You are on page 1of 11

Series

Arthritis 2
Spondyloarthritis
Maxime Dougados, Dominique Baeten

Spondyloarthritis is a group of several related but phenotypically distinct disorders: psoriatic arthritis, arthritis Lancet 2011; 377: 212737
related to inammatory bowel disease, reactive arthritis, a subgroup of juvenile idiopathic arthritis, and ankylosing See Comment page 2067
spondylitis (the prototypic and best studied subtype). The past decade yielded major advances in the recognition of This is the second in a Series of
spondyloarthritis as an entity, the classication of the disease, and understanding of the genetic and three papers about arthritis
pathophysiological mechanisms of disease-related inammation and tissue damage. In parallel, new clinical and Paris-Descartes University,
imaging outcomes have allowed the assessment of various therapeutic modalities. Blockers of tumour necrosis Medicine Faculty, UPRES
EA 4058, AP-HP, Cochin
factor are a major therapeutic advance, but the exact roles of physiotherapy, and treatment with non-steroidal anti- Hospital, Department of
inammatory drugs and other biological treatments are unknown. The major challenges with direct relevance for Rheumatology B, Paris, France
clinical practice for the next decade are the development of techniques for early diagnosis, therapeutic modulation (Prof M Dougados MD); and
of structural damage, and, ultimately, induction of long-term, drug-free remission. University of Amsterdam,
Academic Medical Centre,
Clinical Immunology and
Introduction back pain and peripheral arthritis as major entry Rheumatology, Amsterdam,
In 1974, Moll and colleagues1 established the concept of criteria. Recognition of the drawbacks of criteria Netherlands (D Baeten MD)
a group of inter-related disorders originally termed focused on a specic subtype, the Assessment of Correspondence to:
seronegative spondarthritides. The group of diseases Spondyloarthritis International Society (ASAS) did a Prof Maxime Dougados,
Ren Descartes University,
now called spondyloarthritis consists of psoriatic large cross-sectional study to propose new criteria on Department of Rheumatology,
arthritis, reactive arthritis, arthritis related to the basis of the two main clinical features identied in Hpital Cochin, 27 Rue du
inammatory bowel disease, a subgroup of juvenile daily practiceeg, axial symptoms and peripheral Faubourg Saint Jacques, Paris,
idiopathic arthritis, and ankylosing spondylitisthe involvement. 75014, France
m.doug@cch.aphp.fr
prototype of spondyloarthritis.2 The various clinical In the rst set of criteria focusing on patients
forms include spinal (axial) features, peripheral presenting with axial symptoms (panel),7 the term axial
arthritis, enthesopathy, and extra-articular features such spondyloarthritis was proposed for the entire range of
as uveitis, psoriasis, and inammatory bowel disease. axial diseases irrespective of structural damage. These
The clinical rationale for grouping these diseases is that criteria emphasise three important points: the relevance
they are simultaneously or sequentially identied in the of the clinical features identied whatever the presenting
same patient or in a family member. Furthermore, symptoms, the value of new imaging techniques to
clinical characteristics such as eye involvement and detect sacroiliac changes, and the contribution of
enthesopathy are similar whatever the diagnosis.1,2 A HLA-B27 typing.
strong argument, based on work in animals, in favour One important advance is the use of MRI to assess
of grouping these diseases is that HLA-B27 transgenic sacroiliac changes. Plain radiographs can detect only
rats develop the various clinical features that are noted structural changes such as joint erosion and
in human beings with spondyloarthritis.3
One subject of debate at present is whether the
clinical approach, including diagnosis, classication, Search strategy and selection criteria
and management, should be focused on a specic We searched The Cochrane Library and Medline for work
disease subtype (eg, ankylosing spondylitis) or on the published in the past 5 years (200510), as well as the
overall group of spondyloarthritis. In the 1970s, several abstracts of the American (American College of
sets of criteria were proposed to classify patients with a Rheumatology) and European (European League Against
specic spondyloarthritis subtype, such as the modied Rheumatism) congresses of Rheumatology published during
New York criteria for ankylosing spondylitis.4 These the past 2 years (200910). We used the search terms
criteria have important restrictions in clinical practice: spondyloarthropathy, spondylarthropaty,
they focus exclusively on the axial features, omitting spondyloarthritis, spondylarthritis, ankylosing
the other clinical features of the disease. In 1990, Amor spondylitis, and psoriatic arthritis. We limited our search to
and colleagues5 proposed the rst set of classication published work in English. We also searched the reference lists
criteria for the entire group of spondyloarthritis, of articles identied by this search strategy and in particular
allowing a patient to be classied as having the articles that summarised systematic research on a specic
spondyloarthritis whatever the presenting symptoms. topic. Review articles and book chapters are cited to provide
A dierent set of criteria for the entire group of readers with more details and more references than we can
spondyloarthritis was developed by the European accommodate in this paper.
Spondyloarhropathy Study Group,6 with inammatory

www.thelancet.com Vol 377 June 18, 2011 2127


Series

rheumatological involvement (eg, peripheral arthritis,


Panel: ASAS classication criteria for axial enthesopathy, dactylitis) without axial symptoms.11
spondyloarthritis in patients with back pain for 3 months These criteria (gure 1) also emphasise the importance
or more and age at onset younger than 45 years of the dierent clinical features, HLA-B27 typing, and
Sacroiliitis on imaging* plus one or more features of imaging of sacroiliac joints despite the absence of
spondyloarthritis spinal symptoms. Sacroiliac abnormalities at imaging
or raise the question of which of the investigations should
HLA-B27 plus two or more other features of be done when spondyloarthritis is suspected, whatever
spondyloarthritis the presenting symptoms. Clinicians agree on the use
of HLA-B27 typing, although it is only useful in cases
ASAS=Assessment of Spondyloarthritis International Society. *Active (acute)
with an a-priori high suspicion, and a negative result
inammation on MRI highly suggestive of sacroiliitis associated with spondyloarthritis
or denite radiographic sacroiliitis according to modied New York criteria. does not preclude the presence of spondyloarthritis.
Inammatory back pain, arthritis, enthesitis (heel), uveitis, dactylitis, psoriasis, Crohns The ndings of the study12 used to develop the criteria
disease or ulcerative colitis, good response to non-steroidal anti-inammatory drugs,
family history for spondyloarthritis, HLA-B27, or elevated C-reactive protein
also suggest that in a case of peripheral rheumatological
(a spondyloarthritis feature in the context of chronic back pain). presentation, the systematic radiological (eg, plain
radiographs and MRI) assessment of the sacroiliac
joints might be of interest even in the absence of any
subchondral-bone sclerosis seen at the late stage of the axial features. Similarly, a systematic assessment of
disease; this restriction is also the case for CT, although dierent entheses allows dierentiation between
with higher sensitivity and specicity but greater spondyloarthritis patients and controls even in the
exposure to radiation. Unfortunately, the medical term absence of clinical enthesopathy.1315
used to describe such chronic changes focuses on Enthesopathy, inammation at the bone insertion
inammationeg, sacroiliitisdespite the fact that sites of ligaments and tendons, is an important ASAS
plain radiographs cannot detect inammation. By criterion. The main peripheral clinical location is the
contrast, MRI allows the visualisation of synovial uid, heel (inferior part at the insertion of plantar fascia on
synovitis within the sacroiliac joint, and subchondral- the calcaneus and posterior part at the insertion of
bone oedema. The relevant abnormalities detected with Achilles tendon on the calcaneus). The recognition of
MRI have been described and clearly dened,8 allowing spondyloarthritis and the use of these new criteria
inclusion of active inammatory lesions of sacroiliac should allow clinical trials in patients with early disease
joints with denite bone-marrow oedema and osteitis on and thereby the assessment of treatments to alter the
MRI in the new criteria for axial spondyloarthritis.7 course of the disease. Whether these criteria will also
Whether such a denitioneg, MRI ndings at the shorten the diagnostic delay remains to be investigated
sacroiliac jointsis optimum remains an open question prospectively. Another interesting approach to reduce
since data suggest that inammatory lesions of the the diagnostic delay is the development of early referral
posterior structures of the spine as well as the spinal strategies, since patients with back pain are usually rst
fatty Romanus lesions (fatty changes at vertebral corners) seen by primary care physicians.16 Dening better
are also suggestive of spondyloarthritis.9,10 More strategies and techniques for early diagnosis remains
importantly, however, these criteria were developed in a one of the major challenges in spondyloarthritis for the
well dened cross-sectional study population (eg, age next decade.
<45 years and with back pain for at least 3 months) and
have not yet been validated for diagnostic use in Pathophysiology
prospective studies in clinical practice. Advances in the classication of spondyloarthritis show
The second set of criteria proposed by ASAS is that progress in the understanding of genetics (eg, the
focused on patients presenting with peripheral gene for HLA-B27), the pathophysiology of inammation
(eg, lesions on MRI), and structural damage
(eg, sacroiliitis on plain radiographs) aect clinical
Arthritis, enthesitis, or dactylitis practice in the context of classication and diagnosis.
Basic understanding of the pathophysiology of the
disease is even more relevant for outcome measurement
Plus one of Plus two of
Uveititis Arthritis
and targeted treatment.
Psoriasis Enthesitis Through familial aggregation studies17 investigators
Crohns disease or ulcerative colitis or Dactylitis have estimated that genetic risk factors contribute to
Preceding infection Inammatory back pain ever
HLA-B27 Family history of spondyloarthritis 8090% of the susceptibility to ankylosing spondylitis.
Sacroiliitis on imaging The stronger concordance rates between monozygotic
(5075%) versus dizygotic (15%) twins conrms that
Figure 1: ASAS criteria for peripheral spondyloarthritis in patients with peripheral features only familial aggregation is related to genetic rather than
Adapted from Rudwaleit and colleagues.11 ASAS=Assessment of Spondyloarthritis International Society. environmental factors.

2128 www.thelancet.com Vol 377 June 18, 2011


Series

The major genetic risk factor is HLA-B27, an MHC


Gene Function Associated with
class I molecule. This association is present in many
genetically diverse populations and across all major Ankylosing Psoriasis* Inammatory
spondylitis bowel disease*
HLA-B27 subtypes. Whereas HLA-B2706 and HLA-B2709
have long been thought to be protective, the nding of 6p21.3 HLA-B Antigen presentation Yes

ankylosing spondylitis in carriers of these alleles that 5q15 ERAP1 Aminopeptidase Yes Probable
encode these molecules suggests a hierarchy of 1p31.2 IL23R Cytokine receptor Yes Yes Yes
association of dierent HLA-B27 subtypes with 2p15 Yes
ankylosing spondylitis.18 Whether the eect of the specic 21q22 Yes
aminoacid substitutions in the peptide binding groove of 12p13.2 TNFRSF1A Cytokine receptor Probable Yes
HLA-B2706 and HLA-B2709 can explain the dierential 16q22 TRADD Signalling Probable
association in vivo remains to be established.19 9q32 TNFSF15 Inammatory cytokine Probable Yes
The presence of HLA-B27 in 8090% of patients 2q14 IL1A Inammatory cytokine Probable
with ankylosing spondylitis and the spontaneous 2q12 IL1R2 Cytokine receptor Probable
spondyloarthritis-like disease in HLA-B27 transgenic rats 9q34 CARD9 Innate immune defence Probable
suggest a direct and dominant eect of the gene encoding 4q21.3 ANTXR2 Vascular morphogenesis Probable
this molecule.3 However, only a small proportion of
*Some factors are also associated with psoriasis or inammatory bowel disease.
people in the general population who harbour HLA-B27
(56% in white people) develop ankylosing spondylitis, Table: Overview of the locus, gene, and function of denite and probable genetic risk factors for
and HLA-B27 explains only 2040% of the genetic ankylosing spondylitis
susceptibility to ankylosing spondylitissuggesting the
contribution of additional genes. Genome-wide associ- and skin disease, with environmental factors such as
ations studies (GWASs) have allowed the identication the gut ora and additional genetic factors determining
of several of these additional genes (table). the exact phenotype.27,28
A denite association has been identied with the The traditional pathophysiological framework for
genes for endoplasmic reticulum aminopeptidase 1 spondyloarthritis is the arthritogenic-peptide theory,
(ERAP1), interleukin 23 receptor (IL23R), and the gene which proposes that HLA-B27 presents self-peptides that
deserts on chromosome 2p15 and 21q22.20,21 Besides resemble pathogen-derived peptides to CD8-restricted
these denite associations, GWAS ndings suggested T lymphocytes. Circumstantial evidence for this
potential associations with genes for tumour necrosis hypothesis is provided by the triggering of
factor (TNF) receptor 1 (TNFSF1A), the signalling spondyloarthritis by gastrointestinal or urogenital
molecule TNF receptor 1-associated death domain infections, and the presence of HLA-B27-restricted
protein (TRADD), the TNF superfamily cytokine CD8-T-cell clones that are reactive against bacterial
TNFSF15, interleukin 1 (IL1A), interleukin 1 receptor 2 antigens29 as well as against self-proteins from cartilage30
(IL1R2), the vascular morphogenesis protein gene in the inamed joint. However, this hypothesis has been
anthrax toxin receptor 2 (ANTXR2), and the innate seriously challenged by two independent reports that
immune receptor caspase recruitment domain family, CD8 T cells are not needed for disease in HLA-B27
member 9 (CARD9).2025 Other candidate genes such as transgenic rats.31,32 Moreover, the anticartilage responses
non-B27 MHC genes, the familial Mediterranean fever- in human beings are not disease-specic, suggesting
related MEFV, and signal transducer and activator of common secondary autoimmune responses rather
transcription 3 (STAT3) need additional conrmation. than primary pathophysiological processes. In more
The strong genetic predisposition also applies to other general terms, the scarce evidence for HLA-B27-
spondyloarthritis subtypes as suggested by a recurrence restricted autoimmune-T-cell responses, the absence of
rate of disease in 12% of the rst-degree relatives of shared genetic risk factors for autoimmune diseases
spondyloarthritis patients. Accordingly, genes encoding such as PTPN22 polymorphisms, and the absence of
HLA-B27 and interleukin 23 receptor are associated with known disease-specic autoantibodies question
dierent spondyloarthritis subtypes. Additionally, genes whether spondyloarthritis is a genuine autoimmune
such as IL23R also confer risk for spondyloarthritis- disease driven by T-cell or B-cell reactivity
associated disorders such as Crohns disease and towards self-antigens.
psoriasis (table). The absence of familial clustering of Two additional hypotheses have emerged to explain
distinct phenotypic features of the subtypes suggests a the role of HLA-B27 (gure 2). Both hypotheses argue
dominant shared genetic factor in all spondyloarthritis for an autoinammatory rather than autoimmune
forms, with additional genetic and environmental origin since HLA-B27 has a role in triggering innate
factors contributing to the phenotypic diversity.26 immune responses rather than its canonical role of
Reinforcing this idea, HLA-B27 transgenic rats develop antigen presentation. If correct, this hypothesis might
not only spondylitis but also the full spondyloarthritis have important implications because it predicts, for
clinical range with peripheral arthritis, colitis, uveitis, example, that inammation will happen at sites of

www.thelancet.com Vol 377 June 18, 2011 2129


Series

function.40,41 The role of mechanical stress is emphasised


HLA-B27
?
by imaging and pathological ndings that inammation
ERAP1 happens mainly at the synovio-entheseal complex.42
Taken together with the prominent inltration with
Antigen presentation Heavy-chain homodimers Heavy-chain misfolding
innate immune cells at aected sites,43,44 the stress
Bacterial Mechanical hypothesis proposes that inammation in spondylo-
stress stress arthritis is induced by abnormal innate immune
Autoreactive T cells NK-cell activation Unfolded-protein responses responsiveness to mechanical or bacterial danger signals
and should thus be seen as an autoinammatory rather
than autoimmune disorder.45
TNFR1, TRADD
?
TNF IL-17
? IL23R, IL1A, IL1R2, Two cytokines are of particular interest in the
TNFSF15, CARD9
propagation and perpetuation of inammation in
spondyloarthritis. First, a key role for TNF has been
Autoimmune Autoinammatory
shown through the eectiveness of TNF blockers. This
Figure 2: Potential roles of HLA-B27 in triggering the pathogenesis of spondyloarthritis role ts with the genetic associations with TNFR1 and
The three main hypotheses relate to the presentation of arthritogenic peptides to autoreactive T lymphocytes, the the TNFR1 signalling molecule TRADD; however, how
formation of heavy-chain homodimers (which activate natural killer cells), and the misfolding of HLA-B27 in the TNF drives spondyloarthritis is unclear. Many models of
endoplasmic reticulum leading to an unfolded-protein response. The role of ERAP1 has not been assessed. Upon
TNF overexpression lead to sacroiliitis, with one model
bacterial or mechanical stress, these pathways can lead to the abnormal production of proinammatory cytokines such
as tumour necrosis factor and interleukin 17. Investigators do not completely understand the role of additional genetic giving TNFR1 signalling to stromal cells a prominent
associations. ERAP1=endoplasmic reticulum aminopeptidase 1. NK=natural killer. TNFR1=TNF receptor 1. TRADD=TNF role.46 However, these models dier fundamentally from
receptor 1-associated death domain protein. IL23R=interleukin 23 receptor. IL1A=interleukin 1. IL1R2=interleukin 1 spondyloarthritis by their polyarticular, erosive character
receptor 2. TNFSF15=TNF superfamily cytokine 15. CARD9=caspase recruitment domain family, member 9.
without osteoproliferation. The low titres of soluble TNF
in spondyloarthritis synovitis47 and spinal deformities in
bacterial or mechanical stress, and that T-cell or B-cell mice overexpressing transmembrane TNF48 warrant
directed treatments might not be eective further investigation of the forms of TNF and TNF
in spondyloarthritis. receptors in the disease process.
The rst hypothesis proposes that 2 microglobulin- The second cytokine of interest is interleukin 23.
free HLA-B27 heavy chains can assemble into disulphide- Besides the genetic association with IL23R, evidence is
linked homodimers expressed at the cell surface that emerging that the HLA-B27 induced UPR augments the
can be directly recognised by the killer immunoglobulin- production of interleukin 23.39 Altered interleukin-23
like receptors KIR3DL2 independently of the bound production or signalling in spondyloarthritis could lead
peptide.33 Titres of natural killer and T cells expressing to abnormal interleukin-17 responses, certainly in view
KIR3DL2 are raised in HLA-B27-positive patients and of the data that TNFSF15, CARD9, and the DR3-TRADD
can be directly activated by ligation of the homodimers.34 pathways can also aect responses of T-helper-17 (Th17)
The second hypothesis proposes that the Cys 67 residue cells. Early demonstration of interleukin-17 overexpres-
of the B pocket leads to HLA-B27 heavy-chain misfolding sion in spondyloarthritis49 could, however, not be
in the endoplasmic reticulum before assembly into conrmed by independent studies on blood, synovial
complexes with 2 microglobulin and peptide.35,36 The uid, and gut.47,5052 Keeping in mind that interleukin 23
resulting unfolded-protein response (UPR) induces an has several functions and targets many cells besides
altered responsiveness and cytokine production of Th17 cells, these expression studies need to be extended
inammatory cells to a range of innate immune to functional studies in vitro and in vivo. Investigators
stimuli.3739 However, overexpression of human have yet to clarify the potential role of interleukin 1, as
2 microglobulin to reduce the UPR in HLA-B27 suggested by the genetic associations, and interleukin 6
transgenic rats exacerbated rather than prevented in the induction of a Th17 response, and more generally
arthritis and spondylitis, whereas colitis was in the pathophysiology of spondyloarthritis. Emerging
unchanged.28 Although investigators still debate to what data from clinical trials aiming to block
extent 2 microglobulin overexpression really down- interleukins 1, 6, 17, or 23 will be crucial to understand
regulates UPR, this discrepancy emphasises that the the role of these cytokines.
non-mutually exclusive functions of HLA-B27 might Genetic risk factors, and the related hypotheses, fall
dier between models and between distinct features of short of explaining the second major feature of
spondyloarthritis. spondyloarthritis: the prominent tissue remodelling that
The altered cellular responsiveness induced by the leads to osteoproliferation and ankylosis. Three major
UPR accords with the predilection of spondyloarthritis hypotheses have emerged.
for tissues exposed to either bacterial or mechanical First, the typical structural features cannot be
stress. Bacterial stress is shown by the association with explained by the presumption that the disease is non-
inammatory bowel disease, gastrointestinal infections, erosive. Imaging and histological studies clearly show
and abnormal Toll-like receptor expression and that bone destruction and erosions are prominent

2130 www.thelancet.com Vol 377 June 18, 2011


Series

features of both axial and peripheral spondyloarthritis.53,54


Accordingly, cellular and molecular pathways of Trigger Stromal activation
cartilage and bone destruction are activated at the sites
of pathology and, as in rheumatoid arthritis, are largely Tissue inammation
dependent on TNF.5558
BMP
In line with these ndings, the second emerging
hypothesis is that the structural features of spondylo- TNF Wnt

arthritis relate to important pathways of endochondral-


bone formation. In a model of spontaneous ankylosing Tissue destruction Tissue formation
enthesitis, signalling by bone morphogenetic proteins
was the key pathway driving the structural changes and
active signalling of the proteins was identied in target
Structural outcome
tissues of human spondyloarthritis.59 In TNF transgenic
mice, activation of Wnt signalling by targeting the
inhibitor Dickkopf-related protein 1 reversed the process Figure 3: Connection between inammation and endochondral-bone
formation in spondyloarthritis
of bone destruction and induced fusion of sacroiliac Mechanical or other triggers might cause tissue inammation, which is initially
joints.60,61 Several inhibitors of the Wnt pathway seem to characterised by TNF-driven tissue destruction and inhibition of Wnt-mediated
be dysfunctional in human spondyloarthritis and are repair processes. Resolution of inammation might reduce the inhibition of Wnt
associated with new bone formation.62,63 Further signalling and lead to reactive osteoproliferation. The same initiating trigger might
also directly activate stromal cells and induce an inammation-independent
functional analyses of bone morphogenetic proteins, pathway of endochondral-bone formation, in which BMP signalling is thought to
Wnt, and other tissue-remodelling pathways are of have a key role. The structural outcome as well as the eect of anti-inammatory
paramount importance because they could be attractive treatments on the structural outcome will be determined by the relative
targets for treatment. contributions of these pathways as well as by the type, severity, and duration of
inammation. BMP=bone morphogenic protein. TNF=tumour necrosis factor.
The third emerging possibility is that osteoproliferation
in spondyloarthritis is, at least partly, uncoupled from
inammation. Two hypotheses have been proposed to treatment will prevent structural damage whereas the
account for this uncoupling. The rst hypothesis claims second hypothesis predicts that separate assessment and
that osteoproliferation can be explained by the therapeutic targeting of stromal pathways is needed for
intermittent nature of the inammation.64 In an early optimum management of spondyloarthritis.
disease phase, TNF would simultaneously drive
destruction and inhibit remodelling by the Wnt pathway Outcome assessment
by upregulating Dickkopf-related protein 1 (gure 3). The optimum management of patients necessitates
On downregulation of TNF in a later phase, the brake systematically addressing ve points related to the
on Wnt-mediated remodelling would be released and possible clinical presentations (axial, peripheral,
the early erosions would trigger reactive osteo- enthesopathy, and extra-articular): does the patient
proliferation. The relation between early inammation really have the disease, is the disease active, is the
and subsequent new bone formation is, however, still disease severe, is the disease potentially severe, and is
highly debatable in human ankylosing spondylitis the disease refractory? One of the major challenges in
because although inammation is associated with a spondyloarthritis remains the development of sensitive
greater likelihood new bone formation, most and specic imaging or biological markers for
syndesmophytes are located at sites without detectable early diagnosis.
inammation.65,66 Moreover, this hypothesis cannot Activity in spondyloarthritis is a reference to the
explain why new bone formation is independent of inammation caused by the disease, which is commonly
osteoclasts in various models67,68 or why TNF blockade assessed in daily practice with the Bath ankylosing
does not prevent ankylosing enthesitis.69 The second spondylitis disease activity index (BASDAI).71 This index
hypothesis proposes that direct activation of stromal consists of questions related to the patients self-
pathways, including the pathways of bone morphogenic assessment (eg, fatigue, axial symptoms, peripheral
protein, leads to new tissue formation independent of symptoms, enthesopathy, and duration and intensity of
inammation or early erosive changes.70 Mechanical morning stiness). To improve the objective properties
stress at synovio-entheseal complexes might then of such an index, an ankylosing spondylitis disease
induce distinct and unrelated pathways of inammation activity score (ASDAS) has been developed that includes
and tissue remodelling. not only four questions from the BASDAI, but also the
Although both hypotheses are not mutually exclusive level of acute phase reactants.72 Preliminary data suggest
(gure 3), the relative contribution of both mechanisms that the ASDAS is more discriminative than BASDAI
and the exact relation between inammation and when in assessment of TNF blockers.73 However,
stromal-cell activation has major clinical implications: clinicians must further assess the usefulness of this new
the rst hypothesis predicts that early anti-inammatory composite index in daily practice.

www.thelancet.com Vol 377 June 18, 2011 2131


Series

Apart from clinical assessment, the activity of the without needing to be classied a failure of a disease-
disease could also be assessed by MRI of the spine and modifying antirheumatic drug. For peripheral arthritis,
sacroiliac joints. Several scoring systems at present assess a refractory disease is dened by an active disease despite
the reliability, validity, and responsiveness of the current or past intake of disease-modifying antirheumatic
technique.7476 Investigators are yet to clarify if the drugs.90 To disseminate and facilitate the implementation
inammatory abnormalities of the posterior elements of of such methods, ASAS has recently published a guide
the spine should be included77,78 or even if a whole-body to assess spondyloarthritis.91
MRI should be preferred.79,80 Whatever the scoring
system, axial disease activity measured by MRI as well as Treatment
other imaging modalities (eg, ultrasonography for The objectives of treatment of spondyloarthritis are to
enthesopathy) are useful additional outcome measures improve the condition of the patient (eg, pain, functional
in clinical trials;8186 but their added value in daily clinical disability) as well as to prevent subsequent clinical
practice is unknown. deterioration. ASAS has provided recommendations for
The severity of spondyloarthritis is a reference to both the management of spondyloarthritis in general90
irreversible structural damage caused by the disease, and the use of TNF blockers in particular.92
often due to tissue remodelling and its functional A recent Cochrane systematic review of published
consequences. For clinical studies, several outcomes work93 concluded that an individual home-based or
have been proposed to show severity: death, job loss, supervised exercise programme is better than no
functional impairment, range of motion, and hip intervention, that supervised group physiotherapy is
involvement. Radiological scoring systems assess better than home exercises, and that combined in-patient
structural damage at the axial level (eg, mainly new bone spondyloarthritis-exercise therapy with subsequent group
formation because of syndesmophytes).87,88 The scoring physiotherapy is better than group physiotherapy alone.
system recommended at present is the modied stoke Despite the modality of physiotherapy, another important
ankylosing spondylitis scoring system,88 which consists question is related to the characteristics of the patients
of cervical and lumbar assessments. The addition of the who should benet most from this therapy. In particular,
thoracic spine might improve the sensitivity to change.89 the benet of such therapy during the painful
This new system is very useful in clinical research but it inammatory ares of the disease or at a very early stage
remains unclear whether it should be used routinely in has not been investigated.
clinical practice. NSAIDs are the cornerstone of pharmacological
A further factor in the optimum management of the intervention for ankylosing spondylitis, rapidly reducing
disease relates to prediction of the natural course of the pain and stiness after 4872 h. Despite this dramatic
disease at an early stage in an individual patient. This symptomatic eect, NSAIDs might be also eective on
notion is clinically highly relevant because structural some other outcome measures. These drugs might
damage and functional impairment in spondyloarthritis substantially reduce the level of acute-phase reactants
are largely irreversible. If the hypothesis that early compared with placebo, but with questionable relevance
inammatory and erosive lesions trigger subsequent of the recorded size of eect.94 The investigators of one
osteoproliferation is correct, highly ecient anti- study95 also suggest that NSAIDs can delay radiological
inammatory treatments should be started as early as progression of spine disease when given continuously as
possible. However, preliminary data suggest that axial a daily dose over 2 years, compared with an on-demand
inammatory lesions detected with MRI are not highly treatment schedule.95 Although this nding suggests
predictive for subsequent ossication.63,64 If the structural that a systematically continuous daily intake of NSAIDs
damage is independent of inammation but relates to might be of benet, the converse argument is the
stromal remodelling pathways, it would make a case for potential long-term gastrointestinal and cardiovascular
additional treatments targeting stromal remodelling toxic eects of such therapy, in particular in patients
irrespective of disease activity and before irreversible recognised as having more comorbidities than the
structural damage. Long-term follow-up of patients general population.96
presenting with recent inammatory back pain in Conventional disease-modifying antirheumatic drugs
dierent European cohorts will help to delineate which such as sulfasalazine, methotrexate, and leunomide
patients are at risk for long-term structural damage. which have been shown to be eective in the treatment
Whether the disease is refractory is important to guide of rheumatoid arthritis, have no proven ecacy for either
the decision for second-line treatments such as anti- the axial or enthesopathic features of spondyloarthritis,
TNF. For axial spondyloarthritis, the present recom- but some ecacy for peripheral arthritis and other extra-
mendation from ASAS and the European League Against articular features such as psoriasis, uveitis, and
Rheumatism is to dene a patient as refractory when inammatory bowel disease.9799
active disease persists despite the intake of at least two Thalidomide has some ecacy in axial spondyloarthritis
courses of non-steroidal anti-inammatory drugs in open uncontrolled studies, possibly because of its
(NSAIDs) taken at an optimum dose for at least 2 weeks anti-TNF eect, but is thought too toxic for widespread

2132 www.thelancet.com Vol 377 June 18, 2011


Series

use.100102 Although some clinical eectiveness of anti-inammatory treatments as well as for remission-
pamidronate, a bisphosphonate with potential anti- inducing and structure-modifying therapies.
inammatory and antierosive eects, has been Several other biological agents have been tested in
reported,103 further placebo-controlled studies are needed small proof-of-concept trials125129 in ankylosing
before this treatment can be recommended. spondylitis and psoriatic arthritis on the basis of their
The major clinical and therapeutic advance in ecacy in related diseases such as rheumatoid arthritis
spondyloarthritis care is the successful use of TNF blockade and psoriasis. B-cell depletion by the anti-CD20
in active, refractory disease.104107 Registration studies in antibody rituximab did not show similar ecacy in
early, preradiographic axial spondyloarthritis as well as in ankylosing spondylitis as in rheumatoid arthritis,
undierentiated peripheral spondyloarthritis are ongoing. although some response was noted in TNF-blocker
A couple of issues are of particular relevance for daily naive patients in an open study.125,126 T-cell targeted
clinical practice. First, TNF blockade is highly eective therapies have also not been very successful in
in targeting the dierent disease featureseg, not only spondyloarthritis. Despite the role of T cells in psoriasis,
axial disease but also peripheral arthritis, enthesitis, psoriatic arthritis did not respond to the anti-CD11a
and extra-articular features such as psoriasis or monoclonal antibody efalizumab127 and showed only
uveitis.108111 TNF blockade also has a substantial eect slight clinical improvement with the anti-LFA3 antibody
on general symptoms such as fatigue and substantially alefacept128 and the CTLA4-Ig construct abatacept.129
improves the overall function and quality of life. Long- These ndings are thus consistent with the
term follow-up studies suggest that eectiveness is pathophysiological concept that spondyloarthritis might
maintained for several years of treatment. Second, be an autoinammatory rather than a T-cell or B-cell
short-term and long-term studies suggest a safety driven autoimmune disease.
prole of TNF blockade in spondyloarthritis that is The genetic associations have raised interest in
similar to that of rheumatoid arthritis and inammatory blockade of cytokines other than TNF. Interleukin-1
bowel disease, and reactivation of tuberculosis remains blockade with the interleukin-1 receptor antagonist
the major concern.112 Third, the various TNF blockers anakinra did not show consistent ecacy in ankylosing
seem to be equally potent for the treatment of axial, spondylitis,130 the human anti-interleukin 12/
peripheral, and extra-articular features, with the interleukin 23 monoclonal antibody ustekinumab had
exception that etanercept has no proven ecacy in slight but signicant (p=00002) ecacy in psoriatic
inammatory bowel disease. As to safety, large registries arthritis.131 Trials with interleukin-6 and interleukin-17
suggest that the risk for tuberculosis and possibly also blockade are underway.
herpes zoster might be lower with etanercept than with
the monoclonal anti-TNF antibodies iniximab and Future prospects
adalimumab.113115 Finally, by contrast with rheumatoid From the present state of the art in spondyloarthritis,
arthritis and Crohns disease, there is no recom- several important clinical and pathophysiological issues
mendation at present in axial spondyloarthritis to seem unsolved: the development and validation of better
combine TNF blockers with drugs such as methotrexate clinical or biological markers for early diagnosis and for
or azathioprine. prognosis, clarication whether the subtypes of spondylo-
Despite its major therapeutic eectiveness, TNF arthritis are driven by dierent pathophysiological
blockade also has important limitations. First, 2040% processes or rather represent dierent phenotypes of a
of the patients do not respond well to treatment and single pathological entity, deciphering the functional role
clinical, biological, or imaging ndings that predict a and the interaction of genes emerging from GWAS, in-
better response at the group level lack specicity to depth understanding of the cellular and molecular
make reliable predictions in individual patients.116,117 In mechanisms of tissue remodelling and their interaction
case of failure of a rst TNF blocker, trying a second with inammation, and the development of newer anti-
drug is justied since many patients do still respond to inammatory therapies, including the clinical assessment
a dierent anti-TNF.118,119 Second, TNF blockade does not of interleukin-6 blockade and interleukin-17 blockade, as
induce longlasting remission since almost all patients well as treatments targeting tissue remodelling. In view of
relapse within 612 months of interruption of the present research eorts, our understanding of these
treatment.120 Third, TNF blockade seems to halt joint issues will probably develop rapidly over the coming years.
destruction,121 but fails to substantially slow new bone Contributors
formation in spondyloarthritis.122124 It remains unclear The authors contributed equally to the preparation of this paper.
whether this eect is related to the fact that TNF Conicts of interest
blockade was started too late in the disease course in MD has received consultancy fees, speakers fees, and his department
these studies or to the fact that new bone formation is has received research grants from Abbott, Aventis, Pzer, Roche, Sano,
Shering Plough, and UCB. DB has received consultancy fees, speakers
uncoupled from TNF-driven inammation in fees, and research grants from Abbott, Centocor, Novartis, Pzer,
spondyloarthritis. These three caveats suggest that there Shering Plough, and UCB.
is still an important unmet need for highly eective

www.thelancet.com Vol 377 June 18, 2011 2133


Series

Acknowledgments 21 Australo-Anglo-American Spondyloarthritis Consortium (TASC),


We thank Kris A Reedquist for critical reading of our paper. Reveille JD, Sims AM, et al. Genome-wide association study of
ankylosing spondylitis identies non-MHC susceptibility loci.
References Nat Genet 2010; 42: 12327.
1 Moll JM, Haslock I, Macrae IF, et al. Associations between 22 Sims AM, Timms AE, Bruges-Armas J, et al. Prospective
ankylosing spondylitis, psoriatic arthritis, Reiters disease, the meta-analysis of interleukin 1 gene complex polymorphisms
intestinal arthropathies, and Behcets syndrome. conrms associations with ankylosing spondylitis.
Medicine (Baltimore) 1974; 53: 34364. Ann Rheum Dis 2008; 67: 130509.
2 Dougados M, Landew R. Spondyloarthritis: pathogenesis, clinical 23 Pointon JJ, Harvey D, Karaderi T, et al. The chromosome 16q
aspects and diagnosis. In: JWJ Bijlsma, ed. EULAR compendium region associated with ankylosing spondylitis includes the
on rheumatic diseases. London: BMJ Publishing Group, 2009: candidate gene tumour necrosis factor receptor type 1-associated
92115. death domain (TRADD). Ann Rheum Dis 2010; 69: 124346.
3 Hammer RE, Maika SD, Richardson JA, Tang JP, Taurog JD. 24 Zinovieva E, Bourgain C, Kadi A, et al. Comprehensive linkage
Spontaneous inammatory disease in transgenic rats expressing and association analyses identify haplotype, near to the TNFSF15
HLA-B27 and human 2m: An animal model of HLA-B27- gene, signicantly associated with spondyloarthritis. PLoS Genet
associated human disorders. Cell 1990; 63: 1099112. 2009; 5: e1000528.
4 van der Linden S, Valkenburg HA, Cats A. Evaluation of 25 Pointon JJ, Harvey D, Karaderi T, et al. Elucidating the
diagnostic criteria for ankylosing spondylitis. A proposal for chromosome 9 association with AS; CARD9 is a candidate gene.
modication of the New York criteria. Arthritis Rheum 1984; Genes Immun 2010; 11: 49096.
27: 36168.
26 Said-Nahal R, Miceli-Richard C, dAgostina MA, et al.
5 Amor B, Dougados M, Mijiyawa M. Criteria of the classication of Phenotypic diversity is not determined by independent genetic
spondylarthropathies. Rev Rhum Mal Osteoart 1990; 57: 8589. factors in familial spondylarthropathy. Arthritis Rheum 2001;
6 Dougados M, van der Linden A, Juhlin R, et al. The European 45: 47884.
Spondylarthropathy Study Group preliminary criteria for the 27 Taurog JD, Richardson JA, Croft JT, et al. The germfree state
classication of spondylarthropathy. Arthritis Rheum 1991; prevents development of gut and joint inammatory disease in
34: 121827. HLA-B27 transgenic rats. J Exp Med 1994; 180: 235964.
7 Rudwaleit M, van der Heijde D, Landew R, et al. The 28 Tran TM, Dorris ML, Satumtira N, et al. Additional
development of Assessment of SpondyloArthritis international human 2-microglobulin curbs HLAB27 misfolding and
Society classication criteria for axial spondyloarthritis (part II): promotes arthritis and spondylitis without colitis in male
validation and nal selection. Ann Rheum Dis 2009; 68: 77783. HLAB27transgenic rats. Arthritis Rheum 2006; 54: 131727.
8 Rudwaleit M, Jurik AG, Hermann KG, et al. Dening active 29 Hermann E, Yu DT, Meyer zum Bschenfelde KH, et al.
sacroiliitis on Magnetic Resonance Imaging (MRI) for classication HLA-B27-restricted CD8 T cells derived from synovial uids of
of axial spondyloarthritis: a consensual approach by the ASAS/ patients with reactive arthritis and ankylosing spondylitis. Lancet
OMERACT MRI Group. Ann Rheum Dis 2009; 68: 152027. 1993; 342: 64650.
9 Bochkova AG, Levshakova AV, Bunchuk NV, et al. Spinal 30 Atagunduz P, Appel H, Kuon W, et al. HLA-B27-restricted
inammation lesions as detected by magnetic resonance imaging CD8+ T cell response to cartilage-derived self peptides in
in patients with early ankylosing spondylitis are more often ankylosing spondylitis. Arthritis Rheum 2005; 52: 892901.
observed in posterior structures of the spine. Rheumatology 2010;
49: 74955. 31 May E, Dorris ML, Satumtira N, et al. CD8 T cells are not
essential to the pathogenesis of arthritis or colitis in HLA-B27
10 Bennett AN, Rehman A, Hensor EM, et al. The fatty Romanus transgenic rats. J Immunol 2003; 170: 1099105.
lesion: a non-inammatory spinal MRI lesion specic for axial
spondyloarthropathy. Ann Rheum Dis 2011; 70: 2531. 32 Taurog JD, Dorris ML, Satumtira N, et al. Spondylarthritis in
HLAB27/human 2-microglobulintransgenic rats is not
11 Rudwaleit M, van der Heijde D, Landew R, et al. The Assessment prevented by lack of CD8. Arthritis Rheum 2009; 60: 197784.
of SpondyloArthritis International Society classication criteria
for peripheral spondyloarthritis and for spondyloarthritis in 33 Kollnberger S, Bird LA, Roddis M, et al. HLA-B27 heavy chain
general. Ann Rheum Dis 2011; 70: 2531. homodimers are expressed in HLA-B27 transgenic rodent models
of spondyloarthritis and are ligands for paired Ig-like receptors.
12 Rudwaleit M, van der Heijde D, Landew R, et al. New ASAS J Immunol 2004; 173: 1699710.
classication criteria for peripheral spondyloarthritis.
Ann Rheum Dis 2009; 68 (suppl 3): 127. 34 Chan AT, Kollnberger SD, Wedderburn LR, Bowness P. Expansion
and enhanced survival of natural killer cells expressing the killer
13 dAgostino MA, Aegerter P, Jousse-Joulin S, et al. How to evaluate immunoglobulin-like receptor KIR3DL2 in spondylarthritis.
and improve the reliability of power Doppler ultrasonography for Arthritis Rheum 2005; 52: 358695.
assessing enthesitis in spondyloarthritis. Arthritis Rheum 2009;
61: 6169. 35 Mear JP, Schreiber KL, Mnz C, et al. Misfolding of HLA-B27 as a
result of its B pocket suggests a novel mechanism for its role in
14 de Miguel E, Cobo T, Munoz-Fernandez S, et al. Validity of susceptibility to spondyloarthropathies. J Immunol 1999;
enthesis ultrasound assessment in spondyloarthropathy. 163: 666570.
Ann Rheum Dis 2009; 68: 16974.
36 Dangoria NS, DeLay ML, Kingsbury DJ, et al. HLA-B27
15 Munoz-Fernandez S, de Miguel E, Cobo-Ibanez T, et al. Enthesis misfolding is associated with aberrant intermolecular disulde
inammation in recurrent acute anterior uveitis without bond formation (dimerization) in the endoplasmic reticulum.
spondyloarthritis. Arthritis Rheum 2009; 60: 198590. J Biol Chem 2002; 277: 2345968.
16 Brandt HC, Spiller I, Song IH, et al. Performance of referral 37 Turner MJ, Sowders DP, DeLay ML, et al. HLA-B27 misfolding in
recommendations in patients with chronic back pain and transgenic rats is associated with activation of the unfolded
suspected axial spondyloarthritis. Ann Rheum Dis 2007; protein response. J Immunol 2005; 175: 243848.
66: 147984.
38 Smith JA, Turner MJ, DeLay ML, et al. Endoplasmic reticulum
17 Thomas GP, Brown MA. Genetics and genomics in ankylosing stress and the unfolded protein response are linked to synergistic
spondylitis. Immunol Rev 2010; 233: 16280. IFN- induction via X-box binding protein 1. Eur J Immunol 2008;
18 Ramos M, Alvarez I, Sesma L, et al. Molecular mimicry of an 38: 1194203.
HLA-B27-derived ligand of arthritis-linked subtypes with 39 DeLay ML, Turner MJ, Klenk EI, et al. HLA-B27 misfolding and
chlamydial proteins. J Biol Chem 2002; 277: 3357381. the unfolded protein response augment interleukin-23 production
19 Brown MA, Laval SH, Brophy S, et al. Recurrence risk modelling and are associated with Th17 activation in transgenic rats.
of the genetic susceptibility to ankylosing spondylitis. Arthritis Rheum 2009; 60: 263343.
Ann Rheum Dis 2000; 59: 88386. 40 De Rycke L, Vandooren B, Kruithof E, et al. Tumor necrosis factor
20 Burton PR, Clayton DG, Cardon LR, et al. Association scan of blockade treatment down-modulates the increased systemic and
14,500 nonsynonymous SNPs in four diseases identies local expression of Toll-like receptor 2 and toll-like receptor 4 in
autoimmunity variants. Nat Genet 2007; 39: 132937. spondylarthropathy. Arthritis Rheum 2005; 52: 214658.

2134 www.thelancet.com Vol 377 June 18, 2011


Series

41 Tsui FW, Xi N, Rohekar S, et al. Toll-like receptor 2 variants are 62 Daoussis D, Liossis SN, Solomou EE, et al. Evidence that Dkk-1 is
associated with acute reactive arthritis. Arthritis Rheum 2008; dysfunctional in ankylosing spondylitis. Arthritis Rheum 2010;
58: 343638. 62: 15058.
42 McGonagle D, Lories RJ, Tan AL, et al. The concept of a synovio- 63 Appel H, Ruiz-Heiland G, Listing J, et al. Altered skeletal
entheseal complex and its implications for understanding joint expression of sclerostin and its link to radiographic progression
inammation and damage in psoriatic arthritis and beyond. in ankylosing spondylitis. Arthritis Rheum 2009; 60: 325762.
Arthritis Rheum 2007; 56: 248291. 64 Sieper J, Appel H, Braun J, et al. Critical appraisal of assessment
43 Baeten D, Demetter P, Cuverlier C, et al. Macrophages expressing of structural damage in ankylosing spondylitis: implications for
the scavenger receptor CD163: a link between immune alterations treatment outcomes. Arthritis Rheum 2008; 58: 64956.
of the gut and synovial inammation in spondyloarthropathy. 65 Baraliakos X, Listing J, Rudwaleit M, Sieper J, Braun J. The
J Pathol 2002; 196: 34350. relationship between inammation and new bone formation in
44 Baeten D, Kruithof E, De Rycke L, et al. Inltration of the synovial patients with ankylosing spondylitis. Arthritis Res Ther 2008;
membrane with macrophage subsets and polymorphonuclear 10: R104.
cells reects global disease activity in spondyloarthropathy. 66 Maksymowich WP, Chiowchanwisawakit P, Clare T, et al.
Arthritis Res Ther 2005; 7: R35969. Inammatory lesions of the spine on magnetic resonance
45 McGonagle D, McDermott MF. A proposed classication of the imaging predict the development of new syndesmophytes in
immunological diseases. PLoS Med 2006; 3: e297. ankylosing spondylitis: evidence of a relationship between
46 Armaka M, Apostolaki M, Jacques P, et al. Mesenchymal cell inammation and new bone formation. Arthritis Rheum 2009;
targeting by TNF as a common pathogenic principle in chronic 60: 93102.
inammatory joint and intestinal diseases. J Exp Med 2008; 67 Schett G, Stolina M, Dwyer D, et al. Tumor necrosis factor alpha
205: 33137. and RANKL blockade cannot halt bony spur formation in
47 Vandooren B, Noordenbos T, Ambarus C, et al. Absence of a experimental inammatory arthritis. Arthritis Rheum 2009;
classically activated macrophage cytokine signature in peripheral 60: 264454.
spondylarthritis, including psoriatic arthritis. Arthritis Rheum 68 Lories RJ, Derese I, Luyten FP. Inhibition of osteoclasts does not
2009; 60: 96675. prevent joint ankylosis in a mouse model of spondyloarthritis.
48 Edwards CK, Bendele AM, Reznikov LI, et al. Soluble human p55 Rheumatology 2008; 47: 60508.
and p75 tumor necrosis factor receptors reverse spontaneous 69 Lories R, Derese I, de Bari C, et al. Evidence for uncoupling of
arthritis in transgenic mice expressing transmembrane tumor inammation and joint remodeling in a mouse model of
necrosis factor . Arthritis Rheum 2006; 54: 287285. spondylarthritis. Arthritis Rheum 2007; 56: 48997.
49 Jandus C, Bioley G, Rivals JP, et al. Increased numbers of 70 Lories RJ, Luyten FP, de Vlam K. Progress in spondylarthritis.
circulating polyfunctional Th17 memory cells in patients with Mechanisms of new bone formation in spondyloarthritis.
seronegative spondylarthritides. Arthritis Rheum 2008; 58: 230717. Arthritis Res Ther 2009; 11: 221.
50 Shen H, Goodall JC, Hill Gaston JS. Frequency and phenotype of 71 Garrett S, Jenkinson T, Kennedy LG, et al. A new approach to
peripheral blood TH17 cells in ankylosing spondylitis and dening disease status in ankylosing spondylitis: the BATH
rheumatoid arthritis. Arthritis Rheum 2009; 60: 164756. Ankylosing Spondylitis Disease Activity Index. J Rheumatol 1994;
51 Melis L, Vandooren B, Kruithof E, et al. Systemic levels of IL-23 21: 228691.
are strongly associated with disease activity in rheumatoid 72 Lukas C, Landew R, Sieper J, et al. Development of an
arthritis but not spondyloarthritis. Ann Rheum Dis 2010; ASAS-endorsed disease activity score (ASDAS) in patients with
69: 61823. ankylosing spondylitis. Ann Rheum Dis 2009; 68: 1824.
52 Ciccia F, Bombardieri M, Principato A, et al. Overexpression of 73 van der Heijde D, Lie E, Kvien TK, et al. ASDAS, a highly
interleukin-23, but not interleukin-17, as an immunologic discriminatory ASAS-endorsed disease activity score in patients
signature of subclinical intestinal inammation in ankylosing with ankylosing spondylitis. Ann Rheum Dis 2009; 68: 181118.
spondylitis. Arthritis Rheum 2009; 60: 95565. 74 Braun J, Baraliakos X, Golder W, et al. Magnetic resonance
53 Francois RJ, Gardner DL, Degrave EJ, et al. Histopathologic imaging examinations of the spine in patients with ankylosing
evidence that sacroiliitis in ankylosing spondylitis is not merely spondylitis, before and after successful therapy with iniximab:
enthesitis. Arthritis Rheum 2000; 43: 201124. evaluation of a new scoring system. Arthritis Rheum 2003;
54 McGonagle D, Wakeeld RJ, Tan AL, et al. Distinct topography of 48: 12636.
erosion and new bone formation in achilles tendon enthesitis: 75 Maksymowych WP, Imman RD, Salonen D, et al.
implications for understanding the link between inammation Spondyloarthritis Research Consortium of Canada magnetic
and bone formation in spondylarthritis. Arthritis Rheum 2008; resonance imaging index for assessment of spinal inammation
58: 269499. in ankylosing spondylitis. Arthritis Rheum 2005; 53: 50209.
55 Vandooren B. Kruithof E, Yu DT, et al. Involvement of matrix 76 Madsen KB, Jurik AG. Magnetic resonance imaging grading
metalloproteinases and their inhibitors in peripheral synovitis system for active and chronic spondylarthritis changes in the
and down-regulation by tumor necrosis factor blockade in sacroiliac joint. Arthritis Care Res 2010; 62: 1118.
spondylarthropathy. Arthritis Rheum 2004; 50: 294253. 77 Feydy A, Gossec L, Bazeli R, et al. MR Imaging of the spine and
56 Vandooren B, Cantaert T, Noordenbos T, et al. The abundant sacroiliac joints in ankylosing spondylitis. J Radiol 2010;
synovial expression of the RANK/RANKL/Osteoprotegerin system 91: 14050.
in peripheral spondylarthritis is partially disconnected from 78 Rennie WJ, Dhillon SS, Conner-Spady B, et al. Magnetic
inammation. Arthritis Rheum 2008; 58: 71829. resonance imaging assessment of spinal inammation in
57 Vandooren B, Cantaert T, ter Borg M, et al. Tumor necrosis factor ankylosing spondylitis: standard clinical protocols may omit
drives cadherin 11 expression in rheumatoid inammation. inammatory lesions in thoracic vertebrae. Arthritis Rheum 2009;
Arthritis Rheum 2008; 58: 305162. 61: 118793.
58 Vandooren B, Yeremenko N, Noordenbos T, et al. Mediators of 79 Maksymowych WP, Crowther SM, Dhillon SS, et al. Systematic
structural remodeling in peripheral spondyloarthritis. assessment of inammation by magnetic resonance imaging in
Arthritis Rheum 2009; 60: 353445. the posterior elements of the spine in ankylosing spondylitis.
59 Lories RJ, Derese I, Luyten FP. Modulation of bone Arthritis Care Res 2010; 62: 410.
morphogenetic protein signaling inhibits the onset and 80 Weber U, Maksymowych WP, Jurik AG, et al. Validation of
progression of ankylosing enthesitis. J Clin Invest 2005; whole-body against conventional magnetic resonance imaging for
115: 157179. scoring actue inammatory lesions in the sacroiliac joints of
60 Diarra D, Stolina M, Polzer K, et al. Dickkopf-1 is a master patients with spondyloarthritis. Arthritis Rheum 2009; 61: 89399.
regulator of joint remodelling. Nat Med 2007; 13: 15663. 81 Weber U, Hodler J, Jurik AG, et al. Assessment of active spinal
61 Uderhardt S, Diarra D, Katzenbeisser J, et al. Blockade of inammatory changes in patients with axial spondyloarthritis:
Dickkopf (DKK)-1 induces fusion of sacroiliac joints. validation of whole body MRI against conventional MRI.
Ann Rheum Dis 2010; 69: 59297. Ann Rheum Dis 2010; 69: 64853.

www.thelancet.com Vol 377 June 18, 2011 2135


Series

82 Lambert RG, Salonen D, Rahman P, et al. Adalimumab 102 Wei JC, Chan TW, Lin HS, et al. Thalidomide for severe
signicantly reduces both spinal and sacroiliac joint inammation refractory ankylosing spondylitis: a 6-months open-label trial.
in patients with ankylosing spondylitis: a multicenter, J Rheumatol 2003; 30: 262731.
randomized, double-blind, placebo-controlled study. 103 Maksymowych WP, Jhangri GS, Fitzgerald AA, et al. A six-month
Arthritis Rheum 2007; 56: 400514. randomized, controlled, double-blind, dose-response comparison
83 Maksymowych WP, Salonen D, Inman RD, et al. Low-dose of intravenous pamidronate (60 mg versus 10 mg) in the
iniximab (3 mg/kg) signicantly reduces spinal inammation on treatment of nonsteroidal antiinammatory drug-refractory
magnetic resonance imaging in patients with ankylosing ankylosing spondylitis. Arthritis Rheum 2002; 46: 76673.
spondylitis: a randomized placebo-controlled study. J Rheumatol 104 Van den Bosch F, Kruithof E, Baeten D, et al. Eects of a
2010; 37: 172834. loading dose regimen of three infusions of chimeric monoclonal
84 Pedersen SJ, Sorensen IJ, Hermann KG, et al. Responsiveness of antibody to tumour necrosis factor alpha (iniximab) in
the Ankylosing Spondylitis Disease Activity Score (ASDAS) and spondyloarthropathy: an open pilot study. Ann Rheum Dis 2000;
clinical and MRI measures of disease activity in a 1-year follow-up 59: 42833.
study of patients with axial spondyloarthritis treated with tumour 105 Brandt J, Haibel H, Cornely D, et al. Successful treatment of
necrosis factor alpha inhibitors. Ann Rheum Dis 2010; 69: 106571. active ankylosing spondylitis with the anti-tumor necrosis factor
85 Barkham N, Keen HI, Coates LC, et al. Clinical and imaging alpha monoclonal antibody iniximab. Arthritis Rheum 2000;
ecacy of iniximab in HLA-B27-positive patients with magnetic 43: 134652.
resonance imaging-determined early sacroiliitis. Arthritis Rheum 106 Van den Bosch F, Kruithof E, Baeten D, et al. Randomized
2009; 60: 94654. double-blind comparison of chimeric monoclonal antibody to
86 Dougados M, Combe B, Braun J, et al. A randomised, tumor necrosis factor alpha (iniximab) versus placebo in active
multicentre, double-blind, placebo-controlled trial of etanercept spondylarthropathy. Arthritis Rheum 2002; 46: 75565.
in adults with refractory heel enthesitis in spondyloarthritis: the 107 Braun J, Brandt J, Listing J, et al. Treatment of active ankylosing
HEEL trial. Ann Rheum Dis 2010; 69: 143035. spondylitis with iniximab: a randomised controlled multicentre
87 Wanders AJ, Landew R, Spoorenberg A, et al. What is the most trial. Lancet 2002; 359: 118793.
appropriate radiologic scoring method for ankylosisng 108 van den Bosch F, Kruithof E, De Vos M, et al. Crohns disease
spondylitis? A comparison of the available methods based on the associated with spondyloarthropathy: eect of TNF- blockade
Outcome Measures in Rheumatology Clinical Trials lter. with iniximab on articular symptoms. Lancet 2000;
Arthritis Rheum 2004; 50: 262232. 356: 182122.
88 van der Heijde D, Landew R. Selection of a method for scoring 109 Haibel H, Rudwaleit M, Listing J, et al. Ecacy of adalimumab in
radiographs for ankylosing spondylitis clinical trials, by the the treatment of axial spondylarthritis without radiographically
Assessment in Ankylosing Spondylitis Working Group and dened sacroiliitis: results of a twelve-week randomized,
OMERACT. J Rheumatol 2005; 32: 204849. double-blind, placebo-controlled trial followed by an open-label
89 Baraliakos X, Listing J, Rudwaleit M, et al. Development of a extension up to week fty-two. Arthritis Rheum 2008; 58: 198191.
radiographic scoring tool for ankylosing spondylitis only based 110 Dougados M, Combe B, Braun J, et al. A randomised,
on bone formation: addition of the thoracic spine improves multicentre, double-blind, placebo-controlled trial of etanercept
sensitivity to change. Arthritis Rheum 2009; 61: 76471. in adults with refractory heel enthesitis in spondyloarthritis: the
90 Zochling J, van der Heijde D, Burgos-Vargas R, et al. ASAS/ HEEL trial. Ann Rheum Dis 2010; 69: 143035.
EULAR recommendations for the management of ankylosing 111 Barkham N, Keen HI, Coates LC, et al. Clinical and imaging
spondylitis. Ann Rheum Dis 2006; 65: 44252. ecacy of iniximab in HLA-B27-Positive patients with magnetic
91 Sieper J, Rudwaleit M, Baraliakos X, et al. The Assessment of resonance imaging-determinated early sacroiliitis.
SpondyloArthritis international Society (ASAS) handbook: Arthritis Rheum 2009; 60: 94654.
a guide to assess spondyloarthritis. Ann Rheum Dis 2009; 112 Baeten D, Kruithof E, Van den Bosch F, et al. Systematic safety
68 (suppl 2): ii144. follow up in a cohort of 107 patients with spondyloarthropathy
92 Braun J, Davis J, Dougados M, et al. First update of the international treated with iniximab: a new perspective on the role of host
ASAS consensus statement for the use of anti-TNF agents in patients defence in the pathogenesis of the disease? Ann Rheum Dis 2003;
with ankylosing spondylitis. Ann Rheum Dis 2006; 65: 31620. 62: 82934.
93 Dagnrud H, Kvien TK, Hagen KB. Physiotherapy interventions 113 Tubach F, Salmon D, Ravaud P, et al. Risk of tuberculosis is
for ankylosing spondylitis. Cochrane Database Syst Rev 2008; higher with anti-tumor necrosis factor monoclonal antibody
1: CD002822. therapy than with soluble tumor necrosis factor receptor therapy:
94 Benhamou M, Gossec L, Dougados M. Clinical relevance of the three-year prospective French Research Axed on Tolerance of
C-reactive protein in ankylosing spondylitis and evaluation of the Biotherapies registry. Arthritis Rheum 2009; 60: 188494.
NSAIDs/coxibs treatment eect on C-reactive protein. 114 Dixon WG, Hyrich KL, Watson KD, et al. Drug-specic risk of
Rheumatology (Oxford) 2010; 49: 53641. tuberculosis in patients with rheumatoid arthritis treated with
95 Wanders A, van der Heijde D, Landew R, et al. Nonsteroidal anti-TNF therapy: results from the British Society for
anti-inammatory drugs reduce radiographic progression in Rheumatology Biologics Register (BSRBR). Ann Rheum Dis 2010;
patients with ankylosing spondylitis: a randomized clinical trial. 69: 52228.
Arthritis Rheum 2005; 52: 163436. 115 Strangfeld A, Listing J, Herzer P, et al. Risk of herpes zoster in
96 Kang JH, Chen YH, Lin HC. Comorbidity proles among patients with rheumatoid arthritis treated with anti-TNF-
patients with ankylosing spondylitis: a nationwide agents. JAMA 2009; 301: 73744.
population-based study. Ann Rheum Dis 2010; 69: 116568. 116 Rudwaleit M, Listing J, Brandt J, et al. Prediction of a major
97 Chen J, Liu C. Sulfasalazine for ankylosing spondylitis. clinical response (BASDAI 50) to tumour necrosis factor alpha
Cochrane Database Syst Rev 2005; 2: CD004800. blockers in ankylosing spondylitis. Ann Rheum Dis 2004;
98 Chen J, Liu C. Methotrexate for ankylosing spondylitis. 63: 66570.
Cochrane Database Syst Rev 2006; 4: CD004524. 117 Rudwaleit M, Schwarzlose S, Hilgert ES, et al. MRI in predicting
99 van Denderen JC, van der Paardt M, Nurmohamed MT, et al. a major clinical response to anti-tumour necrosis factor
Double blind, randomized, placebo controlled study of treatment in ankylosing spondylitis. Ann Rheum Dis 2008;
leunomide in the treatment of active ankylosing spondylitis. 67: 127681.
Ann Rheum Dis 2005; 64: 176114. 118 Coates LC, Cawkwell LS, Ng NW, et al. Real life experience
100 Breban M, Gombert B, Amor B, et al. Ecacy of thalidomide in conrms sustained response to long-term biologics and
the treatment of refractory ankylosing spondylitis. switching in ankylosing spondylitis. Rheumatology 2008;
Arthritis Rheum 1999; 42: 58081. 47: 897900.
101 Huang F, Gu J, Zhao W, et al. One-year open-label trial of 119 Coates LC, Cawkwell LS, Ng NW, et al. Sustained response to
thalidomide in ankylosing spondylitis. Arthritis Rheum 2002; long-term biologics and switching in psoriatic arthritis: results
47: 24954. from real life experience. Ann Rheum Dis 2008; 67: 71719.

2136 www.thelancet.com Vol 377 June 18, 2011


Series

120 Baraliakos X, Listing J, Brandt J, et al. Clinical response to 126 Song IH, Heldmann F, Rudwaleit M, et al. Dierent response to
discontinuation of anti-TNF therapy in patients with ankylosing rituximab in tumor necrosis factor blocker-naive patients with
spondylitis after 3 years of continuous treatment with iniximab. active ankylosing spondylitis and in patients in whom tumor
Arthritis Res Ther 2005; 7: R43944. necrosis factor blockers have failed: a twenty-four-week clinical
121 van der Heijde D, Kavanough A, Gladman DD, et al. Iniximab trial. Arthritis Rheum 2010; 62: 129097.
inhibits progression of radiographic damage in patients with 127 Papp KA, Caro I, Leung HM, et al. Efalizumab for the treatment
active psoriatic arthritis through one year of treatment: results of psoriatic arthritis. J Cutan Med Surg 2007; 11: 5766.
from the induction and maintenance psoriatic arthritis clinical 128 Mease PJ, Gladman DD, Keystone EC. Alefacept in combination
trial 2. Arthritis Rheum 2007; 56: 2698707. with methotrexate for the treatment of psoriatic arthritis: results
122 van der Heijde D, Landew R, Einstein S, et al. Radiographic of a randomized, double-blind, placebo-controlled study.
progression of ankylosing spondylitis after up to two years of Arthritis Rheum 2006; 54: 163845.
treatment with etanercept. Arthritis Rheum 2008; 58: 132431. 129 Mease P, Genevose M, Ritchlin C, et al. Abatacept in psoriatic
123 van der Heijde D, Landew R, Baraliakos X, et al. Radiographic arthritis: results of a phase II study Arthritis Rheum 2009;
ndings following two years of iniximab therapy in patients 60 (suppl 10): 1260.
with ankylosing spondylitis. Arthritis Rheum 2008; 58: 306370. 130 Haibel H, Rudwaleit M, Listing J, et al. Open label trial of
124 van der Heijde D, Salonen D, Weissman BN, et al. Assessment anakinra in active ankylosing spondylitis over 24 weeks.
of radiographic progression in the spines of patients with Ann Rheum Dis 2005; 64: 29698.
ankylosing spondylitis treated with adalimumab for up to 2 years. 131 Gottlieb A, Menter A, Mendelsohn A, et al. Ustekinumab, a
Arthritis Res Ther 2009; 11: R127. human interleukin 12/23 monoclonal antibody, for psoriatic
125 Nocturne G, Dougados M, Constantin A, et al. Rituximab in the arthritis: randomised, double-blind, placebo-controlled, crossover
spondyloarthropathies: data of eight patients followed up in the trial. Lancet 2009; 373: 63340.
French Autoimmunity and Rituximab (AIR) registry.
Ann Rheum Dis 2010; 69: 47172.

www.thelancet.com Vol 377 June 18, 2011 2137

You might also like