Ankylosing spondylitis is characterised by inflammation of the spine and the entheses followed by bone formation. Excessive bone formation in ankylosing spondylitis leads to the formation of bone spurs, such as syndesmophytes and enthesiophytes, which contribute to ankylosis of joints and poor physical function. This process is based on increased differentiation of osteoblasts from their mesenchymal precursors, which allows to rapidly build up new bone. Prostaglandins, bone morphogenic proteins and Wnt proteins play an essential role in this process. By contrast, tumour necrosis factor (TNF) does not appear to be the direct trigger for osteophyte formation in ankylosing spondylitis. The article reviews the current knowledge regarding the mechanisms and clinical role of ankylosis and explains strategies on how to prevent it in patients with ankylosing spondylitis.
Introduction
Ankylosing spondylitis (AS) is a chronic inflammatory rheumatic disease of unknown origin affecting the axial skeleton including the sacroiliac (SI) joints and the spine and also the peripheral joints and the entheses . Since AS usually starts in the third decade, it affects people for most of their life. AS has a strong genetic component, with HLA-B27 being the most relevant gene. Clinically, pain and stiffness in the back are the leading complaints. In addition, the patient may suffer from peripheral joint pain or enthesitis, typically of the heel, or from bouts of uveitis. With time, bony ankylosis of the SI joints and the spine develops in many patients. Bony ankylosis at the spine may affect the vertebral bodies (syndesmophyte formation) as well as the facet, the zygapophyseal and the costovertebral joints. It is particularly spinal ankylosis, which results in loss of spinal mobility and changes in posture such as the development of thoracic kyphosis and irreversible stiffness. In a large proportion of patients, the disease course in AS that runs is fluctuating with not only periods of flares but also periods of remission or low disease activity. Some patients, however, suffer from persistently high disease activity.
Excessive bone apposition in AS
Bony overgrowth in AS has traditionally been considered as a structural damage arising from chronic immune activation and inflammation. In contrast to rheumatoid arthritis, where structural changes are of primarily catabolic nature, resulting in a net loss of bone substance in the vicinity of joints, structural changes in AS are dominated by anabolic processes. Bony spur formation, which arises from the cortical bone surface, is a common feature of AS and virtually affects all skeletal compartments that show disease morbidity. In the case of the vertebral column, such lesions are termed ‘syndesmophytes’. These lesions grow usually in a vertical orientation and are bridging vertebral bodies, and are formed by apposition of new bone along the edges of the vertebral bodies and subsequent loss of the intervertebral spaces . When several consecutive vertebrae are affected, it may lead to the appearance of a ‘bamboo spine’. In particular, the insertion sites of the tendons are hot spots of bone apposition in AS and also affect peripheral compartments such as the Achilles tendon or the plantar fascia, resulting in the formation of bony spurs, then termed ‘enthesiophytes’. Moreover, inflammation of the SI joints, the small joints of the spine, such as the facet joints, as well as occasionally the peripheral joints leads to bony appositions followed by ankylosis in patients with AS.
Bony spur formation and ankylosis is pathognomonic for AS, progressively limiting the range of motion of patients and leading to functional impairment. This bone anabolic process does not randomly affect intervertebral spaces and joints but is limited to certain predilection sites. Although the prerequisite for bony spur formation is not completely understood, it is likely that both mechanical and inflammatory components of the disease locally trigger increased bone formation. In the case of mechanical components, the fact that these lesions are often found along the insertion sights of the tendons points to a key role of mechanical stress. On the other hand, inflammatory lesions in the neighbouring bone marrow (osteitis) are considered as a risk factor for syndesmophyte formation, although this association is not complete and bony spurs can also be found at sites where no osteitis is seen and vice versa . The notion, however, that mechanical and inflammatory triggers can induce bony overgrowth has suggested that these lesions might represent a ‘response-to-stress mechanism’ and a type of active repair strategy of the joint rather than damage arising from inflammation.
Excessive bone apposition in AS
Bony overgrowth in AS has traditionally been considered as a structural damage arising from chronic immune activation and inflammation. In contrast to rheumatoid arthritis, where structural changes are of primarily catabolic nature, resulting in a net loss of bone substance in the vicinity of joints, structural changes in AS are dominated by anabolic processes. Bony spur formation, which arises from the cortical bone surface, is a common feature of AS and virtually affects all skeletal compartments that show disease morbidity. In the case of the vertebral column, such lesions are termed ‘syndesmophytes’. These lesions grow usually in a vertical orientation and are bridging vertebral bodies, and are formed by apposition of new bone along the edges of the vertebral bodies and subsequent loss of the intervertebral spaces . When several consecutive vertebrae are affected, it may lead to the appearance of a ‘bamboo spine’. In particular, the insertion sites of the tendons are hot spots of bone apposition in AS and also affect peripheral compartments such as the Achilles tendon or the plantar fascia, resulting in the formation of bony spurs, then termed ‘enthesiophytes’. Moreover, inflammation of the SI joints, the small joints of the spine, such as the facet joints, as well as occasionally the peripheral joints leads to bony appositions followed by ankylosis in patients with AS.
Bony spur formation and ankylosis is pathognomonic for AS, progressively limiting the range of motion of patients and leading to functional impairment. This bone anabolic process does not randomly affect intervertebral spaces and joints but is limited to certain predilection sites. Although the prerequisite for bony spur formation is not completely understood, it is likely that both mechanical and inflammatory components of the disease locally trigger increased bone formation. In the case of mechanical components, the fact that these lesions are often found along the insertion sights of the tendons points to a key role of mechanical stress. On the other hand, inflammatory lesions in the neighbouring bone marrow (osteitis) are considered as a risk factor for syndesmophyte formation, although this association is not complete and bony spurs can also be found at sites where no osteitis is seen and vice versa . The notion, however, that mechanical and inflammatory triggers can induce bony overgrowth has suggested that these lesions might represent a ‘response-to-stress mechanism’ and a type of active repair strategy of the joint rather than damage arising from inflammation.
Cellular and molecular mechanisms of bone formation in AS
A better insight into the molecular regulation of new bone formation is key for defining the optimal intervention strategies to retard or block bony ankylosis in patients with AS. Ankylosis in AS is based on the apposition of new bone along periosteal skeletal sites requiring differentiation of osteoblasts, which are the bone-forming cells. Osteoblasts develop from mesenchymal cell precursor cells, which cover the inactive periosteal bone surface. Growth as well as injury, such as observed during inflammation and also in case of fracture, lead to an activation of the periosteal bone surface and differentiation of osteoblasts, which allow to build up new bone. Osteoblasts synthesise bone matrix, which consists of numerous proteins, the most abundant of which are collagen type I and osteocalcin. Bone formation and ankylosis in AS depends on molecular signals, which regulate differentiation and activity of osteoblasts. Several mediators are of importance for osteoblast differentiation: prostaglandins, such as PGE2, are important local factors; PGE2 has anabolic effects on bone and promotes proliferation and differentiation of osteoblasts, thereby inducing the expression of bone sialoprotein and alkaline phosphatase . Moreover, PGE2 can synergise with bone morphogenic protein (BMP)-2, a member of the TGF/BMP protein family in inducing bone formation . Members of the BMP family, that is, BMP-2, -3 and -7, play a critical role in osteoblast differentiation and induce signalling through Smad proteins upon engaging respective surface receptors on mesenchymal cells . Activation of intracellular Smad signalling indicates an increased activity of BMPs and this process occurs during enthesiophyte formation in AS . Bone anabolic effects of BMPs can be antagonised by noggin, a natural inhibitor of BMP. Haploinsufficiency of noggin enhances BMP activity and has been shown to lead to ankylosis of joints. Wnt proteins have also been identified as potent inducers of new bone formation. Wnt proteins bind to a receptor/co-receptor complex on the plasma membrane, which consists of LRP5/6 and Frizzled proteins. Engagement of this receptor complex by Wnt proteins leads to phosphorylation of beta-catenin, which translocates to the nucleus and induces transcription of genes involved in osteoblast differentiation and bone formation . Increased activity of beta-catenin as a surrogate marker for Wnt activation has been observed in bony spur formation, suggesting that Wnt proteins may indeed contribute to new bone formation and ankylosis of joints. Natural inhibitors of Wnt such as Dkk-1 and sclerostin neutralise Wnt activity and actively prevent new bone formation . These proteins have shown to be differentially expressed in diseases such as rheumatoid arthritis and AS, with high expression in the former but low expression in the latter disease .
It is highly likely that inflammation initiates the process of ankylosis in AS. Thus, AS does not represent a disease manifesting with ankylosis per se or where ankylosis is purely driven by a mechanical trigger, but rather reflects diseases, where a chronic inflammatory process in the spine results in ankylosis. This observation implies that inflammation is molecularly linked to ankylosis and that inflammation triggers the process of new bone formation. One of the key inflammatory cytokines involved in AS, tumour necrosis factor (TNF), however, does not induce but rather inhibits bone formation. TNF is a potent inducer of proteins such as Dkk-1 and sclerostin and thus down-regulates bone formation . It is yet unclear why periosteal bone apposition can still occur in AS patients despite TNF playing a central role in the inflammatory processes of this disease. Importantly, there is a striking difference between cortical bone and trabecular bone in AS: whereas trabecular bone mass decreases and leads to vertebral osteoporosis and increased fracture risk in AS patients, specific sites of the cortical bone start to proliferate and expand . It can be hypothesised that the skeletal effects of TNF, which are down-regulation of bone formation plus enhancement of bone resorption, are reflected by systemic bone loss in the trabecular bone compartment, whereas cortical bone apposition is not linked to the expression of TNF itself. As TNF is not the key trigger for new bone formation in AS, one can hypothesise that other inflammatory mediators drive new bone formation and ankylosis in AS. Classical pro-inflammatory pathways such as interleukin (IL)-1 and IL-17 as well as signalling through the IL-6R/gp130 complex, however, exert negative rather than positive net effects on bone mass, supporting the notion that inflammation and immune activation exert a negative effect on bone formation and promote bone loss rather than bone apposition Fig. 1 .
These three groups of molecules, PGE2, BMP proteins and Wnt proteins manage differentiation of mesenchymal precursor cells into bone-forming osteoblasts. This differentiation process is accompanied by the expression of key transcription factors, such as Cbfa1 and Osterix, which are important for inducing osteoblast-specific genes. It is yet unclear whether bony-spur formation (osteophytes sensu lato ) along peripheral joints (osteophytes sensu strictu ), insertion sites of the tendons (enthesiophytes) and along the vertebral bodies (spondylophytes, syndesmophytes) follows similar molecular pathways. There are principally two pathways known to be relevant for bone formation: one is endochondral ossification and the other is membranous bone formation, including chondroidal metaplasia. Studies in humans and mice have shown morphological features of hypertrophic chondrocytes as a sign of endochondral bone formation in axial and peripheral joints ; also, direct evidence for membranous bone formation and chondroidal metaplasia has been obtained, especially at enthesial sites .
Instruments used for assessing treatment responses in AS
In the past, the cornerstones of treatment in AS have been physiotherapy and non-steroidal anti-inflammatory drugs (NSAIDs). The introduction of anti-TNF agents has resulted in therapeutic effects, which were hitherto unknown to patients with active AS, especially those who had suffered from symptoms despite high doses of NSAIDs. With anti-TNF therapy, at least every second patient achieves a good-to-very-good clinical response and 20–25% of patients achieve clinical remission. On the other hand, traditional disease-modifying anti-rheumatic drugs (DMARDs) and low-to-moderate doses of steroids do not have major therapeutic effects on axial disease in AS .
For the assessment of disease activity in AS, patient-reported outcomes, such as the Bath ankylosing Spondylitis Disease Activity Index (BASDAI), play a major role. BASDAI captures five disease domains (fatigue, back pain, pain in other joints, enthesitis and morning stiffness). Each domain is assessed on a scale from 0 to 10 (or 0–100) and the average of the five domains gives the overall BASDAI score . In general, a BASDAI score higher than 4 is considered to indicate active disease. BASDAI has been used in many clinical trials over the past years and has been proven to be reproducible and sensitive to change. A new composite measure of disease activity in AS is the ankylosing spondylitis disease activity score (ASDAS), which is currently being evaluated .
For measuring response to treatment in AS, absolute changes in BASDAI or, alternatively, certain cut-offs of relative improvement such as BASDAI 50 (50% improvement in BASDAI) have been used in clinical trials. Other well-established response measures are the Assessment of SpondyloArthritis international Society (ASAS) response criteria such as ASAS20, ASAS40 or ASAS partial remission . These response criteria are composite indices comprising the four domains – morning stiffness, patient global, total pain and function as assessed by the Bath Ankylosing Spondylitis Functional Index (BASFI). To achieve ASAS20, an improvement of at least 20% in at least three out of the four domains, without worsening in the potentially remaining fourth domain, is required. Accordingly, ASAS40 requires an improvement by at least 40%, while ASAS partial remission defines a state of very low disease activity with all domains having absolute values of ≤2. The ASAS20 response measure has been developed specifically for typical NSAID trials in AS with up to 50% of NSAID-treated AS patients reaching ASAS20 . For comparison, in AS patients, who have active disease despite NSAIDs (typical anti-TNF trials), an ASAS20 response is reached by 60–70% of patients, an ASAS40 by as many as 40–50% and ASAS partial remission by 20–30% . In general, patients who achieve BASDAI50 are also likely to achieve ASAS40 response because of overlap of domains, which are assessed by these two measures.
Over the years, several instruments, such as the Stoke Ankylosing Spondylitis Spine Score (SASSS), the Bath Ankylosing Spondylitis Radiology Index (BASRI) and the modified SASSS (mSASSS), have been evaluated to measure structural damage in AS on plain radiographs. A detailed analysis of these instruments showed the mSASSS to be the most appropriate instrument to assess change . Therefore, in recent clinical trials, the mSASSS was usually applied. In all three instruments, the thoracic part of the spine is not included because of technical concerns related to superimposition of the lungs. Excluding the thoracic spine may be considered as a true disadvantage because MRI studies have shown that most of the inflammation takes place in the thoracic spine, and, therefore, most of the damage could be expected in that region . This is even more important given the relatively low rate of progression of structural damage in AS (only ∼30% of patients have a detectable change after 2 years). To overcome this shortcoming, a recently proposed instrument captures the lower spine from T10 to T12, thereby aiming to increase sensitivity to change .
Progression of structural damage in the anti-TNF era
Magnetic resonance imaging (MRI) is currently used not only to facilitate early diagnosis , but also to monitor the course of inflammation in AS. In several controlled clinical trials using the anti-TNF agents infliximab, etanercept or adalimumab, serial MRI scans have been performed over time, most of the time imaging the spine. A dramatic improvement of active inflammatory lesions in the spine by up to 80% after 6–24 months of anti-TNF therapy could be observed . These findings have raised expectations that once spinal inflammation disappears during anti-TNF therapy, bony ankylosis may also be prevented. In 2-year follow-up studies using all three anti-TNF agents, however, no clear indication of such inhibitory effect of anti-TNF therapy on bony ankylosis could be detected . In fact, there seemed to be no difference in the rate of progression between anti-TNF-treated patients and patients from an observational cohort receiving conventional NSAIDs treatment (mean rate of progression of 0.8–1.0 mSASSS points in all groups). Even after adjustment for disease activity at baseline, no significant effect was detectable.
Thus, the question was posed as to how inflammation and bone formation are related to each other, particularly in the light of data from animal models suggesting a disconnect between inflammation and bone formation . Subsequent studies have compared MRI findings at baseline with radiographic progression after 2 years. Interestingly, these studies revealed that new syndesmophytes did arise about three times more often at sites where active inflammation was present at baseline. However, the majority of syndesmophytes arose from sites without detectable inflammation on MRI at baseline ( Table 1 ). Interestingly, in one but not in another of these studies, it was found that new syndesmophytes occurred mainly at sites where previous active inflammation has completely resolved, but rarely at sites with persistent inflammation. However, the absolute numbers for the latter analysis were small and more data are needed. Nonetheless, these data suggest that inflammation and bone formation are linked to at least some extent. Apart from issues such as appropriateness of the instruments to measure damage (see above) and the low progression rate in AS in general, it may well be that periods longer than 2 years are needed to detect an inhibitory effect of anti-TNF therapy on radiographic progression. An alternative explanation for the findings is that, although the inhibitory effect of anti-TNF agents on inflammation on MRI is substantial (reduction by around 80%), the inflammation is not yet completely suppressed, allowing low-grade or intermittently occurring inflammation to provide sufficient stimulus for forming new syndesmophytes. Moreover, it may well be that the repair mechanisms, which take place once inflammation abates, are simply exaggerated in AS in comparison to healthy individuals. If this is the case, the question arises whether such repair processes are self-limiting after a certain time or perpetuate themselves .
Related posts:
Inflammatory osteoarthritis of the hand
Management of osteoporosis in a pre-menopausal woman
Prevention and cure of rheumatoid arthritis: Is it possible?
Evidence-based management of rapidly progressing systemic sclerosis
Treatment-resistant inflammatory myopathy
Primary angiitis of the central nervous system: differential diagnosis and treatment
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
