Machine learning identification of thresholds to discriminate osteoarthritis and rheumatoid arthritis synovial inflammation

Advancing precision rheumatology: applications of machine learning for rheumatoid arthritis management

Rheumatoid arthritis (RA) is an autoimmune disease causing progressive joint damage. Early diagnosis and treatment is critical, but remains challenging due to RA complexity and heterogeneity. Machine learning (ML) techniques may enhance RA management by identifying patterns within multidimensional biomedical data to improve classification, diagnosis, and treatment predictions. In this review, we summarize the applications of ML for RA management. Emerging studies or applications have developed diagnostic and predictive models for RA that utilize a variety of data modalities, including electronic health records, imaging, and multi-omics data. High-performance supervised learning models have demonstrated an Area Under the Curve (AUC) exceeding 0.85, which is used for identifying RA patients and predicting treatment responses. Unsupervised learning has revealed potential RA subtypes. Ongoing research is integrating multimodal data with deep learning to further improve performance. However, key challenges remain regarding model overfitting, generalizability, validation in clinical settings, and interpretability. Small sample sizes and lack of diverse population testing risks overestimating model performance. Prospective studies evaluating real-world clinical utility are lacking. Enhancing model interpretability is critical for clinician acceptance. In summary, while ML shows promise for transforming RA management through earlier diagnosis and optimized treatment, larger scale multisite data, prospective clinical validation of interpretable models, and testing across diverse populations is still needed. As these gaps are addressed, ML may pave the way towards precision medicine in RA.

Predicting Rheumatoid Arthritis Development Using Hand Ultrasound and Machine Learning-A Two-Year Follow-Up Cohort Study

BACKGROUND

The early diagnosis and treatment of rheumatoid arthritis (RA) are essential to prevent joint damage and enhance patient outcomes. Diagnosing RA in its early stages is challenging due to the nonspecific and variable clinical signs and symptoms. Our study aimed to identify the most predictive features of hand ultrasound (US) for RA development and assess the performance of machine learning models in diagnosing preclinical RA.

METHODS

We conducted a prospective cohort study with 326 adults who had experienced hand joint pain for less than 12 months and no clinical arthritis. We assessed the participants clinically and via hand US at baseline and followed them for 24 months. Clinical progression to RA was defined according to the ACR/EULAR criteria. Regression modeling and machine learning approaches were used to analyze the predictive US features.

RESULTS

Of the 326 participants (45.10 ± 11.37 years/83% female), 123 (37.7%) developed clinical RA during follow-up. At baseline, 84.6% of the progressors had US synovitis, whereas 16.3% of the non-progressors did (p < 0.0001). Only 5.7% of the progressors had positive PD. Multivariate analysis revealed that the radiocarpal synovial thickness (OR = 39.8), PIP/MCP synovitis (OR = 68 and 39), and wrist effusion (OR = 12.56) on US significantly increased the odds of developing RA. ML confirmed these US features, along with the RF and anti-CCP levels, as the most important predictors of RA.

CONCLUSIONS

Hand US can identify preclinical synovitis and determine the RA risk. The radiocarpal synovial thickness, PIP/MCP synovitis, wrist effusion, and RF and anti-CCP levels are associated with RA development.

Inflammation in osteoarthritis: Our view on its presence and involvement in disease development over the years

Inflammation, both locally in the joint and systemic, is nowadays considered among the mechanisms involved in osteoarthritis (OA). However, this concept has not always been generally accepted. In fact, for long OA has been described as a relatively simple degeneration of articular cartilage as the result of wear and tear only. In this narrative review, we present what our understanding of OA was at the time of the inaugural release of Osteoarthritis and Cartilage about 30 years ago and discuss a set of pivotal papers that changed our view on the role of inflammation in OA development. Furthermore, we briefly discuss the current view on the involvement of inflammation in OA. Next, we use the example of transforming growth factor-β signaling to show how inflammation might influence processes in the joint in a manner that is beyond the simple interaction of ligand and receptor leading to the release of inflammatory and catabolic mediators. Finally, we discuss our view on what should be done in the future to bring the field forward.

The presence of Mott cells in the lymph nodes of rats with experimental autoimmune encephalomyelitis

Mott cells are plasma cells that have multiple spherical Russell bodies packed in their cytoplasm. Russell bodies are dilated endoplasmic reticulum cisternae filled with aggregates of immunoglobulins that are neither secreted nor degraded. Mott cells were observed in our study by light and electron microscope in the lymph nodes of rats with experimental autoimmune encephalomyelitis (EAE), an animal model of multiple sclerosis. Mott cells were detected on hematoxylin and eosin (HE)-stained lymph node sections as vacuolated cells with eccentrically positioned nuclei and large number of faint blue spherical inclusions in the cytoplasm. Electron microscopic investigation revealed the presence of Russell bodies of the "medusa" form inside Mott cells in lymph node ultra-thin sections of EAE animals. Mott cells expressed the plasma cell marker CD138 and either kappa or lambda immunoglobulin light chains, indicating their origin from polyclonally activated B cells. Finally, Mott cells were associated with active EAE, as they were not found in the lymph nodes of EAE-resistant Albino Oxford rats. The presence of Russell bodies implies an excessive production of immunoglobulins in EAE, thus further emphasizing the role of B cells, and among them Mott cells, in the pathogenesis of this animal model of multiple sclerosis.

Mimicking the Human Articular Joint with In Vitro Model of Neurons-Synoviocytes Co-Culture

The development of in vitro models is essential in modern science due to the need for experiments using human material and the reduction in the number of laboratory animals. The complexity of the interactions that occur in living organisms requires improvements in the monolayer cultures. In the work presented here, neuroepithelial stem (NES) cells were differentiated into peripheral-like neurons (PLN) and the phenotype of the cells was confirmed at the genetic and protein levels. Then RNA-seq method was used to investigate how stimulation with pro-inflammatory factors such as LPS and IFNγ affects the expression of genes involved in the immune response in human fibroblast-like synoviocytes (HFLS). HFLS were then cultured on semi-permeable membrane inserts, and after 24 hours of pro-inflammatory stimulation, the levels of cytokines secretion into the medium were checked. Inserts with stimulated HFLS were introduced into the PLN culture, and by measuring secreted ATP, an increase in cell activity was found in the system. The method used mimics the condition that occurs in the joint during inflammation, as observed in the development of diseases such as rheumatoid arthritis (RA) or osteoarthritis (OA). In addition, the system used can be easily modified to simulate the interaction of peripheral neurons with other cell types.

Broken-fat pad sign: a characteristic radiographic finding to distinguish between knee rheumatoid arthritis and osteoarthritis

OBJECTIVES

Diagnostic imaging plays an important role in the pre-treatment workup of knee osteoarthritis (OA) and rheumatoid arthritis (RA). Herein, we identified a useful MRI sign of infrapatellar fat pad (IPFP) to improve diagnosis.

METHODS

Eighty-one age- and sex-matched RA and OA patients each, with pathological diagnosis and pre-treatment MRI were retrospectively evaluated. All randomized MR images were blinded and independently reviewed by two radiologists. The assessment process included initial diagnosis, sign evaluation, and final diagnosis, with a 3-week interval between each assessment. Broken-fat pad (BFP) sign was assessed on sagittal T2-weighted-imaging in routine MRI. The area under the curve and Cohen's kappa (κ) were used to assess the classification performance. Two shape features were extracted from IPFP for quantitative interpretation.

RESULTS

The median age of the study population was 57.6 years (range: 31.0-78.0 years). The BFP sign was detected more frequently in patients with RA (72.8%) than those with OA (21.0%). Both radiologists achieved better performance by referring to the BFP sign, with accuracies increasing from 58.0 to 75.9% and 72.8 to 79.6%, respectively. The inter-reader correlation coefficient showed an increase from fair (κ = 0.30) to substantial (κ = 0.75) upon the consideration of the BFP sign. For quantitative analysis, the IPFP of RA had significantly lower sphericity (0.54 ± 0.04 vs. 0.59 ± 0.03, p < 0.01). Despite larger surface-volume-ratio of RA (0.38 ± 0.05 vs. 0.37 ± 0.04, p = 0.25) than that of OA, there was no statistical difference.

CONCLUSIONS

The BFP sign is a potentially important diagnostic clue for differentiating RA from OA with routine MRI and reducing misdiagnosis.

CRITICAL RELEVANCE STATEMENT

With the simple and feasible broken-fat pad sign, clinicians can help more patients with early accurate diagnosis and proper treatment, which may be a valuable addition to the diagnostic workup of knee MRI assessment.

KEY POINTS

• Detailed identification of infrapatellar fat pad alterations of patients may be currently ignored in routine evaluation. • Broken-fat pad sign is helpful for differentiating rheumatoid arthritis and osteoarthritis. • The quantitative shape features of the infrapatellar fat pad may provide a possible explanation of the signs. • This sign has good inter-reader agreements and is feasible for clinical application.

Machine learning application in autoimmune diseases: State of art and future prospectives

Autoimmune diseases are a group of disorders resulting from an alteration of immune tolerance, characterized by the formation of autoantibodies and the consequent development of heterogeneous clinical manifestations. Diagnosing autoimmune diseases is often complicated, and the available prognostic tools are limited. Machine learning allows us to analyze large amounts of data and carry out complex calculations quickly and with minimal effort. In this work, we examine the literature focusing on the use of machine learning in the field of the main systemic (systemic lupus erythematosus and rheumatoid arthritis) and organ-specific autoimmune diseases (type 1 diabetes mellitus, autoimmune thyroid, gastrointestinal, and skin diseases). From our analysis, interesting applications of machine learning emerged for developing algorithms useful in the early diagnosis of disease or prognostic models (risk of complications, therapeutic response). Subsequent studies and the creation of increasingly rich databases to be supplied to the algorithms will eventually guide the clinician in the diagnosis, allowing intervention when the pathology is still in an early stage and immediately directing towards a correct therapeutic approach.