A polygenic risk score (PRS) that assesses the genetic risk of rheumatoid arthritis in individual patients: Tokyo Medical and Dental University and Tokyo Women's Medical University blazing a new trail in genomic medicine

Originally, the main target of personalized genomic medicine was single-factor diseases caused by singular genetic mutations. However, it is now possible to predict the risk in multifactorial diseases such as rheumatoid arthritis and diabetes by scoring the accumulation of risk genes. As part of a joint research project with the RIKEN Center for Integrative Medical Sciences, the research group of Professor Yuta Kochi of the Medical Research Institute of Tokyo Medical and Dental University and Suguru Honda, a doctoral student at Tokyo Women's Medical University, have, for the first time, developed a polygenic risk score (PRS) that is associated with the progression of radiographic images in rheumatoid arthritis. Professor Kochi says," This is the first step in creating personalized medicine using genomic information for rheumatoid arthritis." The PRS was published in the online version of Arthritis & Rheumatology.

Rheumatoid arthritis is an immune disorder that results in the deterioration of articular bone and cartilage. Despite rapid advances in treatment in recent years, some patients still have progressive joint deterioration. Prediction of these risks may allow the selection of treatment intensity tailored for individual patients to more effectively reduce the progression and severity of joint deterioration. Anti-citrullinated protein antibody (ACPA) and the human leukocyte antigen (HLA) region were previously shown to be associated with joint deterioration; however, their predictive accuracies were inadequate.

PRS is a method of scoring the accumulation of genetic risks in individual patients to predict disease onset and progression. Studies using this score have been actively conducted in recent years. The research group examined whether the PRS was a predictive indicator of joint deterioration. Comparisons of PRS between groups with and without advanced joint deterioration showed the difference in scores, which was especially large in the younger-onset group aged 40 years or younger. In addition, quantification of this score using the receiver operating characteristic (ROC) curves to classify joint deterioration showed its increased ability to classify joint deterioration at a younger age of onset.

When patients were divided into five groups according to PRS to compare the risk of being classified as having severely advanced disease, patients with the highest PRS had approximately twice the risk of being classified with advanced disease compared to those with the lowest PRS. When the analysis was restricted to the younger age group, the highest PRS group had approximately five times the risk of being classified with advanced disease compared to the lowest group. Furthermore, all-patient analyses comparing ACPA and HLA regions showed that the PRS was as discriminatory as the ACPA and superior to ACPA and HLA regions in the younger-onset group. Finally, a multivariate analysis was conducted using a model of logistic regression to determine whether combining PRS with other factors (age-at-onset, sex, smoking history, ACPA, rheumatoid factor, body mass index (BMI), periodontal disease, drugs for rheumatoid arthritis, HLA-domain) improved the discriminatory power.

The results showed that the PRS was an independent risk factor associated with the progression of joint deterioration, with improved discriminatory power in combination with other factors. "PRS analysis is a global trend and indeed, we've obtained clear data. This is thanks to the creation of a more robust clinical database by the doctors in Japanese hospitals. We are now working to improve our prediction accuracy using AI," said Dr. Honda.