Safety and efficacy of subcutaneous tocilizumab in adults ... · 5/5/2016 · body area with an increase of modiﬁ ed Rodnan skin score of at least 2 or two new body areas with

Articles

www.thelancet.com Published online May 5, 2016 http://dx.doi.org/10.1016/S0140-6736(16)00232-4 1

Safety and effi cacy of subcutaneous tocilizumab in adults with systemic sclerosis (faSScinate): a phase 2, randomised, controlled trialDinesh Khanna, Christopher P Denton, Angelika Jahreis, Jacob M van Laar, Tracy M Frech, Marina E Anderson, Murray Baron, Lorinda Chung, Gerhard Fierlbeck, Santhanam Lakshminarayanan, Yannick Allanore, Janet E Pope, Gabriela Riemekasten, Virginia Steen, Ulf Müller-Ladner, Robert Lafyatis, Giuseppina Stifano, Helen Spotswood, Haiyin Chen-Harris, Sebastian Dziadek, Alyssa Morimoto, Thierry Sornasse, Jeff rey Siegel, Daniel E Furst

SummaryBackground Systemic sclerosis is a rare disabling autoimmune disease with few treatment options. The effi cacy and safety of tocilizumab, an interleukin 6 receptor-α inhibitor, was assessed in the faSScinate phase 2 trial in patients with systemic sclerosis.

Methods We did this double-blind, placebo-controlled study at 35 hospitals in Canada, France, Germany, the UK, and the USA. We enrolled adults with progressive systemic sclerosis of 5 or fewer years’ duration from fi rst non-Raynaud’s sign or symptom. Patients were randomly assigned (1:1) to weekly subcutaneous tocilizumab 162 mg or placebo. The primary endpoint was the diff erence in mean change from baseline in modifi ed Rodnan skin score at 24 weeks. This study is registered with ClinicalTrials.gov, number NCT01532869.

Findings We enrolled 87 patients: 43 assigned to tocilizumab and 44 assigned to placebo. The least squares mean change in modifi ed Rodnan skin score at 24 weeks was –3·92 in the tocilizumab group and –1·22 in the placebo group (diff erence –2·70, 95% CI –5·85 to 0·45; p=0·0915). The least squares mean change at 48 weeks was –6·33 in the tocilizumab group and –2·77 in the placebo group (treatment diff erence −3·55, 95% CI −7·23 to 0·12; p=0·0579). In one of several exploratory analyses, fewer patients in the tocilizumab group than in the placebo group had a decline in percent predicted forced vital capacity at 48 weeks (p=0·0373). However, we detected no signifi cant diff erence in disability, fatigue, itching, or patient or clinician global disease severity. 42 (98%) of 43 patients in the tocilizumab group versus 40 (91%) of 44 in the placebo group had adverse events. 14 (33%) versus 15 (34%) had serious adverse events. Serious infections were more common in the tocilizumab group (seven [16%] of 43 patients) than in the placebo group (two [5%] of 44). One patient died in relation to tocilizumab treatment.

Interpretation Tocilizumab was not associated with a signifi cant reduction in skin thickening. However, the diff erence was greater in the tocilizumab group than in the placebo group and we found some evidence of less decline in forced vital capacity. The effi cacy and safety of tocilizumab should be investigated in a phase 3 trial before defi nitive conclusions can be made about its risks and benefi ts.

Funding F Hoff mann-La Roche, Genentech.

IntroductionSystemic sclerosis is a rare connective tissue disorder characterised by fi brosis, infl ammation, and microvascular injury with heterogeneous clinical presentations. Pulmonary, cardiac, gastrointestinal, and renal complications contribute to patient morbidity and decreased survival.1

Increasing evidence supports important roles for interleukin 6 in the pathogenesis of systemic sclerosis,2 including B-cell diff erentiation towards immunoglobulin-secreting plasma cells, T-cell diff erentiation towards Th17 cell types, and transformation of fi broblasts to myofi broblasts leading to synthesis of extracellular matrix.2 Dermal fi broblasts from patients with systemic sclerosis constitutively express more interleukin 6 than those from healthy controls,3 and serum and skin concentrations of interleukin 6 are elevated in patients

with early systemic sclerosis.4,5 In patients with systemic sclerosis with or without interstitial lung disease,

increased interleukin 6 has been associated with higher mortality, more severe skin involvement, and increased incidence of progressive pulmonary decline.5,6 Although the exact cellular mechanisms of the eff ects of interleukin 6 on fi brosis are unknown, myeloid cells are implicated in systemic sclerosis skin pathogenesis.7 mRNA expression of a cluster of macrophage genes, including CD14 in the skin, correlates strongly with modifi ed Rodnan skin score, and CD14 expression is prognostic for progressive skin disease.8 M2-macrophages appear to have an important role in mediating infl ammation and promoting fi brosis through the release of profi brotic factors.9,10

In the bleomycin mouse model, blockade of the interleukin 6 pathway reduced skin fi brosis, α smooth-muscle actin protein expression,11 hydroxyproline

Published OnlineMay 5, 2016http://dx.doi.org/10.1016/S0140-6736(16)00232-4

See Online/Commenthttp://dx.doi.org/10.1016/S0140-6736(16)00622-X

University of Michigan Scleroderma Program, Ann Arbor, MI, USA (Prof D Khanna MD); University College London, London, UK (Prof C P Denton PhD); Genentech, South San Francisco, CA, USA (A Jahreis MD, H Chen-Harris PhD, A Morimoto PhD, T Sornasse PhD, J Siegel MD); University Medical Center Utrecht, Utrecht, Netherlands (Prof J M van Laar MD); University of Utah, Veterans Affairs Medical Center, Salt Lake City, UT, USA (T M Frech MD); University of Liverpool and Aintree University Hospital, Liverpool, UK (M E Anderson FRCP); Jewish General Hospital, Montreal, QC, Canada (M Baron MD); Stanford University School of Medicine and Palo Alto VA Health Care System, Palo Alto, CA, USA (L Chung MD); University of Tübingen, Tübingen, Germany (G Fierlbeck MD); University of Connecticut Health Center, Farmington, CT, USA (S Lakshminarayanan MD); Paris Descartes University and Cochin Hospital, Paris, France (Prof Y Allanore PhD); Schulich School of Medicine and Dentistry, University of Western Ontario, London, ON, Canada (J E Pope MD); Charité University Hospital, German Rheumatism Research Centre, Berlin, Germany (Prof G Riemekasten PhD); Georgetown University, Washington, DC, USA (Prof V Steen MD); Justus-Liebig University Giessen, Kerckhoff

http://crossmark.crossref.org/dialog/?doi=10.1016/S0140-6736(16)00232-4&domain=pdf

Articles

2 www.thelancet.com Published online May 5, 2016 http://dx.doi.org/10.1016/S0140-6736(16)00232-4

Clinic, Bad Nauheim, Germany (Prof U Müller-Ladner MD);

Boston University School of Medicine, Boston, MA, USA

(R Lafyatis MD, G Stifano MD); Roche Products, Welwyn

Garden City, UK (H Spotswood PhD); Roche,

Basel, Switzerland (S Dziadek PhD); and University

of California, Los Angeles, CA, USA (Prof D E Furst MD)

Correspondence to:Prof Dinesh Khanna, Division of Rheumatology, Department of

Internal Medicine, University of Michigan Scleroderma Program,

Ann Arbor, MI 48109, [email protected]

content, and myofi broblast counts.4 Initial data from patients with systemic sclerosis indicated that treatment with tocilizumab improved skin sclerosis and systemic sclerosis-associated polyarthritis.12,13

ENPP2 is an enzyme induced by interleukin 6, is associated with fi brosis development, and is increased in the circulation of patients with systemic sclerosis.14 It catalyses production of lysophosphatidic acid,15 and high concentrations of circulating lysophosphatidic acid are associated with systemic sclerosis pathobiology.16,17 Serum concentration of COMP correlates with skin fi brosis and predicts mortality in patients with systemic sclerosis.18 High serum concentrations of CCL18 are associated with scleroderma-associated pulmonary fi brosis and progression of idiopathic lung disease,19,20 and high serum concentrations of POSTN are associated with the degree of skin fi brosis in patients with systemic sclerosis.21

We did the phase 2 faSScinate trial to investigate the effi cacy and safety of interleukin 6 blockade with tocilizumab in systemic sclerosis and to do exploratory analysis of biomarkers.

MethodsStudy design and participantsWe did this randomised, double-blind, placebo-controlled study in 35 hospitals across Canada, France, Germany, the UK, and the USA. Investigators from each centre enrolled patients aged 18 years or older who met the 1980 American College of Rheumatology criteria for systemic sclerosis22 with no more than 5 years since their fi rst non-Raynaud’s sign or symptom and a modifi ed Rodnan skin score of 15–40 and had clinical skin involvement proximal to the elbows, knees, or both, with or without face involvement. Participants had to have had an increase of

at least 3 on the modifi ed Rodnan skin score at screening compared with the last visit within the previous 1–6 months or new-onset systemic sclerosis diagnosed within 1 year before screening, involvement of one new body area with an increase of modifi ed Rodnan skin score of at least 2 or two new body areas with increase of at least 1, documentation of worsening of skin thickening in the previous 6 months, or at least one tendon friction rub plus at least one laboratory criterion (C-reactive protein ≥10·0 mg/L, erythrocyte sedimentation rate ≥28 mm/h, or platelets ≥330 000/μL). All patients provided written informed consent. Patients or caregivers could administer subcutaneous injections of the investigational drug. Eligible patients had clinically uninvolved skin in at least one body area for study drug injections.

Each site’s institutional review board or ethics committee approved the protocol before the study commenced. The study was done in accordance with the Declaration of Helsinki and with Good Clinical Practice.

Randomisation and maskingPatients were randomly assigned (1:1) using an interactive voice and web response system to receive weekly subcutaneous treatment with tocilizumab 162 mg or placebo for 48 weeks followed by open-label weekly tocilizumab for 48 weeks. Randomisation numbers were generated by the sponsor, and randomisation was stratifi ed by joint involvement at baseline (<four vs ≥four joints on the 28 tender joint count). Investigators, patients, and sponsor personnel were masked to treatment assignment. To prevent unmasking, separate assessors evaluated effi cacy and safety. The effi cacy assessor did not have access to safety data during the

Research in context

Evidence before this studyWe searched PubMed with the terms systemic sclerosis or scleroderma and a combination of systemic sclerosis with any of the following terms: IL-6, biomarkers, CCL-18, review, modifi ed Rodnan skin score, clinical trials, and interstitial lung disease. We also reviewed rheumatology textbooks written in English, and an advisory board with experts in the fi eld of systemic sclerosis was convened. In-vivo studies show high concentrations of interleukin 6 in serum and skin biopsy samples of patients with systemic sclerosis, which are associated with greater disease activity and higher mortality. Serum concentrations of interleukin 6 in patients with early systemic sclerosis can predict the extent of progression of skin disease. In a bleomycin-induced mouse model of scleroderma, an antibody that blocks interleukin 6 reduced dermal sclerosis. Furthermore, the extent of fi brosis was attenuated in interleukin 6 knockout mice. Two treatment-refractory patients with systemic sclerosis had improved skin thickening after treatment with tocilizumab for 6 months.

Added value of this studyThis study is the fi rst placebo-controlled trial in patients with early systemic sclerosis to show a clinically signifi cant—albeit not statistically signifi cant—amelioration of skin sclerosis, and clinically relevant improvement in lung function in patients treated with tocilizumab. The safety profi le was consistent with complications of systemic sclerosis, including osteomyelitis, and with the safety profi le of tocilizumab.

Implications of all the available evidenceGiven the lack of disease-modifying treatment options for patients with systemic sclerosis, combined with the morbidity and mortality associated with this disease, data from our trial provide hope for a potential future treatment. FaSScinate was a phase 2 study; therefore, the effi cacy and safety of tocilizumab should be further investigated in an adequate, randomised, well-controlled, phase 3 trial before defi nitive conclusions can be made about its risks and benefi ts.

Articles


double-blind phase of the trial, but the safety assessor had access to both effi cacy and safety data. Although some sponsor personnel were unmasked after the primary analysis at 24 weeks, personnel interacting with sites and site staff remained masked to treatment assignment until the database lock at 48 weeks.

ProceduresEscape treatment with methotrexate, hydroxychloroquine, or mycophenolate mofetil was permitted after 24 weeks for patients with 20% or more worsening of modifi ed Rodnan skin score from baseline, worsening complications associated with systemic sclerosis such as arthritis, and idiopathic lung disease, as determined by the treating investigator. Safety monitoring for adverse events and serious adverse events and laboratory monitoring were done at least every 8 weeks.

OutcomesThe primary endpoint was the diff erence in mean change in modifi ed Rodnan skin score from baseline to 24 weeks. Secondary endpoints were patient-reported and physician-reported outcomes to 24 weeks and 48 weeks (Health Assessment Questionnaire–Disability Index [HAQ–DI] score, patient global visual analogue scale [VAS, 0–100], physician global VAS, Functional Assess-ment of Chronic Illness Therapy [FACIT]-Fatigue score, and pruritus 5-D Itch score), change in modifi ed Rodnan skin score from baseline to 48 weeks, proportion of patients with change in modifi ed Rodnan skin score from baseline at 48 weeks greater than or equal to change from baseline to week 24, and change from baseline in VAS score at 24 weeks and 48 weeks (intestinal, breathing, Raynaud’s disease, fi nger ulcers, overall disease from the Scleroderma HAQ–DI).

Exploratory endpoints included proportion of patients achieving minimum clinically important diff erence24 (improvement ≥0·22 [defi ned in protocol and statistical analysis plan; based on rhematoid arthritis] and ≥0·14 [defi ned in statistical analysis plan; based on the early diff use systemic sclerosis trial24) in the HAQ–DI at 24 weeks and 48 weeks, proportion of patients with at least 4·7 (minimum clinically important diff erence24) improvement in modifi ed Rodnan skin score at 24 weeks and 48 weeks, change from baseline at 24 weeks and 48 weeks in pulmonary function measured by forced vital capacity (FVC; mL), percent predicted FVC, percent predicted diff using capacity for CO at 48 weeks, change from baseline at 24 weeks and 48 weeks in 28 tender joint count in patients with joint involvement (at least four tender joints at baseline), change from baseline at 24 and 48 weeks in EQ-5D-3L, change from baseline at 24 and 48 weeks in digital ulcer count, and change from baseline at 24 and 48 weeks in tender friction rub. Proportions of patients with 20%, 40%, and 60% improvement in modifi ed Rodman skin score at 24 and 48 weeks and correlation analysis between improvement in modifi ed Rodnan skin

score and change in percent predicted FVC at 48 weeks were post hoc outcomes. Exploratory biomarker analyses included gene expression analysis of skin biopsy specimens collected at baseline and at 24 weeks and COMP, POSTN, autotaxin/ENPP2, and CCL18 serum concentrations measured with immunoassays (appendix pp 4, 5). Additionally, protein expression in the skin was analysed with immuno histochemistry, and leucocyte immunophenotypes were analysed with fl ow cytometry. Pharmacokinetic and pharmacodynamic (tocilizumab, sIL-6R and IL-6) analyses were conducted.

Statistical analysisA sample size of 36 patients per group (86 patients total, allowing for 15% dropout) was determined to provide 80% power to detect a diff erence in means for the change in modifi ed Rodnan skin score of 4·7 from baseline to 24 weeks, based on an estimated common SD of 6·9923 using a two-group t test with a 5% two-sided signifi cance level.

The primary endpoint was analysed with a mixed-model repeated-measures approach. There was no imputation of missing data before these analyses, and all patients were stratifi ed by joint involvement at baseline. Patients discontinuing treatment before the end of the double-blind period underwent their last evaluation at the time of discontinuation. Except for exploratory analysis of pulmonary function, which included all available data, all mixed-model repeated-measures analyses of secondary endpoints at week 48 used data that were censored after initiation of escape treatment. The mixed-model repeated-measures approach assumes that data are missing at random; therefore, sensitivity analyses were done for the primary endpoint to account for data that might not have been missing because of chance (appendix p 11). There was no preplanned statistical analysis of exploratory endpoints; when clinically important exploratory endpoints showed a treatment eff ect, further unplanned analyses were conducted.

Patients with missing modifi ed Rodnan skin score data at 24 weeks or 48 weeks were considered non-responders for analysis of 20%, 40%, and 60% improvement from baseline. Similarly, patients without data were considered non-responders for the minimal clinically important diff erence24 for HAQ–DI, modifi ed Rodnan skin score, and maintenance of modifi ed Rodnan skin score response. The van Elteren test (stratifi ed by joint involvement) was used to compare the treatment eff ect on the cumulative distribution of change from baseline in percent predicted FVC and percent predicted diff using capacity for CO. No adjustment was made for multiplicity in statistical testing for any of the analyses. For serious adverse event rates rates per 100 patient-years, multiple occurrences of events in a patient were counted, and confi dence intervals were based on the Poisson distribution. The effi cacy analyses were done for the modifi ed intention-to-treat population, which included all randomly assigned patients who

See Online for appendix

Articles


received any study drug. No randomly assigned patients were excluded from this population, and patients were analysed by assigned treatment. The safety population included all patients who received study drug and provided at least one safety assessment after treatment, summarised by treatment received.

Role of the funding sourceThe funder designed the study in collaboration with the authors. The sponsor collected, analysed, and interpreted the data, and drafted the report. DK, CPD, AJ, MB, LC, GF, YA, RL, JEP, GR, UM-L, GS, HS, AM, JS, and TS contributed to data interpretation, revised the manuscript, and attest to the accuracy and completeness of the reported data. The corresponding author had full access to all data congregates in the study and made the fi nal decision to submit the manuscript for publication.

ResultsPatients were randomly assigned from March 13, 2012, to June 18, 2013; the last patient completed the week 48 visit on May 20, 2014. 87 patients were enrolled: 44 to the placebo group and 43 to the tocilizumab group (fi gure 1). At 48 weeks, 12 (27%) of 44 patients in the placebo group and six (14%) of 43 in the tocilizumab group had received escape treatment. Similar numbers of patients withdrew in each group. In the tocilizumab group, eight patients withdrew because of safety reasons (fi ve adverse events, three deaths) and fi ve withdrew for non-safety reasons (three patient decisions, one lack of effi cacy, one lost to follow-up). In the placebo group four withdrew because of safety reasons (four adverse events) and seven withdrew for non-safety reasons (fi ve patient decisions, one clinician decision, one non-compliance). Baseline characteristics were balanced between groups (table 1). Mean dose intensity, a measure of compliance (defi ned as the number of doses actually received divided by the expected number of doses for the time on study), was 87·3% (median 97·9%) in the placebo group and 85·6% (97·9%) in the tocilizumab group. Among patients with available data for modifi ed Rodnan skin score at 48 weeks, four (13%) of 32 in the tocilizumab group and nine (27%) of 33 in the placebo group received escape treatment.

The primary endpoint was not met. Least squares mean change in modifi ed Rodnan skin score from baseline to 24 weeks was –3·92 in the tocilizumab group and –1·22 in the placebo group (treatment diff erence –2·70, 95% CI –5·85 to 0·45; p=0·0915; table 2, fi gure 2). The observed mean modifi ed Rodnan skin score at 24 weeks was 21·84 (SD 9·89) for tocilizumab (n=37) and 23·21 (SD 9·26) for placebo (n=38).

The least squares mean change from baseline to 48 weeks was –6·33 in the tocilizumab group and –2·77 in the placebo group (treatment diff erence −3·55, 95% CI −7·23 to 0·12; p=0·0579; table 2, fi gure 2). The observed mean modifi ed Rodnan skin score at 48 weeks was 19·56 (10·08) in the tocilizumab group (n=32) and 22·27

Figure 1: Trial profi le

141 patients screened

87 randomly assigned

43 assigned to tocilizumab

8 withdrew 8 withdrew

44 assigned to placebo

35 completed 24 weeks

6 received escape treatment 12 received escape treatment

5 withdrew 3 withdrew

36 completed 24 weeks

30 completed 48 weeks 33 completed 48 weeks

Placebo group (n=44)

Tocilizumab group (n=43)

Age (years) 48 (12·9) 51 (11·7)

Women 35 (80%) 32 (74%)

White 40 (91%) 38 (88%)

Duration of systemic sclerosis (months)* 19·5 (17·0) 17·6 (13·9)

Anti-RNA polymerase antibody positive 17 (39%) 13 (30%)

Previous biological drugs 0 (0%) 0 (0%)

Immunosuppressive drugs† 19 (43%) 22 (51%)

Concomitant systemic corticosteroid use‡ 17 (39%) 21 (49%)

Anti-topoisomerase antibody positive 20 (45%) 18 (42%)

Patients with tendon friction rub ≥1 22 (50%) 20 (47%)

TJC28 ≥4 21 (49%)§ 20 (47%)

TJC28 7 (8·5)§ 7 (8·9)

Total mRSS 26 (5·9) 26 (7·2)

Overall HAQ–DI score 1 (0·7) 1 (0·6)§

Clinician global VAS 61 (15·2) 64 (15·1)

CRP (mg/L) 10 (13·5)§ 10 (13·5)

Platelet count (10⁹ per L) 308 (88·9) 306 (82·4)

Percent predicted FVC 82% (13)¶ 80% (14)

Percent predicted DLCO 74% (21)§ 73% (19)¶

Patients with ≥1 digital ulcer|| 10 (23%) 11 (26%)

For the safety population. Data are mean (SD) or n (%) unless stated otherwise. CRP=C-reactive protein. DLCO (Hb corr)=diff using capacity for CO corrected for haemoglobin. FVC=forced vital capacity. HAQ–DI=Health Assessment Questionnaire—Disability Index. mRSS=modifi ed Rodnan skin score. TJC28=tender joint count using 28 joints. VAS=visual analogue scale. *Calculated from time of fi rst non-Raynaud’s sign or symptom. †Previous or concomitant. ‡Treatment that ended before fi rst study treatment administration excluded from the summary. Medication class “steroids” excluding the following routes of administration: nasal, nasogastric, topical, and respiratory (inhalation). §n=43. ¶n=42. ||Ulcer at or distal to the proximal interphalangeal joint, with loss of surface epithelialisation.

Table 1: Baseline characteristics

Articles


(SD 8·05) in the placebo group (n=33). Of patients whose modifi ed Rodnan skin score had improved by 24 weeks, more patients in the tocilizumab group (15 [68%] of 22) than in the placebo group (eight [44%] of 18) maintained or had further improvement in modifi ed Rodnan skin score by 48 weeks. At 24 weeks, more patients in the tocilizumab group had modifi ed Rodnan skin score improvement of at least 20% and of at least 4·7 points, but the same number of patients in each group had at least 40% or 60% improvement. At 48 weeks, more patients in the tocilizumab group than in the placebo group had modifi ed Rodnan skin score improvement of at least 20%, 40%, 60%, and of at least 4·7 points (table 3). Eight patients in the placebo group and four in the tocilizumab group had an increase in modifi ed Rodnan skin score of more than 5 and at least 25% from baseline25 at any time to 24 weeks.

For clinician and patient global VAS and FACIT-fatigue scores, the diff erence was not signifi cant at 24 weeks or 48 weeks, but scores were better in the tocilizumab group than in the placebo group (table 2). For the HAQ–DI, we found no signifi cant diff erence between treatment groups (table 2). At 48 weeks, 12 (28%) of 43 patients in the tocilizumab group versus 3 (7%) of 44 in the placebo group achieved an improvement of 0·22 or more in HAQ–DI (p=0·0111), with identical results for improvement defi ned as 0·14 or more. At 24 weeks, 11 (26%) of 43 patients in the tocilizumab group versus 7 (16%) of 44 in the placebo group achieved an improvement of 0·22 or more in HAQ–DI, with identical results for improvement defi ned as 0·14 or more. Overall, with the exception of breathing VAS score, all Scleroderma HAQ–DI VAS scores at 48 weeks had no signifi cant diff erence between groups, although the scores favoured tocilizumab (appendix p 12). At 24 weeks, little treatment benefi t was observed.

20 (47%) of 43 patients in the tocilizumab group and 21 (49%) of 43 in the placebo group had joint involvement. Among these patients, mean tender joint counts changed from baseline by –4·3 (SD 7·3; median –2·5, IQR –10·0 to –1·0; n=16) at 24 weeks and –5·1 (SD 7·3; median –4·5, IQR –12·0 to 0·0; n=10) at 48 weeks with tocilizumab compared with –2·1 (SD 6·3; median –2·0, IQR –4·0 to –1·0; n=17) at 24 weeks and –2·9 (SD 7·1; median –2·5, IQR –5·5 to 2·5; n=12) at 48 weeks in the placebo group.

We showed a signifi cantly smaller decrease in FVC for tocilizumab than for placebo from baseline to 24 weeks (tocilizumab –34 mL vs placebo –171 mL; least square mean diff erence 136 mL, 95% CI 9 to 264; p=0·0368) but from baseline to 48 weeks the diff erence was not signifi cant (tocilizumab –117 mL vs placebo –237 mL; 120 mL, 95% CI –23 to 262; p=0·0990). Fewer patients in the tocilizumab group than in the placebo group had worsening of percent predicted FVC at 24 weeks (p=0·009; fi gure 3A) or at 48 weeks (p=0·037; fi gure 3B). At 24 weeks, one (3%) of 30 patients in the tocilizumab group versus

seven (19%) of 36 in the placebo group had more than 10% (absolute) decreases in percent predicted FVC from baseline. At 48 weeks, the proportion was three (10%) of 30 versus seven (23%) of 31. At 24 weeks the least squares mean change from baseline in percent predicted FVC was –0·7 (95% CI –3·2 to 1·8) in the tocilizumab group versus –4·5 (–7·0 to –1·9) in the placebo group. At 48 weeks, the change was–2·6 (–5·2 to –0·1) versus –6·3 (–8·9 to –3·8). At 48 weeks, the change from baseline in percent predicted diff using capacity for CO did not diff er signifi cantly between placebo and tocilizumab (appendix p 6). In a post-hoc analysis, correlations between improvements in



Diff erence in means (95% CI)

p value

mRSS

Week 24 –1·22 (n=43) –3·92 (n=41) –2·70 (–5·85 to 0·45) 0·0915

Week 48 –2·77 (n=43) –6·33 (n=41) –3·55 (–7·23 to 0·12) 0·0579

HAQ–DI

Week 24 0·118 (n=42) 0·137 (n=41) 0·020 (–0·186 to 0·225) 0·8503

Week 48 0·205 (n=41) –0·002 (n=41) –0·207 (–0·471 to 0·056) 0·1212

Clinician Global VAS

Week 24 –7·25 (n=41) –8·24 (n=39) –0·99 (–9·20 to 7·23) 0·8118

Week 48 –9·39 (n=41) –18·41 (n=40) –9·02 (–19·04 to 1·00) 0·0768

Patient Global VAS

Week 24 1·53 (n=42) –2·33 (n=42) –3·85 (–13·04 to 5·34) 0·4063

Week 48 –2·70 (n=41) –11·00 (n=42) –8·30 (–19·31 to 2·71) 0·1371

FACIT (Fatigue) Score

Week 24 1·26 (n=41) 2·68 (n=42) 1·43 (–2·97 to 5·82) 0·5197

Week 48 0·36 (n=40) 3·11 (n=42) 2·75 (–1·38 to 6·88) 0·1886

5-D Itch Scale

Week 24 –1·73 (n=41) –0·94 (n=41) 0·79 (–0·94 to 2·51) 0·3651

Week 48 –1·08 (n=40) –2·19 (n=41) –1·11 (–3·16 to 0·94) 0·2841

For the modifi ed intention-to-treat population. A negative change from baseline shows improvement for all effi cacy measures, except for the FACIT-Fatigue Scale, where positive change indicates improvement. We did a mixed-model repeated-measures analysis that included the fi xed categorical eff ects of treatment, visit, stratifi cation factor of joint involvement at baseline, and treatment-by-visit interaction, and the continuous covariates of baseline score and baseline score-by-visit interaction. FACIT=Functional Assessment of Chronic Illness Therapy. HAQ–DI=Health Assessment Questionnaire—Disability Index. VAS=visual analogue scale.

Table 2: Least square mean change from baseline for mRSS and in patient-reported and physician-reported outcomes

Figure 2: Change (95% CI) from baseline in modifi ed Rodnan skin score

0 3 8 16 24 32 40 48–9

–8

–7

–6

–5

–4

–3

–2

–1

0

1

Chan

ge fr

om b

asel

ine

Time (weeks)

Placebo groupTocilizumab group

Articles


modifi ed Rodnan skin score and change in percent predicted FVC from baseline to week 48 were similar in each treatment group (tocilizumab, r=–0·311, p=0·121; placebo, r=–0·278, p=0·223; appendix p 7). The results of the analyses of digital ulcer, tendon friction rub, EQ-5D-3L and pharmacokinetic and pharmacodynamic analyses are shown in the appendix (p 2).

Serum concentrations of COMP, POSTN, ENPP2, and CCL18 (fi gure 4) at baseline were all signifi cantly high

compared with those of age-matched and sex-matched healthy controls from Bioreclamation IVT (New York, NY, USA; median age 46 years [30–59], seven male and 18 female; p<0·0001 for all biomarkers; two-tailed t test assuming unequal variance). Serum concentration of CCL18 was signifi cantly lower in the tocilizumab group than in the placebo group but we detected no eff ect of tocilizumab on the serum concentrations of COMP, POSTN, or ENPP2 (fi gure 4).

We explored gene expression using microarray technology on skin biopsy specimens collected from participants at baseline (34 patients in the tocilizumab group vs 39 in the placebo group) and at 24 weeks (28 vs 30), and from 20 age-matched and sex-matched healthy controls from Epistem (Manchester, UK; median age 43 years [22–62] years, six male and 14 female). First, we conducted gene set enrichment analysis26 (using GSEAlm: Linear Model Toolset for Gene Set Enrichment Analysis; R, version 1.30.0) for fi brosis, interferon-α, interleukin 6, TGF-β, M1-macrophage, and M2-macrophage gene sets (appendix pp 4, 13–36) of the microarray data obtained from all available samples. We looked for gene set enrichment among the genes overexpressed in patients with systemic sclerosis versus healthy controls, and among the genes downregulated by tocilizumab at week 24 versus baseline. At baseline, gene sets associated with fi brosis, interferon-α, interleukin 6, TGF-β, M1-macrophage, and M2-macrophage were signifi cantly enriched in the group of genes overexpressed in patients with systemic sclerosis compared with those of healthy controls (p=0·005; appendix p 37). None of these gene sets was signifi cantly enriched in the genes downregulated by tocilizumab when comparing week 24 samples to baseline (appendix p 38). We recorded non-signifi cant enrichment in interferon α (p=0·07), interleukin 6 (p=0·095), and M2-macrophage (p=0·105) gene sets in the genes downregulated by tocilizumab when comparing week 24 with baseline (appendix p 38); we recorded no enrichment in the placebo group when comparing week 24 with baseline (appendix p 39).

Genome-wide analysis of gene expression in the biopsy samples was used to select 83 genes by unsupervised clustering of genes overexpressed in systemic sclerosis and downregulated by tocilizumab, and genes with expression correlated with modifi ed Rodnan skin score (appendix p 5). Confi rmatory analysis was done on these 83 genes by use of nCounter technology (NanoString Technologies; appendix p 5). Of the 83 genes representing the fi brosis, TGF-β, interleukin 6, interferon, and myeloid pathways, 62 transcripts were signifi cantly overexpressed and two were signifi cantly underexpressed in patients with systemic sclerosis compared with healthy controls (t test, Bonferroni correction for multiple comparisons). Clustering analysis of the average gene expression for the placebo and tocilizumab groups at both baseline and 24 weeks yielded nine clusters representing TGF-β, interleukin 6/STAT3, M1-macrophage, M2-macrophage,

24 weeks 48 weeks



p value Placebo group (n=44)


p value

Decrease ≥20% 12 (27%) 16 (37%) 0·36 12 (27%) 17 (40%) 0·26

Decrease ≥40% 6 (14%) 6 (14%) 1·00 3 (7%) 9 (21%) 0·069

Decrease ≥60% 2 (5%) 2 (5%) 1·00 0 (0%) 5 (12%) 0·026

Decrease ≥4·7 units

10 (23%) 16 (37%) 0·16 11 (25%) 16 (37%) 0·025

A change of 4·7 or more is deemed clinically important. Patients can be counted more than once. For example, a patient with a 43% improvement would be counted as having had decrease of both more than 20% and more than 40%.

Table 3: Changes from baseline in modifi ed Rodnan skin score

Figure 3: Cumulative distribution of patients by change in percent predicted FVCDecline is shown as absolute change in percent predicted FVC. From baseline to 24 weeks (A), and from baseline to 48 weeks (B). FVC=forced vital capacity.

0

20

40

60

80

100

p=0·009

Worsening No decline

Patie

nts (

%)

A

Placebo group (n=36)Tocilizumab group (n=30)

–40 –30 –20 –10 0 10 200

20

40

60

80

100

p=0·0373

Patie

nts (

%)

Percent predicted FVC change from baseline

BPlacebo group (n=31)Tocilizumab group (n=30)

Articles


and interferon (appendix pp 8, 9). Analysis of the eff ect of treatment on change in gene expression at 24 weeks identifi ed 16 genes specifi cally downregulated by tocilizumab compared with placebo (uncorrected p<0·05; appendix pp 8, 9). Although most of these genes (12 [75%] of 16) belonged to the M2-macrophage cluster, two belonged to the M1-macrophage cluster, suggesting that tocilizumab might have inhibitory actions on macrophages in general, and M2-macrophages in particular. Gene expression in the interleukin 6/STAT3 cluster was numerically reduced by tocilizumab at 24 weeks compared with placebo, with only two genes (CCL2 and SOCS3) signifi cantly downregulated by tocilizumab, possibly refl ecting the residual activity of other STAT3-activating factors (eg, EGF, interleukin 5, interleukin 6, HGF, LIF, and BMP2, which are not blocked by tocilizumab). None of the genes in the TGF-β-1 or TGF-β-2 clusters was signifi cantly downregulated by tocilizumab compared with placebo, probably because the expression of these genes was reduced at 24 weeks in the skin of patients in both treatment groups, suggesting a tocilizumab-independent reduction in the fi brotic pathway in both groups. Gene expression in the interferon cluster was highly

heterogeneous, limiting our ability to identify any pattern of expression in this gene cluster.

Finally, using the same dataset, we tested the diff erential eff ect of treatment with tocilizumab using a multianalyte, longitudinal, pharmacodynamic biomarker (2GSSc skin biomarker27). This biomarker, which predicts modifi ed Rodnan skin score on the basis of the weighted values for THBS1 and MS4A4A mRNA expression, was validated and has been used in two clinical trials, indicating that it is a robust surrogate outcome measure for the extent of systemic sclerosis skin disease.28 At baseline, the predicted modifi ed Rodnan skin scores were similar between treatment groups (placebo mean 16·4, 95% CI 15·1 to 17·8; tocilizumab mean 15·5, 95% CI 14·1 to 17·0). Comparison of the change in predicted modifi ed Rodnan skin score at 24 weeks indicated that treatment with tocilizumab resulted in a signifi cant decrease compared with treatment with placebo (placebo mean change –0·98, 95% CI –2·99 to 1·03; tocilizumab mean change –4·03, 95% CI –7·58 to –0·49; p=0·0488; appendix p 10). We did not detect any numerically signifi cant eff ect of tocilizumab on protein expression in the skin (aSMA, PTX3, pSTAT3, interleukin 6, and CCL2) or on leucocytes

Figure 4: Concentrations of CCL18, COMP, POSTN, and ENPP2Data points are mean (SD). p value is for the overall treatment eff ect of tocilizumab versus placebo based on two-way ANOVA with factors for treatment, time, and treatment × time.

Number of patientsHealthy controls

TocilizumabPlacebo

Baseline 83 24 48

244143

.. 3842

.. 4142

.. 3436

.. 2628

0

100

200

300

400

p<0·0001

p=0·46

p=0·90

p=0·87

Conc

entr

atio

n (n

g/m

L)

A Serum CCL18

Baseline 83 24 48

254143

.. 3641

.. 4043

.. 3225

.. 2626

0

200

400

600

800

Conc

entr

atio

n (n

g/m

L)

B Serum COMP

Number of patientsHealthy controls

TocilizumabPlacebo

Baseline 83 24 48

243641

.. 3740

.. 3841

.. 3232

.. 2525

Time (weeks)

0

50

100

150

200

Conc

entr

atio

n (n

g/m

L)

C Serum POSTN

Baseline 83 24 48

254241

.. 4038

.. 4338

.. 3534

.. 2625

Time (weeks)

200

250

300

400

500

350

450

Conc

entr

atio

n (n

g/m

L)

D ENPP2

Placebo groupTocilizumab groupHealthy controls

Articles


immunophenotypes (B cells, T cells, and monocytes) in the blood (appendix pp 40, 41).

At 24 weeks, 38 (88%) of 43 patients in the tocilizumab group versus 40 (91%) of 44 in the placebo group had had adverse events (table 4). By 48 weeks, 42 (98%) of 43 versus 40 (91%) of 44 had had adverse events. Similar proportions of patients in each group withdrew because of adverse events (table 4). The most common adverse events were infections, gastrointestinal disorders, skin or subcutaneous disorders, and musculoskeletal or connective tissue disorders (appendix p 42–44).

Serious infections were more common with tocilizumab than placebo, whereas non-infectious serious adverse events, including cardiac disorders, gastrointestinal disorders, and renal and urinary disorders, were more common with placebo (table 4). Osteomyelitis involving proximal interphalangeal joints occurred in two patients in the tocilizumab group and one patient in the placebo group. No anaphylaxis, gastrointestinal perforations, or malignancies were recorded, and subcutaneous injections were well tolerated. Total exposure to study drug before open-label treatment was 34·49 patient-years for tocilizumab and 36·77 patient-years for placebo. The overall rate of serious adverse events was 66·7 (95% CI 42·3 to 100·1) per 100 person-years with tocilizumab versus 76·1 (50·6 to 110·0) per 100 person-years with placebo. For serious infections, the rate was 34·8 (18·0 to 60·8) per 100 person-years versus 10·9 (3·0 to 27·9) per 100 person-years, and for noninfectious serious adverse events the rate was 31·9 (15·9 to 57·1) per 100 person-years versus 65·3 (41·8 to 97·1).

Four deaths were reported by 48 weeks. One patient in the placebo group died of cardiac failure, and one patient each in the tocilizumab group died of arrhythmia, multiorgan failure, and lung infection. One death, caused by lung infection, was considered related to study drug (appendix pp 2, 3).

Laboratory parameters of interest for tocilizumab (high alanine or aspartate aminotransferase concentration, decreased neutrophils, and decreased platelets) were mostly grade 1 or 2 in intensity and were not associated with clinically relevant sequelae such as hepatic events, serious infections, or serious bleeding events (appendix pp 45, 46).

DiscussionThis study is the fi rst, to our knowledge, phase 2 randomised controlled trial of tocilizumab for the treatment of systemic sclerosis. Week 24 was selected for the primary endpoint assessment because it was assumed that the rapid response of tocilizumab seen in clinical trials in patients with rheumatoid arthritis would translate to a rapid response in patients with systemic sclerosis. We detected no signifi cant diff erence for the primary endpoint (modifi ed Rodnan skin score at 24 weeks) or for secondary effi cacy endpoints at 24 weeks

Baseline to 24 weeks Baseline to 48 weeks





Patients with ≥1 adverse event 40 (91%) 38 (88%) 40 (91%) 42 (98%)

Patients with ≥1 infectious adverse event 18 (41%) 17 (40%) 22 (50%) 24 (56%)

Patients with injection site reactions 1 (2%)* 2 (5%)† 2 (5%)* 3 (7%)†

Patients with ≥1 serious adverse event 11 (25%) 9 (21%) 15 (34%) 14 (33%)

Patients with ≥1 infectious serious adverse event

1 (2%) 6 (14%) 2 (5%) 7 (16%)

Patients with ≥1 non-infectious serious adverse event

10 (23%) 5 (12%) 14 (32%) 10 (23%)

Withdrawals because of an adverse event 5 (11%) 4 (9%) 5 (11%) 6 (14%)

Deaths 1 (2%) 1 (2%) 1 (2%) 3 (7%)

Serious adverse events‡

Infections and infestations 1 (2%) 6 (14%) 2 (5%) 7 (16%)

Osteomyelitis 0 2 (5%) 1 (2%) 2 (5%)§

Bronchitis 0 1 (2%) 0 1 (2%)

Cellulitis 1 (2%)¶ 0 1 (2%)¶ 0

Infected skin ulcer 0 1 (2%) 0 1 (2%)

Lower respiratory tract infection 1 (2%)¶ 0 1 (2%)¶ 0

Lung infection 0 1 (2%) 0 1 (2%)

Oesophageal candidiasis 0 1 (2%) 0 1 (2%)

Pneumonia 0 0 0 1 (2%)§

Postprocedural cellulitis 0 0 0 1 (2%)

Sepsis 0 0 0 1 (2%)§

Cardiac disorders 3 (7%) 0 4 (9%) 1 (2%)

Acute myocardial infarction 1 (2%) 0 1 (2%) 0

Cardiac failure 1 (2%) 0 1 (2%) 0

Cyanosis 1 (2%) 0 1 (2%) 0

Arrhythmia 0 0 0 1 (2%)

Atrioventricular block 0 0 1 (2%) 0

Gastrointestinal disorders 2 (5%) 0 4 (9%) 1 (2%)

Colonic pseudo-obstruction 1 (2%) 0 1 (2%) 0

Gastric antral vascular ectasia 1 (2%) 0 1 (2%) 0

Abdominal distension 0 0 1 (2%) 0

Abdominal pain 0 0 1 (2%) 0

Gastritis 0 0 0 1 (2%)

Retroperitoneal fi brosis 0 0 1 (2%) 0

Musculoskeletal and connective tissue disorders

2 (5%) 0 2 (5%) 2 (5%)

Scleroderma or systemic sclerosis (worsening)

2 (5%) 0 2 (5%) 1 (2%)

Osteoarthritis 0 0 0 1 (2%)

Skin and subcutaneous tissue disorders 0 1 (2%) 1 (2%) 2 (5%)

Skin ulcer 0 1 (2%) 1 (2%) 2 (5%)

Vascular disorders 2 (5%) 1 (2%) 2 (5%) 1 (2%)

Hypertension 0 1 (2%) 0 1 (2%)

Hypertensive emergency 1 (2%) 0 1 (2%) 0

Raynaud’s phenomenon 1 (2%) 0 1 (2%) 0

Blood and lymphatic system disorders 1 (2%) 1 (2%) 1 (2%) 1 (2%)

Haemolytic uraemic syndrome 0 1 (2%) 0 1 (2%)

Iron defi ciency anaemia 1 (2%) 0 1 (2%) 0

(Table 4 continues on next page)

Articles


and 48 weeks. However, despite not meeting our prespecifi ed defi nition of minimal clinically important change (≥4·7), the change in modifi ed Rodnan skin score at 24 weeks and 48 weeks was within previously reported minimal clinically important diff erences for early, diff use systemic sclerosis (mRSS improvement of 3·2–5·3 from baseline to week 24 or week 24 to week 48).24 HAQ–DI, clinician global VAS, and other patient-reported outcomes did not diff er signifi cantly between groups, but, with the exception of breathing VAS, they favoured tocilizumab over placebo by week 48. Fewer patients receiving tocilizumab than patients receiving placebo had absolute FVC declines of more than 10%.

According to previous defi nitions,24 we recorded a clinically meaningful decline in modifi ed Rodnan skin score over 48 weeks in the tocilizumab group compared with the placebo group. Selection of patients with early progressive disease might have contributed to this eff ect. Eight patients in the placebo group and four patients in the tocilizumab group had progressive skin disease based on an increase in modifi ed Rodnan skin score of more than 5 and at least 25% worsening25 at any time between baseline and 24 weeks. Skin thickness is a surrogate for internal organ involvement and mortality in systemic sclerosis, and patients with attenuated skin thickness have improved survival and physical function.29,30 Conversely, patients with high modifi ed Rodnan skin score have greater internal organ involvement31 and are at greater risk for death.29

Systemic sclerosis has the highest case fatality among rheumatic diseases, with a cumulative survival of 75% at 5 years from diagnosis.32 Cardiopulmonary involvement is the leading cause of death,1 and FVC is the primary outcome measure in most trials of systemic sclerosis idiopathic lung disease and idiopathic pulmonary fi brosis.33 High concentrations of C-reactive protein have been associated with progressive idiopathic lung disease, and observational cohort studies suggest that high C-reactive protein concentration at baseline predicts long-term FVC decline.6,34 Although our study was not designed to enrol patients with progressive systemic sclerosis idiopathic lung disease, enrolling those with early progressive skin disease and high acute-phase reactants probably enriched for patients at high risk for idiopathic lung disease. Our data suggest that tocilizumab might have a disease-modifying eff ect by slowing the decline in lung function of patients with systemic sclerosis.

Treatment with tocilizumab resulted in the downregulation of skin myeloid-associated genes, including genes associated with M2-macrophage. Perivascular macrophages (resident or recruited) contribute to vascular infl ammation,35 and M2-macrophages might have a role in systemic sclerosis skin pathology through the release of infl ammatory and profi brotic factors.10 It is therefore tempting to speculate that the improvement in skin disease in patients treated with tocilizumab might be attributed to the inactivation

or depletion of skin macrophages in general and M2-macrophages in particular. Furthermore, treatment with tocilizumab resulted in rapid, sustained reductions in serum concentrations of CCL18, a chemokine associated with M2-macrophage, strengthening the hypothesis that tocilizumab might act, at least in part, by modulating the activity of M2-macrophages. Given that serum CCL18 is a prognostic biomarker that identifi es patients at higher risk for the progression of scleroderma lung disease,19 the eff ect of tocilizumab on circulating CCL18 could be related to the positive clinical eff ect of tocilizumab on lung function. These hypotheses will be further explored in the ongoing phase 3 study of tocilizumab in patients with systemic sclerosis.

Tocilizumab tended to inhibit skin expression of genes in the interleukin 6/STAT3 cluster, but only two genes from this cluster (CCL2 and SOCS3) were signifi cantly downregulated by tocilizumab at 24 weeks compared with placebo. This relatively modest eff ect of tocilizumab on the interleukin 6 pathway could be the result of residual activity of the multiple additional factors (eg, EGF, interleukin 5, HGF, LIF, and BMP2) that activate the STAT3 pathway but are not inhibited by tocilizumab. We did not record specifi c eff ects of tocilizumab on genes belonging to the TGF-β clusters or the interferon cluster, suggesting that the clinical eff ect of tocilizumab might not be mediated through these pathways. However, heterogeneity in the expression of genes associated with these pathways might have prevented the detection of subtle eff ects of tocilizumab on these genes. Tocilizumab did not aff ect the serum concentrations of ENPP2, COMP, or POSTN, three biomarkers related to fi brosis, suggesting

Baseline to 24 weeks Baseline to 48 weeks





(Continued from previous page)

General disorders and administration site conditions

0 1 (2%) 0 2 (5%)

Impaired healing 0 1 (2%) 0 1 (2%)

Multiorgan failure 0 0 0 1 (2%)

Nervous system disorders 0 0 2 (5%) 0

Headache 0 0 1 (2%) 0

Subarachnoid haemorrhage 0 0 1 (2%) 0

Renal and urinary disorders 2 (5%) 0 2 (5%) 0

Renal failure acute 1 (2%) 0 1 (2%) 0

Scleroderma renal crisis 1 (2%) 0 1 (2%) 0

Psychiatric disorders 0 1 (2%) 0 1 (2%)

Psychotic disorder 0 1 (2%) 0 1 (2%)

*Grade 1 haematoma (baseline to week 24) and grade 1 injection site bruising (baseline to week 48). †Two patients had grade 1 injection site erythema (one each baseline to week 24 and baseline to week 48), and one patient had grade 2 injection site rash (baseline to week 24). ‡Multiple occurrences of the same event in a patient were counted once. §One patient in the tocilizumab who had osteomyelitis by week 24 subsequently had pneumonia and sepsis by week 48. ¶Both events occurred in the same patient.

Table 4: Adverse events

Articles


that the clinical eff ect of tocilizumab on skin fi brosis might not be mediated by direct inhibition of the fi brosis pathway by tocilizumab. Finally, treatment with tocilizumab resulted in a signifi cant decrease in the two-gene systemic sclerosis skin biopsy-predicted modifi ed Rodnan skin score. The magnitude of change in predicted modifi ed Rodnan skin score was similar to that recorded clinically.

Overall, after 48 weeks of treatment, safety was consistent with the natural history of systemic sclerosis and the safety profi le for tocilizumab. Refl ecting the high morbidity and mortality rates of systemic sclerosis,1 we recorded more serious adverse events, serious infections, and deaths than in clinical trials of tocilizumab for rheumatoid arthritis.

Three patients in the tocilizumab group and one in the placebo group died, and one death was judged by the investigator to be related to tocilizumab. It will be important to assess mortality and causes of death carefully during the phase 3 study of tocilizumab for systemic sclerosis that has just begun. Furthermore, serious infections were more common in the tocilizumab group. These included bronchitis, lung infection, pneumonia, and sepsis, which are similar to the types of events reported during treatment of rheumatoid arthritis with tocilizumab. Small bone osteomyelitis and infected digital ulcers, however, were rare in patients treated with tocilizumab for rheumatoid arthritis and thus could require particular vigilance in patients with systemic sclerosis.

Although the gastrointestinal tract is the most commonly aff ected internal organ in patients with systemic sclerosis,1 no patients in our study had gastrointestinal perforations. Most injection site reactions, increases in alanine and aspartate aminotransferase concentrations, and decreases in neutrophil and platelet counts were grade 1 or 2 and involved no clinically relevant sequelae.

No disease-modifying treatment is approved for systemic sclerosis, and its management is based on organ involvement.36,37 Methotrexate aff ected skin thickening without establishing positive eff ects in patients with other organ manifestations,36,38 whereas oral cyclophosphamide modestly improved lung function and skin thickness in studies of patients with systemic sclerosis idiopathic lung disease.39 Data from a clinical trial40 of haemopoietic stem-cell transplantation suggest a disease-modifying eff ect in patients with systemic sclerosis; thus, haemopoietic stem-cell transplantation can be considered an option for some patients with rapidly progressing and severe systemic sclerosis. Data from the present trial suggest that tocilizumab might have a broader eff ect on systemic sclerosis skin and lung disease than methotrexate and a more favourable risk–benefi t profi le than cyclophosphamide, which is associated with clinically signifi cant risk of adverse reactions.30

The high discontinuation rate should be taken into account when interpreting data from this study. FVC was an exploratory endpoint, and high-resolution CT, which was not done to substantiate the pulmonary function data, is planned for a phase 3 study. Cohort enrichment

by retaining patients with a higher probability of skin progression and high acute-phase reactants might have contributed to the treatment responses we recorded. Although our results did not show a signifi cant treatment diff erence in modifi ed Rodnan skin score at week 24, by week 48 there seemed to be a treatment benefi t across most endpoints, which was suggestive of effi cacy for tocilizumab in systemic sclerosis. There were no previous positive studies in systemic sclerosis that could be used to inform the design, and the observed eff ect size for the primary endpoint was smaller than that used in the power calculation for the study (2·7 vs 4·7 modifi ed Rodnan skin score units). The eff ect size from this phase 2 study and the improvement at week 48 have informed the design of a phase 3 trial (NCT02453256) with the primary endpoint at 48 weeks.

This study shows that skin sclerosis did not diff er signifi cantly between groups, but did decrease more in the tocilizumab group than in the placebo group, and lung function showed a clinically relevant improvement. The safety profi le was consistent with complications of systemic sclerosis and with the safety profi le of tocilizumab. The propensity of patients with systemic sclerosis to develop digital ulcers could increase susceptibility to small bone osteomyelitis and infections of digital ulcers during tocilizumab treatment. Overall, the data warrant further investigation of tocilizumab in a phase 3 trial.ContributorsDK designed the study, recruited patients, advised on data analysis, interpreted the data, and wrote and revised the report. CPD designed the study, recruited participants, collected, analysed, and interpreted data, and wrote the report. AJ designed the study, oversaw the study, interpreted data, and wrote the report. JMvL and JS designed the study and collected data. TMF and SL collected data. MEA designed the study. MB collected and interpreted data. LC collected and interpreted data and wrote the report. GF collected, analysed, and interpreted data and recruited patients. YA and RL collected, analysed, and interpreted data. JEP did a literature search, designed the study, collected, analysed and interpreted data, and wrote the report. GR designed the study, and collected and interpreted data. VS designed the study and collected data. UM-L analysed and interpreted data. GS designed the study and collected, analysed, and interpreted data. HS designed the study and analysed and interpreted data. HC-H collected and analysed data and wrote the report. SD designed the study and collected data. AM and TS collected, analysed, and interpreted data and wrote the report. DEF designed the study, collected and analysed data, and wrote the report. All authors revised the report and approved the fi nal draft for publication.

Declaration of interestsDK has received grants from Bristol-Myers Squibb, Genentech/Roche, National Institute of Allergy and Infectious Diseases, National Institute of Arthritis and Musculoskeletal and Skin Diseases, Patient-Centered Outcomes Research Institute, and the Scleroderma Foundation; and consultancy fees from Actelion, Bayer, Cytori, EMD Serono (Merck), Genkyotex, Gilead, GlaxoSmithKline, Genentech/Roche, Sanofi -Aventis, and Seattle Genetics. CPD has received grants from CSL Behring and GlaxoSmithKline (paid to his institution); consultancy fees from GlaxoSmithKline and Roche (paid to his institution); consultancy fees from Merck-Serono; and speaker fees from Actelion and Bayer. AJ is an employee of and owns stock options in Genentech and has been issued a patent for subcutaneously administered tocilizumab (US 8580264 B2). JMvL has received honoraria from Merck Sharp & Dohme, Pfi zer, Roche, and Eli Lilly. MEA has received advisory board and related fees from

Articles


Actelion and honoraria from Actelion and Bristol-Myers Squibb and has served as principal investigator of clinical trials for Actelion and Roche. LC has served on an advisory board for Gilead and a data monitoring committee for Cytori. YA has received grants from Bristol-Myers Squibb, Roche/Genentech, Inventiva, Pfi zer, Sanofi , and Servier; and personal fees from Actelion, Bayer, Roche/Genentech, Inventiva, Medac, Pfi zer, Sanofi , Servier, and UCB. JEP has received research and consulting fees from Roche and consulting fees from Actelion, Bayer, Biogen, Celgene, and Genentech. GR has received honoraria for lectures and advisory boards from Actelion, Bayer, GlaxoSmithKline, Pfi zer, and Roche. VS has received a grant and advisory board fees from Roche. UM-L has received grants and speaker or advisory board fees from Roche and Chugai. RL has received grants from Shire, Sanofi , Regeneron, Genentech, UCB, Human Genome Sciences, Precision Dermatology, Biogen, Bristol-Myers Squibb, Inception, Stromedix, Pfi zer, Boston University, Bristol-Myers Squibb, and PRISM; and consultancy fees from Shire, Sanofi , Regeneron, Roche/Genentech, Biogen, Lycera, Novartis, Celgene, Bristol-Myers Squibb, Amira, Celdara, Celltex, Dart Therapeutics, Idera, Inception, Intermune, Medimmune, Precision Dermatology, Promedior, Zwitter, PRISM, UCB, Actelion, EMD Serono, Akros, Extera, Reneo, Scholar Rock, and Merck. HS is an employee of Roche Products and owns stock in Roche. HC-H is an employee of Genentech. SD is an employee of Roche Diagnostics/F Hoff mann-La Roche. AM and JS are employees of Genentech. TS is an employee of Genentech, and holds stocks, stock options, or bond holdings in Genentech. DEF has received grants from AbbVie, Actelion, Amgen, Bristol-Myers Squibb, Gilead, GlaxoSmithKline, National Institutes of Health, Novartis, Pfi zer, Roche/Genentech, UCB; consultancy fees from AbbVie, Actelion, Amgen, Bristol-Myers Squibb, Cytori, Janssen, Gilead, GlaxoSmith Kline, National Institutes of Health, Novartis, Pfi zer, and Roche/Genentech, UCB; and speaker fees from AbbVie, Actelion, and UCB. The remaining authors declare no competing interests.

AcknowledgmentsWe thank the investigators in the faSScinate trial (appendix p 1). We also thank Sara Duggan and Meryl Mandle, who provided writing services on behalf of F Hoff mann-La Roche, Sabrina Cheng for her involvement in the trial, and Robyn Rourick, Cécile Holweg, and Jason Hackney for their contributions to the biomarker analyses.

References1 Nikpour M, Stevens WM, Herrick AL, Proudman SM.

Epidemiology of systemic sclerosis. Best Practice Res Clin Rheumatol 2010; 24: 857–69.

2 O’Reilly S, Cant R, Ciechomska M, van Laar JM. Interleukin-6: a new therapeutic target in systemic sclerosis? Clin Transl Immunol 2013; 2: e4.

3 Kadono T, Kikuchi K, Ihn H, Takehara K, Tamaki K. Increased production of interleukin 6 and interleukin 8 in scleroderma fi broblasts. J Rheumatol 1998; 25: 296–301.

4 Desallais L, Avouac J, Frechet M, et al. Targeting IL-6 by both passive or active immunization strategies prevents bleomycin-induced skin fi brosis. Arthritis Res Ther 2014; 16: R157.

5 De Lauretis A, Sestini P, Pantelidis P, et al. Serum interleukin 6 is predictive of early functional decline and mortality in interstitial lung disease associated with systemic sclerosis. J Rheumatol 2013; 40: 435–46.

6 Khan K, Xu S, Nihtyanova S, et al. Clinical and pathological signifi cance of interleukin 6 overexpression in systemic sclerosis. Ann Rheum Dis 2012; 71: 1235–42.

7 Christmann RB, Lafyatis R. The cytokine language of monocytes and macrophages in systemic sclerosis. Arthritis Res Ther 2010; 12: 146.

8 Stifano G, Aff andi AJ, Mathes AL, et al. Chronic Toll-like receptor 4 stimulation in skin induces infl ammation, macrophage activation, transforming growth factor beta signature gene expression, and fi brosis. Arthritis Res Ther 2014; 16: R136.

9 van Lieshout AW, van der Voort R, le Blanc LM, et al. Novel insights in the regulation of CCL18 secretion by monocytes and dendritic cells via cytokines, toll-like receptors and rheumatoid synovial fl uid. BMC Immunol 2006; 7: 23.

10 Higashi-Kuwata N, Makino T, Inoue Y, Takeya M, Ihn H. Alternatively activated macrophages (M2 macrophages) in the skin of patient with localized scleroderma. Exp Dermatol 2009; 18: 727–29.

11 Kitaba S, Murota H, Terao M, et al. Blockade of interleukin-6 receptor alleviates disease in mouse model of scleroderma. Am J Pathol 2012; 180: 165–76.

12 Shima Y, Kuwahara Y, Murota H, et al. The skin of patients with systemic sclerosis softened during the treatment with anti-IL-6 receptor antibody tocilizumab. Rheumatology (Oxford) 2010; 49: 2408–12.

13 Elhai M, Meunier M, Matucci-Cerinic M, et al. Outcomes of patients with systemic sclerosis-associated polyarthritis and myopathy treated with tocilizumab or abatacept: a EUSTAR observational study. Ann Rheum Dis 2013; 72: 1217–20.

14 Castelino FV, George LM, Bain G, Goulet L, Lafyatis R, Tager AM. Autotaxin is overexpressed in systemic sclerosis (SSC) skin, mediates bleomycininduced dermal fi brosis via IL6 and is a target for SSC therapy. Ann Rheum Dis 2014; 73 (suppl 2): 153.

15 Okudaira S, Yukiura H, Aoki J. Biological roles of lysophosphatidic acid signaling through its production by autotaxin. Biochimie 2010; 92: 698–706.

16 Tokumura A, Carbone LD, Yoshioka Y, et al. Elevated serum levels of arachidonoyl-lysophosphatidic acid and sphingosine 1-phosphate in systemic sclerosis. Int J Med Sci 2009; 6: 168–76.

17 Pattanaik D, Postlethwaite AE. A role for lysophosphatidic acid and sphingosine 1-phosphate in the pathogenesis of systemic sclerosis. Discov Med 2010; 10: 161–67.

18 Hesselstrand R, Andreasson K, Wuttge DM, et al. Increased serum COMP predicts mortality in SSc: results from a longitudinal study of interstitial lung disease. Rheumatology (Oxford) 2012; 51: 915–20.

19 Tiev KP, Hua-Huy T, Kettaneh A, et al. Serum CC chemokine ligand-18 predicts lung disease worsening in systemic sclerosis. Eur Respir J 2011; 38: 1355–60.

20 Prasse A, Pechkovsky DV, Toews GB, et al. CCL18 as an indicator of pulmonary fi brotic activity in idiopathic interstitial pneumonias and systemic sclerosis. Arthritis Rheum 2007; 56: 1685–93.

21 Yamaguchi Y, Ono J, Masuoka M, et al. Serum periostin levels are correlated with progressive skin sclerosis in patients with systemic sclerosis. Br J Dermatol 2013; 168: 717–25.

22 Preliminary criteria for the classifi cation of systemic sclerosis (scleroderma). Subcommittee for scleroderma criteria of the American Rheumatism Association Diagnostic and Therapeutic Criteria Committee. Arthritis Rheum 1980; 23: 581–90.

23 Khanna D, Clements PJ, Furst DE, et al. Recombinant human relaxin in the treatment of systemic sclerosis with diff use cutaneous involvement: a randomized, double-blind, placebo-controlled trial. Arthritis Rheumatol 2009; 60: 1102–11.

24 Khanna D, Furst DE, Hays RD, et al. Minimally important diff erence in diff use systemic sclerosis: results from the D-penicillamine study. Ann Rheum Dis 2006; 65: 1325–29.

25 Maurer B, Graf N, Michel BA, et al, EUSTAR co-authors. Prediction of worsening of skin fi brosis in patients with diff use cutaneous systemic sclerosis using the EUSTAR database. Ann Rheum Dis 2015; 74: 1124–31.

26 Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profi les. Proc Natl Acad Sci USA 2005; 102: 15545–50.

27 Rice LM, Ziemek J, Stratton EA, et al. A longitudinal biomarker for the extent of skin disease in patients with diff use cutaneous systemic sclerosis. Arthritis Rheumatol 2015; 67: 3004–15.

28 Rice LM, Padilla CM, McLaughlin SR, et al. Fresolimumab treatment decreases biomarkers and improves clinical symptoms in systemic sclerosis patients. J Clin Invest 2015; 125: 2795–807.

29 Steen VD, Medsger TA Jr. Improvement in skin thickening in systemic sclerosis associated with improved survival. Arthritis Rheum 2001; 44: 2828–35.

30 Clements PJ, Hurwitz EL, Wong WK, et al. Skin thickness score as a predictor and correlate of outcome in systemic sclerosis: high-dose versus low-dose penicillamine trial. Arthritis Rheum 2000; 43: 2445–54.

31 Steen VD, Medsger TA Jr. Severe organ involvement in systemic sclerosis with diff use scleroderma. Arthritis Rheum 2000; 43: 2437–44.

32 Rubio-Rivas M, Royo C, Simeon CP, Corbella X, Fonollosa V. Mortality and survival in systemic sclerosis: systematic review and meta-analysis. Semin Arthritis Rheum 2014; 44: 208–19.

Articles


33 King TE Jr, Bradford WZ, Castro-Bernardini S, et al. A phase 3 trial of pirfenidone in patients with idiopathic pulmonary fi brosis. N Engl J Med 2014; 370: 2083–92.

34 Liu X, Mayes MD, Pedroza C, et al. Does C-reactive protein predict the long-term progression of interstitial lung disease and survival in patients with early systemic sclerosis? Arthritis Care Res 2013; 65: 1375–80.

35 Mathes AL, Christmann RB, Stifano G, et al. Global chemokine expression in systemic sclerosis (SSc): CCL19 expression correlates with vascular infl ammation in SSc skin. Ann Rheum Dis 2014; 73: 1864–72.

36 Kowal-Bielecka O, Landewe R, Avouac J, et al. EULAR recommendations for the treatment of systemic sclerosis: a report from the EULAR Scleroderma Trials and Research group (EUSTAR). Ann Rheum Dis 2009; 68: 620–28.

37 Nagaraja V, Denton CP, Khanna D. Old medications and new targeted therapies in systemic sclerosis. Rheumatology (Oxford) 2015; 54: 1944–53.

38 Pope JE, Bellamy N, Seibold JR, et al. A randomized, controlled trial of methotrexate versus placebo in early diff use scleroderma. Arthritis Rheum 2001; 44: 1351–58.

39 Tashkin DP, Elashoff R, Clements PJ, et al. Cyclophosphamide versus placebo in scleroderma lung disease. N Engl J Med 2006; 354: 2655–66.

40 van Laar JM, Farge D, Sont JK, et al. Autologous hematopoietic stem cell transplantation vs intravenous pulse cyclophosphamide in diff use cutaneous systemic sclerosis: a randomized clinical trial. JAMA 2014; 311: 2490–98.

Documents

Safety and efficacy of subcutaneous tocilizumab in adults ... · 5/5/2016 · body area with an increase of modiﬁ ed Rodnan skin score of at least 2 or two new body areas with