Background/Purpose:Systemic juvenile idiopathic arthritis (SJIA) is one of the most striking forms of juvenile arthritis, required biologics due to failure of corticosteroids (CS) and DMARDs. In Russia until March 2013 no Il‐1 blockers were available. Since tocilizumab (TCZ) was approved in adults with rheumatoid arthritis we used it for treatment of SJIA.The aim of our study was to evaluate outcomes and to find possible criteria, related with different treatment TCZ protocols and achievement TCZ‐off remission.Methods:our retrospective study was included 33 active SJIA children who fall CS, methotrexate (MTX), cyclosporine A (CsA) and their combination. We used TCZ in 2 branches in standard doses: every 2 (Q2W) or every 4 weeks (Q4W). The randomization was based on efficacy which was evaluated in 14th day. If patient had no signs of TCZ inefficacy during next 2 weeks (days 15–29) the patient was referred to Q4W group (n=24), if any clinical or laboratorial signs of inefficacy during first 4 weeks were occurred the patient was treated Q2W (n=9). The protocol of this trial was approved by local Ethic Committee of our University. The term “efficacy” means at least ACR70 improvement and absence of systemic features.Results:The main demographic parameters (Me;IQR) included the age‐10.2 (6.0–12.75) years and delay of TCZ‐36.0 (11.2–97.0) months. Treatment before TCZ were CS‐31 (93.9%), MTX‐29 (87.9%), CsA‐16 (48.5%) and their combination. The macrophage activation syndrome (MAS) before TCZ was in 8 (24.2%). During the trial CS successfully discontinued 19/31 (61.3), CsA 9/16 (56.3%), MTX 8/29 (27.6%) patients. In 5 children TCZ was discontinued due to stable remission with median duration 640 days (3/5 remission off‐medication, 2/5 still on MTX). After TCZ initiation 4 children have experienced MAS, but all of them had MAS before TCZ, so no “new cases” were observed on TCZ. 3 children early withdrew during the trial due to adverse events (infusion reaction, MAS) and 1 child died (severe uncontrolled MAS). Patients which were treated every 4 weeks had milder SJIA course compare with Q2W (table ). Factors, related with milder SJIA course and allowed to use Q4W TCZ were: Hb>10.3 g/dl (p=0.00001), WBC ≤12.6*109/l (p = 0.013), granulocytes ≤8556 cells/μl (p = 0.00037), CRP ≤82.2 mg/l (p = 0.002), ESR ≤26 mm/h (p = 0.02), ferritin ≤605 mg/ml (p = 0.0001) before start of TCZ and granulocytes in 1 week after 1st TCZ ≤8142 cells/μl (p = 0.014). Criteria, associated with TCZ‐off remission were: absence of knee (p = 0.045) and cervical spine involvement (p = 0.03), number of active joints ≤12 (p = 0.05), ESR ≤40 mm/h (p = 0.05), total protein ≤6.6 g/dl (p = 0.04) at the moment of 1st TCZ infusion and granulocytes in 2 week ≤2907/μl (p = 0.038), WBC in 4 week ≤6300/μl (p = 0.026) after 1st TCZ. Parameters Q2W (n = 9) Q4W (n = 24) p Hemoglobin, g/dl 9.8 (9.6–10.3) 11,5 (10,9–13,1) 0,003 Anemia, n (%) 8 (88.9) 10 (41,7) 0,02 WBC,109/l 17.2 (13.6–20.8) 9,5 (7,65–13,2) 0,016 Granulocytes, cells in 1 μl 13728 (11424–18382) 6318 (4822–8184) 0,005 Granulocytes in 1 week, cells 1 μl 8944 (6560–12375) 3314 (1840–7240) 0,015 CRP, mg/l 84.5 (20.6–104.8) 17.6 (6.7–74.5) 0.025 ESR, mm/h 45.0 (42.0–61.0) 25.0 (10.5–47.5) 0.016 Ferritin, mg/ml 858.0 (326.0–1859.0) 128.0 (48.5–238.0) 0.004 LDH, U/l 714.0 (655.5–792.5) 464.0 (360.0–513.0) 0.02 Total protein, g/dl 6.6 (6.2–6.9) 7.1 (6.8–7.7) 0.003 Albumin, g/dl 2.1 (1.7–2.6) 3.1 (2.9–3.3) 0.002 Hepatomegaly 8 (88.9) 9 (37.5) 0.017 Coagulopathy 3 (33.3) 0 (0.0) 0.015 Interstitial lung disease 4 (44.4) 1 (4.2) 0.013 CNS dysfunction 5 (55.6) 0 (0.0) 0.001 Knee arthritis 2.0 (2.0–2.0) 1.5 (0.0–2.0) 0.02 Knee arthritis, n (%) 9 (100.0) 14 (58.3) 0.03 MAS before TCZ 4 (44.4) 4 (16.7) 0.17 MAS during TCZ 4 (44.4) 0 (0.0) 0.003 TCZ efficacy, n (%) 6 (66.7) 23 (95.8) 0.05 SJIA relapses 4/9 (44.4) 1 (4.2) 0.013 TCZ discontinuation. due to: 3 (100.0) 6 (100.0) 0.038 ‐remission 0/3 (0.0) 5/6 (83.3) ‐infusion reaction 0/0 (0.0) 1/6 (16.7) ‐MAS + infusion reaction 2/3 (66.7) 0/6 (0.0) ‐MAS → death 1/3 (33.3) 0/6 (0.0) Me (IQR), Fisher's exact testConclusion:We found clinical and laboratorial criteria for mild/moderate SJIA allowed to use TCZ Q4W and provisional criteria, related to achievement TCZ‐off remission.