Advances in arthritis diagnosis – from biomarkers to AI

By Dr Gehad ElGhazali
MD, PhD, Consultant Physician
SEHA Hospital – SKMC AUH Hospital Laboratory, PureLab

Incidence

Arthritis, a prevalent health issue among people worldwide, afflicts over 350 million individuals and stands as a prominent factor leading to disability.¹ Arthritis poses a significant challenge in the United Arab Emirates (UAE), with research revealing a one-year delay in diagnosis or screening due to low awareness. The condition affects approximately 20% of the population and carries a considerable economic burden on the nation.²

Despite recent advances in interpreting the pathophysiology of the disease, treating arthritis remains a significant problem. Nonetheless, research indicates that early detection of any arthritis can save individuals from the critical consequences of the condition.³

Current diagnostic scenario

The current method of diagnosing arthritis uses several bodily fluids to identify the possible type of arthritis. Commonly determined fluids include joint fluid, blood, and urine. Additionally, imaging techniques such as X-rays and magnetic resonance imaging are employed to confirm the specific arthritis type.³

Antinuclear antibody (ANA): The presence of ANAs and their specific subtypes raises the probability of a systemic autoimmune disorder. The test is widely considered a diagnostic tool and a potential serum marker that could improve the clinical identification of many autoimmune diseases, such as systemic lupus erythematosus, mixed connective tissue disease, Sjogren’s disease, and polymyositis.³

Anti-cyclic citrullinated peptide (anti-CCP) antibodies are specific markers for diagnosing rheumatoid arthritis (RA) and related autoimmune disorders. Their presence in blood tests helps in the early identification and diagnosis of RA, allowing for timely intervention and management.⁴

C-Reactive Protein (CRP) test: CRP is routinely measured as a significant marker of systemic inflammation in RA, psoriatic arthritis, and ankylosing spondylitis.⁴

Synovial fluid analysis: Synovial fluid analysis detects elevated white blood cells and inflammatory markers, aiding in arthritis diagnosis. It distinguishes between osteoarthritis, gout, and infectious arthritis, providing insights into the cause of joint symptoms. ⁴

Human Leukocyte Antigen (HLA) tissue typing: HLA tissue typing is a genetic test for arthritis diagnosis. It identifies specific HLA genes associated with autoimmune forms of arthritis such as Behcet’s disease, ankylosing spondylitis, and RA.⁴

Purelab provides a wide range of standard diagnostic procedures to diagnose arthritis. Purelab offers an extensive array of diagnostic testing to identify the many forms and stages of arthritis, ranging from genetic testing, such as HLA tissue typing, to serum biomarker testing, like the antinuclear antibody test.

The diagnosis of arthritis incorporates a range of laboratory and imaging techniques, yet the absence of established diagnostic criteria or a definitive gold standard test poses a significant challenge.^3,4,5

Innovative approaches in diagnostic assessment have emerged to address these challenges. The utilization of artificial intelligence (AI) has enhanced diagnostic precision for arthritis. These advancements offer promising avenues for more accurate and detailed assessments of joint structures, contributing to improved diagnosis and management of the condition.

Cutting-edge research

AI in the diagnosis of RA
AI has provided promising evidence in the screening, diagnosis, and management of RA, showcasing its potential to revolutionize the field. Here are a few of the tool’s highlights:

• Assessing RA Development Risk: AI models are utilized to analyze various factors and predict the risk of developing RA in individuals.
• Diagnosing RA with Omics, Imaging, Clinical, and Sensor Data: AI systems integrate data from omics (genomics, proteomics, etc.), medical imaging, clinical records, and sensor data to assist in accurate diagnosis of the condition.
• Detecting Patients Within Electronic Health Records (EHR): AI algorithms can sift through vast amounts of EHR to identify individuals exhibiting signs or symptoms of RA.
• Determining Prognosis: Using machine learning, AI can assess various factors to predict the likely course of a condition in everyone, helping clinicians plan long-term care strategies.⁵

Role of novel biomarkers in the diagnosis and treatment of RA

The research identified four promising and novel biomarkers for RA, including CRTAM, PTTG1IP, ITGB2, and MMP13. These biomarkers could serve as potential diagnostic and treatment targets for RA.⁶ The study’s highlights include the following:

Diagnostic performance: The four biomarkers constructed a nomogram that distinguished and diagnosed RA from normal tissues satisfactorily. The nomogram achieved a high area under the curve (AUC) in both the training (AUC = 0.894) and testing (AUC = 0.843) cohorts.

The heterogeneity of RA: The study identified two distinct subtypes of RA patients based on the four biomarkers. This finding suggests that these biomarkers may be involved in the immune infiltration process and contribute to the heterogeneity of RA.

Conclusion

The advancements in AI approach hold immense promise in revolutionizing the diagnosis and treatment of arthritis. AI offers improved risk assessment and personalized care for RA. Future research should address validation challenges and ethical considerations to ensure these innovations benefit patients globally.

References:

1. Global RA Network. About Arthritis and RA. https://globalranetwork.org/project/disease-info/
2. The Emirates Arthritis foundation | About MEAF. Available from: https://arthritis.ae/about-meaf/
3. Ledingham J, Snowden N, Ide Z. Diagnosis and early management of inflammatory arthritis. BMJ. 2017 Jul 27; 358: j3248
4. Heidari B. Rheumatoid Arthritis: Early diagnosis and treatment outcomes. Caspian J Intern Med. 2011 Winter; 2(1): 161-70
5. Momtazmanesh S, Nowroozi A, Rezaei N. Artificial Intelligence in Rheumatoid Arthritis: Current Status and Future Perspectives: A State-of-the-Art Review. Rheumatol Ther. 2022 Oct; 9(5): 1249-1304. doi: 10.1007/s40744-022-00475-4.
6. Liu F, Ye J, Wang S, Li Y, Yang Y, Xiao J, et al. Identification and Verification of Novel Biomarkers Involving Rheumatoid Arthritis with Multimachine Learning Algorithms: An In Silicon and In Vivo Study. Mediators of Inflammation. 2024 Feb 14 [cited 2024 Mar 1]; 2024: e3188216. https://www.hindawi.com/journals/mi/2024/3188216