Disease modification in spondyloarthritis should target the improvement of symptoms and preservation of function. Therefore, inhibition of structural damage caused by the disease processes appears essential. In spondyloarthritis, structural damage results mainly in progressive ankylosis of the spine and peripheral joint destruction. Currently available therapies for the treatment of spondyloarthritis appear effective at inhibiting tissue destruction but, with the exception of celecoxib, do not appear to affect new tissue formation leading to ankylosis. In this article, we discuss clinical and pathophysiological concepts of disease modification in spondyloarthritis, challenges in its evaluation, recent clinical data and new concepts that may help explain structural damage as well as the onset and progression of disease.
The concept of disease modification in arthritis
Major inflammatory joint diseases such as the spondyloarthritides (SpAs) and rheumatoid arthritis (RA) are chronic disorders. The immediate impact of inflammation (pain and loss of function) is complicated by the dimension of time. Any intervention in patients suffering from arthritis, therefore, aims to limit the short- and long-term consequences of disease. The concept of ‘disease modification’ refers to an improvement of symptoms in combination with changes in the course of the illness . The former is perceived as the disease process and at the centre of the patient’s immediate expectations and short-term interventions; the latter is reflected in the disease outcome and the main focus of the rheumatologist who defines a therapeutic strategy.
Drugs may be able to influence the disease process resulting in short-term control of the symptoms caused by inflammation (symptom-modifying drugs). They may also have an impact on the disease course and thereby the long-term outcome (disease-controlling or modifying drugs) . Different instruments have been developed to assess disease activity (evaluating symptom control) and structural damage in SpA and RA but questionnaires and imaging methods do not seem to correlate very well. The combination of symptom control and prevention of permanent damage and disability results in an effect on function for which additional measurements exist. At any point in time during the disease, function is determined by a number of factors including inflammatory symptoms, disability caused by structural damage to the involved tissues and other consequences of the disease such as loss of muscle strength, general fitness, psychological impact and coping style ( Fig. 1 ) . The influence of each of these variables varies over time in individual patients. Disease modification results in improved function and is a complex issue for which many interventions exist.
However, this inclusive model is often narrowed down to the prevention of structural damage as a defining factor to categorise and evaluate the impact of drugs. Most investigations on drug interventions and their effects on long-term outcome of disease have been performed in patients with RA. Several drugs prevent structural damage, which is evaluated by radiographic methods and characterised by joint space narrowing and bone erosions. These drugs include chemical immune modulators such as methotrexate , and leflunomide as well as biological cytokine or cell targeting approaches such as tumour necrosis factor (TNF) blocking agents , abatacept, a modulator of T-cell costimulatory signals and tocilizumab, an interleukin (IL)-6 antagonist . The acronym, DMARD (disease-modifying anti-rheumatic drug), is used to distinguish the chemicals within the first group as opposed to the biologics in the second group. However, the concept of DMARDs applies specifically to RA. It is less clear that structural modification is the sine qua non for disease modification in SpA and the focus on structural damage is not undisputed . Therefore, in this article, we look more specifically at clinical and pathophysiological concepts of disease modification in SpA in a broad context.
Current clinical concepts of disease modification in spondyloarthritis
The SpAs are a group of distinct diagnostic entities that share clinical, genetic and pathological characteristics. The disease cluster includes ankylosing spondylitis (AS), psoriatic arthritis (PsA), inflammatory bowel disease associated arthritis (IBD-SpA), reactive arthritis (ReA), juvenile and undifferentiated SpA. Axial disease with involvement of sacroiliac joints and spine, oligoarthritis frequently confined to the lower limbs, extra-articular enthesitis typically affecting the insertion of the Achilles’ tendon and the fascia plantaris, and extra-articular manifestations such as psoriasis and IBD help to distinguish SpA from RA. These common traits are easily recognised and provide support for a unified disease concept. However, AS and PsA, the most common forms of SpA, can also be very different with the former, mainly a disease of the spine, and the latter, characterised by peripheral arthritis, is often polyarticular and symmetrical with an RA-like appearance. This heterogeneity has stimulated a debate on the position of PsA in the disease concept with pathophysiological studies mainly supporting its inclusion .
Disease modification from a clinical point of view should focus on factors that affect function including inflammatory symptoms and structural damage. The effect and type of structural damage appear very different between patients in whom the axial disease is predominant as compared with those who suffer mainly from peripheral arthritis. In this article, we use AS as a paradigm for axial disease and PsA for peripheral arthritis. However, severe spinal disease can occur in PsA patients and peripheral joint involvement can contribute to morbidity in patients with AS.
AS patients are confronted with stiffness and their axial mobility is limited. Although this is partly reversible, as inflammation also contributes to these manifestations, progressive loss of mobility is associated with new cartilage and bone formation leading to spine and sacroiliac joint ankylosis. Progressive ankylosis of the spine is a major determinant of disability in AS . New bone formation in the peripheral joints or in extra-articular entheses also contributes significantly to disability. In patients with PsA, peripheral arthritis erosive disease is an important determinant of joint function. The impact of structural damage in this subset of patients closely resembles the paradigm of RA . In some forms of PsA, joint destruction can be extreme and grotesque leading to arthritis mutilans with a severe impact on function .
Disease modification in SpA should target rapid improvement of inflammatory symptoms as well as the short- and long-term consequences of both types of structural damage: new cartilage and bone formation and joint destruction.
Current clinical concepts of disease modification in spondyloarthritis
The SpAs are a group of distinct diagnostic entities that share clinical, genetic and pathological characteristics. The disease cluster includes ankylosing spondylitis (AS), psoriatic arthritis (PsA), inflammatory bowel disease associated arthritis (IBD-SpA), reactive arthritis (ReA), juvenile and undifferentiated SpA. Axial disease with involvement of sacroiliac joints and spine, oligoarthritis frequently confined to the lower limbs, extra-articular enthesitis typically affecting the insertion of the Achilles’ tendon and the fascia plantaris, and extra-articular manifestations such as psoriasis and IBD help to distinguish SpA from RA. These common traits are easily recognised and provide support for a unified disease concept. However, AS and PsA, the most common forms of SpA, can also be very different with the former, mainly a disease of the spine, and the latter, characterised by peripheral arthritis, is often polyarticular and symmetrical with an RA-like appearance. This heterogeneity has stimulated a debate on the position of PsA in the disease concept with pathophysiological studies mainly supporting its inclusion .
Disease modification from a clinical point of view should focus on factors that affect function including inflammatory symptoms and structural damage. The effect and type of structural damage appear very different between patients in whom the axial disease is predominant as compared with those who suffer mainly from peripheral arthritis. In this article, we use AS as a paradigm for axial disease and PsA for peripheral arthritis. However, severe spinal disease can occur in PsA patients and peripheral joint involvement can contribute to morbidity in patients with AS.
AS patients are confronted with stiffness and their axial mobility is limited. Although this is partly reversible, as inflammation also contributes to these manifestations, progressive loss of mobility is associated with new cartilage and bone formation leading to spine and sacroiliac joint ankylosis. Progressive ankylosis of the spine is a major determinant of disability in AS . New bone formation in the peripheral joints or in extra-articular entheses also contributes significantly to disability. In patients with PsA, peripheral arthritis erosive disease is an important determinant of joint function. The impact of structural damage in this subset of patients closely resembles the paradigm of RA . In some forms of PsA, joint destruction can be extreme and grotesque leading to arthritis mutilans with a severe impact on function .
Disease modification in SpA should target rapid improvement of inflammatory symptoms as well as the short- and long-term consequences of both types of structural damage: new cartilage and bone formation and joint destruction.
Current pathophysiological concepts of disease modification in spondyloarthritis
Understanding the pathophysiology of SpA requires insights into at least three different phenomena. Inflammation, tissue destruction and tissue formation leading to ankylosis, all determine the disease outcome over time.
Inflammation
Considerable progress has been made in understanding cellular and molecular aspects of inflammation in SpA, but many questions remain. Different tissues play critical roles in SpA: the enthesis, an anatomical zone in which tendons and ligaments insert into the bone, the synovium lining the joint cavity and the bone marrow . Entheseal involvement can be joint-associated or extra-articular and has been proposed as a unifying factor to explain the anatomic localisation, the histomorphological appearance and the initiation of the disease process ( and below). SpA-specific features of synovitis have been demonstrated including infiltration by CD163-positive macrophages . These macrophages are regulatory cells more than a source of the major pro-inflammatory cytokines such as TNF and IL-1 . Some evidence suggests that these cytokines are lower in SpA than in RA synovitis despite a similar degree of overall inflammation . The involvement of the bone marrow has been demonstrated by magnetic resonance imaging (MRI) . Enthesis, synovium and bone marrow are closely connected and communicating tissues . The fibrocartilaginous enthesis is relatively resistant to vascular and cellular invasion, and pro-inflammatory signals from the enthesis are likely to attract cells to immune-privileged sites that are in close proximity and direct contact with the source of the initial trigger, in particular, the synovium and the underlying bone marrow.
Different triggers for SpA and links with infection have been suggested. The genetic association with HLA-B27 and with a number of other immune genes suggests that abnormal immune responses towards conventional antigens may trigger disease. The best evidence for this paradigm is found in patients suffering from ReA, where arthritis typically occurs in association with a urinary tract or gastrointestinal infection . However, this example also demonstrates that there is more than one trigger for arthritis as different pathogens are associated with ReA. The association between IBD and SpA can be explained by intestinal priming of immune cells that recirculate to the joint but also by activation of innate immunity . Micro-damage in the enthesis, which is under continuous mechanical strain, could trigger innate immune mechanisms, for example, by interaction of fibrocartilage neoantigens with Toll-like receptors . This may also explain the specific involvement of the enthesis in SpA .
These or other factors amplify and sustain the inflammatory response. SpA is associated more strongly with activation of innate immunity and little evidence is found for adaptive immune responses such as autoantibodies like in RA . From this perspective, chronicity of disease could be sustained by repeated triggering rather than by the development of an autonomous self-perpetuating auto-immune process. Alternatively, failure to contain or down-regulate a local tissue response associated with inflammation towards antigens or micro-damage could become a self-amplifying and sustaining process without the development of acquired immune responses. Thus, the inflammatory reaction in AS and PsA may have a different course than that seen in RA. In SpA, bouts of inflammation can alternate with short or long quiescent periods .
New bone formation – ankylosis
Understanding the mechanisms of ankylosis is a challenge as patient tissues, in particular, spinal specimens are difficult to obtain. Most data have been obtained in animal models . New bone formation in disease often mimicks embryonic development and fracture healing, even at the cellular and molecular level. An important difference between developmental and SpA bone formation is the heterotopic growth in the latter . Fracture healing results in a skeletal structure that is largely akin to the original one, whereas bone formation in SpA breaks out of the normal boundaries and results in outgrowths that change the original structure. Nevertheless, the molecular mechanisms involved in bone development and fracture healing also play a role in pathologic new bone formation in SpA, but the triggers and regulatory mechanisms may be very different.
Ankylosis appears to originate from the enthesis. Osteophytes in osteoarthritis originate from the border of the articular cartilage and the bone . The latter play a stabilising role after damage to the articular cartilage or ligament structures of the joint has occurred. Most joint tissues contain cells that have the potential to differentiate into specific cell types and can be considered as progenitor cells, including the bone marrow , the synovium , the periosteum , the synovial fluid and the articular cartilage . No such cells have been specifically isolated from the enthesis, but its close anatomic relationship with the periosteum and the synovium suggest that cells capable of differentiating into cartilage and bone are present in the disease environment.
Two molecular signalling pathways have so far received attention. We identified bone morphogenetic protein (BMP) signalling as the critical mediator of entheseal endochondral bone formation , whereas Schett et al. stressed the role of Wnts and their antagonists in ankylosis . BMPs trigger a cascade of endochondral bone formation in progenitor cells, stimulate bone formation by osteoblasts and have an anabolic effect on chondrocytes (for a recent review on their potential roles in arthritis, see also ). We demonstrated that inhibition of BMP signalling by overexpression of the endogenous antagonist noggin is an effective preventive and therapeutic strategy in a mouse model of entheseal ankylosis . In this model, spontaneous peripheral oligo-arthritis occurs in male DBA/1 mice and entheseal progenitor cells proliferate, accumulate and differentiate into chondroblasts, and pre- and hypertrophic chondrocytes . In the final stages, the cartilage matrix is invaded by vessels and replaced by bone. BMP signalling is activated in progenitor cells in the initial disease stages, thereby identifying BMPs as an early target. A similar activation pattern of BMP signalling has been identified in SpA patients with extra-articular enthesitis . A recent study, available only in abstract form, suggests that noggin can also inhibit spine ankylosis in a spondylitis model . Some data on levels of BMPs in AS patients are available and suggest that serum levels of BMP2 and BMP7 are elevated with BMP2 levels correlated to the Bath Ankylosing Spondilytis Disease Activity Index (BASDAI) disease activity index and BMP7 levels correlated to the Bath Ankylosing Spondylitis Radiology Index (BASRI) radiological severity index .
Diarra et al., demonstrated that inhibition of Dickkopf-1 (DKK1), a secreted Wnt-receptor antagonist, leads to new bone formation in the peripheral joints of human TNF transgenic mice who normally develop destructive arthritis . The anabolic features of the joint disease were characterised by activation of Wnt signals. These effects are not limited to the peripheral joints and also play a role in the sacroiliac joints . DKK1 levels are low in AS patients compared with RA and controls in studies using a functional assay in which DKK1 binds to Wnt-coreceptor low-density lipoprotein receptor-related protein 5 (LRP5) . Conversely, recent direct analysis suggests that DKK1 levels are even higher in AS patients as receptor binding is reduced or impaired . Sclerostin is another Wnt antagonist of which the expression is confined to osteocytes. A recent report demonstrated that osteocytes from AS patients have lower amounts of sclerostin as determined by immunohistochemistry and that low serum levels of sclerostin are associated with radiographic progression in AS .
Joint destruction
The molecular and cellular mechanisms underlying bone erosions have been well defined in the recent years . The main cells involved are osteoclasts, multinucleated giant cells that differentiate from monocyte precursor cells. Osteoclasts play an important role in the striking bone destruction that is sometimes seen in patients with PsA . However, synovial tissue analysis of osteoclast differentiation and signalling pathways has not shown specific differences between RA, AS and PsA .
Osteoclasts have also been identified in the spine and hips of patients with AS . Proof of principle has been provided that osteoclasts can be specifically targeted in arthritis ; however, as osteoclast maturation and differentiation appear directly linked to inflammation in these diseases, the specific need for such therapies remains to be defined from the patients’ perspective. Nevertheless, such an approach appears useful for patients with arthritis mutilans.
Bone erosion and cartilage loss are associated with the transformation of synovial fibroblasts into so-called pannus cells. The synovial fibroblast population has remarkable features such as in vitro anchorage-independent growth and multipotentiality . Their transformation, including the occurrence of somatic mutations and different epigenetic mechanisms, has been demonstrated mainly in tissues from patients with RA . The data with regard to SpA are much more limited and current evidence does not support or exclude similar mechanisms in AS or PsA. Despite similar degrees of inflammation, SpA patients show less destructive disease with the aforementioned exception of some forms of PsA. One explanation may be the lower amount of cytokines such as IL-1 and TNF in the SpA synovium . However, synovial matrix metalloproteinase (MMP) expression is not different among RA, AS and PsA . An alternative hypothesis has, therefore, been proposed. As patients with SpA show increased activation of chondro- and osteogenesis stimulating pathways, activation of such pathways in the articular cartilage could contribute to resistance towards direct damage with some evidence supporting a higher anabolic status of the SpA cartilage as compared to that in RA .
Linking inflammation, ankylosis and destruction
Interactions between inflammation, ankylosis and destruction are crucial for SpA outcome ( Fig. 2 ). Molecular links between pro-inflammatory cytokines such as TNF and destruction by tissue-destructive enzymes and osteoclasts have been well established. Links between inflammation and new cartilage and bone formation are more difficult to understand. Inflammation has an inhibitory effect on cartilage and bone differentiation . Cross-talk between BMPs and Wnts is complex and both pathways are directly influenced by TNF . TNF induces BMP2 expression in different mesenchymal cell types including periosteal and synovial cells . TNF also induces DKK1 . Wnts and BMPs are secreted molecules with specific affinities for extracellular matrix components. Therefore, many downstream events could be determined by the local balances of specific ligands, antagonists and receptors.
The structural outcome of SpA is not merely a consequence, but an integral part of the disease. Ankylosis is often described as a repair process but this has been challenged. Tissue responses towards stress or damage can take different forms . The ideal situation is restoration in which tissue integrity and homeostasis have fully recovered. Repair typically results in provisional or surrogate tissue that, more or less, compensates for the damage that has occurred. The repair tissue does not lead to the full restoration of homeostasis and may fail in the long-term. Remodelling indicates structural changes including new tissue formation. Some remodelling efforts may be beneficial in the short and long-term such as stabilising osteophytes in osteoarthritis, but in diseases such as AS extensive remodelling contributes to pathology, symptoms and loss of function.
Challenges in the evaluation of disease-modifying therapies in spondyloarthritis
Clinical research communities such as Assessment of Spondyloarthritis international Society (ASAS) and Group for Research and Assessment of Psoriasis and Psoriatic Arthritis (GRAPPA) have identified core data sets to determine the effect of therapeutic interventions on symptoms and disease outcome. Structural damage can be further evaluated using radiographic methods . The use of advanced imaging methods such as MRI has recently been reviewed extensively and goes beyond the scope of this article .
Radiographic progression of erosive joint destruction in PsA is commonly assessed by a PsA-modified Van der Heijde–Sharp score that includes both erosions and joint space narrowing . However, this evaluation does not take into account specific features of entheseal new tissue formation and ankylosis. The PsA ratingen score that includes both bone destruction and formation shows only a weak correlation between destruction and new bone formation and could be a useful alternative or complementary method.
Progression of spinal ankylosis is best studied by the modified Stoke Ankylosing Spondilytis Spinal Score (SASSS) method . However, ankylosis is a fairly slow process requiring a follow-up of at least 2 years before change can be reliably detected and not all parts of the spine are included in the evaluation. Relatively large number of patients are required in trials to ensure sufficient power to detect differences.
As outlined above, disease modification and its evaluation cannot be restricted to radiographic progression in AS. Function over time is determined by different variables of which structural damage is only one. We, therefore, support the use of a broad view on improvement in function and prognosis in the evaluation of specific therapies.
Related posts:
What are the spondyloarthropathies?
Update on the genetics of spondyloarthritis – ankylosing spondylitis and psoriatic arthritis
Classification criteria for psoriatic arthritis and ankylosing spondylitis/axial spondyloarthritis
Soluble and Tissue Biomarkers in Ankylosing Spondylitis
Therapeutic considerations in spondyloarthritis patients who fail tumour necrosis factor antagonists
Bacterial agents in spondyloarthritis: a destiny from diversity?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
