Pre-arthritis stages of rheumatoid arthritis: from pathology to precision prevention Authors: A H M van der Helm-van Mil, K Raza Published: 2026 Over the last decade, the outlook for patients with rheumatoid arthritis (RA) has improved, but there is growing interest in preventing RA by treating individuals in pre-arthritic phases of disease. Although clinical trials in people at increased risk of RA have shown promising initial results, complete prevention of RA is not yet possible and several questions remain unanswered. Just as RA is a heterogeneous disease, consisting of at least autoantibody-positive and autoantibody-negative subsets, and perhaps additional subsets within this dichotomy, the pre-RA phase also appears heterogeneous. A range of different genetic and environmental risk factors (including smoking and professional dust) have been implicated. Clinical phenotypes in the pre-arthritis phases vary, with some people experiencing significant symptoms for prolonged periods of time, in some cases associated with imaging evidence of tenosynovitis and synovitis, while in other people clinically apparent arthritis develops without a prolonged clinical prodrome. However, not everyone with risk factors and symptoms of clinically suspect arthralgia (CSA) actually develops RA. The development of RA, at least in most people, occurs over years, with various processes occurring sequentially and the risk of RA gradually increasing over time. Yet, all observational studies and clinical trials show that a large proportion of people at risk ultimately do not develop clinically apparent synovial inflammation. Understanding which processes are at play at which points during the pre-RA phase is an important step towards precision prevention, where the goal is to alter the deleterious environmental events or to administer the right medication to the right patient at the right time in the at-risk setting, thereby reducing symptom burden in the at-risk phase and/or preventing RA or modifying the trajectory of the RA that does eventually develop so that its articular and extra-articular impacts are lessened. The articles in this thematic collection contribute to the ultimate goal of precision prevention of RA. First, Kastbom et al provide a timely overview of immune dysregulation in the pathogenesis of RA,1 including the possibility of a breach of self-tolerance at mucosal surfaces, abnormalities of the adaptive immune system in the systemic compartment and the infiltration of the synovium by T cells in the pre-RA phase. While Kastbom et al focus on abnormalities of the adaptive immune system in the development of autoantibody-positive RA,1 this collection also includes two studies on synovial fibroblasts in anticitrullinated peptide antibodies (ACPA)-positive, at-risk patients. Atzinger et al used positron emission tomography-CT (PET-CT) to visualise activated fibroblasts (expressing fibroblast activation protein (FAP)). They demonstrated the presence of the FAP-expressing activated fibroblasts in joints in the pre-RA phase, which was more prominent in people who developed RA over time.2 While previous studies have suggested that subclinical inflammation initially occurs in the small joints, particularly of the hand,3 this study by Atzinger et al suggests that fibroblasts in large joints including the knee are also dysregulated early in disease development. Although the value of PET as a prognostic tool remains uncertain, this observation is valuable for understanding disease pathogenesis and is reminiscent of earlier research in which knee biopsies were taken from ACPA-positive individuals in the pre-RA phase, where histological findings were also already present before clinically apparent arthritis occurred.4 Another study from this topic collection further explores the synovium in the pre-RA phase, demonstrating DNA damage not only in synovial T cells but also in fibroblast-like synovial cells.5 Thus, these two articles suggest that activated fibroblasts play a role in the early stages of rheumatoid synovitis. Kastbom et al described the different stages during the development of ACPA-positive RA and that the transitions between these stages likely have their own risk factors, with most studies to date focused on environmental exposure during adulthood. It has previously been reported that smoking primarily plays a role in the asymptomatic risk phase, for example, by promoting the development of autoantibodies and the progression of the autoantibody response, while shared human leucocyte antigen epitopes are not a risk factor for the development of autoantibodies but are associated with the subsequent maturation of the autoimmune response.6 Dusser et al explored childhood exposures focusing on pets (cats/dogs) and farming environments and reported that exposure to pets in early life (specifically between the ages of 1 and 2 years) might protect against RA development. Further assessment of environmental exposure, including to pets and other animals across the life course (including in early childhood), is warranted.7 Precision prevention requires not only molecular understanding but also proper risk stratification. European Alliance of Associations for Rheumatology/American College of Rheumatology risk stratification criteria for RA development have recently been published for people with arthralgia.8 Such criteria are not yet available for groups of people for whom the presence of symptoms is not their defining characteristic. In this context, Aymon et al’s study in first-degree relatives of people with RA is of great importance.9 1.7% of ACPA-positive first-degree relatives developed RA after an average of 7 years. They clearly show that the presence of autoantibodies (rheumatoid factor in particular) and the presence of CSA symptoms are good predictors for the later development of RA. However, predicting at the individual level in this setting remains difficult, given the low positive predictive value, which is likely related to the low frequency of the endpoint. This study is a major achievement and a step forward in identifying risk of future disease in first-degree relatives. Another interesting finding was that 41% of the people in this study who developed RA were seronegative. Seronegative RA also has a pre-RA phase, and understanding the characteristics of this is just as important as seropositive RA, but potentially even more challenging because autoantibodies are, by definition, absent.9 In addition, an unexplored area is the development of seronegative RA in relatives of seropositive RA. Finally, treatment is a crucial element of precision prevention. This is a completely new field requiring new measurement methods. Using data from the TREAT EARLIER trial10 of secondary prevention with methotrexate, Claassen et al investigated two different issues. First, they assessed what improvement in function (measured with the HAQ) is clinically relevant in the risk phase of CSA.11 This proved to be 0.7, slightly lower than the well-known minimally clinically important difference of 0.9 known for the early arthritis phase. They also investigated predictors for a good treatment response. These appear to be mainly imaging factors; people with two or more inflamed tendon sheaths often showed a good response to methotrexate use.12 Tenosynovitis is therefore not only a potent predictor of the development of RA but also of a good response to methotrexate. Several things remain on the research agenda to ensure precision prevention becomes possible, and to assign the right drug to the right patient. While known risk-stratification criteria contribute to selecting the right patient, the right drug is not yet clear. Preventing ACPA-positive RA is particularly challenging, as methotrexate, rituximab and abatacept delayed diagnosis for varying periods but ultimately failed to prevent it.10 13 14 Drugs that target processes or cells that specifically play a key role in the progression from CSA to RA may be needed. This is an unmet need for the near future, which large-scale studies of markers in the systemic circulation and relevant tissues (including the synovium or tenosynovium) can hopefully resolve. Another unmet need is to develop outcome measures that can be used in future prevention trials that allow response to be measured during treatment and without, for example, waiting many years for the primary endpoint of RA to develop. Whether responses early in the treatment course correlate with the important long-term outcome of RA prevention will need to be explored, but given the clinical impact of symptoms and signs in the pre-RA phase, the effect of treatments on this burden must be assessed as an important outcome in its own right. This topic collection highlights the directions in which, and the extent to which, the field of RA prevention is evolving, offering hope that the burden of RA will be reduced for future generations.