Efficacy and Safety of Tocilizumab in a Real-Life Observational Cohort of Patients With Polyarticular Juvenile Idiopathic Arthritis

Minna-Maija Grönlund; Terhi Remes-Pakarinen; Liisa Kröger; Kati Markula-Patjas; Maria Backström; Anne Putto-Laurila; Kristiina Aalto; Paula Vähäsalo


Rheumatology. 2020;59(4):732-741. 

In This Article


Study Design and Patients

All consecutive pJIA patients aged 2–17 years attending one of the existing five tertiary university hospitals in Finland or one of the secondary central hospitals (Vaasa Central Hospital) for whom the treating physician had decided to commence tocilizumab after July 2013 were included in a multicentre retrospective chart review. In addition, the electronic JIA databases of each hospital were used to collect the JIA core outcome variables (see below). Since tocilizumab was a new type of biologic treatment for pJIA it is very unlikely that many pJIA patients would have been treated in the other secondary central hospitals not included in this study.

This study yielded 62 JIA patients with seropositive or seronegative polyarthritis or extended oligoarthritis (i.e. pJIA) according to the ILAR classification criteria.[1] Patients had started i.v. tocilizumab between October 2013 and September 2017. Only patients starting tocilizumab for the first time according to the official dose instructions (dose 8 mg/kg every 4 weeks in patients weighing ≥30 kg and 10 mg/kg every 4 weeks in patients weighing <30 kg) were included. The dose could be rounded to the closest practical amount but no more than 15% deviation from the official dose was allowed. Six patients were excluded from the study because of the dose deviation, leaving a final study population of 56 patients.

Assessment and Outcome Parameters

Data at clinical visits closest to 0, 3, 6, 12, 18 and 24 months after starting tocilizumab were included. The follow-up ended at the last available visit (patients continuing tocilizumab) or the last recorded visit closest to the discontinuation of tocilizumab. Median follow-up was 23.6 months [interquartile range (IQR) 19.0–25.3] for those continuing the treatment and 6.7 months (IQR 5.8–16.8) for those who discontinued the treatment.

Demographic and clinical characteristics of the patients collected at initiation of tocilizumab treatment included: age at diagnosis, disease duration, JIA category, prevalence of anterior uveitis and previous treatment of JIA (sDMARDs, bDMARDs, systemic and intra-articular glucocorticoids).

Treatment patterns of tocilizumab (dose and dose interval, reasons for dose modifications, proportion of patients discontinuing tocilizumab and reasons for discontinuing, proportion of patients on tocilizumab monotherapy at entry and during follow-up), concomitant sDMARD therapy and use of intra-articular glucocorticoids were recorded.

The JIA core outcome variables (number of joints with active arthritis, joints with limited range of motion, ESR, CRP, physician's global assessment of disease activity and patient's/parent's global assessment of well-being, both on 100-mm visual analogue scale (VAS), the Childhood Health Assessment Questionnaire-Disability Index (CHAQ-DI), range 0–3, 0 = no disability) were recorded.

Efficacy of the treatment was evaluated by drug survival and 10-joint Juvenile Arthritis Disease Activity Score (JADAS-10).[17] The new Finnish JADAS-10 cut-off values of low disease activity (LDA) (≤3.9) and inactive disease (≤0.7) were used.[18] In addition, the number of patients with clinically inactive disease was determined by Wallace's preliminary criteria, since they take into account the activity of uveitis.[19]

Assessment of safety included rates of adverse events (AEs) and serious adverse events (SAEs). AEs are defined as any untoward medical occurrence in a patient administered a pharmaceutical product, even without a causal relationship with the treatment. SAEs are defined as events that are fatal or immediately life-threatening, or require inpatient care or prolong hospitalization, or result in persistent or significant disability/incapacity, congenital anomaly, or spontaneous or elective abortion, or require medical or surgical intervention to prevent a serious outcome. Rates of abnormal laboratory results in total blood count, neutrophil count, transaminases, cholesterol, low-density lipoprotein, high-density lipoprotein and triglyceride values found in routine laboratory tests were recorded. Abnormal laboratory results were defined as values leading to adjustment in tocilizumab treatment or concomitant medication. The cut-off values for neutrophils and transaminases (alanine aminotransferase) were ≤1.0 ×109/l and ≥70–100 U/l, respectively. Injuries and flares of JIA were not reported as AEs. AEs were coded according to the Medical Dictionary for Regulatory Activities, version 21.1 (https://www.meddra.org).


The parents were given verbal and written information about the study and they provided written informed consent. Permission for the study was received from the Ethics Committee of the Hospital District of South-West Finland and the ethical review board of the paediatric departments of each hospital.


Data are descriptive and mostly summary statistics are presented as mean and S.D., median and range, or IQR as relevant. The Mann–Whitney U test was used to compare JADAS values in patients with monotherapy and combination therapy at 12 months of follow-up. The comparison was done only between those patients who had been on monotherapy or combination therapy for the whole 12-month period. Friedman's test was used to analyse the time effect between all of the follow-up points on continuous variables and Cochrane's Q test for categorical variables. No post hoc analysis was done. SPSS version 25 (IBM, Armonk, NY, USA) was used for the analysis.

The numbers of patients included at different time points were: commencement of tocilizumab, 56 patients; 3 months, 54 patients (2 patients had no clinical visit); 6 months, 55 patients (1 patient discontinued earlier); 12 months, 47 patients (9 discontinued earlier); 18 months, 39 patients (10 discontinued earlier, 7 followed up for shorter duration); and 24 months, 29 patients (16 discontinued earlier, 11 followed up for shorter duration). In addition, due to missing information on core set variables, JADAS was missing in three to eight patients per visit (8–15% of patients included at each time point). We compared the baseline characteristics included in Table 1 between patients with complete JADAS values (n = 43) and those with some missing JADAS values at one or more clinical visits (n = 13) using the Mann–Whitney U test and χ 2 test or Fisher's Exact test, as relevant. No statistically significant differences were found in the baseline characteristics between these two groups.