Introduction

Adult-onset Still’s disease (AOSD) is an inflammatory disease of unknown etiology which commonly affects young adults [1]. AOSD is a rare condition, interesting at all ages, and its expression and progression are highly variable [2]. Usually, it is characterized by high spiking fever, arthritis, and an evanescent, macular, and salmon-colored rash, appearing on the trunk and the extremities. Although some patients have a unique flare without any recurrence, most of the patients experience several episodes which progress to a chronic form [3, 4].

Two chronic forms are described in literature: a systemic form with fever, asthenia, and other general symptoms, and an arthritic form with polyarthritis which could be erosive in up to 50 % of patients [1–4].

AOSD pathogenesis is still poorly understood and controversial.

The immunological signature represents a key feature in AOSD pathogenesis. High levels of pro-inflammatory cytokines such as interleukin-1 (IL-1), IL-6, IL-18, tumor necrosis factor-α (TNF-α), and interferon-γ were found to be significantly increased in patients with active AOSD when compared to normal subjects [5]. It might be the result of a global reactive syndrome with activation of macrophage, natural killer cells, and B lymphocytes leading to a predominant cell-mediated immune response [5]. A predominance of Th1 cytokines was described in the peripheral blood and tissues of patients with active untreated AOSD.

To date, no specific laboratory tests are available for AOSD. Clinical diagnosis could be performed excluding important mimickers such as infections, malignancies, and others systemic autoimmune disorders.

The most widely accepted diagnostic criteria have been presented by Yamaguchi and colleagues including major and minor criteria [6]. More recently, Fautrel and colleagues proposed classification criteria that include diagnostic markers such as serum ferritin and glycosylated ferritin [7].

Treatment of AOSD patients is largely empiric and based on the results of case reports or small case series due to rarity of this disease. It includes nonsteroidal anti-inflammatory drugs, corticosteroids, and disease-modifying antirheumatic drugs (DMARDs) such as methotrexate (MTX) for severe and potentially disabling manifestations [8].

In this setting, commercially available biologic agents for rheumatoid arthritis (RA) have been employed in AOSD with variable success [9]. Recently, small series and retrospective studies have shown improvement in some patients with AOSD following biologic therapy with TNF-α antagonists [10, 11] or anakinra, an antagonist of IL-1 [12, 13]. Nevertheless, efficacy of these biologic agents is inconstant and severe side effects have also been reported [9–13].

IL-6 blockade might be an attractive therapeutic option for AOSD because this cytokine contributes to the pathogenesis of major AOSD symptoms including fever, leukocytosis, acute-phase protein production and bone resorption [14]. Tocilizumab (TCZ) is a humanized anti-IL-6 receptor antibody that recognizes both the membrane-bound and the soluble form of IL-6 receptor and specifically blocks IL-6 mediated pathways. The results obtained in short series with limited follow-up have been already published [14–19].

Furthermore, growing evidences described the efficacy and safety of TCZ in children with systemic juvenile idiopathic arthritis (SJIA), which is considered to share many common pathogenetic pathways with AOSD. SJIA is characterized by chronic arthritis and systemic manifestations such as spiking fever, rash, hepatosplenomegaly, lymphadenopathy, serositis, and elevated inflammatory markers [20]. IL-6 levels are particularly higher in this juvenile disorder correlating with disease activity. Thus, IL-6 is a rational target for the treatment of SJIA and TCZ is approved since 2008 in Japan and 2011 by the US Food and Drug Administration and European Medicine Agency for kids with SJIA [21, 22].

In this work, we reported our experience concerning both the safety and the efficacy of the therapy with TCZ in patients with AOSD refractory to corticosteroids and MTX.

Patients and methods

Eleven consecutive patients with AOSD referring to the Rheumatology Clinic at University of L’Aquila between January 2009 and January 2012 were treated with TCZ. The treatment was approved by the local ethics committee, and written informed consent was obtained from all patients according to the Declaration of Helsinki. All patients fulfilled the criteria proposed by Yamaguchi for AOSD [14]. All patients underwent chest X-ray and serological analyses to exclude B and C viral hepatitis, latent TBC, and/or other severe infectious diseases prior the biological treatment.

Patients showing more than six tender and/or swollen joints and/or systemic involvement such as fever and/or loss of body weight more than 10 % in the last 3 months and/or inflammatory anemia not responding to conventional therapies for 3 months or evidence of drug toxicity received TCZ (8 mg/kg every 4 weeks) for 12 months and successively underwent to 6 months of follow-up after discontinuation of the treatment. Due to the lack of validated measures for joint involvement in AOSD, we assessed the joint involvement of our patients by Disease Activity Score 28 and EULAR improvement criteria [23], as reported in other series [10, 11, 19]. Pain was evaluated using a 100-mm visual analogue scale. The number of active joints was defined by the presence of swelling or restriction of motion accompanied by pain or tenderness or both. Clinical records were collected at the baseline and at 3, 6, and 12 months of active treatment and at 18 months [26].

The improvement of systemic symptoms was considered the primary outcome in our cohort. In this regard, the presence or absence of fever and patient global health using a 100-mm visual analogue scale (VAS) were evaluated.

Furthermore, routine laboratory tests including erythrocyte sedimentation rate (ESR), C-reactive protein (CRP), ferritin serum levels, and hemoglobin were reported. These values were used as surrogate markers of IL-6 function to assess the response to the treatment. Adverse events were recorded in a specific form.

After 12 months of active treatment, patients discontinued biologic treatment and they were followed up to the next 6 months to evaluate the activity of the disease.

Statistical analysis was performed with GraphPad 5.0 software. Our results are expressed as median (range). One way repeated measures ANOVA and Bonferroni multiple comparison post hoc test were calculated. P values less than 0.05 were considered significant.

Results

Patients’ clinical characteristics are summarized in Table 1. All enrolled patients assumed prednisone and eight out of 11 patients assumed MTX. At radiological examination, seven patients (four in combination therapy and three with prednisone alone) showed joint erosions.

Table 1 Clinical characteristics of the 11patients with adult-onset Still’s disease treated with tocilizumab
Full size table

In all patients, TCZ was administrated at 8 mg/kg every 4 weeks. Eight patients were treated in combination with MTX plus prednisone, and three patients received TCZ plus prednisone without MTX. These later discontinued MTX because of severe hypertransaminasemia and megaloblastic anemia. The median dosage of MTX remains stable during all period study 15 (0–25). Patients’ therapies are reported in Table 2. All patients successfully completed the treatment and the follow-up period until the 18 months.

Table 2 Methotrexate and corticosteroid therapies of the 11patients with adult-onset Still’s disease treated with tocilizumab
Full size table

Median Disease Activity Score in 28 joints (DAS28) decreased from 5.62 (3.75–8.28) at baseline to 2.31 (1.25–6.28) at month 3, 1.88 (0.63–3.57) at month 6, and 1.61 (0.49–3.5) at month 12 (Fig. 1). EULAR remission (DAS28 <2.6) was achieved in 45.45 % (five out of 11) at 3 months. EULAR remission was reached in 63.63 % (seven out of 11) at 6 months and 81.82 % (9 out of 11) at 12 months. A good EULAR response (DAS28 <3.2) was achieved by other patient during 12 months of active treatment.

Fig. 1
figure 1

Clinical activity scores at baseline, at 3, 6, 12 months of treatment with tocilizumab therapy, and after 6 months of follow-up (18 months) in adult Still’s disease patients. a Disease Activity Score in 28 joints (DAS28), b tender joints (Tj) number, c swollen joints (Sj) number, d visual analogue scale (VAS) of patient global health expressed as millimeters (mm). *p < 0.05; **p < 0.01; ***p < 0.001

Full size image

Tender joint and swollen joint count displayed a progressive reduction during active therapy study period (Fig. 1). A statistically significant improvement of joint assessment was observed after 6 and 12 months of treatment when compared to baseline (p < 0.05 for any comparison).

Eight patients had fever at the beginning of biologic treatments. Remission of fever and improvement of systemic symptoms including skin involvement were observed in all patients during the study period. Furthermore, the results of median patient VAS global health were 75 (80–50) mm at baseline, 35 (0–70) mm at 3 months, 20 (10–60) mm at 6 months, and finally, 0 (0–30) mm at 12 months. The improvement in patient VAS global health values was statistically significant (p < 0.005) for any comparison between baseline and each time point during active treatment (Fig. 1).

Prednisone dosage was progressively tapered according to clinician judgment during scheduled visits. The median dosage was 50 (25–100) mg/day at baseline, 12.5 (12.5–25) mg/day at month 3, 6.25 (5–12.5) mg/day at month 6, and 0 (0–12.5) mg/day at month 12 (Table 2) showing an higher corticosteroid-sparing activity of TCZ; in fact, after 12 months, eight out of 11 patients (72.72 %) discontinued corticosteroid therapy.

The ESR, CRP, ferritin, and hemoglobin values of patients in our cohort are summarized in Table 3.

Table 3 Laboratory parameters of 11 patients with adult-onset Still’s disease treated with tocilizumab
Full size table

Concerning TCZ safety profile in AOSD, no severe adverse events including death was observed. Only minor adverse events were observed. Specifically, one patient developed upper respiratory tract infection needing antibiotic therapy and two patients reported injection-site reaction after one infusion of TCZ. In addition, no significant cytopenias were observed throughout the study.

The analysis of clinical conditions after discontinuation of biological therapy at month 18 showed that eight out of 11 patients (72.72 %) were still maintaining clinical remission, and all these patients were still receiving only MTX. The three patients that experienced a flare, showing both systemic and joints symptoms, needed an increasing dose of corticosteroids, and interestingly, they were not assuming MTX for previous intolerance. These patients reported an increase number of tender and swollen joint, the onset of systemic symptoms, and a worsening of patient VAS global health. Furthermore, laboratory tests showed higher serum concentrations of ESR, CRP, and ferritin.

Discussion

Growing evidences on the role of IL-6 in AOSD pathogenesis provide the basis for TCZ therapy in this disease.

In agreement with previous data [14–19], we observed a dramatic improvement of all disease-related clinical and serological findings after TCZ therapy in the totality of AOSD patients which were refractory to combination therapy with MTX and prednisone.

In particular, we observed a decrease of swollen and tender joint counts as well as of DAS28 after TCZ treatment. Furthermore, such improvement was rapid and sustained until 12 months. In addition, disease remission according to EULAR criteria was achieved in 81.82 % of our patients.

Of note, our work was planned to evaluate the TCZ as first line therapy in patients resistant at conventional combination therapy with corticosteroid plus MTX. In fact, several studies has been recently published exploring the possibility to treat AOSD by biologics after the failure of TNF-α blockers or anakinra and employing TCZ as rescue therapy [19, 24]. The different selection of patients among published papers could explain the differences in the results obtained in our study when compared to other published trials. Probably, our patients could be considered less severe than patients enrolled after many biologic therapies, but, in our opinion, this is not a limit of our study but an important tool because of the large amount of patients reaching remission Achieving and maintaining remission are presently considered the most important goal in many rheumatic diseases, and as we showed in this short but homogeneous population, this target may be reached in a large percentage of patients when electively treated by TCZ.

Interestingly, our data are in agreement with results raised from clinical trials performed in patients with RA [25]. In addition, a randomized placebo-controlled trial demonstrated the efficacy and safety of TCZ in children with SJIA. In this study, children with SJIA refractory to conventional treatment were treated with TCZ achieving and maintaining clinical and laboratory remission [21] as observed in our patients.

Taking into account the lack of validated radiological scores for AOSD joint involvement, we did not perform any radiological examination during the study period. However, we could speculate as reported in clinical trial for RA or SJIA that TCZ therapy in AOSD might inhibit progression of joint radiological damage [25, 26].

Furthermore, we observed a reduction of patient VAS global health that might be related in our experimental group to the resolution of systemic symptoms, including fever and skin eruption which was sustained during the entire study period and probably leading to a significant improvement of health-related quality of life. Our results concerning systemic involvement are in agreement with previous studies involving TCZ-treated RA patients [27]. TCZ [34], ESR, CRP, and ferritin values were reduced by TCZ treatment at month 3 and such decrease remained stable during the study period. The decrease of the inflammatory markers paralleled to clinical improvement. It is well known that anemia due to chronic inflammation, neutrophilic leukocytosis, and reactive thrombocytosis are commonly observed during active AOSD [4]. These abnormalities were also documented in our cohort of patients, and TCZ treatment led to a progressive normalization of all parameters. Serum ferritin level correlates with AOSD disease activity, and disease remission is associated to its normalization [28–30]. In our cohort of AOSD patients, we found a strong correlation between ferritin serum values and activity of the disease with a progressive and sustained reduction of these levels until normalization. Furthermore, we observed a significant improvement of hemoglobin levels in our patients. It is well known that the effect of TCZ in increasing the hemoglobin in patients with higher levels of inflammatory cytokines, mainly IL-6, via the inhibition of hepcidin pathway, and this increase could be related with the improvement is some systemic symptom such as fatigue and adynamia [31].

Corticosteroid long-term therapy is a cornerstone in AOSD treatment [4]. However, it is well known that corticosteroids are weighted by several deleterious side effects such as hyperglycemia, glaucoma, and systemic hypertension. To note, the higher the initial dose, the higher the risk to develop such effects [32]. We found a statistically significant reduction of corticosteroid daily intake from baseline to 12 months, confirming that TCZ displays a strong and very effective corticosteroid-sparing effect. It must be pointed out that at month 12 72.72 % of our patients discontinued corticosteroid therapy, maintaining a good clinical response. This result mirrors what observed in previous studies reporting that corticosteroid treatment was successfully tapered and discontinued in AOSD patients treated with TCZ [33, 35].

Our study suggests that TCZ might be considered as a helpful therapeutic option in those patients needing a long-term therapy with corticosteroid in order to reach a good clinical response.

In a recent study enrolling 112 children aged 2 to 17 years, with active and persistent SJIA, TCZ was reported to be effective in reducing signs and symptoms of disease. At week 52, a large percentage of the patients who received TCZ had no fever and no joints with active arthritis and had discontinued oral glucocorticoids [36]. As far as RA patients receiving a combination therapy of TCZ and DMARDs are concerned, it has been demonstrated that TCZ monotherapy displays a similar efficacy profile [37].

Concerning the safety issue in our study, no severe side events including death were observed during the study period. Only few minor adverse events, one episode of upper respiratory tract infection and infusion site reaction, were reported during our study period. Such findings mirror the results arising from clinical trials involving RA patients in which a generally good safety profile of TCZ has been demonstrated [31]. However, the follow-up duration of the present study do not allow any speculation on long-term safety profile in AOSD patients. Again, the different therapeutic background of our patients could explain the better safety profile of our study.

The analysis of clinical conditions at 6 months after discontinuation of biological therapy showed that 72.72 % were still maintaining the clinical remission. These data in our patients mirror those observed in SJIA patients treated with TCZ in which after the discontinuation of therapy only a small percentage of children experimented a flare of the disease [21].

It remains unclear how long TCZ should be administered. The lack of published guidelines concerning treatment with biologics for AOSD [38] and how long to treat patients in other rheumatic diseases such as RA and spondyloarthropathies lead us to discontinued therapy after reaching a persistent remission. AOSD is a chronic disease with long course and late relapses, and on this basis, we have chosen to avoid the risk of a long period of exposition to a strong immunosuppressive therapy in our patients. In this setting, some authors suggest that the decrease of serum IL-6 during TCZ treatment might indicate disease remission and might be a guide for discontinuation [14]. In this regard, ESR, CRP, and ferritin serum levels could be measured to assess the disease response as a surrogate marker of IL-6 function.

In conclusion, our study demonstrated that blocking IL-6 with TCZ could be a therapeutic option for AOSD patients in which overproduction of IL-6 plays a pathological role. Although we cannot draw certain conclusions due to the limited number of cases, our results demonstrated the potential therapeutic effect of this treatment. Future clinical studies investigating the safety and the efficacy will elucidate the clinical benefits of IL-6 blockade for AOSD patients.