Tocilizumab in refractory adult Still's disease




There is an unmet need for the treatment of adult Still's disease (ASD), the pathogenesis of which may involve interleukin-6 (IL-6). We report the first series of patients with ASD treated with tocilizumab (TCZ), a humanized anti–IL-6 receptor antibody.


All ASD patients treated with TCZ in France between July 2006 and July 2009 after failure to all available therapies were included in this cohort study. The main outcome measures were the European League Against Rheumatism (EULAR) improvement criteria and resolution of systemic symptoms at the 3- and 6-month followup periods.


Fourteen patients with refractory ASD were included. At the start of TCZ treatment, despite a mean prednisone dosage of 23.3 mg/day, based on a 28-joint count, mean tender joints were 10.5, mean swollen joints were 7.9, and the mean Disease Activity Score in 28 joints was 5.61. Recurrent systemic involvement, including fever and rash, was present in 7 patients. TCZ was administered at 5–8 mg/kg every 2 or 4 weeks (8 mg/kg/month, n = 9). Eleven patients successfully completed the 6-month study; 1 withdrew due to necrotizing angiodermatitis, another due to chest pain at each TCZ infusion, and a third due to systemic flare. A good EULAR response was observed in 64% of patients (9 of 14) at 3 months and EULAR remission was observed in 57% (8 of 14) at 6 months. Systemic symptoms were resolved in 86% of patients (6 of 7). Moreover, corticosteroid dose was reduced by 56%. No other severe adverse effects occurred.


TCZ is a promising new treatment for ASD.


Adult Still's disease (ASD) is a systemic disorder characterized by high spiking fever, evanescent rash, arthritis, leukocytosis, and increased concentrations of acute-phase reactants, including serum ferritin level. Arthritis occurs in nearly 90% of patients. Up to 50% of patients may develop progressive and destructive arthritis (1). In many patients, the arthritis is refractory to therapies, including corticosteroids and methotrexate. Recently, small series and modest-scale retrospective studies have shown improvement in some patients using tumor necrosis factor α (TNFα) antagonists (2, 3) or anakinra (4, 5). Nevertheless, efficacy of these biotherapy approaches is inconstant (2–5), and severe side effects have been reported (3, 5).

The pathogenesis of ASD remains unclear. However, cytokine-mediated inflammation may be responsible for the features of the disease, and a role of interleukin-6 (IL-6) has been suggested (6–9). A few isolated reports have described some degree of improvement after treating patients with ASD with the anti–IL-6 receptor antibody tocilizumab (TCZ) (10–15). We describe a cohort of patients whose ASD, refractory to all therapies, was treated with TCZ.

Patients and Methods

Patient selection.

To be eligible to receive off-label TCZ for ASD in France between July 2006 and July 2009, patients had to be enrolled in a French Agency for the Safety of Health Products protocol. All of the patients enrolled in this protocol were included in the study. The study population consisted of patients who fulfilled the Yamaguchi criteria for ASD (16). Patients had to be at least 18 years of age, although disease onset may have occurred during childhood. To be eligible for the protocol, patients must have had persistent active arthritis and/or systemic involvement resistant to therapies, including corticosteroids, methotrexate, anakinra, and anti-TNFα drugs. Delay before institution of TCZ treatment was determined by the physician responsible for care. Because it was anticipated that affected patients may show substantial disability due to recurrent flares of systemic features and/or arthritis despite high-dose corticosteroids, we believe that it would not have been appropriate to require a washout period. Informed consent was obtained in all cases before starting treatment.

Assessment of TCZ efficacy and tolerance.

Data were collected prospectively from physicians in charge of the patients. Physicians were asked to fill in a standardized questionnaire sent online with the support of the Club Rhumatismes et Inflammation (online at All of the patients receiving at least one infusion of TCZ were evaluated. Data were analyzed for the baseline assessment visit and at 3 and 6 months. The Disease Activity Score in 28 joints (DAS28) was used to assess arthritis activity (17). Clinical improvement in arthritis activity was evaluated using the European League Against Rheumatism (EULAR) improvement criteria (17). Improvement in systemic features was defined as the resolution of systemic symptoms. Routine laboratory indicators of disease activity, including hematology profile, erythrocyte sedimentation rate (ESR), C-reactive protein (CRP) level, and ferritin serum levels, were collected. Adverse events were recorded in a specific questionnaire.


Characteristics of ASD.

Fourteen patients with intractable refractory ASD were included in the cohort. Their clinical characteristics are shown in Table 1. All of the patients had chronic arthritis; radiologic examination showed irreversible joint damage in 8 patients. All of the patients had experienced failure or intolerance with methotrexate and anakinra and 12 with at least one anti-TNF drug. At baseline, all of the patients were receiving prednisone at a mean dosage of 23.3 mg/day (median 15.5, range 5–80) (Table 2). Based on a 28-joint count, mean tender joints were 10.5, mean swollen joints were 7.9, and the mean DAS28 was 5.61. At baseline, in addition to arthritis, recurrent systemic involvement, including fever and rash, was present in 7 patients. The median ESR was 36.5 mm/hour and the median CRP level was 5.2 mg/dl (Table 2).

Table 1. Clinical characteristics of the 14 patients with adult Still's disease treated with tocilizumab*
  • *

    Values are the number (percentage) unless otherwise indicated. MTX = methotrexate; anti-TNFα = anti–tumor necrosis factor α; IVIG = intravenous immunoglobulin.

Women9 (64)
Age, mean (range) years38.4 (23–68)
Still's disease onset 
 Childhood onset5 (36)
 Adult onset9 (64)
Disease duration, mean (range) years13.6 (3–27)
Chronic arthritis14 (100)
Destructive arthritis8 (57)
Past history of recurrent systemic flares11 (79)
 Ever used14 (100)
 Ongoing8 (57)
Anakinra, ever used14 (100)
Anti-TNFα drugs, ever used12 (86)
 3 previous4 (29)
 2 previous6 (43)
 1 previous2 (14)
Abatacept, rituximab, or IVIG ever used7 (50)
Long-term corticosteroids treatment14 (100)
Table 2. Outcome measures for patients with adult Still's disease treated with tocilizumab*
 Baseline assessment (n = 14)3-month evaluation (n = 14)6-month evaluation (n = 14)
  • *

    Values are the mean ± SD unless otherwise indicated. VAS = visual analog scale; DAS28 = Disease Activity Score in 28 joints; ESR = erythrocyte sedimentation rate; CRP = C-reactive protein.

  • Including 1 patient who withdrew from the study at 2 months.

  • Including 3 patients who withdrew before the 6-month end point.

  • §

    Tender and swollen joint counts on the 28 articles included in the DAS28 (17).

Recurrent systemic flares, no. (%)7 (50)1 (7)1 (7)
Tender joint count§10.5 ± 7.44.8 ± 6.73.9 ± 5.9
Swollen joint count§7.9 ± 6.84.4 ± 4.42.9 ± 3.6
Patient global assessment by 100-mm VAS61.9 ± 26.933.1 ± 23.224.6 ± 25.9
DAS285.61 ± 1.493.21 ± 1.762.92 ± 2.17
Prednisone dosage, mg/day23.3 ± 20.913.0 ± 6.510.3 ± 4.9
ESR, median (range) mm/hour36.5 (2–120)8.0 (1–98)12.5 (1–91)
CRP level, median (range) mg/dl5.2 (0.3–27.2)0.5 (0.1–14.7)0.6 (0.02–23.0)
Leukocyte count, gm/liter10.65 ± 5.569.80 ± 6.009.03 ± 3.01
Serum ferritin level, ng/ml1,939 ± 4,685132 ± 105209 ± 144

Efficacy of TCZ.

TCZ was administered at 8 mg/kg every 4 weeks in 9 patients, the registered dosage in rheumatoid arthritis. In 4 patients, TCZ was started at 8 mg/kg every 2 weeks, 1 of whom was then given injections at 3-week intervals. The remaining patient received 5 mg/kg infusions monthly. Eight patients were treated in combination with methotrexate, 2 with leflunomide, and 4 with TCZ as a monotherapy. Eleven patients successfully completed the 6-month study; 2 withdrew due to side effects and a third due to systemic flare. The mean DAS28 dropped from 5.61 to 3.21 and 2.91 at the 3- and 6-month followup visits, respectively, including data from the 3 of 14 patients who withdrew from the study (Figure 1). A good EULAR response was achieved in 64% of patients (9 of 14) at 3 and 6 months. EULAR remission (DAS28 <2.6) was achieved in 36% of patients (5 of 14) at 3 months and in 57% (8 of 14) at 6 months. Improvement of all disease activity scores was observed. At 6 months, there was a 60% improvement in the number of tender joints, the number of swollen joints, and the mean visual assessment score for patient global health. Resolution of systemic symptoms, including fever and eruption, was observed for 86% of patients (6 of 7) at 3 and 6 months. Moreover, the prednisone dosage was reduced to a mean of 13.0 mg/day (median 13.5, range 0–25) at 3 months and to a mean of 10.3 mg/day (median 11.0, range 0–20) at 6 months, including data from the 3 of 14 patients who withdrew from the study. Seven patients had a corticosteroid dosage of 10 mg/day or less at 6 months. No correlation was found between the TCZ dose and the achievement of arthritis or systemic remission or with the decrease of corticosteroids dose.

Figure 1.

Individual disease activity score and mean prednisone dose before and after tocilizumab therapy in adult Still's disease. Each line shows the progression of the individual disease activity score (Disease Activity Score in 28 joints [DAS28]) (17). Solid lines show the patients still undergoing treatment with tocilizumab. Broken lines correspond to the patients who stopped therapy before the 6-month end point. The threshold of 2.6 indicated corresponds to the limit below which the patient is considered to be in remission. Bars show the mean prednisone dosage at the different end points of the study.

Adverse events.

A 56-year-old woman with high blood pressure and diabetes mellitus had a 9-year history of ASD with persistent active arthritis despite successive treatment with anakinra, infliximab, etanercept, and rituximab associated with methotrexate and high-dose corticosteroids. After 3 infusions of TCZ, arthritis improved, but necrotizing angiodermatitis was observed, leading to discontinuation of treatment. The cutaneous outcome was favorable despite arthritis flare after TCZ withdrawal. A 30-year-old woman had severe disabling polyarthritis for 3 years due to ASD, which had been complicated with macrophage activation syndrome. Before TCZ, she had received successively but without success anakinra, etanercept, adalimumab, and abatacept in addition to methotrexate and corticosteroids. At each TCZ infusion, she experienced chest pain and chills, leading to discontinuation of treatment at month 2. A third patient age 68 years was treated with TCZ for ASD of a duration of 7 years. At the initiation of TCZ, he had ongoing episodes of high fever lasting 3 days every 2 weeks despite unsuccessful treatment with anakinra in association with leflunomide and corticosteroids. TCZ was initiated but stopped after 3 months due to increased arthralgia and CRP values. Mild hyperlipidemia was observed in one patient and increased alanine aminotransferase levels (2 times the upper normal range) were observed in another patient. No infection was observed.


In this cohort consisting of all ASD patients treated with TCZ in France during the study period, for whom all other available therapeutic options, including methotrexate, anakinra, and anti-TNFα drugs, had previously failed, we observed rapid resolution of systemic manifestations and arthritis in most cases.

First-line treatment for ASD includes nonsteroidal antiinflammatory drugs, but long-term corticosteroids are often necessary. Many alternative medications have been tried to provide a steroid-sparing effect, such as methotrexate, thalidomide, intravenous immunoglobulin, azathioprine, anakinra, and anti-TNFα drugs. These drugs have variable effects on signs and symptoms and severe side effects have been reported repeatedly. None of these drugs have been evaluated in a controlled study.

The basis of using the anti–IL-6 receptor antibody TCZ in ASD stems from our understanding of the role of IL-6 in rheumatoid arthritis and in this disease. IL-6 is a pleiotropic cytokine and acts as a B cell differentiation factor, induces T cell growth and cytotoxic T cell differentiation, increases proliferation of multipotential hemopoietic progenitors, induces terminal macrophage differentiation, and stimulates hepatocytes to produce acute-phase proteins such as CRP (18). IL-6 transgenic mice have hypergammaglobulinemia, thrombocytosis, infiltration of inflammatory cells into the tissues, and splenomegaly and lymphadenopathy (18). Overproduction of IL-6 may account for the major symptoms of ASD, inducing fever, leukocytosis, thrombocytosis, acute-phase reactant production, and bone resorption. Furthermore, high levels of IL-6 have been reported in the blood of patients with ASD and have been shown to correlate with disease activity (6–9).

Only a few case reports on TCZ in ASD have been published showing promising preliminary results (10–15). A randomized placebo-controlled trial demonstrated efficacy of TCZ in systemic juvenile idiopathic arthritis, which shows some similarities with ASD (19). In our study, the infusion schemes varied in terms of TCZ dose or infusion interval because no data on TCZ therapy were available in ASD and no recommendation could be done at the institution of the cohort. The majority of the patients had TCZ treatment according to the therapeutic sequence now registered in rheumatoid arthritis (8 mg/kg monthly).

In most of our patients with refractory disease, response to TCZ treatment was rapid and sustained. The effectiveness of treatment was assessed by treatment responses (a good EULAR response and resolution of systemic symptoms) and by the proportion of patients achieving arthritis remission (EULAR remission). The former reflects the magnitude of change in disease activity, and the latter reflects whether disease activity is suppressed below a certain threshold. The good EULAR response observed in two-thirds of patients at 3 months was due to improvement of all disease activity scores. The high observed rate (57%) of arthritis remission at 6 months is of particular interest in these patients. Of the 7 patients with active systemic features at the start of TCZ treatment, all but one showed resolution of these symptoms. Notably, we observed these improvements despite a 56% mean reduction in corticosteroid dose after starting treatment. We believe such a reduction to be clinically relevant. Therefore, TCZ had a corticosteroid-sparing effect. This is particularly important in these patients who have a high cumulative dose of corticosteroids and a high prevalence of side effects. The observed side effects were acceptable and did not obviously differ from those observed in patients treated with TCZ or with TNF antagonists for rheumatoid arthritis.

Although this is an observational and uncontrolled study, the strength of the study is the collection of data from all of the ASD patients in France who received at least one TCZ infusion during a 3-year period. Patients were followed according to good clinical practice and were enrolled in a specific off-label protocol. The population-based nature of this cohort, the number of patients treated, as well as the use of well-defined indices of clinical response that capture both systemic and musculoskeletal features of ASD, add strength to the study. Moreover, treatment of this orphan disease is often challenging and the observed reduction of DAS28 and rates of remission were impressive. Furthermore, to our knowledge, this is the first series of patients being treated after failure of methotrexate, anakinra, and TNF inhibitors and the first series to evaluate this new treatment option in ASD patients. Although the optimal therapeutic sequence of TCZ for patients with refractory ASD remains to be determined, we propose that initial monthly infusions of 8 mg/kg be started in these patients. Evaluation should be performed at 3 and 6 months using markers that capture both systemic and musculoskeletal features of ASD such as those used in this cohort.

In summary, findings from this cohort demonstrate that as an IL-6–inhibiting therapeutic approach, TCZ was effective against systemic involvement of the disease in almost all of the patients, and led to arthritis remission in half of the patients, showing a marked corticosteroid- sparing effect and an acceptable tolerance profile. Given the limited treatment options and the potential benefit of an IL-6–inhibiting strategy in this disease, the use of TCZ should be further investigated in a multicenter randomized controlled trial. A French nationwide registry is currently being implemented to collect additional information on the long-term efficacy and tolerance of TCZ in ASD in clinical practice.


All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Puéchal had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Study conception and design. Puéchal, de Bandt, Sibilia.

Acquisition of data. Puéchal, de Bandt, Berthelot, Breban, Dubost, Fain, Kahn, Lequen, Longy-Boursier, Perdriger, Schaeverbeke, Toussirot, Sibilia.

Analysis and interpretation of data. Puéchal, de Bandt, Sibilia.


The authors thank Philippe Ravaud, MD, PhD, Paris Descartes University, Paris, France, for critical reading of the manuscript.