Published: August 2010
Ankylosing spondylitis is classified along with the seronegative spondyloarthritic diseases. Ankylosing spondylitis is a chronic, systemic, inflammatory disease that affects primarily the sacroiliac joints and spine. Certain peripheral joints and tendons can also be affected, and extra-articular manifestations may be present. The disease typically affects young adults, and there are strong genetic features.
The spondyloarthritides, as a subset, have several distinguishing features that are shared among them. The diseases are seronegative by definition. There are common genetic factors including the human leukocyte antigen (HLA) B27 gene. Axial involvement, including sacroiliitis, can be seen in all and is essential for the diagnosis of ankylosing spondylitis. Enthesitis, or inflammatory changes of the tendon and ligament attachment to bone, is the primary pathologic phenomenon. The spondyloarthritides share many extra-articular features including uveitis, dermatitis, and colitis.
The etiology of ankylosing spondylitis remains unclear. There is solid evidence pointing to a genetic component and risk. The HLA B27 gene is commonly present, and there is a strong familial association. Infective mechanisms have been proposed but seem less evident than in the reactive arthritides.
The prevalence of ankylosing spondylitis is 0.1% to 1.4%, depending on the population studied.1 The disease is more common in Caucasians than in other races. Typically, but not universally, the prevalence reflects the associated prevalence of the HLA B27 gene in that population.2 Ankylosing spondylitis is discovered to be the etiology in 4% to 5 % of patients with chronic low back pain.3
In North American Caucasians, the HLA B27 gene is found in 7% of the normal population and in more than 90% of patients with ankylosing spondylitis.2 A person who is HLA B27 positive has a 5% to 6% chance of developing the disease.4 There is a fivefold to 16-fold increase in having ankylosing spondylitis if a first-degree relative has the disease.5 The HLA B27 factor makes up about a third of the overall genetic risk.
In the past, ankylosing spondylitis was thought to be an overwhelmingly male disease, but the actual male-to-female ratio is closer to 3:1. The median age of onset is 23 years. Ankylosing spondylitis rarely has its initial manifestation after age 40 years; however, there can be a delay in diagnosis.
The pathogenesis of ankylosing spondylitis remains unclear to date. It is assumed to be immune mediated. There is an obvious cytokine role, because patients show improvement with anti–tumor necrosis factor α (anti–TNF-α) agents. As discussed earlier, there is also a genetic component, and the HLA B27 gene is found in more than 90% of patients with ankylosing spondylitis, although the incidence varies some depending on the population studied. Familial aggregation is seen even without the HLA B27 gene. An infectious trigger for the disease remains an intriguing hypothesis. Studies with animal models show that rodents that have HLA B27 and are raised in a germ-free environment do not develop disease; however, in a regular environment, they develop a disease similar to ankylosing spondylitis.6 To date however, no infectious trigger has been unequivocally established, as with reactive arthritis.
Enthesitis, in both the axial and the appendicular skeleton, is the primary pathologic feature of the spondyloarthritides.7 The enthesis is where the tendon or ligament attaches to bone. The progression is typically edema of bone followed by erosion, then ossification, then finally ankylosis. The sacroiliac joints exhibit inflammation followed by ankylosis. In the spine, one can see inflammation at the junction of the annulus fibrosis of the disk cartilage with the margin of vertebral bone. Ultimately, this leads to formation of syndesmophytes, with bridging that leads to the radiographic appearance of a bamboo spine (see the discussion of imaging, later). The spinal facet joint can exhibit synovitis followed by ankylosis. In the peripheral skeleton, one sees synovitis and enthesitis.
Spinal and sacroiliac symptoms are typically early and the most prominent. Low back pain is the first symptom in more than 75% of patients. In some patients, the symptoms are more in the buttock. Over time, the patient experiences limited spinal mobility. One of the keys to the diagnosis of ankylosing spondylitis is identifying features of inflammatory low back pain.3, 8 This includes onset before the age of 40 years, insidious onset, chronic (>3 months) pain, morning stiffness for longer than 30 minutes, improvement with exercise, awakening with pain in the second half of the night, and alternating buttock pain (Box 1).
|Box 1 Diagnosing Inflammatory Low Back Pain|
|Onset before the age of 40 years|
|Greater than 3 months in duration (chronic)|
|Morning stiffness >30 minutes|
|Improvement with exercise but not with rest|
|Awakening with pain in the second half of the night|
|Alternating buttock pain|
Enthesopathic symptoms can be spinal or peripheral. Axial enthesitis includes costovertebral, manubriosternal, sternal clavicular, and costochondral. Some patients experience decreased chest expansion. Examples of extraspinal enthesisits include dactylitis (sausage digit), Achilles tendinitis, and plantar fasciitis.
Inflammatory arthritis likewise can be axial or appendicular. The hips and shoulders are commonly involved. In the peripheral skeleton, patients typically experience lower extremity and asymmetric patterns of involvement.
Finally, extra-articular manifestations may be present (Box 2). Eye involvement is the most common, with anterior uveitis (or iritis) seen in 25% to 40% of patients who have ankylosing spondylitis.9 This precedes the spinal symptoms in some cases.
|Box 2 Extra-articular Manifestations of Ankylosing Spondylitis|
Cardiovascular involvement includes valvular heart disease, aortitis (almost always ascending thoracic), and conduction disturbances. Pulmonary manifestations include decreased chest expansion with extrathoracic restrictive lung disease and fibrotic upper lobe disease.10 The bowel can be involved even in cases without frank inflammatory bowel disease. Mild colitis can be seen in upwards of 50% to 60% of cases, and it can be asymptomatic.11 A small percentage of patients progress to inflammatory bowel disease. Conversely, one can see an ankylosing spondylitis type of picture with frank inflammatory bowel disease. Neurologically, one can see cord compression and, rarely, cauda equina syndrome.12 Although there is syndesmophyte formation, the bone itself becomes osteoporotic. One should be vigilant for compression fractures and subluxation with even minimal trauma. Atlantoaxial instability and subluxation, similar to that seen with rheumatoid arthritis, has been demonstrated with ankylosing spondylitis.13
Ankylosing spondylitis is a systemic disease, and fatigue and malaise are not uncommon. Fever and weight loss are rarely seen, however.
On the physical examination of the spine, in general, there is decreased range of motion and tenderness. When examining the neck, one sees decreased range of motion and increased occiput-to-wall distance. In the thoracic spine, one sees decreased chest expansion, kyphosis, and costochondral tenderness. In the lumbar spine, a decreased Schober’s index is seen (Fig. 1).
Several maneuvers elicit pain in the sacroiliac joints: the FABER (flexion abduction external rotation), or Patrick’s, test; Gaenslen’s test; anteroposterior and lateral pelvic compression; and tenderness of the sacroiliac joint with direct pressure (Fig. 2).
Enthesitis can be demonstrated on the axial and peripheral examination. Common locations are the costochondral and Achilles regions.
Peripheral arthritis can be observed. The hip joints are most common, but one can also see an asymmetric arthritis, which is more common in the lower extremities.
The diagnosis of ankylosing spondylitis is made on the basis of a combination of suggestive symptoms, physical examination findings, and imaging. The most common classification system is the modified New York criteria14 (Box 3); however, these have never been validated for clinical use. A common dilemma in making the diagnosis arises in early cases with normal plain radiographs (see the discussion of imaging, later). From a more practical standpoint, a useful strategy for diagnosisng ankylosing spondylitis is identifying features of inflammatory low back pain as noted previously (see Box 1).3, 8
|Box 3 Modified New York Criteria (1984) for Diagnosing Ankylosing Spondylitis|
|Low back pain and stiffness for ?3 months that improves with exercise but not with rest|
|Limitation of lumbar spine mobility in both the sagittal and frontal planes|
|Limitation in chest expansion as compared with normal range for age and sex|
|Unilateral sacroiliitis of grade 3-4 or|
|Bilateral sacroiliitis of grade ≥2|
|Definite ankylosing spondylitis if the radiological criterion is associated with at least one clinical criterion|
|Probable ankylosing spondylitis if|
From van der Linden S, Valkenburg H, Cats A: Evaluation of diagnostic criteria for ankylosing spondylitis. A proposal for modification of the New York criteria. Arthritis Rheum 1984;27:361-368.
Although there are several formal tools for assessing the activity and severity of disease, these can be cumbersome in clinical practice.15 For functional, in-office clinical purposes, a minimal set of indicators includes global pain, axial pain, degree and duration of morning stiffness, limitation of activities, and the physician’s global assessment.
The HLA B27 gene is not a diagnostic test. The sensitivity and specificity vary depending on race or ethnicity.2 For example, in the Caucasian population, about 90% of ankylosing spondylitis patients and 6% to 10% of the entire population have the HLA B27 gene. In certain cases, testing for the HLA B27 gene may be useful in determining the need for further evaluation. Acute-phase reactants, including erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP), may be variably elevated. This may be more highly associated with peripheral arthritic activity. The rheumatoid factor and antinuclear antibody are negative because the spondyloarthritides are seronegative by definition.
Plain films of the sacroiliac joints are the standard initial test for diagnosing ankylosing spondylitis (Figs. 3 and 4). Sacroiliitis is the cardinal feature and is essential for diagnosis based on the 1984 modified New York criteria.14 The findings are classically symmetric and radiographicaly graded as 1 through 4: suspicious, minimal, moderate, ankylosis. Plain films of the lumbar, thoracic, and cervical spine can demonstrate syndesmophytes, squaring of vertebral bodies, ossification of ligament, and disease of the facet joints (Figs. 5 and 6). Radiographic damage may be assessed using the modified Stokes Ankylosing Spondylitis Spine Score [SASSS].16 Sacroiliitis identified by plain x-rays of the sacroiliac joints is still the gold standard for diagnosis and is a relatively fast and cheap test. However, the findings of sacroiliitis on plain x-ray can lag for years behind symptoms and are subject to varied interpretation.17
Because of this, other imaging modalities have been suggested as being superior, especially with earlier disease.18 Nuclear medicine bone scanning is not favored at this time. With this modality, the sacroiliac joints are sometimes difficult to assess, especially if there is bilateral disease. It is typically a sensitive, but highly nonspecific tool. Computed tomography (CT) scanning of the sacroiliac joints can identify the margins of the bony cortex, and therefore one might be able to see erosions better.19 Chronic changes may be easier to identify by CT, but this tool is less helpful in identifying inflammatory joint and bone changes. There is also the risk of significant radiation exposure.
Magnetic resonance imaging (MRI) of the sacroiliac joints and bony pelvis includes T1- and T2-weighted images taken in coronal and axial planes. T1 with gadolinium or fat-suppressed T2 images are the most helpful for seeing inflammatory features (Figs. 7 and 8). With active, inflammatory sacroiliitis, one sees increased T1 gadolinium-enhanced or fat-suppressed T2 signal from the adjacent bone and bone marrow, suggesting edema.20 Synovitis and erosions can also be visualized. Although standard radiographs are still considered the image of first choice, MRI may be most useful in the case of high clinical suspicion and normal or equivocal radiographs.21 MRI has the highest sensitivity and specificity of all imaging techniques for sacroiliac joint inflammation.22 One can assess both acute and chronic changes at once, and there are no radiation concerns. MRI can also be used to assess spinal inflammation. MRI has several disadvantages: it is expensive, it is time consuming, and it requires specialized interpretation.
At this time, there is no known curative treatment for ankylosing spondylitis. Goals of treatment are to reduce pain and stiffness, slow progression of the disease, prevent deformity, maintain posture, and preserve function. Physical therapy, exercise, and medications are the main forms of treatment. There are rare opportunities for surgical intervention.
Physical therapy and exercise regimens are prescribed for all patients with ankylosing spondylitis. This has been shown to improve measures of pain, spinal mobility, patient functioning, and well-being. Supervised programs have been found to be better than at-home versions.23 Patient education programs can increase understanding and compliance. Cigarette smoking has been associated with a poor functional outcome,24 and patients should be encouraged to quit.
Nonsteroidal anti-inflammatory drugs (NSAIDs) should be prescribed for all patients unless contraindicated. They appear to provide benefits in both pain and function. Generally, regardless of the NSAID used, the maximum approved dosage is required for efficacy. In a pooled analysis of NSAIDs versus placebo over 6 weeks, NSAIDs significantly improved spinal pain, peripheral joint pain, and function. Cyclooxygenase-2 inhibitors were equally effective, although effect on peripheral arthritis was not investigated.25 Their disease-modifying properties remain unclear. Possible radiographic slowing has been demonstrated with continuous versus on-demand use.26 Side effects are common, typically mild, and mainly gastrointestinal. Rarely, there can be more significant gastrointestinal, renal, hepatic, and cardiovascular toxicities.
Systemic corticosteroids are typically not necessary and have only a limited role. However, injectable steroid preparations can be used for enthesitis and peripheral arthritis. Injection of long-acting steroids into the sacroiliac joints has been found to be beneficial in some.27
The use of disease-modifying antirheumatic drugs (DMARDs) has been adapted mainly from use in rheumatoid arthritis. Several agents have been employed, including gold salts, methotrexate, sulfasalazine, hydroxychloroquine, and leflunomide.
In a Cochrane meta-analysis of 11 randomized, controlled trials, sulfasalazine demonstrated some benefit in reducing ESR, morning stiffness and peripheral arthritis (two trials) and no benefit in improving physical function, pain, spinal mobility, enthesitis, or patient and physician global assessment.28 Patients with early disease, elevated ESR, and peripheral arthritis might benefit from a trial of sulfasalazine. Side effects are typically dose related, mild, and mainly gastrointestinal. Rarely there are more serious allergic, hematologic, and renal toxicities.
A similar meta-analysis of two trials with methotrexate versus naproxen demonstrated no difference in outcomes of pain, function, peripheral arthritis, enthesitis, morning stiffness, patient and physician global assessement, ESR, or CRP.29 Additional randomized, controlled studies with higher dosages for longer periods of time are necessary.
Four anti–TNF-α agents in the United States (infliximab, golimumab, etanercept, and adalimumab) have been evaluated and approved by the U.S. Food and Drug Administration (FDA) for use in ankylosing spondylitis. 80% of patients respond to at least one of the agents, and the response is typically rapid and sustainable.30 The larger clinical trials data for all four drugs demonstrate relatively similar results. Using standard clinical assessment tools, 50% to 60% of patients achieved more than 20% clinical trial response (Ankylosing Spondylitis Assessment Study Group [ASAS 20]) at 24 weeks.31, 32, 33, 41 Observational studies have demonstrated continued response at 2 years.34, 35 In a pooled analysis of six of the studies, TNF antagonists have demonstrated moderate to large effects on spinal pain, peripheral joint pain, and function.25 Lack of response to one of the agents does not predict response to another. Predictors of positive response include shorter disease duration, better functional index, higher disease activity, and higher CRP.36 These drugs are not available in oral form and are quite expensive. Side effects include infections (re-activation of latent tuberculosis or fungi), demyelination, and injection or infusion reactions. The disease-modifying capabilities of these agents are still being assessed. In infliximab-treated patients, the MRI activity score was significantly reduced.37 However, in another study, radiographic progression continued, but at a slower rate than in traditional cohorts.38
Total hip replacement is the most common surgery performed in patients with ankylosing spondylitis.39, 40 Revision surgery may be necessary, because these patients typically present at a young age and when they are still active. Heterotopic ossification following joint replacement is a well-described but still unusual complication following hip replacement. Cervical fusion is indicated for the rare patient with neurologic complications of atlantoaxial subluxation. This is managed similarly to rheumatoid arthritis.
Ultimately, each patient’s treatment must be individualized. Considerations must be made for the level of disease activity, degree of functional impairment, concomitant illnesses, tolerance of medications, and patient and physician expectations.