Which scoring method depicts spinal radiographic damage in early axial spondyloarthritis best? Five-year results from the DESIR cohort

An optimal deep learning model for the scoring of radiographic damage in patients with ankylosing spondylitis

Background

Detecting vertebral structural damage in patients with ankylosing spondylitis (AS) is crucial for understanding disease progression and in research settings.

Objectives

This study aimed to use deep learning to score the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) using lateral X-ray images of the cervical and lumbar spine in patients with AS in Asian populations.

Design

A deep learning model was developed to automate the scoring of mSASSS based on X-ray images.

Methods

We enrolled patients with AS at a tertiary medical center in Taiwan from August 1, 2001 to December 30, 2020. A localization module was used to locate the vertebral bodies in the images of the cervical and lumbar spine. Images were then extracted from these localized points and fed into a classification module to determine whether common lesions of AS were present. The scores of each localized point were calculated based on the presence of these lesions and summed to obtain the total mSASSS score. The performance of the model was evaluated on both validation set and testing set.

Results

This study reviewed X-ray image data from 554 patients diagnosed with AS, which were then annotated by 3 medical experts for structural changes. The accuracy for judging various structural changes in the validation set ranged from 0.886 to 0.985, whereas the accuracy for scoring the single vertebral corner in the test set was 0.865.

Conclusion

This study demonstrated a well-trained deep learning model of mSASSS scoring for detecting the vertebral structural damage in patients with AS at an accuracy rate of 86.5%. This artificial intelligence model would provide real-time mSASSS assessment for physicians to help better assist in radiographic status evaluation with minimal human errors. Furthermore, it can assist in a research setting by offering a consistent and objective method of scoring, which could enhance the reproducibility and reliability of clinical studies.

Sensitivity to change of structural outcomes in axial spondyloarthritis after 10 years of follow up. Data from the DESIR cohort

OBJECTIVE

To evaluate the sensitivity to change in structural imaging outcomes over 10 years of follow-up in patients with axial spondyloarthritis (axSpA).

METHODS

Patients with axSpA from the Devenir des Spondyloarthropathies Indifferénciées Récentes cohort were included. Radiographs and MRIs of the sacroiliac joints (SIJ) and spine were obtained at baseline and at 1, 2, 5 and 10 years. The yearly rate of change of each structural outcome was analysed using generalised estimating equation models, including all patients with ≥1 score from ≥1 reader from ≥1 reading wave, using the time (years) as an explanatory variable and adjusting for reader and wave. All outcomes were standardised, and the relative standardised rate of change was calculated (ie, the standardised rate of an outcome divided by the rate of a reference outcome).

RESULTS

A total of 659 patients (46% males and mean age 33.6 years) were included. The most sensitive outcome to change in the SIJ (both MRI and radiographs) was the presence of ≥3 fatty lesions at a specific timepoint, with a relative standardised rate of change per year of 5.28 using the modified New York criteria as reference.Similarly, the most sensitive to change (in both MRI and radiographs) outcome in the spine was the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS; relative standardised yearly change 1.76) using ≥1 syndesmophyte as reference.

CONCLUSION

MRI structural outcomes in the SIJ (ie, fatty lesions) are more sensitive to change than radiographic outcomes. Conversely, the mSASSS remains the most sensitive method, even when compared with MRI of the spine.

Good construct validity of the CT Syndesmophyte Score (CTSS) in patients with radiographic axial spondyloarthritis

OBJECTIVES

To assess construct validity of the CT Syndesmophyte Score (CTSS) for the measurement of structural spinal damage in patients with radiographic axial spondyloarthritis.

METHODS

Low-dose CT and conventional radiography (CR) were performed at baseline and 2 years. CT was assessed with CTSS by two readers and CR with modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS) by three readers. Two hypotheses were tested: (1) syndesmophytes scored with CTSS are also detected with mSASSS at baseline or 2 years later; (2) CTSS is non-inferior to mSASSS in correlations with spinal mobility measures. Presence of a syndesmophyte was determined per reader per corner for all anterior cervical and lumbar corners on CT at baseline and CR at baseline and 2 years. Correlations of CTSS and mSASSS with six spinal/hip mobility measurements plus Bath Ankylosing Spondylitis Metrology Index (BASMI) were tested.

RESULTS

Data from 48 patients (85% male, 85% HLA-B27+, mean age 48 years) were available for hypothesis 1 and 41/48 were available for hypothesis 2. At baseline, syndesmophytes were scored with CTSS in 348 (reader 1, 38%) and 327 (reader 2, 36%) corners out of 917. Of these, depending on reader pairs, 62%-79% were also seen on CR at baseline or after 2 years. CTSS correlated well (r_s 0.46-0.73), and with higher correlation coefficients than mSASSS (r_s 0.34-0.64), with all spinal mobility measures and BASMI.

CONCLUSIONS

The good agreement between syndesmophytes detected by CTSS and mSASSS and the strong correlation of CTSS with spinal mobility support the construct validity of the CTSS.

Spinal radiographic progression is correlated with preclinical atherosclerosis in spondyloarthritis

BACKGROUND

A higher prevalence of cardiovascular risk was observed in spondyloarthritis (SpA). The relationship between disease-related factors structural damage and subclinical atherosclerosis is still unknown.

OBJECTIVE

The aim of our study was to evaluate the association of subclinical atherosclerosis with radiographic structural damage in patients with SpA.

METHODS

Forty-seven SpA patients who fulfilled the ASAS criteria were enrolled in a case-control study conducted over 12 months and compared with 47 age and sex-matched healthy controls. None of the subjects had a previous history of cardiovascular diseases or cardiovascular risk factors. Demographic and disease characteristics were recorded. Structural lesions were evaluated using plain radiography, and two scoring tools were used to spine (BASRI and mSASSS). Subclinical atherosclerosis was assessed using ultrasound measurements of flow-mediated dilation (FMD) and carotid intima-media thickness (cIMT).

RESULTS

The median age of patients was 36 years. The sex ratio was 2.35. The median BASRI total score was 3 (IQR 2-4), median mSASSS score was 10 (IQR 415). cIMT was significantly increased in SpA patients compared to controls (p< 0.0001), and FMD was significantly lower in patients than in healthy subjects (p= 0.008). cIMT was significantly associated with ankylosis of the facet joints (p= 0.035) and Romanus spondylitis (p= 005). FMD was negatively associated with vertebral squaring (p= 0049), bridging syndesmophytes (p= 0031) and mSASSS score (p= 0.047).

CONCLUSION

Our result supports the association of radiographic structural damage and subclinical atherosclerosis assessed using cIMT and FMD. This finding highlights the importance of earlier treatment in order to prevent radiographic damage progression and atherosclerotic events.

Associations between syndesmophytes and facet joint ankylosis in radiographic axial spondyloarthritis patients on low-dose CT over 2 years

OBJECTIVES

In radiographic axial spondyloarthritis (r-axSpA), spinal damage manifests as syndesmophytes and facet joint ankylosis (FJA). We evaluated whether the presence of one lesion increased the risk of the other lesion.

METHODS

Patients with r-axSpA underwent low-dose CT (ldCT) and MRI of the whole spine at baseline and 2 years. On ldCT, vertebrae were scored for presence and size of syndesmophytes; facet joints were assessed for ankylosis. MR images were assessed for inflammation. Two hypotheses were tested: (i) presence of FJA is associated with new syndesmophyte(s) on the same vertebral unit (VU) 2 years later, and (ii) presence of bridging syndesmophyte(s) is associated with new FJA on the same VU 2 years later. Two generalized estimating equations models were tested per hypothesis using increase of FJA/syndesmophytes (model A) or presence of FJA/syndesmophytes (model B) as outcome, adjusted for inflammation at baseline. Secondary analyses tested the hypotheses with outcomes on adjacent VUs and dose-response effects.

RESULTS

Fifty-one patients were included (mean age 49, 84% male, 82% HLA-B27+). Baseline bridging syndesmophytes occurred more often (range: 10-60% per VU) than FJA (range: 8-36%). Odds ratios (ORs) (95% CI) for presence of bridging syndesmophytes on development of FJA were 3.55 (2.03, 6.21) for model A and 3.30 (2.14, 5.09) for model B. ORs for presence of baseline FJA on new syndesmophytes were 1.87 (1.20, 2.92) for model A and 1.69 (0.88, 3.22) for model B. Secondary analyses yielded positive ORs for both hypotheses.

CONCLUSIONS

Bone formation in vertebrae and in facet joints influence each other's occurrence, with the effect of syndesmophytes being larger than that of FJA.

Biological DMARDs and disease modification in axial spondyloarthritis: a review through the lens of causal inference

Axial spondyloarthritis (axSpA) is a chronic rheumatic disease characterised by inflammation predominantly involving the spine and the sacroiliac joints. In some patients, axial inflammation leads to irreversible structural damage that in the spine is usually quantified by the modified Stoke Ankylosing Spondylitis Spinal Score (mSASSS). Available therapeutic options include biological disease-modifying antirheumatic drugs (bDMARDs), which have been proven effective in suppressing inflammation in several randomised controlled trials (RCT), the gold standard for evaluating causal treatment effects. RCTs are, however, unfeasible for testing structural effects in axSpA mainly due to the low sensitivity to change of the mSASSS. The available literature therefore mainly includes observational research, which poses serious challenges to the determination of causality. Here, we review the studies testing the effect of bDMARDs on spinal radiographic progression, making use of the principles of causal inference. By exploring the assumptions of causality under counterfactual reasoning (exchangeability, positivity and consistency), we distinguish between studies that likely have reported confounded treatment effects and studies that, on the basis of their design, have more likely reported causal treatment effects. We conclude that bDMARDs might, indirectly, interfere with spinal radiographic progression in axSpA by their effect on inflammation. Innovations in imaging are expected, so that placebo-controlled trials can in the future become a reality. In the meantime, causal inference analysis using observational data may contribute to a better understanding of whether disease modification is possible in axSpA.

Which imaging outcomes for axSpA are most sensitive to change? A 5-Year analysis of The DESIR Cohort

Objective

To compare the sensitivity to change of different imaging scoring methods in patients with early axial spondyloarthritis (SpA).

Methods

Patients from the Devenir des Spondylarthropathies Indifferérenciées Récentes (DESIR) cohort fulfilling the Assessment of SpondyloArthritis international Society criteria for axial SpA were included. Radiographs and magnetic resonance imaging (MRI) of the sacroiliac (SI) joints and spine were obtained at baseline, 1, 2, and 5 years. Each image was scored by 2 or 3 readers in 3 separate reading waves. The rate of change of outcomes measuring inflammation of the spine and SI joints (e.g., Spondyloarthritis Research Consortium of Canada [SPARCC] score) and structural damage on MRI (e.g., ≥3 fatty lesions) and radiographs (e.g., modified New York grading) was assessed using multilevel generalized estimating equation models (taking all readers and waves into account). To allow comparisons across outcomes, rates were standardized (difference between the individual's value and the population mean divided by the SD).

Results

In total, 345 patients were included. Inflammation detected on MRI of the SI joints (MRI‐SI joints) (standardized rate range –0.278, –0.441) was more sensitive to change compared to spinal inflammation (range –0.030, –0.055). Structural damage in the SI joints showed a higher standardized rate of change on MRI‐SI joints (range 0.015, 0.274) compared to radiography of the SI joints (range 0.043, 0.126). MRI‐SI joints damage defined by ≥3 fatty lesions showed the highest sensitivity to change (0.274). Spinal structural damage slowly progressed over time with no meaningful difference between radiographic (range 0.037, 0.043) and MRI structural outcomes (range 0.008, 0.027).

Conclusion

Structural damage assessed in pelvic radiographs has low sensitivity to change, while fatty lesions detected on MRI‐SI joints are a promising alternative. In contrast, MRI of the spine is not better than radiography of the spine in detecting structural changes in patients with early axial SpA.

Inflammation of the Sacroiliac Joints and Spine and Structural Changes on Magnetic Resonance Imaging in Axial Spondyloarthritis: Five-year Data From the DESIR Cohort

OBJECTIVE

To test the impact of inflammation on structural changes occurring in the sacroiliac (SI) joints and the spine detected on magnetic resonance imaging (MRI).

METHODS

Patients with early axial spondyloarthritis (SpA) from the Devenir des Spondylarthropathies Indifferérenciées Récentes (DESIR) cohort were included. MRIs of the SI joints (MRI-SI joints) and spine (MRI-spine), obtained at baseline, 2 years, and 5 years, were scored by 3 central readers. Inflammation and structural damage on MRI-SI joints and MRI-spine were defined by the agreement of ≥2 of 3 readers (binary outcomes) and by the average of 3 readers (continuous outcomes). The effect of inflammation (MRI-SI joints/MRI-spine) on damage (MRI-SI joints/MRI-spine, respectively) was evaluated in 2 models: 1) a baseline prediction model (the effect of baseline inflammation on damage assessed at 5 years); and 2) a longitudinal model (the effect of inflammation on structural damage assessed during a 5-year period).

RESULTS

A total of 202 patients were included. Both the presence of bone marrow edema on MRI-SI joints and on MRI-spine at baseline were predictive of 5-year damage (≥3 fatty lesions) on MRI-SI joints (odds ratio [OR] 4.2 [95% confidence interval (95% CI) 2.4, 7.3]) and MRI-spine (OR 10.7 [95% CI 2.4, 49.0]), respectively, when adjusted for C-reactive protein level. The association was also confirmed in longitudinal models (when adjusted for Ankylosing Spondylitis Disease Activity Score) both in the SI joints (OR 5.1 [95% CI 2.7, 9.6]) and spine (OR 15.6 [95% CI 4.8, 50.3]). Analysis of other structural outcomes (i.e., erosions) on MRI-SI joints yielded similar results. In the spine, a significant association was found for fatty lesions but not for erosions and bone spurs, which occurred infrequently over time.

CONCLUSION

We found a predictive and longitudinal association between inflammation detected on MRI and several types of structural damage detected on MRI in patients with early axial SpA, which adds to the evidence for a causal relationship.

Effect of Therapy on Radiographic Progression in Axial Spondyloarthritis: A Systematic Review and Meta-Analysis

OBJECTIVE

To investigate the effect of therapies on radiographic progression in patients with axial spondyloarthritis (SpA).

METHODS

A comprehensive database search for studies assessing radiographic progression in axial SpA (particular treatment versus no treatment of interest) was performed. Study-specific standardized mean differences in treatment outcomes at 2 and ≥4 years were estimated and combined using random-effects models.

RESULTS

Twenty-four studies in patients with axial SpA were identified, of which 18 involved tumor necrosis factor inhibitors (TNFi), 8 involved nonsteroidal antiinflammatory drugs (NSAIDs), and 1 involved secukinumab. Spinal radiographic progression, as measured by the modified Stoke Ankylosing Spondylitis Spine Score (mSASSS), was not significantly different between TNFi-treated and biologics-naive patients at 2 years (mSASSS difference -0.73 [95% confidence interval (95% CI) -1.52, 0.12], I2 = 28%) and ≥4 years (mSASSS difference -2.03 [95% CI -4.63, 0.72], I2 = 63%). Sensitivity analyses restricted to studies with a low risk of bias showed a significant difference in spinal radiographic progression between TNFi-treated and biologics-naive patients at ≥4 years (mSASSS difference -2.17 [95% CI -4.19, -0.15]). No significant difference in spinal radiographic progression was observed between NSAID-treated and control patients (mSASSS difference -0.30 [95% CI -2.62, 1.31], I2 = 71%) or between secukinumab-treated and biologics-naive patients (mSASSS difference -0.34 [95% CI -0.85, 0.17]). With regard to treatment differences in patients with nonradiographic axial SpA or in patients with radiographic progression measured using the sacroiliac joint score, an insufficient number of studies were available for analysis.

CONCLUSION

Although no significant protective effect of TNFi treatment on spinal radiographic progression was seen over the course of 2 years or ≥4 years in patients with axial SpA, our analysis restricted to studies with a low risk of bias showed a protective effect of TNFi after ≥4 years. Therefore, long-term TNFi exposure might confer beneficial effects on spinal radiographic progression in axial SpA. No difference in radiographic progression at 2 years was seen in either the NSAID or secukinumab treatment groups compared to their controls. Future studies should explore the effects of biologic treatment on radiographic progression, as well as the effects of long-term biologics exposure, in patients with early axial SpA or those with nonradiographic axial SpA.

Spinal Radiographic Progression in Early Axial Spondyloarthritis: Five-Year Results From the DESIR Cohort

OBJECTIVE

To analyze the progression of spinal radiographic damage in patients with early axial spondyloarthritis (SpA).

METHODS

Axial SpA patients from the DESIR (Devenir des Spondylarthropathies Indifférenciées Récentes) cohort with 5-year spinal (cervical and lumbar) radiographs available (n = 549) were included. Two- and 5-year modified Stoke Ankylosing Spondylitis Spine Score (mSASSS) progression and development of new syndesmophytes (net change: the number of patients with positive change minus the number of patients with negative change divided by the total number of patients) were assessed in subgroups defined at baseline according to the Assessment of SpondyloArthritis international Society axial SpA criteria and its arms, modified New York criteria (mNYC) and the presence of syndesmophytes.

RESULTS

Mean ± SD mSASSS progression was 0.2 ± 0.9 at 2 years and 0.4 ± 1.8 at 5 years. Five-year progression was higher in the imaging arm (mean ± SD 0.6 ± 2.3), magnetic resonance imaging (MRI)+/mNYC+ (mean ± SD 1.3 ± 4.0), than in the clinical arm only (mean ± SD 0.1 ± 0.7), and highest in patients with syndesmophytes (mean ± SD 2.7 ± 5.0). At 5 years, 7% of all patients had a net change of any new syndesmophyte; this value was 10% for the imaging arm (mNYC+/MRI+ with 18%), 17% for mNYC+ patients, and 42% for patients with syndesmophytes.

CONCLUSION

Spinal radiographic progression, although limited in early axial SpA, can be captured after 2 years. Inflammation and damage in the sacroiliac joint are associated with higher radiographic progression. The presence of baseline syndesmophytes already strongly predicts the development of further structural damage early in the disease.