Active and chronic sacroiliitis, spondylitis and enthesitis, How specific are imaging findings for axial spondyloarthritis?

The initiating event for this article has been the increasing number of publications on ‘background noise’ related to a relatively high prevalence of MRI findings suggestive of sacroiliitis and spondylitis in subjects with no evidence of axial spondyloarthritis (axSpA) (Fig. 1) by different groups [1–5]. AxSpA is a chronic rheumatic disease that is characterized by inflammatory back pain and several other disease manifestations and comorbidities [6]. The 2009 Assessment of SpondyloArthritis international Society (ASAS) classification criteria [7] have led to a differentiation between the classical AS or radiographic axSpA, and non-radiographic axSpA based on the presence or absence of definite radiographic changes in the sacroiliac joints (SIJs). The two major novel items introduced by the SpondyloArthritis international Society criteria were, for the first time, HLA B27 and, in addition to the already long established conventional radiographs, MRI of the SIJs. Later on, definitions for a ‘positive MRI’ of the SIJs ‘suggestive of axSpA’ published by the SpondyloArthritis international Society [8] have been used in many studies and in clinical practice. The current situation in relation to the new findings of ‘positive MRIs’ in the axial skeleton in individuals without features of axSpA is now that, on the one hand, there is increasing uncertainty related to the diagnosis and classification of axSpA, and, on the other hand, it reminds us of an experience more than a decade ago, when immune responses in patients with ReA were studied, based on substantial clinical and epidemiological experience [9–14]. Even though this research had been quite successful based on even the recognition of specific epitopes of bacteria and self-antigens by T cell clones raised from synovial fluid of affected patients, no definite explanations for the pathogenesis of the disease were achieved. One of the reasons for this is the comparatively good prognosis of ReA with the vast majority of patients being free of symptoms in less than a year [15]. However, >10% of patients have been shown to develop chronic disease including axSpA, and almost all of these chronic patients carry the HLA B27 gene [9–11]. Since epidemiological studies on ReA outbreaks after food poisoning have shown that HLA B27 was not associated with susceptibility to develop arthritis in the first place [12–14], HLA B27 has been regarded rather as a marker for severity and chronicity of ReA [9–11]. MRI of the SIJs of a 29-year-old male patient with the clinical diagnosis of axSpA BME is found bilaterally and with a ‘deep’, intense signal of significant size (arrowheads). AxSpA: axial spondyloarthritis; BME: bone marrow oedema; SIJ: sacroiliac joint; STIR: short tau inversion recovery sequence. The structural consequences of sacroiliitis and spondylitis such as erosions, syndesmophytes and ankylosis have long been considered as being specific for axSpA. However, these are structural changes that develop because of severity and chronicity of the disease rather than susceptibility. We believe that the experience now made with the incidence and prevalence of MRI findings suggestive of sacroiliitis and spondylitis in non-axSpA subjects supports the view, which goes very much in a similar direction as was made with ReA. Thus, such early and often light inflammatory features may well be a lot less specific for axSpA than previously thought and represent rather frequent events that occur likely due to mechanical stress, as suggested by their high prevalence in postpartal females (Fig. 2), runners and soldiers [1–4]. In one of these studies [2] there was only one MRI finding that differentiated the mostly HLA B27+ axSpA patients from all controls: a ‘deep’ or extensive bone marrow oedema (Fig. 1). These data are consistent with an earlier study on the prognostic significance of HLA B27 in connection to the degree of inflammation as assessed by MRI in the SIJs of patients under suspicion of early axSpA [16]. In this study, the combination of both extensive sacroiliitis and HLA B27 had the strongest impact on the development of chronic structural changes and AS. Very much in line with this it was recently confirmed that the prevalence of SIJ erosions in a population with no evidence of axSpA is rather low [17]. MRI of the SIJs of a 45-year-old HLA B27 positive female patient with the clinical diagnosis of osteitis triangularis condensansBME is found mostly on the left SIJ, located in the lower and anterior third of the joint (arrowheads). BME: bone marrow oedema; SIJ: sacroiliac joint; STIR: short tau inversion recovery sequence. Those most likely mechanical stress-induced minor bone marrow oedema findings in subjects with no evidence of SpA are likely to simply disappear in the majority of cases, once the mechanical load is diminished or gone. However, if this is not the case, this may well lead to chronic and structural changes—many such changes of osteoarthritic nature such as spondylophytes or enthesophytes are known to occur frequently in many individuals. Thus, the prevalence of enthesophytes and osteophytes in an elderly population was found to be in the range of 60% and 70%, respectively [18]. However, in the absence of a conducive genetic background, no signs and symptoms of SpA will develop in the majority of such cases. This remains to be confirmed in prospective studies but the hypothesis that HLA B27 is mainly responsible for the chronicity and severity of SpA is clearly backed by the new data. In addition, there is some evidence from basic scientific [19] and epidemiological [20, 21] studies that mechanical stress is an important factor for the development of enthesitis [22], axial inflammation [1–4] and axSpA. Similarly, enthesitis was identified as a rather characteristic sign of SpA about 20 years ago [23], and in consequence, a whole concept of the enthesis organ as the central location of SpA pathology has been developed [24]. Only recently, enthesitis has been highlighted as the key feature of psoriatic arthritis and both axial and peripheral SpA [22]. However, even in the early study, enthesitis also occurred in the control group with rheumatoid arthritis, although less frequently [23], and similar findings in controls have also been reported in patients with plantar fasciitis [25] and enthesitis at the Achilles tendon [26] but no evidence of SpA, and even in healthy controls [27]. Furthermore, in a recent study using arthroscopy in early knee or ankle arthritis, no major difference between RA and SpA was found [28]. Thus, enthesitis, although clearly a typical clinical feature of SpA, is less specific for SpA than previously thought, and, again, HLA B27 is the major risk factor for chronicity [26, 27]. However, it is not only HLA B27 that promotes a chronic course of disease but, as shown for oligoarthritis of the knee, also psoriasis, another complex genetic disease, can take that role [29]. In summary, sacroiliitis, spondylitis, enthesitis and ReA occur in patients not only with SpA (Fig. 1) but also with other conditions (Fig. 2) and even in the normal population. If affected individuals happen to be HLA B27+, they are at higher risk of developing chronic and structural disease than other individuals, and, therefore, we see those patients in our rheumatological offices much more frequently. However, the pathomechanism behind this development of chronic muscular disease on a genetic background has, several decades after the discovery of the association of AS with HLA B27, remained unclear. Nevertheless, there is evidence from large genetic trials (reviewed in [30]) for a role of an antigen-driven immune response in the pathogenesis of axSpA based on the function of MHC class I molecules in connection with polymorphisms of the endoplasmic reticulum aminopeptidase 1 (ERAP I) whose function is to trim peptides to a length suitable to fit into the groove of the MHC molecule (reviewed in [31]). However, it is unclear how mechanical stress and the immune response are linked. In summary, the finding of rather frequent events of sacroiliitis, spondylitis, enthesitis and ReA in the normal population rather than indicating an anatomical site specifically designed to cause axSpA is likely a step forward in our understanding of the pathogenesis of this disease. The doubtless interesting results of animal experiments such as the highly published study showing that IL-23 is essential for the development of enthesitis by acting on resident T cells in mice [32] may be misleading in the way that they target susceptibility rather than severity and chronicity of SpA. We think that there is increasing evidence for common mechanisms such as mechanical stress to initiate inflammatory reactions in entheses and SIJs but for a chronic disease with structural changes the immune system is needed, and, to better understand the pathophysiology of SpA, we should rather concentrate on the question of how chronic rather than acute inflammation develops. Funding: No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this manuscript. Disclosure statement: The authors have declared no conflicts of interest.