It is widely recognized that treatment intervention in patients with rheumatoid arthritis (RA) at early stages is associated with improved outcomes.1 Hence, research efforts have been devoted to investigation of whether intervention starting at the preclinical stage will prevent the clinical onset of RA.2-5 In these studies, individuals at risk for RA were identified mainly by the detection of anti-citrullinated protein antibodies (ACPAs) and/or the rheumatoid factor (RF), which may precede the clinical symptoms of RA patients.6 The outcomes of these prevention studies ranged from no effect to a delay in the onset of RA in some individuals.2-5 Type 1 diabetes (T1D), another autoimmune disease, has a similar natural course to RA. Autoantibodies are present before the onset of clinical T1D, and these autoantibodies can be used to screen those at risk. Prevention of T1D onset has been achieved in 50% of at-risk individuals who received a monoclonal antibody to CD3, teplizumab, compared to 22% of those who received a placebo after a median follow-up of 742 days.7, 8 Based on these results, the US Food and Drug Administration (FDA) has approved teplizumab for use in preventing T1D in at-risk populations. As the success in preventing T1D is being evaluated, there are lessons that can be applied to screening at-risk individuals and preventing the onset of RA. The first important step is to develop an assay that can detect multiple autoantibodies, thereby increasing the efficiency of screening. In addition to ACPA and RF, anti-carbamylated protein antibodies (ACarPAs) have also been found in the preclinical phase of patients who eventually developed RA,9 and the presence of all three antibodies (ACPA, RF, and ACarPA) in at-risk individuals has an even stronger predictive value for the onset of RA.10 Furthermore, anti-acetylated protein antibodies (AAPAs) have been discovered as a group of autoantibodies specific to RA, adding value in distinguishing early RA from other undifferentiated inflammatory arthritis.11-13 Although it has not yet been investigated, it is expected that AAPAs will be present in the preclinical phase of patients with RA. In T1D screening, a high-throughput multiplex electrochemiluminescence (ECL) assay has been successfully applied to measure six anti-islet autoantibodies in the Autoimmunity Screening for Kids (ASK) program. This assay is highly efficient, low cost, and uses a low volume of serum.14 Currently, each of these RA-associated autoantibodies is detected in separate single assays, which is not cost-effective for screening use. A similar multiplex ECL assay to detect ACPAs, ACarPA, AAPAs, and RF in a single assay is desirable and yet to be developed to create a more efficient and cost-effective screening program for at-risk individuals of RA, as well as for dynamic monitoring of changes in these autoantibodies. Second, it is important to thoroughly investigate the binding activity and daynamically follow-up on levels of autoantibodies.15 For example, the anti-zinc transporter 8 (ZnT8A) antibody is widely used as a biomarker for predicting T1D, but only those with high-affinity anti-ZnT8A have predictive value.16 Similarly, not all ACPAs have the same predictive value for RA. The variable domain glycosylation (VDG) of ACPA IgG is increased before RA onset, and increased VDG positively correlates with ACPA IgG levels and the binding breadth of citrullinated protein epitopes.17 Therefore, measuring the VDG of ACPAs dynamically may improve the predictive value of ACPA in the at-risk population.18 It is yet to be determined whether increased VDG is associated with increased ACPA binding avidity to citrullinated proteins in vivo and whether VDG also occurs in ACarPA and AAPAs. It is of great interest to know if VDG renders ACPAs more pathogenic, although this has not yet been tested. Third, genetic information should be integrated into the autoantibody screening scheme. In some T1D screening programs, susceptible genetic screening is included along with anti-islet autoantibodies in the general population.19 Similarly, for RA, the major histocompatibility complex (MHC) DRB1 locus associated with RA can be integrated into the screening with multiple autoantibodies. The MHC DRB1 locus remains the strongest association among multiple genes associated with RA.20 A previous study has shown that carrying the shared epitope MHC DRB1 locus, along with the presence of ACPAs and/or RF, has a very high relative risk for future development of RA. This increased risk is thought to be due to the synergy between the MHC DRB1 locus and autoantibodies.21 Investigations are needed to determine whether the combination of other autoantibodies and the MHC DRB1 locus will further increase the prediction value in at-risk populations. Fourth, in the T1D prevention study with teplizumab, the administration of teplizumab in terms of dosage and duration is unique: daily with escalating dosing and for a total of 14 days.7, 8 This regimen is distinctively different from what has been tried in RA prevention studies.2-5 The protective effect of teplizumab in T1D is attributed to its induction of partially exhausted CD8+ T cells.8 It is not known if the mechanism of action of teplizumab against CD3 in T1D will also apply in individuals at risk for RA but should be explored. Fifth, a more aggressive approach to immunosuppression is being explored in the prevention of T1D through a clinical trial (NCT03929601), where rituximab and abatacept will be administered. In individuals at risk for T1D, rituximab will be administered via intravenous infusion at a dose of 375 mg/m2 weekly for 4 weeks, followed by subcutaneous injections of abatacept weekly for 2 years. While targeting B cells with rituximab or T cells with abatacept has been tested separately in individuals at risk for RA,2, 5 sequential targeting of both B and T cells has not been investigated. If it is proven safe in those at risk for T1D, a trial of sequential targeting of B and T cells with rituximab and abatacept would be a valuable approach to explore for at-risk individuals for RA. Last, for the implementation of preventive scheme in at-risk individuals, it is crucial to develop an accurate prediction model to estimate a risk at a personal level. Subjects with arthralgia and ACPAs and/or RF will have approximately a 30% chance of developing RA within 1 year.22 Thus, not all of those in the preclinical phase or at risk of RA will progress to clinical RA. However, building an accurate prediction model for risk of RA is challenging and becoming sophisticated because of the complexity of the interplay between genetic factors, environmental risk factors, and the host immune response.23 By the same token, a similar challenge also exists in predicting the development of T1D in at-risk children. However, recent advances in artificial intelligence have shown promising results in constructing prediction models for both RA and T1D. For instance, by integrating anti-islet autoantibody status and other risk factors in longitudinal follow-up, Kwon et al. revealed distinct autoimmune trajectories that progress to clinical T1D in at-risk individuals.24 Similarly, by combining autoantibody status (ACPA and RF positive) and serum proteomic information, O'Neil and colleagues found that in first-degree family members of RA patients, Janus kinase-signal transducer and activator of transcription (JAK-STAT) signaling is upregulated for those at the highest risk of developing clinical RA.25 Since many etiologic and pathophysiologic features of RA and T1D overlap, it is possible that some components for the construction of a prediction model may be shared for the development of a clinically applicable prediction model.26 However, further research is necessary to determine the extent to which the two diseases share common predictive factors and to develop accurate and reliable predictive models for each disease. The progress made in the prevention of clinical onset of T1D in at-risk individuals represents a significant step forward in the field of autoimmune disease prevention.7, 8 While the population at risk for T1D differs from that for RA, the screening strategies and immune-based interventions used for T1D prevention can be adapted to develop effective prevention strategies for RA. Cong-Qiu Chu wrote and revised the manuscript. Cong-Qiu Chu's work is supported by an Innovative Research Award by Rheumatology Research Foundation and by VA Merit Review grant (I01BX005195). Cong-Qiu Chu is employee of Innovent Biologics (USA). Not applicable. Not appliable.