, a CTC Grade 3 AST elevation was seen in 1 patient. No patients had elevations in liver enzymes or hepatic events indicative of Hy’s law.
• Mean values of total, direct, and indirect bilirubin concentrations remained within normal ranges throughout study treatment.
• In general, mean lipid parameters including total cholesterol, low density lipoprotein (LDL)-cholesterol, high density lipoprotein (HDL)-cholesterol, and triglyceride concentrations were within normal ranges for the duration of the study; 34.6% and 11.4% experienced a post-baseline elevation of their total cholesterol value to ≥ 170 mg/dL and LDL-cholesterol value to ≥ 110 mg/dL at a time during study treatment.

Conclusions:

The primary efficacy endpoint of the study was met as there was a clinically meaningful and a statistically significant difference in JIA ACR30 flare rate between Week 16 and Week 40 in patients receiving placebo and those receiving RoActemra/Actemra. Overall, RoActemra/Actemra improved or maintained JIA ACR responses to Week 40 in patients with pJIA.

The long-term use of RoActemra/Actemra was well tolerated and there were no additional safety concerns noted in this pJIA patient population. The AEs observed were consistent with the known safety profile for RoActemra/Actemra. RoActemra/Actemra, at a dose of 10 mg/kg IV every 4 weeks in patients weighing < 30 kg and 8 mg/kg IV every 4 weeks in patients weighing ≥ 30 kg, was effective for the treatment of pJIA, with an acceptable safety profile.

Date of report:

01.04.2012

Click here for the protocol registry listing of this trial.