Criteria for Ankylosing Spondylitis and Axial Spondyloarthritis

The diagnosis of AS is based on clinical features. The disease is “primary” or “idiopathic” if no associated disorder is present; it is “secondary” if the disease is associated with psoriasis or chronic inflammatory bowel disease. In daily practice, a presumptive clinical diagnosis of AS is usually supported by evidence of sacroiliitis on conventional pelvic radiographs; indeed, many think of AS as symptomatic sacroiliitis. The presence of sacroiliitis does not necessarily indicate the presence of AS, however. Moreover, although radiographic sacroiliitis is frequent in AS, it is by no means an early or obligate manifestation of the disease.¹⁰ Contrary to the New York criteria, radiographic evidence of sacroiliitis is not obligatory to fulfill the Rome criteria (Table 75-4). Both sets were primarily intended for use in epidemiologic studies. Lack of either sensitivity or specificity led to a modification of the New York criteria for AS¹¹ (see Table 75-4). Two criteria—limitation of lumbar spine motion and limitation of chest expansion—appear to reflect disease duration; they are usually not present in early disease.¹² It should be stressed that classification criteria are usually not useful for early diagnosis owing to a lack of sensitivity. In particular, in the early phase of AS, conventional SI radiographs may be normal. MRI is useful to diagnose the disease with predominantly axial manifestations before the presence of radiographic sacroiliitis.¹³ To encompass both preradiographic AS and classic AS (according to the modified New York criteria), the Assessment of SpondyloArthritis international Society (ASAS) has proposed classification criteria for axial spondyloarthritis for patients with back pain of at least 3 months’ duration and age at onset of complaints before 45 years (Table 75-5). The term axial spondyloarthritis has been introduced as an umbrella for the entire spectrum of spondyloarthritis patients with predominant axial involvement, irrespective of the presence of structural damage on radiographs. Thus axial spondyloarthritis includes nonradiographic axial spondyloarthritis and classical AS (fulfilling the modified New York criteria). According to the ASAS classification criteria for axial spondyloarthritis,⁷ a patient with chronic back pain and age at onset before age 45 can be classified as having axial spondyloarthritis if sacroiliitis on imaging (radiographs or MRI) is present plus at least one further spondyloarthritic feature, or, in the absence of sacroiliitis on imaging, if HLA-B27 plus at least two further spondyloarthritis features are present (see Table 75-5). The sensitivity of the entire set of ASAS criteria for axial spondyloarthritis was 83%, and the specificity was 84%.⁷ In the ASAS study on these new criteria, 30% of all patients diagnosed as having axial spondyloarthritis had definite radiographic sacroiliitis and fulfilled the modified New York criteria for AS. Therefore two-thirds were classified as nonradiographic axial spondyloarthritis.⁷

Table 75-4 Criteria for Ankylosing Spondylitis

Table 75-5 ASAS Classification Criteria for Axial Spondyloarthritis (SpA) (in Patients with Back Pain ≥ 3 Months and Age at Onset < 45 Years)

^* Elevated CRP is considered a SpA feature in the context of chronic back pain.

Prevalence

The prevalence of AS closely parallels the frequency of HLA-B27. This holds true for those B27 subtypes that are associated with the disease, but it is not true for populations in which the HLA-B27*06 subtype that lacks a strong association with AS occurs frequently such as the Indonesian population.^14–16

Among whites, the estimated prevalence rate of AS as defined by the modified New York criteria ranges from 68 per 100,000 population older than 20 years in the Netherlands to 197 per 100,000 in the United States.^8,17 The prevalence of clinical AS in France is 150 per 100,000 adults, whereas in Norway it is 210 per 100,000 adults.^18,19 The prevalence of the disease in Finland is similar, with a figure of 150 per 100,000 people.²⁰

Higher prevalence rates have been reported in central Europe. An epidemiologic study from Berlin reported a prevalence figure of 0.86%.²¹ In the general population, AS is likely to develop in about 1% to 2% of HLA-B27⁺ adults who have a disease-associated B27 subtype, although there may be regional or geographic differences. For example, in northern Norway, AS may develop in 6.7% of HLA-B27⁺ people.^8,22

The disease is much more common among HLA-B27⁺ first-degree relatives of HLA-B27⁺ AS patients; roughly 10% to 30% of them have signs or symptoms of AS.⁸ In fact, a positive family history of AS is a strong risk factor for the disease.

Incidence

There is no adequate evidence that the incidence of AS has changed in the past few decades. Clinical features, age of onset, and survival time have remained stable.²³ One study revealed an overall age and gender-adjusted incidence of 7.3 per 100,000 person-years. This U.S. figure compares quite well with the Finnish study, which revealed a stable incidence of 8.7 (95% confidence interval [CI], 6.4 to 11.0) per 100,000 people aged 16 or older.¹⁸

Racial Distribution

AS occurs in all parts of the world, but there are race-related differences in prevalence. This might reflect differences in the distribution of HLA-B27 among races. Approximately 90% of white patients with AS possess HLA-B27, whereas AS and HLA-B27 are nearly absent (prevalence of B27 < 1%) in African blacks and Japanese. In African-Americans, owing to racial admixture with whites, 2% possess B27, but only about 50% of black patients with AS possess B27. Correspondingly, African-Americans are affected far less frequently than American whites.

Burden of Disease

AS is associated with a considerable burden to the patient and society. Apart from the axial and articular manifestations, extra-articular manifestations such as enthesitis and acute anterior uveitis and comorbidities such as inflammatory bowel disease and psoriasis contribute to the burden of disease. In addition, a large proportion of patients has spinal osteoporosis, leading to vertebral fractures and thoracic kyphosis. All these features result in a decreased quality of life. Disease status scores for physical functioning and disease activity correlate clearly with psychologic scores for anxiety and depression.²⁴ The impact of AS can also be seen in various aspects of employment, ranging from requiring assistance at work to increased sick leave and withdrawal from the workforce.^25,26 Apart from the impact on labor force participation, AS patients have an important impact on health care and non–health care resource utilization, resulting in mean total costs (direct and productivity) of about $6700 to $9500 per year per patient when applying the human capital approach to calculate productivity costs.^27–29

The burden of illness increases with duration of disease. Because the burden due to AS reduces quality of life, and because all types of costs associated with AS result from loss of function and disease activity, early diagnosis and treatment are necessary to prevent or reduce functional decline and improve patient outcome.³⁰

HLA-B27

The dominant role of genetic factors is highlighted by data demonstrating disease concordance in 75% of monozygotic twins compared with 13% of nonidentical twins,³¹ familial aggregation,³² and population data demonstrating associations with HLA-B27.³³ About 90% of white AS patients are HLA-B27⁺. HLA-B27 is the first described and major genetic risk factor for AS but, despite the near-pervasive nature of this association, it has been estimated that B27 contributes only 16% of the total genetic risk. In humans, there appears to be a hierarchy of association of the more than 45 as yet known subtypes of HLA-B27 with AS, ranging from strong association with HLA-B*2705 to weak association with HLA-B*2709. The direct pathogenic role of HLA-B27 is evidenced by the spontaneous spondyloarthritis-like disease in rats overexpressing HLA-B27.^34–36

The main function of human leukocyte antigen (HLA) class I molecules such as HLA-B27 is to present peptides to CD8⁺ T cells and, accordingly, the arthritogenic peptide hypothesis proposes that CD8⁺ T cells are activated by bacterial antigens, for example in the gut, and after recirculation are reactivated in the joint by cartilage or other autoantigens. A number of studies have provided some partial evidence for this concept.^37,38 This hypothesis has been thoroughly questioned by the fact that HLA-B27 transgenic rats develop severe spondyloarthritis-like disease even in the absence of CD8⁺ T cells.^39,40 Two alternative hypotheses have been proposed more recently, suggesting that HLA-B27 has other specific features independent of antigen presentation that may directly contribute to disease. One theory indicates that HLA-B27 has a special tendency to form heavy chain homodimers that, when present on the cell surface, can trigger direct activation of natural killer (NK) cells through recognition via killer immunoglobulin receptor (KIR)-like receptors.^41,42

Alternatively, HLA-B27 has a special propensity to misfold in the endoplasmic reticulum and thereby to promote an unfolded protein response, which in turn modulates the functional behavior and cytokine production of myeloid cells.^43–45

The latter two hypotheses imply a role for altered innate immune responses rather than autoantigen-driven acquired immunity in the pathogenesis of AS.

Population and family studies have shown that HLA-B60 increases susceptibility to AS.⁴⁶ This applies to both HLA-B27⁺ and HLA-B27⁻ persons.

Non–Human Leukocyte Antigen Genes

Two observations clearly suggest a role for non-HLA genetic risk factors. First, there is an increased risk for disease in B27⁺ first-degree relatives of AS probands (10% to 20%) compared with B27⁺ individuals in the general population (2% to 5%).⁸ Second, disease concordance is 75% in identical twins versus 27% in HLA-B27 concordant dizygotic twins.³¹ Genome-wide association studies have now allowed the identification of several of these factors.^47,48 There is a robust association with single-nucleotide polymorphisms (SNPs) in endoplasmic reticulum aminopeptidase-1 (ERAP-1), an enzyme involved in the trimming of peptides for loading in MHC molecules. There is now emerging evidence for a gene-gene interaction between ERAP-1 and HLA-B27, but it remains unknown how the polymorphisms may affect antigen presentation, HLA-B27 homodimer formation, and misfolding. A second strong association is found with SNPs in IL-23R, a genetic feature shared with Crohn’s disease and psoriasis. Together with the suggested but not yet confirmed association with STAT3⁴⁹ and JAK2 polymorphisms, this suggests the potential involvement of an interleukin (IL)-17 response in AS. Definite association was also found with gene deserts on chromosome 2p15 and 21q22, which warrants further investigation of noncoding ribonucleicacid (RNA) and epigenetic effects in the pathophysiology of this disease. Besides these definite associations, genome-wide association studies have suggested potential associations with tumor necrosis factor (TNF) receptor 1 (TNFR1), the signaling molecule TNFR1-associated death domain protein (TRADD), the TNF superfamily cytokine TNFSF15, IL-1A and the IL-1 receptor 2 (IL-1R2), IL-12B, the vascular morphogenesis protein gene anthrax toxin receptor 2 (ANTXR2), and the innate immune receptor caspase recruitment domain family, member 9 (CARD9).⁵⁰

These associations point toward a potential pathogenic role for altered innate immune responses, TNFR1 signaling, and IL-1.

The importance of TNF to the pathogenesis of AS is highlighted in the phenotype of a transgenic mouse model that overexpresses TNF. These animals develop sacroiliitis characterized by the formation of osteoclasts and granulation tissue.

Autoimmunity versus Autoinflammation

Both the alternative roles for HLA-B27 besides antigen presentation and the genetic risk factors suggest that AS may not be primarily driven by a canonical autoantigen-specific T and/or B cell reactivity. Indeed, AS does not share genetic risk factors such as PTPN22 or cytotoxic T lymphocyte–associated antigen-4 (CTLA-4) with other autoimmune diseases and lacks disease-specific autoantibodies. On the basis of the predilection of the disease for sites rich in cartilage, it has been proposed that cartilage antigens may be the primary target of an autoimmune response in AS. This hypothesis is supported by several animal models of peripheral and axial arthritis on the basis of the induction of autoimmunity to antigens present in cartilage and fibrous tissue such as aggrecan and versican.^51,52 T cell responses toward the G1 domain of aggrecan have also been observed in human AS, but similar responses were seen in other inflammatory joint disorders, implying a nonspecific response to joint damage.⁵³ Finally, biologic therapies targeting B (such as anti-CD20) or T (such as CTLA4-Ig) cell pathways have limited to no efficacy in AS in proof-of-concept trials.^54,55 Taken together, little strong evidence supports a primary T cell– or B cell–driven autoimmune process in AS.

An alternative hypothesis proposes that AS would be driven by abnormal reactivity of innate immune cells such as macrophages, neutrophils, and mast cells. This hypothesis is supported by several lines of evidence. First, genetic associations with CARD9 and the IL-1 pathway strongly suggest involvement of the inflammasome. Second, histologic studies demonstrated increased infiltration with innate immune cells, but not T, B, or dendritic cells in the peripheral joint and gut of patients with AS.^56–58 Third, human AS and its experimental models have a predilection for sites exposed to either microbial or mechanical stress. The latter aspect has led to the concept that AS and other forms of spondyloarthritis are primarily characterized by inflammation of the so-called synovio-entheseal complex.⁵⁹ In this perspective, it is interesting to note that the fibrocartilage protein versican, which could be released during mechanical stress and induces spondylitis in BALB/c mice, is also a ligand for the innate immune receptor TLR2. Formal proof that innate immune alterations may trigger AS remains, however, awaited.

Independently of the exact origin of the inflammation, it is clear that several proinflammatory cytokines are pivotal in the downstream effector mechanisms. TNF plays a crucial role in the disease process as evidenced by the successful introduction of TNF blockade for this disease. Accordingly, transgenic overexpression of TNF in mice leads not only to severe, destructive polyarthritis but also to sacroiliitis.⁶⁰ However, it remains incompletely understood which cells are the main producers of TNF in AS, in which form (soluble vs. transmembrane) and through which receptor (TNFR1 or TNFR2) TNF exerts its pathogenic effects, and which are the main target cells. In line with the genetic data, experimental data suggest a role for TNFR1 signaling in stromal cells.⁶¹ Besides TNF, the genetic data and emerging data form functional studies on the unfolded protein response implicate IL-23 and thus possibly IL-17 as important cytokines in the pathogenesis of AS.⁶²

Clinical trials targeting the IL-23/IL-17 pathway in AS are currently being performed.

Structural Remodeling and Ankylosis

A crucial and largely unexplained aspect of the pathogenesis of AS is the occurrence of structural remodeling and new bone formation, ultimately leading to ankylosis. Human and experimental studies have revealed three important concepts. First, the remodeling phenotype of AS cannot be explained by the absence of joint destruction because erosive disease is evidenced by histology and radiology and because the cellular and molecular machinery for cartilage and bone destruction is present and operative at the sites of disease.^63,64

Second, structural remodeling and osteoproliferation are dependent on endochondral bone formation. This process is governed by several molecular pathways including bone morphogenetic protein (BMP) and Wnt signaling.^65,66

There is emerging evidence that several inhibitors of the Wnt pathway are dysfunctional in human spondyloarthritis and that this correlates with new bone formation.⁶⁷ Third, osteoproliferation is not critically dependent on inflammation because, despite clinical efficacy on signs and symptoms, TNF blockade fails to halt radiologic progression in AS and both processes can be uncoupled in experimental models.^68,69

On the other hand, vertebral sites showing signs of inflammation at baseline have a higher risk for subsequent osteoproliferation on follow-up.⁷⁰ The exact relationship between inflammation and osteoproliferation needs to be further clarified because it may have important consequences for treatment.

Axial Skeleton

Detailed histopathologic studies of axial involvement in AS are limited owing to the inaccessibility of biopsy material of both SI joints and the spine. A controlled study of SI biopsies from AS patients at various stages of disease and controls showed cellular infiltration with lymphocytes, macrophages, and plasma cells in the synovium and subchondral marrow as the earliest features of disease.⁷¹ Later features include the development of pannus extending from both synovium and subchondral bone marrow, with erosion of articular cartilage and its replacement by granulation tissue. Osteoclast formation and erosion of subchondral bone account for the typical widening of the joint spondyloarthritisce seen on plain radiography. Enthesitis is also evident in later stages of disease at the insertion of the posterior capsule. Reparative changes include cartilage metaplasia at sites of active inflammation, followed by its calcification and then replacement by endochondral bone, leading to obliteration of the joint space by ankylosis. Para-articular changes include bone sclerosis and fat replacement of bone marrow. Regarding the spine, pathologic data are limited except for a number of recent studies of apophyseal joints. Immunohistologic analysis shows subchondral lymphocyte infiltrates with CD4⁺ and CD8⁺ T cells, together with hypervascularization and foci of CD68⁺ osteoclastic cells.⁷²

Peripheral Skeleton

Regarding peripheral joint manifestations of AS, histopathologic studies have assessed surgical samples of affected hip joint, synovial biopsies of knee and ankle joint, and enthesitis. Involvement of the hips is characterized by subchondral granulation tissue and osteoclast formation in the femoral heads and acetabulum that is associated with degradation of overlying articular cartilage.⁷³ Synovial studies have revealed that the type of inflammation is strongly different from that observed in RA, with increased vascularity, increased infiltration with innate immune cells, and absence of specific features of T and B cell autoimmunity.^74–77

Interestingly, these studies revealed a similar pattern of inflammation in AS in comparison with other subtypes of spondyloarthritis.⁷⁸

Entheseal involvement most frequently occurs at sites rich in fibrocartilage such as the Achilles tendon. Inflammation and chronic cellular infiltration of soft tissues are relatively sparse but may be extensive within the adjacent subchondral bone, particularly in B27⁺ individuals.⁷⁹A comparative study of the subchondral marrow from knee and hip joint entheses showed that AS patients clearly differ from those with RA and osteoarthritis with respect to the frequency of marrow inflammation, infiltration with CD8⁺ T cells, and presence of hyperosteoclastic erosive lesions.

Skeletal Manifestations

Extraskeletal Manifestations

Constitutional symptoms such as fatigue, weight loss, and low-grade fever occur frequently. Other extraskeletal manifestations are more localized.

Spinal Mobility

To arrive at an early diagnosis, the physician must perform a thorough physical examination. On examination of the spine, there may be some limitation of motion of the lumbar spine as elicited by forward flexion, hyperextension, or lateral flexion. Early loss of the normal lumbar lordosis is often the first sign and is easily assessed on inspection.

The Schober test (or its modifications) is useful to detect any limitation of forward flexion of the lumbar spine, although it is typically normal in early disease. As the patient stands erect, one mark is placed with a pen on the skin overlying the fifth lumbar spinous process (usually at the level of the posterosuperior iliac spine or the “dimple of Venus,” and another mark is placed 10 cm above in the midline. The patient is then asked to bend forward maximally without bending the knees. In healthy people, the distance between the two marks on the skin should increase as the skin stretches. If the distance between both marks does not reach 15 cm, this indicates reduced lumbar spine mobility. Lateral flexion may also be diminished, and spinal rotation may cause pain.

Chest Expansion

Mild to moderate reduction of chest expansion is often detectable in early stages of AS. Normal values are age and sex dependent, and there is considerable overlap between normal values and those obtained from AS patients. Reduction below 5 cm in young persons with an insidious onset of chronic, inflammatory low back pain strongly suggests AS. Chest expansion should be measured on maximal inspiration after forced maximal expiration at the level of the fourth intercostal spondyloarthritisce in males and just below the breasts at the xiphisternal level in females.

Enthesitis

Examination of the ischial tuberosities, greater trochanters, spinous processes, costochondral and manubriosternal junctions, supraspinatus insertion, and iliac crests can determine the presence of enthesitis. Heel pain, especially when getting out of bed, is a characteristic manifestation of Achilles and plantar fasciitis enthesitis.

Sacroiliitis

Direct pressure over the SI joints may elicit pain, as may special testing maneuvers, although the latter lack specificity and sensitivity. These signs may also be negative in early disease or may become negative in late stages as inflammation is replaced by fibrosis or bony ankylosis.

Posture

Over the course of the disease, the patient may lose normal posture. Involvement of the cervical spine is manifested by pain and limitation of neck movement. A forward slope of the neck can be detected by having the patient stand against a wall and try to position his or her occiput against it.

After many years of progression in patients with severe disease, the entire spine may become increasingly stiff, with loss of normal posture from gradual loss of lumbar lordosis and the development of thoracic kyphosis.^98,99 The abdomen becomes protuberant; breathing is primarily by diaphragmatic action. These typical deformities usually evolve after disease duration of 10 years or more.

Conventional Radiography

The typical radiographic changes of AS are seen primarily in the axial skeleton, especially in the SI, discovertebral, apophyseal, costovertebral, and costotransverse joints. They evolve over many years, with the earliest, most consistent, and most characteristic findings seen in the SI joints. However, otherwise typical AS has been described in the absence of radiographic evidence of sacroiliitis.⁸ The radiographic findings of sacroiliitis are usually symmetric and consist of blurring of the subchondral bone plate, followed by erosions and sclerosis of the adjacent bone. The changes in the synovial portion of the joint (i.e., the lower two-thirds of the joint) result from inflammatory synovitis and osteitis of the adjacent subchondral bone.¹⁰⁵ The cartilage covering the iliac side of the joint is much thinner than that covering the sacral side. Therefore the erosions and subchondral sclerosis are typically seen first and tend to be more prominent on the iliac side.

In the upper one-third of the SI joint, where strong intra-articular ligaments hold the bones together, the inflammatory process may lead to similar radiographic abnormalities. Progression of the subchondral bone erosions can lead to pseudowidening of the SI joint space. Over time, gradual fibrosis, calcification, interosseous bridging, and ossification occur. Erosions become less obvious, but the subchondral sclerosis persists, becoming the most prominent radiographic feature.

Ultimately, usually after several years, there may be complete bony ankylosis of the SI joints, with resolution of bony sclerosis. It is practical to grade radiographic sacroiliitis according to the New York criteria (Table 75-8).

Table 75-8 Grading of Sacroiliitis: New York Criteria

Grade 0, normal Grade 1, suspicious Grade 2, minimal sacroiliitis Grade 3, moderate sacroiliitis Grade 4, ankylosis < div class='tao-gold-member'> Only gold members can continue reading. Log In or Register to continue Share this: Click to share on Twitter (Opens in new window) Click to share on Facebook (Opens in new window) Related Related posts: Assessment of Health Outcomes Etiology and Pathogenesis of Juvenile Idiopathic Arthritis Immunosuppressive Drugs Hemophilic Arthropathy Antinuclear Antibodies Enteropathic Arthritis Stay updated, free articles. Join our Telegram channel Join Tags: SPEC- Kelleys Textbook of Rheumatology 6 Month Subscription Jul 3, 2016 | Posted by admin in RHEUMATOLOGY | Comments Off on Ankylosing Spondylitis Full access? Get Clinical Tree Get Clinical Tree app for offline access Get Clinical Tree app for offline access