UvA-DARE is a service provided by the library of the University of Amsterdam (http://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Spondyloarthritis: From disease phenotypes to novel treatments

Paramarta, J.E.

Link to publication

Citation for published version (APA):Paramarta, J. E. (2014). Spondyloarthritis: From disease phenotypes to novel treatments.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s) and/or copyright holder(s),other than for strictly personal, individual use, unless the work is under an open content license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, please let the Library know, statingyour reasons. In case of a legitimate complaint, the Library will make the material inaccessible and/or remove it from the website. Please Askthe Library: https://uba.uva.nl/en/contact, or a letter to: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam,The Netherlands. You will be contacted as soon as possible.

Download date: 16 Aug 2020

10

Anti-interleukin-17A monoclonal antibody secukinumab in treatment

of ankylosing spondylitis: a randomised, double-blind, placebo-controlled trial

Dominique Baeten1, Xenofon Baraliakos2, Jürgen Braun2,

Joachim Sieper3, Paul Emery4, Désirée van der Heijde5,

Iain McInnes6, Jacob M. van Laar7, Robert Landewé1,8,

Paul Wordsworth9, Jürgen Wollenhaupt10,

Herbert Kellner11, Jacqueline Paramarta1, Jiawei Wei12,

Arndt Brachat13, Stephan Bek13, Didier Laurent13, Yali Li14,

Ying A Wang14, Arthur P Bertolino13, Sandro Gsteiger15,

Andrew M. Wright15, Wolfgang Hueber13

1Department of Clinical Immunology and Rheumatology, Academic Medical Center/University of Amsterdam, Amsterdam, The Netherlands,

2Rheumazentrum Ruhrgebiet, Herne, Germany, 3Medicine/Rheumatology Department,

Charité Campus Benjamin Franklin, Berlin, Germany, 4Leeds MSK Biomedical Research Unit,

University of Leeds, Leeds, United Kingdom, 5Rheumatology Department, Leiden University

Medical Center, Leiden, The Netherlands, 6Glasgow Biomedical Research Centre,

University of Glasgow, Glasgow, United Kingdom 7Musculoskeletal Research Group, Institute of Cellular Medicine,

Newcastle University, Newcastle upon Tyne, United Kingdom, 8Atrium Medical Center, Heerlen, The Netherlands,

9National Institute for Health Research, Oxford Musculoskeletal Biomedical Research Unit, and Nuffield Department of Orthopaedics,

Rheumatology and Musculoskeletal Sciences, Nuffield Orthopaedic Centre, Oxford, United Kingdom,

10Division of Rheumatology, Schön Klinik Hamburg Eilbeck, Hamburg, Germany,

11Division of Rheumatology, Centre for Inflammatory Joint Diseases, Munich, Germany,

12Novartis Pharma AG, Shanghai, China, 13Novartis Institutes for BioMedical Research, Basel, Switzerland,

14Novartis Institutes for BioMedical Research, Cambridge, MA, United States, 15Novartis Pharma AG, Basel, Switzerland

Lancet. 2013 Nov 23;382(9906):1705-13

SUMMARY

Background

Ankylosing spondylitis is a chronic immune-mediated inflammatory disease characterised by

spinal inflammation, progressive spinal rigidity, and peripheral arthritis. Interleukin 17 (IL-17) is

thought to be a key inflammatory cytokine in the development of ankylosing spondylitis, the

prototypical form of spondyloarthritis. We assessed the efficacy and safety of the anti-IL-17A

monoclonal antibody secukinumab in treating patients with active ankylosing spondylitis.

Methods

We did a randomised double-blind proof-of-concept study at eight centres in Europe (four in

Germany, two in the Netherlands, and two in the UK). Patients aged 18–65 years were randomly

assigned (in a 4:1 ratio) to either intravenous secukinumab (2 × 10 mg/kg) or placebo, given

3 weeks apart. Randomisation was done with a computer-generated block randomisation list

without a stratification process. The primary efficacy endpoint was the percentage of patients

with a 20% response according to the Assessment of SpondyloArthritis international Society

criteria for improvement (ASAS20) at week 6 (Bayesian analysis). Safety was assessed up to

week 28. This study is registered with ClinicalTrials.gov, number NCT00809159.

Findings

37 patients with moderate-to-severe ankylosing spondylitis were screened, and 30 were

randomly assigned to receive either intravenous secukinumab (n=24) or placebo (n=6). The final

efficacy analysis included 23 patients receiving secukinumab and six patients receiving placebo,

and the safety analysis included all 30 patients. At week 6, ASAS20 response estimates were

59% on secukinumab versus 24% on placebo (99.8% probability that secukinumab is superior to

placebo). One serious adverse event (subcutaneous abscess caused by Staphylococcus aureus)

occurred in the secukinumab-treated group.

Interpretation

Secukinumab rapidly reduced clinical or biological signs of active ankylosing spondylitis and

was well tolerated. It is the first targeted therapy that we know of that is an alternative to

tumour necrosis factor inhibition to reach its primary endpoint in a phase 2 trial.

134

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

INTRODUCTIONAnkylosing spondylitis is a chronic immune-mediated inflammatory disease, with an estimated

prevalence of 0.2–0.5% worldwide. It is characterised by spinal inflammation, progressive

spinal ankylosis due to new bone formation, peripheral arthritis and enthesitis, extra-articular

manifestations, familial clustering, and a strong genetic association with human leucocyte

antigen (HLA)-B27.1 Early onset of the disease in young adults, chronic spinal and extraspinal

inflammation, and progressive irreversible structural damage can lead to important morbidity,

functional deterioration, and socioeconomic burden.

Non-steroidal anti-inflammatory drugs (NSAIDs) and physiotherapy are the cornerstones

of treatment for ankylosing spondylitis. Conventional disease-modifying anti-rheumatic

drugs (DMARDs) are not effective for the axial symptoms of the disease. For patients with

inadequate response to NSAIDs, treatment with tumour necrosis factor (TNF)-blockers has

been recommended.2 However, no approved alternative therapies are available for the roughly

40% of patients who do not tolerate or respond to TNF-blockers. Thus, there is an unmet need

for medicines with new modes of action.

Interleukin 17A (IL-17A) has emerged as a novel therapeutic target for ankylosing spondylitis.

First, the disease shows a strong genetic association with a series of protective polymorphisms

in the IL-23 receptor (IL23R) gene, including rs11209026 (Arg381Gln).3 Functional analyses have

indicated that the Arg381Gln polymorphism impairs IL-23–dependent IL-17 production by Th17

cells, suggesting that genetically determined down-modulation of the IL-23/IL-17 axis could provide

protection against ankylosing spondylitis.4 Second, intracellular HLA-B27 misfolding leads to an

unfolded protein response that potentiates abnormal IL-23 production,5 and ankylosing spondylitis

macrophages produce increased levels of IL-23.6 Third, the number of circulating CD4+IL-17+ cells is

expanded in ankylosing spondylitis;7 this includes KIR3DL2-expressing T cells that respond to cell-

surface HLA-B27 homodimers8 and IL-17–producing γ/δ T cells.9 Fourth, inflamed target tissues

indicate an ankylosing spondylitis-specific increase in IL-17–producing innate immune cells.10,11 Fifth,

spondyloarthritis in HLA-B27 transgenic rats and experimental ankylosing enthesitis in (BXSB×NZB)

F1 mice were associated with an expansion of Th17 cells and with up-regulated IL-17 expression,

respectively.12 Moreover, IL-23 overexpression induces a spondyloarthritis-like disease in B10.RIII

mice via RAR-related orphan receptor γt(+)CD3(+)CD4(–)CD8(–) T cells that produce IL-17 and

IL-22.13 Finally, trials with anti-IL-12/IL-23 and anti-IL-17 agents have shown significant clinical efficacy

in psoriasis, a disease closely related to ankylosing spondylitis.14-19

We therefore undertook a proof-of-concept study to assess the efficacy and safety of

secukinumab, a fully human anti-IL-17A monoclonal antibody, in patients with active ankylosing

spondylitis.

METHODS

Study design

We did a 28-week multicentre, randomised, double-blind, placebo-controlled study between

March, 2009, and May, 2011, at eight centres in Europe (four in Germany, two in the Netherlands,

135

and two in the UK; appendix). After a 4-week screening period, patients were randomly

assigned (in a 4:1 ratio) to receive secukinumab or placebo at days 1 and 22.

The randomisation plan was generated by Novartis Drug Supply Management using a

validated system. It was reviewed and approved by the Biostatistics Quality Assurance group of

Novartis and locked after approval. Randomisation was done by using a computer-generated

block randomisation list without a stratification process. Data were accessible only to authorised

personnel until database lock. The unblinded pharmacist or designated person (eg, study nurse)

at each site received treatment allocation cards. When an eligible patient at a participating

study site was to be entered into the study, the unblended pharmacist or designated person

completed a randomization request by email to Novartis. The assigned randomisation number

was returned to the pharmacist within 24 h of the request. After inclusion of each new patient,

the pharmacist returned the email to confirm the patient’s randomisation number and the date

the patient received the first administration of study medication.

Efficacy and safety assessments were done up to week 28, with week 6 being the primary

endpoint. All week-6 analyses were designated as interim analyses. All centres received approval

from independent ethics committees or institutional review boards, and the study was done in

accordance with the principles of the Declaration of Helsinki. Signed informed consent for the

main and pharmacogenetic studies was obtained from each patient before any study-related

procedures were undertaken. The full study protocol is available from the sponsor.

Patients

The study included 30 patients, 18–65 years of age, with definite ankylosing spondylitis as defined

by the modified New York criteria,20 a score of at least 4 on the Bath ankylosing spondylitis disease

activity index (BASDAI), and a total back pain and nocturnal pain score of 40 or more on the visual

analogue scale (0–100 mm), irrespective of the maximum tolerated doses of NSAIDs.

Major exclusion criteria included total spinal ankylosis, concomitant fulfilment of criteria

for psoriatic arthritis, and presence of severe acute or subacute anterior uveitis. Patients with a

history of or current stable psoriasis or inflammatory bowel disease were allowed to participate.

Patients with active tuberculosis, hepatitis B or C, HIV, or any active systemic infection

within the 2 weeks before baseline were excluded. Patients with latent tuberculosis had to

complete their treatment and be considered cured before study entry. Patients with a history of

malignancy (except basal cell carcinoma or adequately treated carcinoma in situ of the cervix)

or significant cardiac, renal, neurological, psychiatric, endocrinologic, metabolic, or hepatic

disease were also excluded.

Previous use of DMARDs, cyclosporine, azathioprine, or TNF-blockers was allowed. However,

a maximum of ten patients with previous TNF-blocking therapy were allowed, to account for the

possibility that response rates for secukinumab in such patients could be lower, thereby biasing

the mean efficcacy readout. Washout periods of 3 months for adalimumab and certolizumab and

2 months for etanercept and infliximab were required before baseline. Patients were allowed

to continue any of the following medications if the dosage remained stable for at least 4 weeks

before the baseline visit and during the study: sulfasalazine (up to 3 g a day), methotrexate (up to

25 mg a week), prednisone or prednisone equivalent (up to 10 mg a day), and NSAIDs.

136

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

Study medication and dosage

The intravenous secukinumab dose of 10 mg/kg was based on previous pharmacokinetic and

pharmacodynamic experience with secukinumab in rheumatoid arthritis.15 A treatment regimen

of two infusions of 10 mg/kg given intravenously 3 weeks apart was used to achieve high

systemic exposure for induction of response, and because the 4-week toxicology coverage

available at study start did not allow a longer treatment period.

Endpoints

The primary efficacy endpoint was the percentage of patients with a 20% response according to

the Assessment of SpondyloArthritis international Society criteria for improvement (ASAS20)

at week 6.21 Secondary efficacy endpoints included ASAS40 (40% response according to ASAS

criteria for improvement) and ASAS5/6 responses (improvement in five of six domains: pain,

patient global assessment, function, inflammation, spinal mobility, C-reactive protein [acute-

phase reactant] without deterioration in the 6th domain), BASDAI,22 and assessment with the

ankylosing spondylitis quality-of-life (ASQoL) measure.23 All of these assessments were done at

screening, baseline, days 8, 15, and 29, and weeks 6, 8, 10, 12, 16, 20, 24, and 28.

Biological endpoints, which were measured in secukinumab-treated patients, included

erythrocyte sedimentation rate and serum C-reactive protein (CRP). Additionally, analyses

of S100-proteins A8, A9, and A12 in secukinumab patients were done by multiplexed liquid

chromatography–mass spectrometry multiple reaction monitoring. Sagittal MRI of the entire

spine in secukinumab-treated and placebo-treated patients was done using short tau inversion

recovery and T1-weighted sequencing to assess bone marrow oedema at baseline, week 6, and

week 28. MRI studies were analysed by an independent reader, who was blinded to treatment

allocation and chronology of images, using the Berlin score.24 To further assess the relevance of

targeting IL-17A, 13 genetic polymorphisms that are either associated with ankylosing spondylitis

or directly related to IL-17 were tested for association with response to secukinumab (appendix).25-

27 All genotyping was done by Novartis: single-nucleotide polymorphism (SNP) using TaqMan,

and HLA genotyping using sequencespecific oligonucleotide probes. ERAP1 transcript levels in

whole blood were determined using Affymetrix DNA microarrays (Santa Clara, CA, USA).

Adverse events and vital signs were assessed at every visit for safety evaluations. Laboratory

measurements were done at screening; baseline; days 2, 8, 15, 23, and 29; and weeks 6, 8, 10, 12,

16, 20, 24, and 28. Randomisation and blinding details are provided in the appendix.

Statistical analysis

The primary endpoint was analysed using Bayesian methods, as described previously.28 For the

placebo ASAS20 response rate, data from eight previous trials of ankylosing spondylitis (N=533)

were included.29 The information contained in these trials was transformed into an informative

prior distribution by means of the meta-analytic-predictive methodology described by

Neuenschwander and colleagues.30 This approach accounts for between-trial heterogeneity,

since the data from the previous trials are downweighted compared with the data collected

in this trial. The resulting prior distribution for the placebo response rate was equivalent to 43

patients (appendix). For the secukinumab response rate, an uninformative prior was used.

137

The predefined criterion to declare a positive proof-of-concept required a posterior probability

of at least 95% that the ASAS20 response rate on secukinumab is larger than that on placebo.

With data from 20 patients on secukinumab and five patients on placebo, the study showed

a 90% probability of achieving proof-of-concept, assuming true response rates of 25% on

placebo and 60% on secukinumab (appendix). Secondary and exploratory endpoint analyses

were non-Bayesian and are summarised using descriptive statistics. Exploratory (post-hoc)

analyses of MRI total scores and protein biomarkers were done to compare the change from

baseline and post-treatment measurements (Wilcoxon signed-rank test).

Patients who dropped out of the study were considered non-responders in all planned ASAS

assessments, irrespective of the reason for discontinuation. Comparisons between baseline

and post-treatment measurements, as well as between ASAS responders and non-responders,

applied Mann-Whitney tests.

An exploratory analysis of genetic data (13 polymorphisms shown to be associated with

ankylosing spondylitis or in the target IL-17A gene region) was done with a subgroup of

consenting patients taking secukinumab (N=22; appendix). The polymorphisms were tested

individually for association with ASAS20 and ASAS40 response at week 6 using logistic

regression models and with BASDAI using a linear regression model, with number of copies of

the minor (less common) allele carried by the individual as the predictor and baseline BASDAI

score as a covariate. A permutation test was done to account for small sample size and multiple

comparisons of polymorphisms, with genotypes randomly permuted 250 000 times.

Role of the funding source

An academic advisory committee was involved in study design and data interpretation, together with

authors from Novartis Institutes for Biomedical Research. The corresponding author had full access to

all the data in the study and had final responsibility for the decision to submit for publication.

RESULTSOf 37 patients with active ankylosing spondylitis who were screened, 30 were enrolled in the study

and randomly assigned to receive intravenous 2 × 10 mg/kg secukinumab (n=24) or placebo (n=6) on

day 1 and day 22 (figure 1). Baseline demographics were similar between the two groups (table 1). 22

of the 24 (92%) secukinumab-treated patients and three of the six (50%) placebo-treated patients

reached week 6. Of the 25 patients still in the study at week 6, 15 in the secukinumab group and all

three in the placebo group completed the study up to week 28. Patients who discontinued the study

were considered to be non-responders for the efficacy analysis. Reasons for discontinuation were

mainly unsatisfactory therapeutic effect, withdrawal of consent, and adverse events. Additionally, a

dosing error led to one patient in the secukinumab group being excluded from the efficacy analysis

but not from the safety analysis (as predefined by the protocol). Thus the efficacy analysis included

23 secukinumab patients and six placebo patients.

The primary efficacy outcome, ASAS20 response at week 6, was met by 14 (61%) of the

23 patients in the secukinumab group and one (17%) of the six patients in the placebo group.

Calculated response rates were 59% for the secukinumab group versus 24% for the placebo

138

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10group. The observed placebo response rate was similar to the results obtained from the meta-

analysis of eight comparable trials (26.5%).29 Bayesian analysis (table 2) indicated that the

probability of secukinumab inducing greater response rates than placebo was 99.8%.

The secondary efficacy endpoints, ASAS20, ASAS40, and ASAS5/6 response rates, and

BASDAI scores up to week 28 are summarised in the appendix and shown for the secukinumab-

treated patients in figure 2. Response rates for all ASAS measures and BASDAI were seen at week

6. Consistent with the loading dose regimen, there was a decreasing response rate by week 28.

Clinically relevant improvement in quality of life, defined as a change of two or more points in

ASQoL, was seen in 11 of 21 (52%) patients in the secukinumab group versus two of six (33%)

patients in the placebo group at day 29. Additional ASQoL data are shown in the appendix.

Mean serum CRP levels decreased from 14.4 mg/L (SD 17.60) at baseline to 6.1 mg/L (9.23)

at week 6 (p=0.03; figure 3A). Similarly, mean erythrocyte sedimentation rate decreased from

35.5 mm/h (SD 24.81) at baseline to 22.4 mm/h (16.27) at week 6 after secukinumab treatment

(p=0.11; appendix). Mean levels of the inflammatory biomarker protein S100A8 decreased from

1818.7 ng/mL (SD 716.85) at baseline to 1286.2 ng/mL (478.37) at week 6 (p=0.01; figure 3B).

Mean levels of the inflammatory biomarker protein S100A9 decreased from 2012.6 ng/mL (SD

835.46) at baseline to 1513.3 ng/mL (445.28) at week 6 (p=0.01; figure 3C). Additionally, ASAS20

response at week 6 was significantly correlated with both baseline levels of S100A8 (R=0.55)

and A9 (0.39) and changes between baseline and week 6 levels of S100A8 (0.43) and A9 (0.28).

Even stronger correlations were seen with the ASAS40 response at week 6 (R=0.40, 0.59, 0.52,

Figure 1. Trial profile

139

Table 1. Baseline characteristics

Safety analysis dataset Secukinumab† (N=24) Placebo (N=6) Total (N=30)

Age (years) 41.1±10.10 45.0±9.96 41.9±10.03

Men — no. (%) 14 (58) 5 (83) 19 (63)

Race — no. (%)

White 20 (83) 6 (100) 26 (87)

Black 1 (4) - 1 (3)

Asian 1 (4) - 1 (3)

Other 2 (8) - 2 (7)

Height (cm) 170.6±8.00 178.3±3.50 172.2±7.92

Weight (kg) 78.9±15.52 80.2±14.80 79.2±15.14

Body-mass index 27.1±4.81 25.3±5.07 26.7±4.83

Efficacy analysis dataset‡

Disease duration (years) 10.1±12.20 10.2±12.02 10.1±11.95

HLA-B27 — no. (%) 16 (70) 5 (83) 21 (72)

History of extra-axial involvement — no. (%)

Uveitis 7 (30) 2 (33) 9 (31)

Psoriasis 3 (13) 1 (17) 4 (14)

Inflammatory bowel disease 3 (13) 1 (17) 4 (14)

Dactylitis 1 (4) 1 (17) 2 (7)

Peripheral arthritis 12 (52) 4 (67) 16 (55)

Concomitant DMARD — no. (%) 8 (35) 3 (50) 11 (38)

Methotrexate 4 (17) - 4 (14)

Leflunomide 1 (4) - 1 (3)

Sulfasalazine 5 (22) 3 (50) 8 (28)

Other immunosuppressant§ 2 (9) - 2 (7)

NSAID — no. (%) 22 (96) 6 (100) 28 (97)

Steroid — no. (%) 3 (13) - 3 (10)

Oral 2 (9) - 2 (7)

Topical 1 (4) - 1 (3)

Prior TNF-blocker — no. (%) 10 (43) 3 (50) 13 (45)

ASQoL 12.1±3.07 10.3±0.57 11.7±3.72

BASDAI 7.1±1.40 7.2±1.76 7.1±1.45

BASMI 4.3±1.86 3.7±1.11 4.2±1.73

BASFI 6.4±1.83 5.3±3.27 6.2±2.18

MASES 4.3±3.54 1.7±1.63 3.8±3.40

LEI 1.2±1.67 0.2±0.41 1.0±1.55

CRP (mg/L) 13.3±17.23 13.2±15.72 13.3±16.65

Values are means ±SD. DMARD=disease-modifying anti-rheumatic drug, NSAID=non-steroidal anti-inflammatory drug, TNF=tumor necrosis factor, ASQoL=Ankylosing Spondylitis Quality of Life, BASDAI=Bath Ankylosing Spondylitis Disease Activity Index, BASMI=Bath Ankylosing Spondylitis Metrology Index, BASFI=Bath Ankylosing Spondylitis Functional Index, MASES=Maastricht Ankylosing Spondylitis Enthesis Score, LEI=Leeds Enthesis Index, CRP=C-reactive protein. † Secukinumab 2x10 mg/kg. ‡ Efficacy analysis dataset included only 29 participants, since one patient on secukinumab was excluded due to a dosing error. § Azathioprine or cyclosporine.

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

and 0.64, respectively), indicating that clinical improvement was paralleled by a decrease in

inflammatory parameters (appendix).

Findings from direct visualisation of the axial inflammatory lesions by MRI (figure 4) were

consistent with the clinical effects and results seen for the protein measures. The mean MRI

scores decreased from 9.2 (SD 8.87) at baseline to 6.6 (6.56) at week 6 (p=0.10) and 5.7 (6.20) at

week 28 (p=0.16) in the secukinumab group (appendix). The mean MRI scores were unchanged in

the placebo group: 20.6 (SD 20.18) at baseline, 21.0 (24.56) at week 6, and 19.0 (19.33) at week 28.

For the genetic biomarkers, a non-synonymous SNP of the ERAP1 gene, rs30187, which

leads to significantly reduced endopeptidase activity in vitro and is associated with protection

against ankylosing spondylitis,27,31 had a highly significant association with ASAS40 at week 6 in

secukinumab-treated patients (nominal p=8.14 × 10-⁵ and empirical p=0.004 in permutation test

after correction for multiple comparisons). The association of SNP rs30187 with the ASAS40

response was consistent over time (figure 5A). The analysis also yielded a significant p value when

restricted to the HLA-B27 positive subset of patients (n=17, p=0.0022 for association between

rs30187 and ASAS40 at week 6). A similar, but weaker association with ASAS20 was seen for

Figure 2. Primary and secondary endpoints over time in patients receiving secukinumab 2 × 10 mg/kg. Arrows below x-axis denote drug administration. Error bars in ASAS graphs represent mean (SE). Error bars in BASDAI graph represent mean (SD). ASAS=Assessment of SpondyloArthritis international Society criteria. BASDAI=Bath ankylosing spondylitis disease activity index.

141

Table 2. Primary endpoint Bayesian analysis of ASAS20 responders at week 6

TreatmentResponders,

n (%) Response rate*Difference

vs Placebo)†95% credibility

interval† Probability

Secukinumab‡ 14 (60.9) 59.2% 34.7% 11.5%-56.4% 99.8%

Placebo 1 (167) 24.5%

ASAS=Assessment of SpondyloArthritis international Society criteria. *Means from the posterior beta (0.5 + x, 1 + n − x) distribution for secukinumab and beta (11 + x, 32 + n − x) distribution for placebo, where x represents the number of responders and n – x represents the number of non-responders in the corresponding treatment group. †Difference in response rates simulated from the posterior probability distributions of secukinumab and placebo. ‡Secukinumab 2 × 10 mg/kg. §The efficacy dataset included only 23 of 24 patients in the secukinumab group, since one patient was excluded due to a dosing error.

Figure 3. Biological markers of response. C-reactive protein (A), S100A8 (B), and S100A9 (C) levels at week 6, after secukinumab (2 × 10 mg/kg) administration at weeks 0 and 3.

142

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

Fig

ure

4. R

educ

tio

n o

f in

f am

mat

ion

in t

ho

raci

c (T

2–T3

) an

d (

T7–

T8)

unit

s at

bas

elin

e an

d a

t w

eek

6 in

a p

atie

nt

trea

ted

wit

h s

ecuk

inum

ab.

143

Figure 5. Association of the rs30187 genotype for the ERAP1 gene with mean ASAS40 response rates (A), mean ASAS20 response rates (B), and mean change in BASDAI from baseline (C) in patients treated with secukinumab. Error bars in ASAS graphs represent mean (SE). Error bars in BASDAI graph represent mean (SD). Polymorphisms were tested individually for association with ASAS40 and ASAS20 at week 6 using a logistic regression model, and with BASDAI at week 24 using a linear regression model, with number of copies of the minor (less common) allele carried by the individual as the predictor and baseline BASDAI score as a covariate. Minor allele/AS risk allele=T. Major allele=C. ASAS=Assessment of SpondyloArthritis international Society criteria. BASDAI=Bath ankylosing spondylitis disease activity index. CRP=C-reactive protein. SE=standard error.

144

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

SNP rs30187 (figure 5B), but only a minor association was seen with BASDAI changes (figure 5C,

appendix). Supporting the notion that these SNPs have a biological effect, the carriers of the ERAP1

risk alleles of rs30187 or rs27434 typically showed higher levels of ERAP1 gene expression, whereas

homozygous non-carriers mostly showed lower transcript levels (appendix). Additionally, an

association with BASDAI changes over time was seen for IL23R polymorphisms rs11209032 and

rs2201841. However, no association of IL23R polymorphisms with ASAS40 was seen (appendix).

Data for all randomised patients were used for the safety analysis (table 3). Since this study was

not powered for statistical comparisons between the secukinumab (N=24) and placebo groups

(N=6), only descriptive analyses are provided. The numbers and rates of patients with at least one

adverse event were similar between secukinumab (23 patients [96%]) and placebo (6 [100%]).

The most frequently reported adverse events on secukinumab were nasopharyngitis and

headache (table 3). The incidence of infections was higher for secukinumab than for placebo

(table 3). One serious adverse event was seen in the secukinumab group: at week 9 a patient

developed an abscess of the foot caused by Staphylococcus aureus that required antibiotic

therapy and hospitalisation for surgical drainage, and the patient was discontinued from the

study. Several cases of leucopenia and neutropenia were reported; all were graded Common

Terminology Criteria grade 1, with no apparent temporal relation to infections. No evidence of

formation of antidrug antibodies or clinical signs of immunogenicity were noted.

DISCUSSIONSecukinumab was associated with rapid and significant reduction in the signs and symptoms of

active ankylosing spondylitis in patients with inadequate response to NSAIDs. 60% of patients

in the secukinumab group reached the primary endpoint of ASAS20 responses at week 6,

indicating a 99.8% probability that secukinumab was more effective than placebo (table 2, figure

2). Favourable changes in secondary clinical outcomes, including quality-of-life measures,

further supported clinical efficacy.

Secukinumab treatment was associated with a rapid and significant decrease in acute-

phase parameters. Since CRP and S100 proteins were only weakly correlated, S100 proteins

might have added value as markers of inflammation. There were also decreased MRI scores

for spinal inflammatory lesions. The clinical response to secukinumab was also significantly

associated with genetic polymorphisms in ERAP1, and showed a trend towards association with

polymorphisms in IL23R. Although these data suggest that both genes could be associated with

a better clinical response to secukinumab, future experiments are needed to clarify the potential

role of these disease-risk alleles in the response to an anti-IL-17 therapy. Nevertheless, our

findings indicate a strong association between pathway-related genetic risk factors and clinical

response to therapeutic targeting of the pathway in immune-mediated inflammatory disease.

Larger and more extensive analysis of the biomarker value of ERAP1 and IL23R SNPs for the

prediction of clinical response are needed. Taken together, these data provide evidence that

IL-17 blockade modulates clinical and biological signs of inflammation and might therefore be a

valid therapeutic approach in ankylosing spondylitis. Biomarkers related to antigen processing

and presentation, as well as to the IL-23/IL-17 axis, could be useful for prediction of outcomes.

145

Table 3. Most frequent adverse events seen in more than 5% of patients on secukinumab

Secukinumab* N=24 Placebo N=6

All adverse events

Patients with an adverse event 23 (95.8) 6 (100.0)

Most frequent adverse events

Headache 7 (29.2) —

Nasopharyngitis 7 (29.2) —

Nausea 4 (16.7) 1 (16.7)

Diarrhea 4 (16.7) —

Pyrexia 4 (16.7) —

Back pain 3 (12.5) 1 (16.7)

Fatigue 3 (12.5) 1 (16.7)

Paresthesia 3 (12.5) 1 (16.7)

Arthralgia 3 (12.5) —

Influenza 3 (12.5) —

Oropharyngeal pain 3 (12.5) —

Dizziness 2 (8.3) 1 (16.7)

Oral herpes 2 (8.3) 1 (16.7)

Pruritus 2 (8.3) 1 (16.7)

Abdominal pain, upper 2 (8.3) —

Bronchitis 2 (8.3) —

Cough 2 (8.3) —

Dysgeusia 2 (8.3) —

Gastroenteritis 2 (8.3) —

Mouth ulceration 2 (8.3) —

Myalgia 2 (8.3) —

Rhinitis 2 (8.3) —

Adverse events related to infections

Nasopharyngitis 7 (29.2) —

Influenza 3 (12.5) —

Oral herpes 2 (8.3) 1 (16.7)

Gastroenteritis 2 (8.3) —

Data are n (%). * Secukinumab 2 x 10 mg/kg.

To the best of our knowledge, this study is the first to provide evidence for clinical efficacy of a

non-TNF-blocker targeted therapy in ankylosing spondylitis (panel). Recent trials with abatacept,

IL-6 blockade, and the small molecule apremilast, failed to show significant clinical efficacy in

ankylosing spondylitis.25 Rituximab plus methylprednisolone showed modest clinical effects

in a small open study.32 Ustekinumab, a fully human monoclonal antibody directed against the

146

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

p40 subunit shared by IL-12 and IL-23, is currently in phase 2 trials (ClinicalTrials.gov number

NCT01330901). Several other compounds (including tofacitinib, a Janus kinase 3 inhibitor [not yet

listed in ClinicalTrials.gov]) have entered clinical trials, with no or mixed results reported so far.25

IL-17 blockade is currently being explored as a therapeutic strategy for various immune-

mediated inflammatory diseases. Randomised controlled trials have reported consistent

clinical efficacy of this approach in psoriasis,15-19 a disease that shares certain IL23R genetic

associations with Crohn’s disease and ankylosing spondylitis. In Crohn’s disease, however,

anti-IL-17A blockade with secukinumab was not superior to placebo,28 and two trials with an

IL-17R antagonist (AMG 827) were stopped (ClinicalTrials.gov numbers NCT01199302 and

NCT01150890), indicating that clinical outcomes of IL-17 inhibition could vary among conditions

that share genetic risk factors related to the IL-23/IL-17 axis. Therefore, there was good reason

to design and undertake a proof-of-concept trial to investigate the safety and potential efficacy

of secukinumab in patients with active ankylosing spondylitis.

To make the best use of the existing information on placebo response rates (based on eight

previous trials of biological agents in active ankylosing spondylitis), these historical data were

integrated into the primary end point analysis using Bayesian methods, allowing us to reduce

the number of placebo patients required.29 The number of treated patients (N=24) was similar

to the numbers included in other proof-of-concept randomised trials with TNF-blockers in

ankylosing spondylitis (N=35) and spondyloarthritis (N=20).2,33 On the basis of previous data

in psoriasis and rheumatoid arthritis,15 we set the primary endpoint at week 6 (after only two

administrations of drug or placebo). A set of inflammatory, imaging, and pharmacogenetic

outcomes were included in the study to substantiate clinical data with biological evidence.

Study limitations included the inability to undertake sufficiently powered statistical

comparisons for the secondary efficacy endpoints because of the limited placebo sample size.

Tender and swollen joint counts and enthesitis scores were planned as secondary endpoints,

but were seen at such low frequencies that we were unable to assess the potential treatment

effect of secukinumab. The exploratory biomarker analyses in this study were also based on

small sample sizes; hence the possibility that the observed differences occurred by chance

cannot be excluded. Small sample sizes could also explain the different strength of associations

at different timepoints, and between assessments. For the MRI analyses, the higher baseline

scores for inflammation reported in the placebo group might represent a confounding factor

in interpreting the treatment effect seen for secukinumab. However, recent studies in patients

receiving anti-TNF agents34,35 have indicated that higher baseline MRI scores do not identify

therapeutic non-responders, but rather identify those patients who are candidates for better

response to anti-inflammatory treatment, a hypothesis supported by our results. A confounding

effect of background medication on transcriptional profiles cannot be ruled out. A post-hoc

analysis of patient subgroups according to previous TNF inhibitor exposure was done but was

compromised due to an inadequate number of patients (and thus not included in this paper).

The preliminary safety profile from our trial was similar to those reported in larger phase

2 studies with secukinumab in other indications, in regards to both the type and frequency of

adverse events.15,18,19,28,36 However, the small size of our study cohort did not permit definitive

conclusions regarding the safety of secukinumab. Larger and longer-term studies are needed

to confirm our findings.

147

PANEL: RESEARCH IN CONTEXT

Systematic review

We searched the PubMed database, using the terms “ankylosing spondylitis”, “biologic

therapy”, and “IL-17”, for English-language articles published up to April, 2013. IL-17 has been

proposed to be a key inflammatory cytokine in ankylosing spondylitis,3-13 the prototypical

form of spondyloarthritis. The IL-17 blockers secukinumab (anti-IL-17A), ixekizumab (anti-

IL-17A), and brodalumab (anti-IL-17RA) have shown efficacy in phase 2 trials in psoriasis,15-

19 a spondyloarthritis-related inflammatory skin disease. A trend towards clinical efficacy

was recently reported for secukinumab in psoriatic arthritis, a distinct but related form of

spondyloarthritis. The current report describes the first clinical trial assessing the safety and

efficacy of IL-17 blockade in ankylosing spondylitis.

Interpretation

Treatment with secukinumab induced a significant and clinically relevant reduction of disease activity

in active ankylosing spondylitis as early as 6 weeks after initiation of treatment. IL-17 blockade could

be the first therapeutic alternative to TNF blockade in NSAID-resistant ankylosing spondylitis.

Acknowledgments We thank Vasundhara Pathak, Novartis Healthcare, Hyderabad, India,

for support in the writing of this manuscript. Editorial assistance was provided by BioScience

Communications, New York, NY, USA, supported by Novartis Pharma AG, Basel, Switzerland.

REFERENCES1 Dougados M, Baeten D. Spondyloarthritis. Lancet

2011; 377: 2127–37.

2 Braun J, Brandt J, Listing J, et al. Treatment of active ankylosing spondylitis with infliximab: a randomised controlled multicenter trial. Lancet 2002; 359: 1187–93.

3 Burton PR, Clayton DG, Cardon LR, et al. Association scan of 14,500 nonsynonymous SNPs in four diseases identifies autoimmunity variants. Nat Genet 2007; 39: 1329–13.

4 Sarin R, Wu X, Abraham C. Inflammatory disease protective R381Q IL23 receptor polymorphism results in decreased primary CD4+ and CD8+ human T-cell functional responses. Proc Natl Acad Sci USA 2011; 108: 9560–65.

5 DeLay ML, Turner MJ, Klenk EI, Smith JA, Sowders DP, Colbert RA. HLA-B27 misfolding and the unfolded protein response augment interleukin-23 production and are associated with Th17 activation in transgenic rats. Arthritis Rheum 2009; 60: 2633–43.

6 Zeng L, Lindstrom MJ, Smith JA. Ankylosing spondylitis macrophage production of higher levels of interleukin-23 in response to lipopolysaccharide without induction of a significant

7 unfolded protein response. Arthritis Rheum 2011; 63: 3807–17.

8 Shen H, Goodall JC, Hill Gaston JS. Frequency and phenotype of peripheral blood Th17 cells in ankylosing spondylitis and rheumatoid arthritis. Arthritis Rheum 2009; 60: 1647–56.

9 Bowness P, Ridley A, Shaw J. Th17 cells expressing KIR3DL2+ and responsive to HLA-B27 homodimers are increased in ankylosing spondylitis. J Immunol 2011; 186: 2672–80.

10 Kenna TJ, Davidson SI, Duan R, et al. Enrichment of circulating interleukin-17-secreting interleukin-23 receptor-positive γ/δ T cells in patients with active ankylosing spondylitis. Arthritis Rheum 2012; 64: 1420–29.

11 Appel H, Maier R, Wu P, et al. Analysis of IL-17+ cells in facet joints of patients with spondyloarthritis

148

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

suggests that the innate immune pathway might be of greater relevance than the Th17-mediated adaptive immune response. Arthritis Res Ther 2011; 13: R95.

12 Noordenbos T, Yeremenko N, Gofita I, et al. Interleukin-17-positive mast cells contribute to synovial inflammation in spondylarthritis. Arthritis Rheum 2012; 64: 99–109.

13 Glatigny S, Fert I, Blaton MA, et al. Proinflammatory Th17 cells are expanded and induced by dendritic cells in spondylarthritis-prone HLA-B27-transgenic rats. Arthritis Rheum 2011; 64: 110–20.

14 Sherlock JP, Joyce-Shaikh B, Turner SP, et al. IL-23 induces spondyloarthropathy by acting on ROR-γt+ CD3+CD4-CD8- entheseal resident T cells. Nat Med 2012; 18: 1069–76.

15 Krueger GG, Langley RG, Leonardi C, et al, for the CNTO 1275 Psoriasis Study Group. A human interleukin-12/23 monoclonal antibody for the treatment of psoriasis. N Engl J Med 2007; 356: 580–92.

16 Hueber W, Patel DD, Dryja T, et al. Effects of AIN457, a fully human antibody to interleukin-17A, on psoriasis, rheumatoid arthritis, and uveitis. Sci Transl Med 2010; 2: 52ra72.

17 Papp KA, Leonardi C, Menter A, et al. Brodalumab, an anti-interleukin-17-receptor antibody for psoriasis. N Engl J Med 2012; 366: 1181–89.

18 Leonardi C, Matheson R, Zachariae C, et al. Anti-interleukin-17 monoclonal antibody ixekizumab in chronic plaque psoriasis. N Engl J Med 2012; 366: 1190–99.

19 Rich P, Sigurgeirsson B, Thaci D, et al. Secukinumab induction and maintenance therapy in moderate-to-severe plaque psoriasis: a randomized, double-blind, placebo-controlled, phase II regimen-finding study. Br J Dermatol 2013; 168: 402–11.

20 Papp KA, Langley RG, Sigurgeirsson B, et al. Efficacy and safety of secukinumab in the treatment of moderate-to-severe plaque psoriasis: a randomized, double-blind, placebo-controlled phase II dose-ranging study. Br J Dermatol 2013; 168: 412–21.

21 Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984; 27: 361–68.

22 Anderson JJ, Baron G, van der Heijde D, Felson DT, Dougados M. Ankylosing spondylitis assessment group preliminary definition of short-term improvement in ankylosing spondylitis. Arthritis Rheum 2001; 44: 1876–86.

23 Garrett S, Jenkinson T, Kennedy LG, Whitelock H, Gaisford P, Calin A. A new approach to defining disease status in ankylosing spondylitis: the Bath

Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994; 21: 2286–91.

24 Doward LC, Spoorenberg A, Cook SA, et al. Development of the ASQoL: a quality of life instrument specific to ankylosing spondylitis. Ann Rheum Dis 2003; 62: 20–26.

25 Lukas C, Braun J, van der Heijde D, et al, for the ASAS/OMERACT MRI in AS Working Group. Scoring inflammatory activity of the spine by magnetic resonance imaging in ankylosing spondylitis: a multireader experiment. J Rheumatol 2007; 34: 862–70.

26 Kiltz U, Heldmann F, Baraliakos X, Braun J. Treatment of ankylosing spondylitis in patients refractory to TNF-inhibition: are there alternatives? Curr Opin Rheumatol 2012; 24: 252–60.

27 Australo-Anglo-American Spondyloarthritis Consortium (TASC). Genome-wide association study of ankylosing spondylitis identifies non-MHC susceptibility loci. Nat Genet 2010; 42: 123–27.

28 Evans DM, Spencer CC, Pointon JJ, et al. Interaction between ERAP1 and HLA-B27 in ankylosing spondylitis implicates peptide handling in the mechanism for HLA-B27 in disease susceptibility. Nat Genet 2011; 43: 761–67.

29 Hueber W, Sands BE, Lewitzky S, et al, for the Secukinumab in Crohn’s Disease Study Group. Secukinumab, a human anti-IL-17A monoclonal antibody, for moderate to severe Crohn’s disease: unexpected results of a randomised, double-blind placebo-controlled trial. Gut 2012; 61: 1693–1700.

30 McLeod C, Bagust A, Boland A, et al. Adalimumab, etanercept and infliximab for the treatment of ankylosing spondylitis: a systematic review and economic evaluation. Health Technol Assess 2007; 11: 1–158, iii–iv.

31 Neuenschwander B, Capkun-Niggli G, Branson M, Spiegelhalter DJ. Summarizing historical information on controls in clinical trials. Clin Trials 2010; 7: 5–18.

32 Kochan G, Krojer T, Harvey D, et al. Crystal structures of the endoplasmic reticulum aminopeptidase-1 (ERAP1) reveal the molecular basis for N-terminal peptide trimming. Proc Natl Acad Sci USA 2011; 108: 7745–50.

33 Song IH, Heldmann F, Rudwaleit M, et al. Different response to rituximab in tumor necrosis factor blocker–naive patients with active ankylosing spondylitis and in patients in whom tumor necrosis factor blockers have failed: a twenty-four–week clinical trial. Arthritis Rheum 2010; 62: 1290–97.

34 Van Den Bosch F, Kruithof E, Baeten D, et al. Randomized double-blind comparison of chimeric monoclonal antibody to tumor necrosis

149

factor alpha (infliximab) versus placebo in active spondylarthropathy. Arthritis Rheum 2002; 46: 755–65.

35 Braun J, Landewé R, Hermann KG, et al, for the ASSERT Study Group. Major reduction in spinal inflammation in patients with ankylosing spondylitis after treatment with infliximab: results of a multicenter, randomized, double-blind, placebo-controlled magnetic resonance imaging study. Arthritis Rheum 2006; 54: 1646–52.

36 Rudwaleit M, Schwarzlose S, Hilgert ES, Listing J, Braun J, Sieper J. MRI in predicting a major clinical response to anti-tumour necrosis factor treatment in ankylosing spondylitis. Ann Rheum Dis 2008; 67: 1276–81.

37 Genovese MC, Durez P, Richards HB, et al. Efficacy and safety of secukinumab in patients with rheumatoid arthritis: a phase II, dose-finding, double-blind, randomised, placebo controlled study. Ann Rheum Dis 2013; 72: 863–69.

150

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

SUPPLEMENTARY APPENDIX

METHODS

Study design

The study was sponsored by Novartis Pharma AG. Academic advisors and Novartis personnel

designed the study. ClinBay (Genappe, Belgium) conducted the data analyses, under the

direction of the sponsor, and all authors had access to the data and vouch for the completeness

and accuracy of the data and data analyses. The first draft of the manuscript was written by

Novartis personnel, with inputs from all authors. All authors reviewed and provided feedback

on the subsequent versions and agreed to submit the manuscript for publication.

Statistical analysis

The predefined criterion for declaring a positive “Proof of Concept” (PoC) required a posterior

probability of at least 95% that the ASAS20 (20% response according to Assessment of

SpondyloArthritis international Society criteria) response rate for secukinumab patients would

be higher than that for placebo patients. Sample size calculations were based on the number

of subjects required to meet the above PoC criterion at the final analysis when assuming true

underlying response rates of 25% and 60% for placebo and secukinumab, respectively, and were

done by simulation. With data from 20 patients on secukinumab and 5 patients on placebo, the

study provided at least a 90% probability of achieving PoC under the above assumptions. An

additional 5 subjects were included in the study to allow for dropouts and incomplete data. A

4:1 randomization ratio was chosen to reduce the number of placebo-treated patients while

maintaining a double-blinded study design and to allow a limited study size in the absence of

knowledge of the risk/benefit of secukinumab in ankylosing spondylitis (AS).

The primary analysis of the study was a Bayesian analysis of the ASAS20 response rate after 6

weeks of treatment. An informative prior was used for the proportion of responders in the 4 placebo

treatment group. The prior distribution was a Beta distribution with parameters 11 and 32. The prior

for the active treatment group was also a Beta distribution with parameters 0.5 and 1. The derivation

of these prior distributions was based on a meta-analysis of historical data as described below.

From a review of anti–tumor necrosis factor (TNF)-α treatment in ankylosing spondylitis,1

historical data were available from eight randomized placebo-controlled clinical trials in

ankylosing spondylitis patients. The earliest timepoint assessed in this review was 12 weeks after

dosing. Assuming a stable placebo response rate between weeks 6 and 12, these data were used

in the derivation of the historical data prior.

A random effects meta-analysis2 of the 8 historical trials was performed assuming

exchangeable placebo response rates on the logit scale. Using this model, the predictive

distribution for the proportion of responders on placebo in a new study was derived, leading

to an estimated response rate of 25% (and a 95% credible interval of 13% to 40%). For ease of

use and interpretation, this predictive distribution was approximated by a Beta density with

matching mean and standard deviation. The result of this process was the Beta (shape1=11,

shape2=32) distribution, which was used as the historical data prior for the placebo group in the

151

1 McLeod C, Bagust A, Boland A, et al. Adalimumab, etanercept and infliximab for the treatment of ankylosing spondylitis: a systematic review and economic evaluation. Health Technol Assess 2007;11(28):1–158,iii–iv.

2 Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian Approaches to Clinical Trials and Health-Care Evaluation. John Wiley & Sons, New York City, 2004.

3 Australo-Anglo-American Spondyloarthritis Consortium (TASC), Reveille JD, Sims AM, Danoy P, et al. Genome-wide association study of ankylosing spondylitis identifies non-MHC susceptibility loci. Nat Genet 2010;42(2):123–7.

4 Burton PR, Clayton DG, Cardon LR, et al. Association scan of 14,500 nonsynonymous SNPs

in four diseases identifies autoimmunity variants. Nat Genet 2007;39:1329–37.

5 Evans DM, Spencer CC, Pointon JJ, et al. Interaction between ERAP1 and HLA-B27 in ankylosing spondylitis implicates peptide handling in the mechanism for HLA-B27 in disease susceptibility. Nat Genet 2011;43:761–7.

6 Sáfrány E, Pazár B, Csöngei V, et al. Variants of the IL23R gene are associated with ankylosing spondylitis but not with Sjögren syndrome in Hungarian population samples. Scand J Immunol 2009;70:68–74.

7 Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP. Summaries of Affymetrix GeneChip probe level data. Nucleic Acids Res 2003;31(4):e15.

primary analysis. The information content of this prior was equivalent to observing 11 responders

in a sample of size 43. In other terms, the equivalent sample size of this prior was 43 patients.

The prior distribution for the proportion of responders in the active group was also a Beta

distribution. One of the parameters was set to 1. The other parameter was chosen such that

there was an approximately 50:50 chance that the responder rate on active treatment would

be 5 greater than the responder rate on placebo (based on the prior distributions only). Thus a

Beta (shape1=0.5, shape2=1) distribution was chosen.

An exploratory analysis of genetic data was performed on a subgroup of consenting

secukinumab patients (N=22). Thirteen genetic polymorphisms previously shown to be either

associated with AS3-6 or in the target IL-17A gene region were selected to be tested for association

with response to secukinumab. All genotyping was performed by Novartis; single nucleotide

polymorphism (SNP) genotyping was done using TaqMan, and human leukocyte antigen (HLA)

genotyping was done using sequence-specific oligonucleotide (SSO). The 13 polymorphisms were

tested individually for association with ASAS20 and ASAS40 at week 6 using logistic regression

models and with Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) scores using a linear

regression model, with the efficacy endpoint as the dependent variable, the number of copies of

the minor (i.e., less common) allele carried by the individual as the predictor, and baseline BASDAI

score as a covariate. All statistical tests were two-tailed, with no imputation for missing data. A

permutation test was performed to account for small sample size and multiple comparisons of 13

polymorphisms, with genotypes randomly permuted 250,000 times.

ERAP1 gene transcript levels in whole blood were determined using Affymetrix DNA

microarrays (Human Genome 133 Plus 2.0). The Affymetrix ‘.cel’ files were normalized using the

RMA procedure7 and analyzed with the Partek Genomics Suite 6.6 beta software.

REFERENCES

152

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

Supplementary Table 1. Participating sites and Principal Investigators for secukinumab Study CAIN457A2209 (NCT00809159)

Participating Site Principal Investigator No. Patients

Germany

Universitaetsklinikum Benjamin FranklinMed. Klinik und Poliklinik I/RheumatologieHindenburgdamm 3012203 Berlin

Joachim Sieper 10

Rheumazentrum RuhrgebietSt.-Josefs-KrankenhausLandgrafenstr. 1544652 Herne

Jürgen Braun 4

Klinikum Eilbek - Schön KlinikenDehnhaide 12022081 Hamburg

Jürgen Wollenhaupt 3

Netherlands

EULAR & FOCIS Center of ExcellenceAcademic Medical CenterUniversity of Amsterdam, F4-2181100 22700 Amsterdam

Dominique Baeten 10

Maastricht University Medical CentreDepartment of Medicine, Division of RheumatologyMaastricht University Medical Centre6202 AZ5800 Maastricht

Robert Landewé 1

United Kingdom

Institute of Cellular MedicineSchool of Clinical Medical Sciences Newcastle University 4th Floor, Catherine Cookson Building The Medical School, Framlington PlaceNE2 4HH Newcastle upon Tyne

Jacob M. van Laar 2

153

Sup

ple

men

tary

Tab

le 2

. Pha

rmac

oge

neti

c an

alys

is: a

sso

ciat

ion

wit

h cl

inic

al r

esp

ons

e

Var

ian

tG

ene

Alle

les

(m

inor

/maj

or)

Posi

tio

n

Wee

k 4

P-v

alue

sW

eek

6 P-

valu

esW

eek

24 P

-val

ues

ASA

S20

ASA

S40

BASD

AI

ASA

S20

ASA

S40

BASD

AI

ASA

S20

ASA

S40

BASD

AI

rs11

2090

32IL

-23R

A/G

Do

wns

trea

m0

.40

0.4

50

.41

0.2

70

.81

0.2

90

.98

0.5

53.

65E-

03

rs22

018

41IL

-23R

G/A

Intr

oni

c0

.26

0.7

80

.58

0.2

40

.39

0.4

60

.56

0.4

58.

52E-

03

rs11

2090

26IL

-23R

A/G

Non

syno

nym

ous

0.4

50

.83

0.7

30

.71

0.9

70

.77

0.9

00

.20

0.6

9

rs10

8653

31—

A/G

Gen

om

ic0

.89

0.4

90

.86

0.6

40

.75

0.2

40

.67

1.0

00

.41

rs23

1017

3IL

-1R2

A/C

Do

wns

trea

m0

.29

0.5

90

.75

0.18

0.4

60

.70

0.0

190

.42

0.3

7

rs43

3313

0A

NTX

R2C

/TIn

tro

nic

0.2

10

.99

0.4

50

.69

0.5

60

.71

0.6

00

.59

0.0

80

rs22

4294

4—

A/G

Gen

om

ic0

.89

0.3

00

.41

0.12

0.3

40

.28

0.3

20

.55

0.9

5

rs27

434

ERA

P1A

/GSy

nony

mo

us0

.33

0.0

830

.67

0.2

14.

75E-

03

0.4

70

.54

0.7

10

.44

rs30

187

ERA

P1T/

CN

onsy

nony

mou

s0

.029

4.45

E-0

30

.26

0.2

28.

14E-

05

0.2

20

.46

0.4

40

.61

rs19

7422

6IL

-17A

T/C

3’ U

tr0

.20

0.9

00

.23

0.5

60

.87

0.6

40

.60

0.14

0.2

0

rs77

4790

9IL

-17A

A/G

3’ U

tr0

.60

0.2

00

.24

0.4

00

.36

0.6

80

.68

0.6

40

.74

HLA

-DRB

1*0

4H

LA-D

RB1

  

0.7

40

.65

0.3

60

.25

0.6

20

.49

0.9

40

.22

0.6

1

HLA

-B*2

7H

LA-B

  

0.3

10

.58

0.9

60

.61

0.10

0.5

30

.74

0.5

50

.83

ASA

S d

eno

tes

Ass

essm

ent

of

Spo

ndyl

oA

rthr

itis

inte

rnat

iona

l So

ciet

y cr

iter

ia, B

ASD

AI d

eno

tes

Bath

Ank

ylo

sing

Sp

ond

ylit

is D

isea

se A

ctiv

ity

Ind

ex. B

* 27 c

onf

ers

the

grea

test

kno

wn

risk

for

anky

losi

ng s

po

ndyl

itis

.

154

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

Sup

ple

men

tary

Tab

le 3

. Sec

ond

ary

effica

cy a

naly

sis,

all

tim

epo

ints

Vis

it

ASA

S20

ASA

S40

ASA

S5/6

BASD

AI

ASQ

oL

Secu

kin

umab

* (N

=23)

n (

%)

Plac

ebo

(N

=6)

n (

%)

Secu

kin

umab

(N

=23)

n (

%)

Plac

ebo

(N

=6)

n (

%)

Secu

kin

umab

(N

=23)

n (

%)

Plac

ebo

(N

=6)

n (

%)

Secu

kin

umab

(N

=23)

n, m

ean

±SD

Plac

ebo

(N

=6)

n, m

ean

±SD

Secu

kin

umab

(N

=23)

n, m

ean

±SD

Plac

ebo

(N

=6)

n, m

ean

±SD

Day

89/

23 (

39)

2/6

(33)

4/23

(17

)0/

6 (0

)5/

23 (

22)

1/6

(17)

23, 5

.27±

2.31

6, 5

.97±

2.90

Day

159/

23 (

39)

1/5

(20

)4/

23 (

17)

0/5

(0)

5/23

(22

)0/

5 (0

)23

, 5.19

±2.2

95,

6.4

8±2.

48

Day

29

11/2

3 (4

8)0/

6 (0

)8/

23 (

35)

0/6

(0)

8/23

(35

)0

/ 6

(0

)23

, 4.7

5±2.

636,

6.2

8±2.

1923

, 9.0

±5.13

6, 9

.70

±6.5

3

Wee

k 6

14/2

3 (6

1)1/

6 (1

7)7/

23 (

30)

1/6

(17)

8/23

(35

)0/

6 (0

)22

, 5.19

±2.4

03,

4.4

9±1.9

9

Wee

k 8

8/23

(35

)1/

6 (1

7)5/

23 (

22)

0/6

(0)

7/23

(30

)0/

6 (0

)21

, 5.6

2±2.

183,

4.3

9±0

.54

Wee

k 10

9/23

(39

)0/

6 (0

)5/

23 (

22)

0/6

(0)

6/23

(26

)0/

6 (0

)20

, 5.4

3±2.

243,

4.8

4±0

.44

Wee

k 12

9/23

(39

)1/

6 (1

7)3/

23 (

13)

0/6

(0)

3/23

(13

)0/

6 (0

)20

, 5.6

5±2.

263,

4.3

7±1.

7820

, 9.6

0±3

.85

3, 8

.30

±0.5

8

Wee

k 16

8/23

(35

)1/

6 (1

7)3/

23 (

13)

1/6

(17)

4/23

(17

)0/

6 (0

)20

, 5.5

8±2.

233,

4.3

4±2.

18

Wee

k 20

6/23

(26

)1/

6 (1

7)2/

23 (

9)1/

6 (1

7)2/

23 (

9)1/

6 (1

7)18

, 5.6

9±1.

203,

4.18

±1.7

8

Wee

k 24

7/23

(30

)0/

6 (0

)3/

23 (

13)

0/6

(0)

3/23

(13

)0/

6 (0

)17

, 5.5

8±2.

333,

4.5

4±0

.63

Wee

k 28

/ en

d of

stu

dy7/

23 (

30)

1/6

(17)

2/23

(9)

0/6

(0)

1/23

(4)

0/6

(0)

23, 6

.26±

2.31

5, 5

.53±

1.72

22, 1

0.6

±4.8

04,

9.3

0±5

.91

* Se

cuki

num

ab:

2x10

mg/

kg.

ASA

S d

eno

tes

Ass

essm

ent

of

Spo

ndyl

oA

rthr

itis

int

erna

tio

nal

Soci

ety

crit

eria

, BA

SDA

I d

eno

tes

Bath

Ank

ylo

sing

Sp

ond

ylit

is D

isea

se

Act

ivit

y In

dex

, ASQ

oL

den

ote

s A

nkyl

osi

ng S

po

ndyl

itis

Qua

lity

of L

ife.

155

Supplementary Table 4. MRI scores at baseline, week 6, and week 28/end of study

Secukinumab 2x10mg/kg Placebo

Baseline Week 6 Week 28* Baseline Week 6 Week 28

Number of patients 22 22 16 5 3 5

ASAS20 responders (n) — 14 6 — 1 1

Mean Berlin score† 9.2±8.87 6.6±6.56 5.7±6.20 20.6±20.18 21.0±24.56 19.0±19.33

P value (vs. baseline) — 0.10 0.16 — 0.50 0.25

* Data from 6 patients who discontinued prior to week 28 because of lack of response were not available for analysis.† Plus–minus values are means ±SD.

Supplementary Figure 1. Panel A shows ESR at baseline (Week 0) and Week 6 following intravenous secukinumab 2x10 mg/kg at Weeks 0 and 3. Panel B shows baseline serum concentrations of S100A8 (left), S100A9 (center), and S100A8 and S100A9 combined (right). Measurements are shown for individual patients and grouped by their week 6 ASAS40 response. Responders are depicted as dark circles, nonresponders as lighter circles.

A

B

156

Secu

kin

um

ab a

nti-iL-17a

ther

apy

in a

S

10

Supplementary Figure 2. Panel A shows ERAP1 gene expression and genotype and association with clinical response at week 6 in ASAS40 responders and nonresponders. Representations of ERAP1 gene expression levels and ERAP1 ankylosing spondylitis risk allele genotypes (rs30187 and rs27434) are combined. Dots represent individual patients from the secukinumab arm at baseline (pre-treatment). Week 6 ASAS40 nonresponders are represented on the left and responders on the right. Panel B shows the association of rs11209032 (left panel; minor allele/AS risk allele=A, major allele=G) and rs2201841 (right panel; minor allele/AS risk allele=C, major allele=T) genotypes in the IL23R gene with mean BASDAI change from baseline (±SD) at different time points from first dosing of secukinumab. The polymorphisms were tested individually for association with BASDAI at week 24 using a linear regression model, with number of copies of the minor (less common) allele carried by the individual as the predictor and baseline BASDAI score as a covariate. ASAS denotes Assessment of SpondyloArthritis international Society criteria; BASDAI denotes Bath Ankylosing Spondylitis Disease Activity Index; SD denotes standard deviation.

A

B

157