Relevance of Osteoproliferation as an Outcome Parameter in Ankylosing Spondylitis

Heiner Appel; Martin Rudwaleit; Joachim Sieper


Nat Clin Pract Rheumatol. 2008;4(11):578-579. 


Ankylosing spondylitis (AS) has a prevalence of around 0.5% in the general population and affects more males than females.[1] Disease onset generally occurs in the second decade of life, and in 90% of patients is first seen in the sacroiliac joint.[1] Further on in the course of the disease, the whole spine can be affected with spondylitis, spondylodiscitis and arthritis of the small intervertebral joints. Subsequent to inflammation, new bone formation occurs via the development of syndesmophytes, which is regarded as the hallmark of the disease. Although it is unclear whether ankylosis of the sacroiliac joints has any functional relevance, the formation of syndesmophytes between vertebral bodies and the ossification of the zygapophyseal joints is generally regarded as a major determinant for long-term outcome and disability. Not all AS patients, however, have or develop syndesmophytes; in an AS cohort, with symptom duration less than 10 years, syndesmophytes were found in about 25% of patients,[2] which increased to about 50%[3] in patients with more-advanced disease.

The modified Stoke Ankylosing Spondylitis Spine Score (mSASSS) is currently the most frequently used scoring method for radiographic progression of the spine.[4] It scores the anterior sites of the lumbar and cervical spine on a lateral view. All four corners of each vertebra are scored in the following way: 1 for erosion, sclerosis and/or squaring; 2 for a syndesmophyte; and 3 for total bony bridging at each site. The maximum possible score is 72. At a cohort level, a notable change in this score does not occur before 2 years after symptom onset, but individual patients might progress more rapidly. Mean increases in this score of between 0.4 and 1.5 have been reported in several AS cohorts, with and without therapeutic intervention, over an observation period of 2 years.[3,5,6]

Patient physical function is probably the most relevant outcome parameter in ankylosing spondylitis and other chronic rheumatic diseases. Not unexpectedly, impairment of physical function, analysed by the Bath Ankylosing Spondylitis Functional Index and Dougados Functional Index, is independently associated with patient-reported disease activity and the level of structural damage of the spine as determined by the mSASSS method.[7] Early in the course of the disease, the degree of impairment is mostly determined by the level of reversible inflammation in the spine—which continues to have a contributing role throughout the majority of the disease course—whereas later, irreversible structural damage normally prevails.

The use of tumor necrosis factor (TNF) blockers to treat active AS in patients who are refractory to conventional therapy has led to a breakthrough in the management of this chronic inflammatory disease. About half of patients treated with any of the approved TNF-blocking agents show a 50% improvement in disease activity.[8,9] The proportion goes up to 70–80% in patients with shorter disease duration (<10 years), and is not higher than about 20% if the disease duration is longer than 20 years.[8,9] In patients administered short-term and long-term TNF-blocker treatment, a continuous reduction of active inflammation of the sacroiliac joints and the spine, as shown by MRI, can be seen over an observation period of 2 years, as well as normalization of C-reactive protein level. If TNF-blocker treatment is continued after this status is reached, physical function and spinal mobility further improve and may remain stable at a low level for many years.[1]

In light of this impressive efficacy, it has become an important question whether TNF blockers can prevent structural damage in AS and how structural damage should be defined. Two studies have shown continued formation of syndesmophytes in AS patients, despite treatment with etanercept[5] or infliximab over 2 years, at a rate similar to that observed in a historical control group of matched patients who have never received TNF blockers (the OASIS cohort). We have argued that the sequence of structural damage in AS is different from that seen in rheumatoid arthritis, which is typified by persistent erosive destruction.[10] In AS, inflammation is initially responsible for erosive structural damage and the resultant bony defects are filled up with fibrous repair tissue, which eventually becomes ossified. Ossification is made easier if inflammation is no longer present at the site because the local presence of TNF, which inhibits osteoblast activity, is reduced.[11] Conversely, the suppression of TNF, using TNF blockers, stimulates osteoblast activity and new bone formation. Accordingly, TNF blockers do not inhibit the ossification of damaged bone or cartilage areas once the erosive defect is already present.[10] In all the treatment trials of TNF blockers, the mean AS disease duration was longer than 10 years, which meant that many patients already had erosive structural damage to their vertebral bodies. Consequently, the absence of major inhibition of new bone formation is unsurprising in patients treated for 2 years. However, as outlined above, very effective suppression of inflammation and impressive improvement of disease activity, physical function and spinal mobility were achieved after TNF-blocker therapy.[1] If effective and early treatment of inflammation also prevents erosive structural damage, this should also result in long-term prevention of syndesmophyte development. This hypothesis needs to be tested in AS patients treated with TNF-blockers and followed up for longer than 2 years.

Thus, with respect to the reasoning presented here, what is the role of osteoproliferation as an outcome parameter in AS? This question has two aspects. First, the clinical importance of the slow growth of syndesmophytes in AS patients is unclear. Furthermore, it takes many years before syndesmophytes develop in a considerable proportion of patients who have this feature. By analyzing a cohort of patients with AS with a symptom duration of less than 10 years, we have gathered evidence that the levels of disease activity and pain do not correlate well with the degree of radiographic chronic bone damage seen in these patients, because the amount of new bone formation is also determined by factors other than inflammation.[2] However, radiographic damage of the spine does correlate with restriction of spinal mobility and impairment of physical function, especially in patients with longstanding disease. Accordingly, measuring the growth of syndesmophytes in the spine remains a relevant outcome parameter, but, compared with measuring structural damage in rheumatoid arthritis, for example, disease activity and physical function are probably more important, mostly because the growth of syndesmophytes is slow and physical function is determined over a long period of time, typically by inflammation. The second aspect to be addressed is that osteoproliferation (i.e. growth of syndesmophytes) cannot be inhibited by short-term treatment with TNF blockers, especially when erosive structural damage is already present. However, we predict that TNF blockade retards osteoproliferation in the long-term and that syndesmophytes are prevented if inflammation is treated sufficiently and early enough. In this case, disease activity and physical function are more relevant outcome parameters to judge the efficacy of TNF blockers in AS. Whether the new formation of syndesmophytes in patients treated with TNF blockers is clinically meaningful enough to justify an additional treatment approach that targets osteoblast activity remains to be seen.

CLICK HERE for subscription information about this journal.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.