Article Text
PDF
Abstract
Axial spondyloarthritis (axSpA) is a chronic inflammatory rheumatic disease that manifests primarily in the axial skeleton, initially mostly in the sacroiliac joints (SIJ), usually later spreading to the spine. The disease is characterised by inflammation and new bone formation which are mainly assessed by conventional radiography (CR) and magnetic resonance imaging (MRI). Tumour necrosis factor inhibitors (TNFi) and interleukin-17 antagonists have been shown to be efficacious and efficient in patients with axSpA. This treatment seems to also inhibit structural damage, for example, retard radiographic progression. Indeed, a reduction of new bone formation in the spine, as assessed by CR, has been reported to occur after at least 2 years of therapy with TNFi. Recently, a reduction of erosions and ankylosis in the SIJ has also been observed in axSpA patients treated with etanercept and filgotinib. In this narrative review, we discuss the limited significance of such findings.
- magnetic resonance imaging
- spondylitis
- ankylosing
- tumour necrosis factor inhibitors
- etanercept
- biological therapy
Statistics from Altmetric.com
- magnetic resonance imaging
- spondylitis
- ankylosing
- tumour necrosis factor inhibitors
- etanercept
- biological therapy
Introduction
Axial spondyloarthritis (axSpA) is a chronic inflammatory rheumatic disease1 2 that covers both, radiographic axSpA (r-axSpA)—which is almost equivalent3 to the classical ankylosing spondylitis (AS)—and non-raxSpA (nr-axSpA). This distinction, which is more important for classification than diagnosis,4 is on the one hand based on the 2009 ASAS-classification and the 1984 New York (NY)-classification criteria for axSpA5 and AS,6 respectively. On the other hand, it is historically grown because the first biological diseas- modifying antirheumatic drugs (bDMARDs) had been approved for AS, now r-axSpA, first, and—to get an approval for the whole spectrum, a second approval for nr-axSpA had to be obtained. Only the fifth tumour necrosis factor inhibitors (TNFi) in line, certolizumab, tried to get both approvals at a time7—which has been difficult for various reasons but finally succeded.
AxSpA is characterised by inflammation of the sacroiliac joints (SIJ), the spine, peripheral joints and entheses as well as extramusculoskeletal pathology affecting eye, skin, gut and heart.2 The main musculoskeletal findings are sacroiliitis, spondylitis, synovitis and enthesitis. Sacroiliitis is associated with inflammatory back pain localised to the buttocks that may alternate between both sides.1
The chronic inflammation in the axial skeleton may lead to erosions and new bone formation in the SIJ and the spine.1 2 While inflammation—mostly in form of a bone marrow oedema —can only be detected by MRI, structural changes such as erosions and ankylosis can also be detected by conventional radiography (CR) and computed tomography (CT).8 There is evidence that MRI can also detect structural changes including fat metaplasia that is not detected by any other imaging technique.9 10 Assessment in SpondyloArthritis international Society (ASAS) recommendations for description and diagnosis of MRI changes in the SIJ have been recently updated.11
Neglecting the substantial genetic impact on the pathology of axSpA—which is not only due to HLA B2712—and concentrating on the established link between inflammation and new bone formation which has been defined several years ago13 14 the hypothesis of the sequence of pathophysiologically relevant events is: inflammation15 16 associated with osteoclastic activity17—repair and tissue transformation mechanisms such as fat metaplasia9 18— and new bone formation.19 However, the significance of that sequence is limited to the presence of fat metaplasia in the joint space that has also been called ‘backfill’ in the SIJ, findings that may relate to subsequent ankylosis.20 21 Indeed, SIJ ankylosis and fat metaplasia but not inflammatory lesions increased the propensity for spinal radiographic progression.20
BDMARDs such as the TNFi are clinically efficacious, recommended22 and they can inhibit structural changes in the spine—with the main target syndesmophytes—of patients with r-axSpA when given over a period of several years.23 In the last years, several studies have reported that the incidence of erosions in the SIJ can be reduced by treatment with TNFi—by both CR24–29 and MRI,21 30 as compared with a more natural course.31–33
The aim of this narrative review is to discuss the significance of such findings.
Detection of structural changes in the SIJ
Ever since the NY criteria (Rome, 1961) were first published in 1963,34 structural changes in the SIJ have played a central role for identification of patients with AS. Even though it was tried to precisely define the SIJ changes seen on radiographs,35 it was never easy to judge whether definite structural changes were present or not and even a special training did not lead to better results.36 In studies aimed for an approval of bDMARDs for nr-axSpA, disagreement on the degree of SIJ structural changes occurred frequently.37
In an early German spondyloarthritis inception cohort (GESPIC), the first transition rate from no definite change to definite change in the SIJ or from nr-axSpA to r-axSpA was reported to be 12% within 2 years.31 In the French early axSpA cohort “Devenir des Spondyloarthrites Indifférenciées Récentes (DESIR), the progression rate was lower: only 5% in 5 years.32 One methodological problem with these studies is that the observed structural changes in the SIJ are not unidirectional38 implying that there is also ‘improvement’, which is pathophysiologically unlikely to occur. Another problem is that many severe patients who develop structural changes early within the first 2 years of back pain can already be classified as r-axSpA before they are even diagnosed which has indeed been reported39—and it still takes an average of 5 years to be diagnosed with axSpA.40
There is increasing evidence that structural changes in the SIJ are also well detected by MRI.9 41 The main pathologies observed are erosions and ankylosis. The diagnostic role of fat metaplasia in the bone or in the joint space (the latter also known as ‘backfill’) is less clear.42 In all studies performed so far, the changes observed have been rather small. However, several studies have reported such small improvements. While it seems possible that erosions improve, there is so far no evidence that ankylosis can ‘really’ improve. Furthermore, the ‘backfill’ of an erosion has also to be considered as new bone formation which cannot really be considered as a relevant aim for therapeutic strategies. The methodological question whether spurious changes can be reliably differentiated from true changes in these MRI studies can only be answered no, we can’t but we can come close to statistical truth by using clear definitions, several experienced readers, providing a good imaging quality and adequate statistical methods.
More recent data have even shown that the yield may be even better with MRI as compared with the historical routine standard CR.43 A major argument backing this data is the fact that the latter is two-dimensional whereas MRI, similar to CT,44 provides a three-dimensional image with much better anatomical insights. In addition, fat metaplasia cannot be detected by CR but may indicate a pathological change indicative of axSpA.9 In summary, structural changes in the SIJ can also be detected by CR and MRI but, as shown in a recent direct comparison to MRI and CR, CT had the best accuracy for diagnosing axSpA.45 Nevertheless, although CT can still be considered as the gold standard to detect structural changes in the SIJ, new methods seem to increasingly challenge this role.46 47
Detection of structural changes in the spine
The detection of structural changes in the spine by CR is usually quantified using the modified Stokes AS Spinal Score (mSASSS)—a widely used and established scoring system,48 however, with some limitations: the most frequently affected thoracic spine is not included,49 and only the anterior part of the spine is being assessed, while the posterior part and the facet joints are left out.
There is limited evidence on the significance of vertebral erosions50 51 in axSpA. On the one hand, they are only infrequently found (about 1% of all vertebral bodies assessed) but, on the other hand, they seem to be somewhat predictive of syndesmophyte formation.51 Since syndesmophytes are detected much more frequently in the spine than erosions, they are clearly in the centre of research and clinical care.52 Low-dose CT seems to be a way to detect structural changes in the spine with better sensitivity and specificity.53 54
In summary, inflammation and new bone formation are the central pathological events in axSpA. Even though the exact sequence of events is not entirely clear, it seems likely that inflammation comes first, then erosions follow, mainly in the SIJ, while in the spine this is not often detected, and finally new bone formation comes in. All these events seem to possibly occur in parallel and fat metaplasia occurs rather between inflammation and bone formation but may also remain the only lesion.
Prediction of new bone formation in the axial skeleton
In GESPIC, the presence of syndesmophytes at baseline, elevated levels of acute-phase reactants, and cigarette smoking were all independently associated with spinal radiographic progression in patients with early axial SpA,55 while there also seem to be differences between male and female patients with less damage in the latter.56 Whether the degree of radiographic changes in the SIJ predicts syndesmophyte formation seems likely but has not been shown to date.
Therefore, the question arises whether structural MRI changes in the SIJ are important for the prediction of syndesmophytes, implying that a possible reduction of structural MRI changes in the SIJ may also be associated with the prevention of new bone formation in the SIJ and functionally more important, in the spine. This relates to the hypothesis that early anti-inflammatory interventions in axSpA—similar to RA—will prevent future radiographic damage and inhibit new bone formation.
Structural changes in the SIJ in axSpA patients on anti-TNF therapy
Four recent studies evaluated the significance of structural changes in the SIJ. In the first study, “Effect of Etanercept on Symptoms and Objective Inflammation in nr-axSpA” (EMBARK) 28, erosion and fat metaplasia of the SIJ were each scored from 0 to 8 per slice for five MRI slices (total score 0–40). Backfill and ankylosis were each scored from 0 to 4 per slice for five slices (total score 0–20). From baseline to 12 weeks, change in mean scores was significantly greater for etanercept than placebo for erosion (–0.57 vs –0.08, respectively) and backfill (0.36 vs 0.06) i.
In the second study on EMBARK data,30 using CR, there was a slightly positive change (worsening) in the total SIJ score for the control group (DESIR) vs a slightly negative change (improvement) in the etanercept group after 104 weeks (least squares mean change: 0.08 (95% CI −0.04 to 0.20) vs −0.14 (95% CI −0.26 to −0.01)).
In the third study on EMBARK data29, again MRI was used, with the same scoring as above. On etanercept, the erosion mean change was − 0.81 (95% CI − 1.09 to –0.53); for the control group (DESIR), erosion mean change was − 0.23 (95% CI − 0.64 to 0.18). The net percentage of patients with a decrease in erosion was significantly greater for etanercept vs controls: 23.9% (95% CI 15.7% to 32.2%) vs 5.3% (95% CI − 6.8% to 17.3%).
In the last one,57 similar results were published for a targeted synthetic (ts) DMARD, the Janus kinase inhibitor (JAK) inhibitor filgotinib. Similar studies on other compounds such as the IL-17 inhibitor ixekizumab are on the way.
The pathophysiological general question whether improvement of erosions and ankylosis is possible at all—and if these are likely to occur at the same time—cannot definitely answered at present. As already mentioned, the observed changes are rather small but they point in the same direction in different studies. We do not think that these reports justify a claim of structural disease modification. Thus, currently the main point is that the likelihood that these changes make a difference in terms of function is rather low.
Functional consequences of bone formation in the axial skeleton
There is no reasonable doubt that both, inflammation and new bone formation contribute to impaired function in patients with axSpA.58 AxSpA patients participating in GESPIC had rather slow progression rates of structural damage in the spine.59 Since these were patients in early disease stages, disease activity was a stronger determinant of function than radiographic damage, while this is different in later disease stages when bone formation is more important. Thus, only a minority of early patients suffers from rapid radiographic spinal progression which is different in later disease stages.60
When will structural damage become clinically relevant with functional impairment? In 10-year study on TNFi treated patients function, as assessed by the Bath AS functional index (BASFI) and spinal mobility, as assessed by the Bath AS mobility index (BASMI), remained stable over time - despite radiographic spinal progression, and no association between the change in mSASSS and BASFI was found, while there was some effect of mSASSS on BASMI changes over time. The data of that study also indicated that, over time, an increase of 20 and 12 mSASSS points would be responsible for an increase of one BASFI and one BASMI point, respectively.61 However, that was a mixed population of patients with r-axSpA and nr-axSpA. Earlier studies on ‘pure’ AS populations have indicated that most of the functional decline occurred within the first 10 years.62
In another GESPIC paper63 statistical analyses adjusted for structural damage in the spine (assessed by mSASSS), disease activity, as assessed by the Bath AS disease activity index (BASDAI), and gender, revealed an independent association of a sacroiliitis sum score with BASFI and BASMI. Change by one radiographic sacroiliitis grade in one joint was associated with BASFI/BASMI worsening by 0.10/0.12 points, respectively, independently of disease activity and structural damage in the spine. These data show that there is a minimal almost neglectable effect of structural SIJ changes on function and mobility. In comparison, structural changes in the spine are much more relevant for function and mobility. As recently discussed in much detail,64 function is a major clinical outcome for patients with axSpA which needs to be put into perspective, also in comparison to treat-to-target strategies that concentrate more on disease activity.
Conclusion
In summary, there is some evidence that the clinical significance of very small changes in the structure of SIJ in patients with axSpA is rather limited. The low frequency of vertebral erosions makes it unlikely that this finding will ever be clinically relevant for spinal disease. Future research should focus on MRI changes in both, the SIJ and the spine with short intervals of examinations and long follow ups to shed more light on the fascinating pathology that can be observed in axSpA. The most important question whether the course of the disease can be substantially changed if anti-inflammatory interventions are set in place very early remains exiting. Predicting a more severe course of disease is one of the main challenges in that regard.
Finally, we like to clearly state that the rather small effect of b-DMARDS and ts-DMARDS on structural SIJ changes as assessed by MRI recently reported is not sufficient to make a claim for structural modification of any of the drugs tested. In our opinion, effects on new bone formation in axSpA need to be shown in the spine.
Ethics statements
Patient consent for publication
References
Footnotes
Handling editor Josef S Smolen
Contributors All three authors have contributed to the manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.