International Journal of Rheumatic Diseases 2016

ORIGINAL ARTICLE

Efficacy and safety of etanercept in patients from Latin

America, Central Europe and Asia with early non-
radiographic axial spondyloarthritis
James Cheng-Chung WEI,1,2,3 Wen-Chan TSAI,4 Gustavo CITERA,5 Sameer KOTAK6 and
Lyndon LLAMADO7
1
Division of Allergy, Immunology and Rheumatology, Chung Shan Medical University Hospital, 2Institute of Medicine, Chung Shan
Medical University, 3Institute of Integrated Medicine, China Medical University, Taichung, 4Division of Rheumatology, Kaohsiung
Medical University Hospital, Kaohsiung, Taiwan, 5Instituto de Rehabilitacion Psicofsica, Buenos Aires, Argentina, 6Pfizer, New
York, New York, USA, and 7Pfizer, Makati, Philippines

Abstract
Aim: To evaluate etanercept in patients from Latin America, Central/Eastern Europe, and Asia with non-radio-
graphic axial spondyloarthritis (nr-axSpA).
Methods: A subset analysis was performed on nr-axSpA patients from Argentina, Colombia, the Czech Republic,
Hungary, Russia and Taiwan who were enrolled in EMBARK (NCT01258738). Patients received either etanercept
50 mg or placebo once weekly. The primary endpoint was proportion of patients achieving 40% improvement
from baseline based on Assessment of SpondyloArthritis International Society (ASAS) criteria. Secondary end-
points included other efficacy assessments, health-related quality of life (HRQoL) and safety.
Results: Of the 117 patients in this subset, 59 were treated with etanercept and 58 received placebo. At week 12,
numerically greater improvements from baseline were observed for all efficacy endpoints in etanercept-treated
patients compared with those receiving placebo. Statistically significant differences between the two treatment
groups were observed for proportion of patients achieving ASAS40 (P = 0.0413, at week 8), ASAS5/6
(P = 0.0126), Ankylosing Spondylitis Disease Activity Score - C-reactive protein (CRP) inactive disease
(P = 0.0093), Spondyloarthritis Research Consortium of Canada magnetic resonance imaging of sacroiliac joint
scores (P = 0.0014), high-sensitivity CRP (P=0.032), and erythrocyte sedimentation rate (P = 0.0082). Statisti-
cally significant improvements in the etanercept-treated group compared with placebo group were observed for
nocturnal back pain (P = 0.040), total back pain (P = 0.025), physician global assessment of disease
(P = 0.023), and Work Productivity and Activity Impairment Questionnaire percent impairment while working
(P = 0.047). Adverse events were similar between the two treatment groups.
Conclusions: In this subset of patients with nr-axSpA from Latin America, Central/Eastern Europe, and Asia,
treatment with etanercept, compared with placebo, resulted in improved disease symptoms and patient HRQoL.
Etanercept was well tolerated.
Key words: ankylosing spondylitis, etanercept, spondyloarthritis, tumor necrosis factor.

Correspondence: Dr James Cheng-Chung Wei, Division of INTRODUCTION

Allergy, Immunology and Rheumatology, Chung Shan Medi-
cal University Hospital, No. 110, Sec. 1, Jianguo N. Road,
South District, Taichung City 40201, Taiwan.
Email: jccwei@gmail.com
With the relatively recent update on classification of
spondyloarthritides developed by the Assessment of
SpondyloArthritis international Society (ASAS),13 this

2016 Asia Pacific League of Associations for Rheumatology and John Wiley & Sons Australia, Ltd
J. C.-C. Wei et al.
heterogeneous group of diseases is now categorized into
axial and peripheral diseases. Many patients with axial
spondyloarthritis (axSpA) experience permanent struc-
tural changes of the spine leading to progressive disabil-
ity, significant pain, fatigue, impaired physical function,
limited spinal mobility, work disability and an overall
diminished health-related quality of life (HRQoL).46
The burden of disease can be quite extensive for the
patient.5 Ankylosing spondylitis (AS) is the most com-
mon type of axSpA and diagnosis is based on the pres-
ence of radiographic evidence of sacroiliitis. Under the
new guidelines,13 a new classification of non-radio-
graphic axSpA (nr-axSpA) has been introduced to
include patients with axSpA but without radiographic
evidence of sacroiliitis. The difference between these two
subcategories of axSpA (with and without radiographic
evidence of sacroiliitis) is still questioned since radio-
graphic imaging is the central criterion and this tech-
nique is not quite infallible.7 Furthermore, patients with
nr-axSpA may develop inflammatory changes in the
sacroiliac joints and the spine that are characteristic of
AS, that is nr-axSpA may be an early form of AS. Ini-
tially, evaluation of treatments for axSpA focused almost
exclusively on patients with radiographic sacroiliitis.811
However, several studies demonstrated that some
patients with no evidence of sacroiliitis, based on radio-
graphs, experience levels of axSpA disease, pain and fati-
gue similar to those patients with observable
radiographic sacroiliitis.1216 Thus, it is only in the past
few years that studies have specifically focused on identi-
fying and treating patients with nr-axSpA.
Anti-tumor necrosis factor alpha (anti-TNF-a) agents
have long been used to treat patients with axSpA.811
Anti-TNF-a agents have also been shown to be effective
in improving disease symptoms, spinal mobility, physi-
cal function and HRQoL in patents with nr-axSpA.1721
Although anti-TNF-a agents are now recommended for
treatment of patients with nr-axSpA, especially when
they are unresponsive to non-steroidal anti-inflamma-
tory drugs (NSAIDs),22 these recommendations are lar-
gely based on studies in the USA and Western Europe.
It is not clear whether all patients will respond in a sim-
ilar manner since there are ethnic and geographical
variations in epidemiological and clinical presentation
of disease.1,23,24 The objective of this analysis was to
evaluate the efficacy of one anti-TNF-a agent, etaner-
cept, to improve disease symptoms and HRQoL in
patients with nr-axSpA from countries under-repre-
sented in studies leading to guidelines using the
EMBARK (NCT01258738, formerly known as AS
EARLY) trial data.25
2 International Journal of Rheumatic Diseases 2016
METHODS
Study design
EMBARK, the details of which have been published,25
was a randomized, double-blind, two-period, multicen-
ter Phase 3b study conducted in 14 countries in Europe,
Asia and Latin America, approved by locally constituted
Ethics Committees, and in accordance with the World
Medical Associations Declaration of Helsinki. This trial
compared the efficacy and safety of etanercept versus
placebo (with a background of NSAID treatment) in
patients with active axSpA. Informed consent was
obtained from all patients enrolled in the study. In the
first period of the trial (double-blind), patients were
randomized 1 : 1 to receive etanercept 50 mg/week
subcutaneously or placebo for 12 weeks. In the second
period of the trial (open-label), patients treated with
etanercept continued to receive the drug, and patients
who received placebo were switched to receive etaner-
cept 50 mg/week subcutaneously for an additional
92 weeks. Results from the first period are reported.
Patients
Detailed eligibility requirements have been previously
published.25 Briefly, eligible patients were aged 18 to
< 50 years, satisfied ASAS classification criteria for
axSpA, with non-radiographic sacroiliitis defined as
those patients who did not meet the modified radio-
graphic 1984 New York criteria for AS.2 Patients had
symptom duration of > 3 months to < 5 years, active
disease defined by a Bath Ankylosing Spondylitis Dis-
ease Activity Index (BASDAI) score 4, and had failed
at least two NSAIDs. All patients had inflammatory
back pain with an inadequate response to at least two
NSAIDs taken at the maximum tolerated doses sepa-
rately, for a total combined duration of more than
4 weeks; they were receiving a stable, maximally toler-
ated dose of an NSAID for at least 14 days prior to
study baseline. In this post hoc subset analysis, the
responses of patients from Latin America, Central/East-
ern Europe and Asia were evaluated.
Assessments
The primary efficacy endpoint was the proportion of
patients who achieved 40% improvement from baseline
according to ASAS criteria (ASAS40) at week 12 (modi-
fied intent-to-treat [mITT]). Secondary efficacy end-
points included the proportion of patients who
achieved ASAS20, ASAS5/6, ASAS partial remission,
Ankylosing Spondylitis Disease Activity Score (ASDAS)
C-reactive protein (CRP) inactive disease and

BASDAI50, changes from baseline in Bath Ankylosing
Spondylitis Functional Index (BASFI), Bath Ankylosing
Spondylitis Metrology Index (BASMI), BASDAI and
magnetic resonance imaging (MRI) score for the spine
and sacroiliac joints using the Spondyloarthritis
Research Consortium of Canada (SPARCC) method.
Inflammation was measured as change from baseline in
high-sensitivity CRP (hsCRP) and erythrocyte sedimen-
tation rate (ESR). Quality of life outcomes were mea-
sured using Ankylosing Spondylitis Quality of Life
(ASQoL), Ankylosing Spondylitis Work Instability
Index (AS-WIS), Bath Ankylosing Spondylitis Patient
Global Assessment Score (BAS-G), EuroQol EQ-5D
Health State Profile (EQ-5D), Hospital Anxiety and
Depression Scale (HADS), Medical Outcomes Study
(MOS) Sleep Scale II, Minimal Clinically Important
Improvement (MCII), Multidimensional Fatigue Inven-
tory (MFI), Patient Acceptable Symptom State (PASS),
patient assessment of total and nocturnal pain, physi-
cian global assessment of disease (PGA), 36-Item Short-
Form Health Survey (SF-36) and Work Productivity and
Activity Impairment Questionnaire (WPAI). Adverse
events were recorded and categorized according to Med-
ical Dictionary for Regulatory Activities (MedDRA),
v15.0.
Statistical analyses
For the global study, the sample size was calculated
based on findings of previously conducted clinical stud-
ies in patients with radiographic axSpA. For the primary
endpoint of ASAS40, a target sample size of 100
patients per treatment group was estimated to provide
90% statistical power based on two-sided testing at
alpha = 0.05 and assuming a 25% difference in
response rates between the etanercept and placebo
groups (e.g., 30% in the placebo group and 55% in the
etanercept group).
The mITT population was used for analyses of all effi-
cacy endpoints using the last-observation-carried-for-
ward (LOCF) approach for missing data; patient-
reported outcomes (PROs) were analyzed using the
observed case approach; adverse events were recorded
for all patients who received at least one dose of etaner-
cept.
Categorical endpoints were analyzed using the
Cochran-Mantel-Haenszel Chi-square test, stratified by
MRI sacroiliitis (SI) positive/negative and geographic
region. Continuous endpoints were analyzed using an
analysis of covariance (ANCOVA) model with treatment,
MRI SI positive/negative, and geographic region as fac-
tors and baseline score as a covariate.
International Journal of Rheumatic Diseases 2016
Etanercept in nr-axSpA patients
RESULTS
Patients
Of the 225 patients randomized in the original
EMBARK trial, 117 patients met the criteria for this sub-
set analysis, that is, were from Latin America (Argen-
tina, n = 3; Colombia, n = 14), Central/Eastern Europe
(the Czech Republic, n = 23; Hungary, n = 42; Russia,
n = 6), and Asia (Taiwan, n = 29). Of these 117
patients, 59 received etanercept 50 mg/week and 58
received placebo. There were no statistical differences in
the baseline demographics and disease characteristics
between the two treatment groups (Table 1).
Efficacy
The proportion of patients achieving ASAS40 steadily
increased from baseline to week 12 in both treatment
groups (Fig. 1a). A higher percentage of patients trea-
ted with etanercept achieved ASAS40 at all time
points, although a statistically significant difference
between the two treatment groups was observed only
at week 8 (P = 0.041). Similarly, the proportion of
patients achieving ASAS20 (Fig. 1b), ASAS5/6
(Fig. 1c), ASAS partial remission (Fig. 1d), ASDAS-
CRP inactive disease (Fig. 1e), and 50% improvement
in BASDAI (BASDAI50; Fig. 1f) steadily increased
from baseline to week 12 for both treatment groups;
there was a decrease from week 2 to week 4 for ASAS
partial remission (Fig. 1d) and no difference between
week 8 and week 12 for ASDAS-CRP inactive disease
(Fig. 1e). There was no statistical difference between
the two treatment groups with respect to ASAS20
(Fig. 1b), ASAS partial remission (Fig. 1d) and BAS-
DAI50 (Fig. 1f). Statistically significant differences in
favor of etanercept treatment were observed from
week 4 onward for both ASAS5/6 (Fig. 1c) and
ASDAS-CRP inactive disease (Fig. 1e).
Steady improvements over time were also observed in
both treatment groups in total BASDAI (Fig. 2a), BASFI
(Fig. 2b) and BASMI (Fig. 2c) scores. Improvements
were numerically greater in patients receiving etanercept
for all of these assessments, although there was no sta-
tistically significant difference between the two treat-
ment groups for any of these parameters at any time
point. Improvements from baseline to week 12 were
also observed in the SPARCC MRI spinal and sacroiliac
joint scores in both treatment groups (Fig. 3). There
was no statistically significant difference between the
two treatment groups for improvements in the SPARCC
MRI spinal score (Fig. 3). Improvement in the SPARCC
MRI sacroiliac joint score was significantly greater in
3
J. C.-C. Wei et al.

Table 1 Patient demographics and disease characteristics at patients treated with etanercept than in those receiving
baseline placebo (P = 0.001; Fig. 3).
Parameters ETN 50 mg Placebo Statistically significant improvements from baseline
QW n = 57 to week 12 were observed in patients treated with etan-
n = 54 ercept compared with those receiving placebo for both
inflammation markers (Fig. 3), hsCRP (P = 0.032), and
Age, years, mean (SD) 32.0 (6.8) 32.2 (8.7)
Gender, n (%)
ESR (P = 0.008). Patients receiving placebo experienced
Female 16 (29.6) 21 (36.8) a worsening in their hsCRP levels (Fig. 3).
Male 38 (70.4) 36 (63.2)
Ethnicity, n (%)
Quality of life assessments
White 40 (74.1) 35 (61.4) Statistically significant improvements from baseline to
Asian 12 (22.2) 17 (29.8) week 12 were observed in patients treated with etaner-
Other 2 (3.7) 5 (8.8) cept compared with those receiving placebo (Table 2)
Body mass index, kg/m2, 25.7 (5.4) 24.0 (4.3) for PGA ( 2.4 vs. 1.6, P = 0.023), nocturnal back
mean (SD) pain ( 1.8 vs. 0.9, P = 0.040), total back pain ( 1.8
HLA-B27 positive, n (%) 36 (66.7) 47 (82.5)
vs. 0.9, P = 0.025) and WPAI percent impairment
CRP, mg/L, mean (SD) 4.9 (6.9) 7.6 (12.5)
while working due to problems ( 24.7 vs. 12.8,
Elevated hsCRP (> 3 mg/L), 22 (40.7) 24 (42.1)
n (%)
P = 0.047), while WPAI % overall work impairment
Disease symptom duration, years, 2.3 (1.5) 2.4 (1.6) due to problems and WPAI % activity impairment due
mean (SD) to problems approached levels of statistical significance
BASDAI total score, mean (SD) 5.9 (1.9) 6.3 (1.7) (P = 0.086 and P = 0.096, respectively). There was no
BASFI score, mean (SD) 4.2 (2.5) 4.1 (2.5) statistically significant difference between the two treat-
ASDAS-CRP, mean (SD) 2.9 (0.9) 3.1 (1.0) ment groups in WPAI % work time missed due to prob-
ASDAS-ESR, mean (SD) 3.3 (1.0) 3.5 (0.9) lems. Numerically greater improvements from baseline
MRI sacroiliitis present, n (%) 47 (87.0) 47 (82.5) to week 12 were observed in patients treated with etan-
SPARCC sacroiliac joint score, 7.4 (8.4) 6.3 (6.9) ercept compared with those receiving placebo in all
mean (SD)
other quality of life assessments (ASQoL, AS-WIS, EQ-
SPARCC spinal score, mean (SD) 5.0 (8.1) 3.4 (5.5)
5D, HADS, MFI mental fatigue, MFI reduced activity
Total back pain, cm, mean (SD) 5.0 (2.7) 5.3 (2.3)
Inflammatory back pain present, 45 (83.3) 46 (80.7)
and SF-36), but did not reach statistical significance
n (%) (Table 2). At week 12, 18/36 (50.0%) of patients trea-
Enthesitis present, n (%) 25 (46.3) 29 (50.9) ted with etanercept experienced minimal clinically
Dactylitis present, n (%) 4 (7.4) 7 (12.3) important improvement compared with 18/43 (41.9%)
Swollen joint count, mean (SD) 1.0 (2.3) 1.1 (2.6) of patients receiving placebo (P = 0.190). The propor-
Tender joint count, mean (SD) 3.7 (5.0) 5.4 (6.9) tion of patients who reported PASS was 77.8% (28/36)
Patient global assessment of 5.6 (2.3) 5.8 (2.1) in the etanercept-treated group compared with 76.7%
disease, cm, mean (SD) (33/43) of patients receiving placebo (P = 0.886).
Physician global assessment of 5.5 (2.1) 5.0 (2.2)
disease, cm, mean (SD) Safety
ASQoL total score, mean (SD) 8.4 (5.0) 7.8 (5.2)
There were no statistically significant differences
EQ-5D score, mean (SD) 0.6 (0.3) 0.6 (0.3)
SF-36 PCS, mean (SD) 40.6 (8.6) 39.2 (8.0)
between patients treated with etanercept and those
Concomitant DMARD use, n (%) 10 (18.5) 13 (22.8) receiving placebo in the incidence of adverse events,
Concomitant NSAID use, n (%) 54 (100.0) 56 (98.3) incidence of serious adverse events, or incidence of
adverse events leading to treatment discontinuation
ASDAS, Ankylosing Spondylitis Disease Activity Score; ASQoL, Anky-
losing Spondylitis Quality of Life; BASDAI, Bath Ankylosing Spondyli- (Table 3). The most common treatment-emergent
tis Disease Activity Index; BASFI, Bath Ankylosing Spondylitis adverse events for both treatment groups were infec-
Functional Index; CRP, C-reactive protein; DMARD, disease-modifying tions, musculoskeletal and connective tissue disorders,
anti-rheumatic drug; EQ-5D, EuroQol EQ-5D Health State Profile;
ESR, erythrocyte sedimentation rate; ETN, etanercept; HLA, human gastrointestinal disorders, and skin and subcutaneous
lymphocyte antigen; hsCRP, high-sensitivity C-reactive protein; MRI, tissue disorders, most of which were mild in severity.
magnetic resonance imaging; NSAID, nonsteroidal anti-inflammatory Among patients receiving etanercept, the most common
drug; QW, once a week; SD, standard deviation; SF-36, 36-Item Short-
Form; SPARCC, Spondyloarthritis Research Consortium of Canada. treatment-emergent adverse events were injection site

4 International Journal of Rheumatic Diseases 2016

 Etanercept in nr-axSpA patients

(a) 30 (b) 60
Etanercept Placebo Etanercept Placebo
achieving ASAS40 (%)

*
Proportion of patients

achieving ASAS20 (%)

Proportion of patients
20 40

10 20

0 0
0 4 8 12 0 4 8 12
Weeks treated Weeks treated

(c) 40 (d)

achieving ASAS partial remission (%)

20
Etanercept Placebo Etanercept Placebo
*
achieving ASAS5/6 (%)

**
Proportion of patients

15

Proportion of patients
30
**

20 10

10 5

0 0
0 4 8 12 0 4 8 12
Weeks treated Weeks treated

(e)
50 (f) 50
ASDAS-CRP inactive disease (%)

Etanercept Placebo Etanercept Placebo

Proportion of patients achieving

* **
achieving BASDAI50 (%)

40 40
Proportion of patients

30 30

*
20 20

10 10

0 0
0 4 8 12 0 4 8 12
Weeks treated Weeks treated

Figure 1 Proportion of patients achieving (a) Assessment of SpondyloArthritis International Society (ASAS)40, (b) ASAS20, (c)
ASAS 5/6, (d) ASAS partial remission, (e) Ankylosing Spondylitis Disease Activity Score C-reactive protein (ASDAS-CRP) inactive
disease state and (f) Bath Ankylosing Spondylitis Disease Activity Index (BASDAI)50.

reaction, headache and rash (n = 2 for each). Among DISCUSSION

patients receiving placebo, the most common treat-
ment-emergent adverse events were upper respiratory Several studies have demonstrated that anti-TNF-a
tract infection (n = 4), diarrhea, nasopharyngitis and agents were not only effective for the treatment of
headache (n = 2 for each). axSpA,810 but also improved disease symptoms, spinal

International Journal of Rheumatic Diseases 2016 5

J. C.-C. Wei et al.

(a) Weeks treated 5

Mean ( SEM) change from baseline

0 4 8 12
0.0 hsCRP ESR MRI Spinal MRI Sacroiliac
Mean ( SEM) change from baseline:

0
0.5
Total BASDAI score

1.0 5 NS
* **

1.5
10

2.0 Etanercept
Placebo
**
15
2.5 Etanercept Placebo

Figure 3 Change from baseline in high-sensitivity C-reactive

(b) protein (hsCRP), erythrocyte sedimentation rate (ESR),
0.5
Spondyloarthritis Research Consortium of Canada magnetic
Mean ( SEM) change from baseline:

Weeks treated resonance imaging (SPARCC MRI) spinal and SPARCC MRI
4 8 12
0.0 sacroiliac scores.
Total BASFI score

0.5 where studies evaluating the efficacy and safety of TNF-

a inhibitors to treat nr-axSpA primarily have been con-
1.0 ducted. In addition, the cost of biologic agents can
often limit access to these drugs. Consequently, there is
1.5 a critical scarcity of data on the utility of TNF-a inhibi-
tors for the management of nr-axSpA in these popula-
2.0 Etanercept Placebo tions, resulting in a lack of implementation of
evidence-based recommendations. In this subset analy-
(c) sis, we investigated whether treatment with etanercept
0.2
was effective in patients with nr-axSpA from Latin
Mean ( SEM) change from baseline:

Weeks treated
4 8 12 America, Central/Eastern Europe and Asia.
0.0
In this subset of patients from the EMBARK study,
Total BASMI score

treatment with etanercept showed numerically greater

0.2
0.4
0.6
Etanercept
0.8
Figure 2 Change from baseline in (a) Bath Ankylosing
Spondylitis Disease Activity Index (BASDAI), (b) Bath Anky-
losing Spondylitis Functional Index (BASFI) and (c) Bath
Ankylosing Spondylitis Metrology Index (BASMI).
mobility, physical function and HRQoL in patients with
nr-axSpA.1721 Indeed, anti-TNF-a agents are now rec-
ommended for the treatment of patients with nr-axSpA
after they have failed at least two NSAID regimens.22
Populations from Latin America, Eastern/Central Eur-
ope and Asia diverge ethnically and socioeconomically
from those in North America and Western Europe
improvements in all efficacy parameters compared with
patients receiving placebo. The global study was pow-
ered assuming a 25% difference in the response rates
between patients treated with etanercept and those trea-
ted with placebo.25 The study was not designed to be
powered similarly for analyses of subpopulations. In
Placebo
this subset analysis, although the difference in response
rates between the two treatment groups was only
10.6%, it nevertheless demonstrated that patients
administered etanercept fared better than those receiv-
ing placebo suggesting that the primary endpoint was
met, a result consistent with that observed in the global
study.25 The difference between the two treatment
groups in the proportion of patients achieving ASAS40
reached statistical significance at week 8. Furthermore,
there was no statistically significant difference between
the two treatment groups for other measures of efficacy,
including the proportion of patients achieving ASAS20,
ASAS partial remission and BASDAI50. Consistent
statistically significant differences in the responses to

6 International Journal of Rheumatic Diseases 2016

 Etanercept in nr-axSpA patients

Table 2 Change from baseline in health-related quality of life Table 3 Summary of treatment-emergent adverse events by
outcomes MedDRA classification
Parameter Change from baseline at 12 weeks ETN 50 mg Placebo
Mean (SEM) QW n = 59
n = 58
ETN Placebo P for
n = 54 n = 57 significance Any adverse event, n (%) 22 (37.9) 23 (39.0)
Serious adverse event, n (%) 1 (1.7) 1 (1.7)
Physician global 2.4 (0.3) 1.6 (0.3) 0.023
Adverse event leading to 2 (3.4) 1 (1.7)
assessment
discontinuation, n (%)
BAS-G 1.6 (0.4) 1.2 (0.3) 0.260
Treatment-emergent adverse events, n (%)
Nocturnal back 1.8 (0.4) 0.9 (0.4) 0.040
Blood and lymphatic system 0 (0) 1 (1.7)
pain
disorders
Total back pain 1.8 (0.4) 0.9 (0.4) 0.025
Endocrine disorders 0 (0) 1 (1.7)
ASQoL 1.5 (0.8) 0.8 (0.7) 0.367
Gastrointestinal disorders 3 (5.2) 2 (3.4)
AS-WIS 1.7 (0.8) 0.7 (0.7) 0.210
General disorders and 2 (3.4) 1 (1.7)
EQ-5D VAS 6.7 (4.3) 0.9 (3.7) 0.092
administration site conditions
EQ-5D utility score 0.1 (0.1) 0.1 (0.0) 0.278
Hepatobiliary disorders 2 (3.4) 0 (0)
HADS-Anxiety 1.4 (0.6) 0.9 (0.6) 0.437
Immune system disorders 1 (1.7) 0 (0)
HADS-Depression 0.6 (0.6) 0.1 (0.5) 0.322
Infections and infestations 6 (10.3) 10 (16.9)
MFI general fatigue 0.7 (0.6) 0.7 (0.6) 0.943
Injury, poisoning, and procedural 0 (0) 2 (3.4)
MFI mental fatigue 0.4 (0.7) 0.2 (0.6) 0.371
complications
MFI physical fatigue 0.2 (0.6) 0.5 (0.5) 0.591
Investigations 2 (3.4) 3 (5.1)
MFI reduced 0.8 (0.5) 0.5 (0.5) 0.500
Metabolism and nutrition 0 (0) 1 (1.7)
activity
disorders
MOS sleep index II 3.9 (2.8) 4.9 (2.4) 0.727
Musculoskeletal and connective 4 (6.9) 3 (5.1)
SF-36 MCS 2.2 (1.6) 2.1 (1.4) 0.964
tissue disorders
SF-36 PCS 4.7 (1.2) 2.6 (1.0) 0.090
Nervous system disorders 2 (3.4) 2 (3.4)
WPAI: % work time 0.6 (4.1) 3.9 (3.9) 0.395
Renal and urinary disorders 0 (0) 1 (1.7)
missed due to
Reproductive system and breast 0 (0) 1 (1.7)
health problem
disorders
WPAI: % 24.7 (6.1) 12.8 (5.9) 0.047
Respiratory, thoracic, and 1 (1.7) 1 (1.7)
impairment while
mediastinal disorders
working due to
Skin and subcutaneous tissue 3 (5.2) 2 (3.4)
problem
disorders
WPAI: % overall 23.9 (6.5) 13.1 (6.2) 0.086
Vascular disorders 1 (1.7) 1 (1.7)
work impairment
due to problem ETN, etanercept; MedDRA, Medical Dictionary for Regulatory Activi-
WPAI: % activity 14.8 (5.0) 6.1 (4.3) 0.096 ties; QW, once a week.
impairment due to
problem EMBARK population, in which patients treated with
etanercept experienced significantly better outcomes in
ASQoL, Ankylosing Spondylitis Quality of Life; AS-WIS, Ankylosing
Spondylitis Work Instability Index; BAS-G, Bath Ankylosing Spondyli- all efficacy measures than those receiving placebo.25
tis Patient Global Assessment Score; EQ-5D, EuroQol EQ-5D Health The reasons for these differences are unclear, although
State Profile; ETN, etanercept; HADS, Hospital Anxiety and Depression it just might be that this subpopulation does not have
Scale; MCS, Mental Health Component Score; MFI, Multidimensional
Fatigue Inventory; MOS, Medical Outcomes Study; PCS, Physical enough power to demonstrate statistical significance.
Health Component Score; QW, once a week; SEM, standard error of Since the number of patients from each of these coun-
the mean; SF-36, 36-Item Short-Form Health Survey; VAS, visual ana- tries in this subset was too small for statistical analyses,
log scale; WPAI, Work Productivity and Activity Impairment Question-
naire. further studies with sufficient statistical power focusing
on the individual subgroups will be needed to elucidate
this issue.
etanercept or placebo were observed when efficacy was Radiographic imaging is a core feature for the diagno-
measured as the proportion of patients achieving sis of nr-axSpA. Patients treated with etanercept experi-
ASAS5/6 or ASDAS-CRP inactive disease. These data are enced a significant benefit over those receiving placebo
slightly different from those observed in the full based on the results of MRI scoring for the sacroiliac

International Journal of Rheumatic Diseases 2016 7

J. C.-C. Wei et al.
joints using the SPARCC method. Although the
improvement in SPARCC MRI scores for the spine was
numerically greater for patients treated with etanercept
compared with those receiving placebo, the difference
was not statistically significant. These data are similar to
those reported for the full EMBARK population25 in
which the differences in SPARCC MRI scores between
the two treatment groups were greater for the sacroiliac
joints than for the spine. It is possible that this subset
population did not have enough power to demonstrate
statistical significance between the two treatment arms
for the SPARCC MRI scores for the spine. These data
suggest that the sacroiliac joints may be more sensitive
to treatment and/or changes in the joints may be more
easily detectable than in the spine. This hypothesis will
need to be tested.
Statistically significant improvements from baseline
to week 12 in patients treated with etanercept compared
with those receiving placebo were observed for both
hsCRP and ESR. These data are consistent with those
reported for the full EMBARK population25 and for
other studies on axSpA.2628 These data suggest that
inflammation markers may need to be considered as
potential indicators of treatment efficacy and/or disease
status.
Similar to the efficacy outcomes, the HRQoL data
mostly showed numerically greater, but not statistically
significant, improvements in patients treated with etan-
ercept compared with those receiving placebo. The pri-
mary exceptions to this pattern were the PROs of
nocturnal and total back pain and WPAI % impairment
while working due to problems, all of which showed
statistically significant benefit for patients treated with
etanercept compared with those receiving placebo.
The safety data for this subpopulation are consistent
with those for the whole EMBARK population,25 other
studies on etanercept,8,18,2830 and other studies with
anti-TNF-a agents.11,17,1921 These data suggest that
etanercept is well tolerated in this subpopulation and
that there appear to be no new safety concerns.
Taken together, the data from this subpopulation
suggest that a composite disease metric that includes
elements of efficacy (such as ASAS5/6 and ASDAS-
CRP inactive disease), radiography (such as SPARCC
MRI score for sacroiliac joints), inflammation (such
as hsCRP and ESR), and PROs (such as back pain)
may provide a way to effectively monitor disease sta-
tus and/or progression. Based on these parameters
alone, the data reported here demonstrate that etan-
ercept was clinically beneficial to patients with nr-
8 International Journal of Rheumatic Diseases 2016
axSpA from Latin America, Central/Eastern Europe
and Asia.
ACKNOWLEDGEMENTS
The authors thank Ron Pedersen, Pfizer, for his statisti-
cal analysis of the data and review of the manuscript.
Medical writing support was provided by Mukund Nori,
PhD, MBA, CMPP, of Engage Scientific Solutions and
funded by Pfizer, New York, NY, USA.
COMPETING INTERESTS
Dr. Wei has received research grants or consultation fees
from AbbVie, BMS, Celgene, Chugai, Eisai, Janssen,
Novartis, Pfizer, Sanofi-Aventis, TSH Taiwan and UCB
Pharma. Dr. Tsai has received consultation fees from
Pfizer, AbbVie and Roche. Dr. Citera has been on the
speakers bureau, and an advisor and/or investigator for
Pfizer, AbbVie, Bristol Myers Squibb and Roche. Drs.
Llamado and Kotak are employees of Pfizer. This study
was sponsored by Pfizer.
REFERENCES
1 Raychaudhuri SP, Deodhar A (2014) The classification
and diagnostic criteria of ankylosing spondylitis. J Autoim-
mun 4849, 12833.
2 Rudwaleit M, Landewe R, van der Heijde D et al. (2009)
The development of Assessment of SpondyloArthritis
international Society classification criteria for axial
spondyloarthritis (part I): classification of paper patients
by expert opinion including uncertainty appraisal. Ann
Rheum Dis 68, 7706.
3 Rudwaleit M, van der Heijde D, Landewe R et al. (2011)
The Assessment of SpondyloArthritis International Society
classification criteria for peripheral spondyloarthritis and
for spondyloarthritis in general. Ann Rheum Dis 70 (1),
2531.
4 Braun J, Sieper J (2007) Ankylosing spondylitis. Lancet
369, 137990.
5 Cho H, Kim T, Kim TH, Lee S, Lee KH (2013) Spinal
mobility, vertebral squaring, pulmonary function, pain,
fatigue, and quality of life in patients with ankylosing
spondylitis. Ann Rehabil Med 37, 67582.
6 Golder V, Schachna L (2013) Ankylosing spondylitis: an
update. Aust Fam Physician 42, 7804.
7 Braun J, Kiltz U, Baraliakos X, van der Heijde D (2014)
Optimisation of rheumatology assessments the actual
situation in axial spondyloarthritis including ankylosing
spondylitis. Clin Exp Rheumatol 32 (5 Suppl 85), S-96
104.

8 Davis JC Jr, van der Heijde DM, Braun J et al. (2008) Effi-
cacy and safety of up to 192 weeks of etanercept therapy
in patients with ankylosing spondylitis. Ann Rheum Dis
67, 34652.
9 Braun J, Baraliakos X, Listing J et al. (2008) Persistent clin-
ical efficacy and safety of anti-tumour necrosis factor alpha
therapy with infliximab in patients with ankylosing
spondylitis over 5 years: evidence for different types of
response. Ann Rheum Dis 67, 3405.
10 Haibel H, Rudwaleit M, Brandt HC et al. (2006) Adali-
mumab reduces spinal symptoms in active ankylosing
spondylitis: clinical and magnetic resonance imaging
results of a fifty-two-week open-label trial. Arthritis Rheum
54, 67881.
11 Braun J, Baraliakos X, Heldmann F, Kiltz U (2014)
Tumor necrosis factor alpha antagonists in the treatment
of axial spondyloarthritis. Expert Opin Investig Drugs 23,
64759.
12 Fianyo E, Wendling D, Poulain C, Farrenq V, Claudepierre
P (2014) Non-radiographic axial spondyloarthritis: what
is it? Clin Exp Rheumatol 32 (1), 14.
13 Wallis D, Haroon N, Ayearst R, Carty A, Inman RD (2013)
Ankylosing spondylitis and nonradiographic axial spondy-
loarthritis: part of a common spectrum or distinct dis-
eases? J Rheumatol 40, 203841.
14 Bennett AN, McGonagle D, OConnor P et al. (2008)
Severity of baseline magnetic resonance imaging-evident
sacroiliitis and HLA-B27 status in early inflammatory back
pain predict radiographically evident ankylosing spondyli-
tis at eight years. Arthritis Rheum 58, 34138.
15 Heuft-Dorenbosch L, Landewe R, Weijers R et al. (2006)
Combining information obtained from magnetic reso-
nance imaging and conventional radiographs to detect
sacroiliitis in patients with recent onset inflammatory back
pain. Ann Rheum Dis 65, 8048.
16 Oostveen J, Prevo R, den Boer J, van de Laar M (1999)
Early detection of sacroiliitis on magnetic resonance imag-
ing and subsequent development of sacroiliitis on plain
radiography. A prospective, longitudinal study. J Rheuma-
tol 26, 19538.
17 Sieper J, van der Heijde D, Dougados M et al. (2013) Effi-
cacy and safety of adalimumab in patients with non-radio-
graphic axial spondyloarthritis: results of a randomised
placebo-controlled trial (ABILITY-1). Ann Rheum Dis 72,
81522.
18 Song IH, Weiss A, Hermann KG et al. (2013) Similar
response rates in patients with ankylosing spondylitis and
non-radiographic axial spondyloarthritis after 1 year of
treatment with etanercept: results from the ESTHER trial.
Ann Rheum Dis 72, 8235.
International Journal of Rheumatic Diseases 2016
Etanercept in nr-axSpA patients
19 Landewe R, Braun J, Deodhar A et al. (2014) Efficacy of
certolizumab pegol on signs and symptoms of axial
spondyloarthritis including ankylosing spondylitis: 24-
week results of a double-blind randomised placebo-con-
trolled Phase 3 study. Ann Rheum Dis 73 (1), 3947.
20 Barkham N, Keen HI, Coates LC et al. (2009) Clinical and
imaging efficacy of infliximab in HLA-B27-Positive
patients with magnetic resonance imaging-determined
early sacroiliitis. Arthritis Rheum 60, 94654.
21 Callhoff J, Sieper J, Wei A, Zink A, Listing J (2015) Effi-
cacy of TNFalpha blockers in patients with ankylosing
spondylitis and non-radiographic axial spondyloarthritis:
a meta-analysis. Ann Rheum Dis 74, 12418.
22 van der Heijde D, Sieper J, Maksymowych WP et al.
(2011) 2010 Update of the international ASAS recommen-
dations for the use of anti-TNF agents in patients with
axial spondyloarthritis. Ann Rheum Dis 70, 9058.
23 Braun J, Bollow M, Remlinger G et al. (1998) Prevalence
of spondylarthropathies in HLA-B27 positive and negative
blood donors. Arthritis Rheum 41 (1), 5867.
24 Dean LE, Jones GT, MacDonald AG, Downham C, Stur-
rock RD, Macfarlane GJ (2014) Global prevalence of anky-
losing spondylitis. Rheumatology 53, 6507.
25 Dougados M, van der Heijde D, Sieper J et al. (2014)
Symptomatic efficacy of etanercept and its effects on
objective signs of inflammation in early nonradiographic
axial spondyloarthritis: a multicenter, randomized, dou-
ble-blind, placebo-controlled trial. Arthritis Rheumatol 66,
2091102.
26 Benhamou M, Gossec L, Dougados M (2010) Clinical rele-
vance of C-reactive protein in ankylosing spondylitis and
evaluation of the NSAIDs/coxibs treatment effect on C-
reactive protein. Rheumatology 49, 53641.
27 Vastesaeger N, van der Heijde D, Inman RD et al. (2011)
Predicting the outcome of ankylosing spondylitis therapy.
Ann Rheum Dis 70, 97381.
28 Braun J, van der Horst-Bruinsma IE, Huang F et al. (2011)
Clinical efficacy and safety of etanercept versus sul-
fasalazine in patients with ankylosing spondylitis: a ran-
domized, double-blind trial. Arthritis Rheum 63, 154351.
29 Braun J, McHugh N, Singh A, Wajdula JS, Sato R (2007)
Improvement in patient-reported outcomes for patients
with ankylosing spondylitis treated with etanercept 50 mg
once-weekly and 25 mg twice-weekly. Rheumatology
(Oxford) 46, 9991004.
30 Hammoudeh M, Zack DJ, Li W, Stewart VM, Koenig AS
(2013) Associations between inflammation, nocturnal
back pain and fatigue in ankylosing spondylitis and
improvements with etanercept therapy. J Int Med Res 41,
11509.
9