Antigen-specific immune reactions by expanded CD8+ T cell clones from HLA-B*27-positive patients with spondyloarthritis.

Spondyloarthritis (SpA) is a chronic inflammatory disease that is tightly linked to HLA-B*27 but the pathophysiological basis of this link is still unknown. It is discussed whether either the instability of HLA-B*27 molecules triggers predominantly innate immune reactions or yet unknown antigenic peptides presented by HLA-B*27 induce adaptive autoimmune reactions by CD8+ T cells. To analyze the pathogenesis of SpA, we here investigated the T cell receptor (TCR) usage and whole transcriptomes of CD8+ single cells from synovial fluid of HLA-B*27-positive SpA patients and HLA-B*27-negative controls. In HLA-B*27-positive patients, we confirmed preferential expression of several TCR ß-chain families, found even more restricted usage of particular TCR α-chains, assigned matching TCR αß-chain pairs with homologous CDR3-sequences, and detected identical TCR-chains in different patients. Gene expression analyses by single cell mRNAseq revealed that genes specific for the tissue resident memory phenotype, exhaustion, and apoptosis were particularly highly expressed in expanded clonotypes from HLA-B*27-positive SpA patients. Together, several independent lines of evidence argue in favor of an (auto)antigenic peptide related pathogenesis.

Clinical characteristics of patients with spondyloarthritis and inflammatory bowel disease versus inflammatory bowel disease-related arthritis.

The purpose of this study was to clarify the clinical characteristics of spondyloarthritis (SpA) patients with inflammatory bowel disease (IBD) compared to those without IBD. Furthermore, among patients with SpA and IBD, we aimed to clarify what clinical characteristics lead rheumatologists to diagnose "IBD-related arthritis." Utilizing SpA and psoriatic arthritis (PsA) patients' data from an international, cross-sectional, observational study, we analyzed information on demographics and disease characteristics, dichotomizing patients by IBD status. The presence or absence of IBD was determined based on data collection of treating rheumatologists. Patients with SpA (including PsA) and IBD were also categorized based on treating rheumatologists' definitive diagnosis in regard to SpA type, and compared by whether the patients had IBD-related arthritis or not. Among 4465 SpA patients, 287 (6.4%, 95%CI 5.7-7.2%) were identified with IBD. Compared to SpA patients without IBD, patients with SpA and IBD had a longer diagnostic delay (5.1 vs. 2.9 years, p < 0.001). In patients with SpA and IBD, 111 (38.7%, 95%CI 33.0-44.6%) were diagnosed with IBD-related arthritis. Multivariable analyses showed that HLA-B27 positivity [OR = 0.35, (95%CI 0.15-0.80)], psoriasis [OR = 0.14, (95%CI 0.04-0.50)], IBD as first symptom of SpA [OR = 3.32, (95%CI 1.84-6.01)], and need for IBD-specific treatment [OR = 5.41, (95%CI 2.02-14.50)] were independently associated with the definitive diagnosis of IBD-related arthritis. Collaboration with gastroenterologists is needed to shorten the diagnostic delay in patients with SpA and IBD. The recognition of the factors for the diagnosis of "IBD-related arthritis" may lead to the elucidation of the pathogenesis.

Sustained clinical response and safety of etanercept in patients with early axial spondyloarthritis: 10-year results of the ESTHER trial.

AIMS: Long-term data on TNFi treatment in patients with axSpA is scarce. The objective of this analysis was to assess long-term clinical efficacy of etanercept in early axSpA [including both non-radiographic and radiographic axSpA forms], who participated in the long-term (until year 10) extension of the ESTHER-trial. METHODS: In the previously reported ESTHER-trial, patients with early active axSpA were randomized to treatment with etanercept (n = 40) or sulfasalazine (n = 36) during the first year. Patients in remission discontinued their therapy and were followed up until the end of year 2; in case of remission-loss, etanercept was (re)-introduced and continued until the end of year 10. If remission was not achieved at year 1, patients continued receiving (or were switched to) etanercept for up to 10 years. RESULTS: A total of 19 patients (12 with r-axSpA and 7 with nr-axSpA at baseline) out of the initial 76 patients (= 25%) completed year 10 of the study. In the entire group, a sustained clinical response was seen over 10 years of follow up in the as-observed analysis. Completers were significantly more often male and showed lower values of patient and physician global assessments of disease activity, Ankylosing Spondylitis Disease Activity Score (ASDAS), and Ankylosing Spondylitis Quality of Life questionnaire (ASQoL) scores at baseline as compared with non-completers. When analyzing clinical data of the completers, mean Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) and Bath Ankylosing Spondylitis Functional Index (BASFI) values were constantly below 2 and mean ASDAS below 2.1 during follow up with no statistically significant differences between the r-axSpA and nr-axSpA subgroups. A total of 39 serious adverse events were documented over the 10 years, while six of them were seen as possibly associated with the etanercept treatment, which led in five patients to treatment discontinuation. CONCLUSION: A sustained clinical response was observed over the 10 years of the study with comparable response and drop-out rates between r-axSpA and nr-axSpA. Etanercept was well tolerated across the entire treatment period and showed a good safety profile with no new safety signals.

Deep learning for detection of radiographic sacroiliitis: achieving expert-level performance.

BACKGROUND: Radiographs of the sacroiliac joints are commonly used for the diagnosis and classification of axial spondyloarthritis. The aim of this study was to develop and validate an artificial neural network for the detection of definite radiographic sacroiliitis as a manifestation of axial spondyloarthritis (axSpA). METHODS: Conventional radiographs of the sacroiliac joints obtained in two independent studies of patients with axSpA were used. The first cohort comprised 1553 radiographs and was split into training (n = 1324) and validation (n = 229) sets. The second cohort comprised 458 radiographs and was used as an independent test dataset. All radiographs were assessed in a central reading session, and the final decision on the presence or absence of definite radiographic sacroiliitis was used as a reference. The performance of the neural network was evaluated by calculating areas under the receiver operating characteristic curves (AUCs) as well as sensitivity and specificity. Cohen's kappa and the absolute agreement were used to assess the agreement between the neural network and the human readers. RESULTS: The neural network achieved an excellent performance in the detection of definite radiographic sacroiliitis with an AUC of 0.97 and 0.94 for the validation and test datasets, respectively. Sensitivity and specificity for the cut-off weighting both measurements equally were 88% and 95% for the validation and 92% and 81% for the test set. The Cohen's kappa between the neural network and the reference judgements were 0.79 and 0.72 for the validation and test sets with an absolute agreement of 90% and 88%, respectively. CONCLUSION: Deep artificial neural networks enable the accurate detection of definite radiographic sacroiliitis relevant for the diagnosis and classification of axSpA.

Assessment of radiographic sacroiliitis in anteroposterior lumbar vs conventional pelvic radiographs in axial spondyloarthritis.

OBJECTIVE: The aim was to investigate the reliability and validity of radiographic sacroiliitis assessment in anteroposterior (AP) lumbar radiographs compared with conventional pelvic radiographs in patients with axial spondyloarthritis (axSpA). METHODS: Patients from the German Spondyloarthritis Inception Cohort were selected based on the availability of pelvic and AP lumbar radiographs with visible SI joints at baseline and year 2. Two readers scored the images independently in a random order according to the modified New York criteria. The sacroiliitis sum score was calculated as the mean of both readers. Patients were classified as radiographic (r-)axSpA if radiographic sacroiliitis of grade ≥2 bilaterally or grade ≥3 unilaterally was present in the opinion of both readers and as non-radiographic (nr-)axSpA otherwise. The reliability and validity of sacroiliitis assessment in AP lumbar radiographs was assessed using intraclass correlation coefficients (ICCs), absolute agreement and κ statistics. RESULTS: A total of 226 sets of radiographs were scored from 113 patients included in the study. The ICC for the sacroiliitis sum score was 0.91 at both baseline and year 2. A total of 62 (54.9%) and 55 (48.7%) patients were classified as r-axSpA at baseline and 65 (57.5%) and 60 (53.1%) patients at year 2 based on evaluation of pelvic and AP lumbar radiographs, respectively. The absolute agreement between the methods on the classification was 84.9 and 85.0% at baseline and year 2, respectively, with the κ of 0.70 at both time points. CONCLUSION: Radiographic sacroiliitis can be assessed in AP lumbar radiographs with a similar reliability to conventional pelvic radiographs.

Comparison of an online self-referral tool with a physician-based referral strategy for early recognition of patients with a high probability of axial spa.

OBJECTIVES: The diagnostic delay in axial spondyloarthritis (axial SpA) remains unacceptably high, with one of the reasons being a late referral. Structured physician-based referral programs are able to improve early diagnosis, but lack of implementation is still an issue. The objective of this study was to evaluate an online self-referral (OSR) tool for patients with back pain and to compare it to an established physician-based referral tool. METHODS: Patients with back pain were included if they either fulfilled the requirements of the OSR tool or were referred by a physician using the Berlin referral tool. Rheumatologists in the specialized center performed a structured assessment in all patients that resulted in the final diagnosis of axial SpA / no axial SpA. Furthermore, we attempted to optimize the OSR tool in terms of maximizing the specificity constrained by a sensitivity of at least 90% of the original strategy. RESULTS: 361 consecutive patients (180 via the OSR and 181 via the Berlin referral tool) were included in the study. A total of 35 patients (19.4%) in the self-referral group and 71 patients (39.2%) in the physician-referral group were finally diagnosed with axial SpA. Axial SpA patients from the OSR group were more often HLA-B27 negative, females, and were more frequently at a non-radiographic stage as compared to axial SpA patients who came via the physician-based tool. Both groups had, however, a similar disease burden. According to the pre-defined selection criterion we identified an optimized combination of ≥2 IBP parameters and ≥1 other SpA parameters (in addition to both stem parameters). CONCLUSIONS: Despite the better performance of the physician-based referral strategy, the proportion of axial SpA among self-referred patients (19.4%) was clearly higher than the assumed 5% prevalence of axial SpA in patients with chronic back pain. Based on our data driven approach the performance of the OSR strategy could be further improved if at least two IBP parameters plus one additional SpA parameter had to be present in addition to the stem parameters. The OSR tool can be used in specialized centers in addition to a physician-based referral strategy to improve early diagnosis and to increase awareness of axial SpA.

Clinical and imaging characteristics of osteitis condensans ilii as compared with axial spondyloarthritis.

OBJECTIVES: Osteitis condensans ilii (OCI) has become an important differential diagnosis for axial spondyloarthritis (axSpA). The objective of this matched case-control study was to investigate demographic, clinical, laboratory and MRI characteristics of OCI as compared with axial spondyloarthritis (axSpA). METHODS: A total of 60 patients diagnosed with OCI were included in the final analysis. From 27 of these patients, MRIs of the sacroiliac joints were available. OCI patients were matched with a 1:1 ratio by back pain duration to patients with definite axSpA in order to compare clinical, laboratory and MRI characteristics. RESULTS: The OCI patients were nearly all females (96.7 vs 46.7%), had a significantly lower prevalence of inflammatory back pain (39.5 vs 88.9%), a significantly lower percentage of HLA-B27 positives (35.2 vs 80.0%) and a lower prevalence of the majority of other SpA features as compared with axSpA patients. Interestingly, there was no difference in the prevalence of osteitis in the sacroiliac joints (92.6 vs 85.2% in OCI and axSpA, respectively, P = 0.44), but there was a difference in the prevalence of erosions (7.4 vs 66.7%, respectively, P = 0.0001). In addition, in OCI nearly all lesions were localized in the anterior part of the sacroiliac joints while in axSpA lesions were localized predominantly in the middle part of the joint (for osteitis: 96 vs 4% in OCI and 28.6 vs 71.4% in axSpA; P = 0.0002 for the inter-group difference). CONCLUSION: Clinical and imaging features of OCI compared with axSpA are described that should help in differential diagnosis.