Key points

Introduction

The concept of preclinical disease has gained increasing attention over the past decade as researchers and providers aspired to prevent and cure chronic disease. In seropositive autoimmune diseases, autoantibodies have been identified well before the onset of disease or even symptoms not only in type 1 diabetes and celiac disease, but also in rheumatic diseases, including rheumatoid arthritis (RA), antiphospholipid syndrome, and systemic lupus erythematosis. Of the rheumatic diseases, prediagnosis and preclinical RA clinics are becoming more prevalent because RA is the most common of the inflammatory arthritides, affecting approximately 1% of the general population. These clinics follow those individuals at risk for RA in whom risk may include the presence of autoantibodies without symptoms, symptoms without signs of synovitis, a first-degree relative (FDR) with RA, and/or a propensity to RA based on belonging to a specific at-risk population (eg, first nations). These clinics are generally aimed at identifying predictors of developing arthritis. Moreover, as the importance of the preclinical phase of inflammatory arthritis has gained prominence, the terminology describing persons in this preclinical phase has evolved. In a recent publication, the European League Against Rheumatism (EULAR) propose that prospective studies should use the following terminology to describe individuals at risk of RA having genetic risk factors for RA, having environmental risk factors for RA, having systemic autoimmunity associated with RA, having symptoms without clinical arthritis, and having unclassified arthritis. These recommendations further suggest that “pre-RA” should refer to the period of time or phase, assessed retrospectively, before RA can be classified. Studies examining the pre-RA phase usually compare patients who were recently classified as RA and those not classified as RA but followed over a similar period of time. These studies compare risk factors in each group. The term “individuals at risk” implies a longitudinal study design in which individuals with risk factors for RA are followed to determine the probability of, or time course to, developing classifiable RA given a set of risk factors. This article focuses on individuals at risk who have symptoms and/or one or more other risk factors for RA but who have not developed clinical arthritis it also looks at studies that examined patients in the pre-RA phase.

Despite an increased interest in the preclinical phases of RA, there are few studies designed to examine the symptoms in this period. This article therefore extrapolates findings about symptoms in the preclinical phase from existing studies and summarizes these from studies that included information about clinical symptoms. Included are studies examining (1) screening tools to detect RA, and (2) physical features examined in cohorts of people with undifferentiated arthritis to predict who developed classifiable RA or persistent arthritis. This article also highlights the signs and symptoms being captured prospectively among patients followed in cohort studies of people at risk for RA, including cohorts of people who screened positively for increased risk of RA (eg, FDRs of individuals with RA, or those who carry antibodies associated with RA, or have symptoms without clinical arthritis [eg, arthralgia without synovitis]). Underscoring the need to understand symptoms in the presynovitis phase, a study by van de Stadt and colleagues identified that joint symptoms are important components of predicting future RA in autoantibody-positive individuals initially without arthritis.

Introduction

The concept of preclinical disease has gained increasing attention over the past decade as researchers and providers aspired to prevent and cure chronic disease. In seropositive autoimmune diseases, autoantibodies have been identified well before the onset of disease or even symptoms not only in type 1 diabetes and celiac disease, but also in rheumatic diseases, including rheumatoid arthritis (RA), antiphospholipid syndrome, and systemic lupus erythematosis. Of the rheumatic diseases, prediagnosis and preclinical RA clinics are becoming more prevalent because RA is the most common of the inflammatory arthritides, affecting approximately 1% of the general population. These clinics follow those individuals at risk for RA in whom risk may include the presence of autoantibodies without symptoms, symptoms without signs of synovitis, a first-degree relative (FDR) with RA, and/or a propensity to RA based on belonging to a specific at-risk population (eg, first nations). These clinics are generally aimed at identifying predictors of developing arthritis. Moreover, as the importance of the preclinical phase of inflammatory arthritis has gained prominence, the terminology describing persons in this preclinical phase has evolved. In a recent publication, the European League Against Rheumatism (EULAR) propose that prospective studies should use the following terminology to describe individuals at risk of RA having genetic risk factors for RA, having environmental risk factors for RA, having systemic autoimmunity associated with RA, having symptoms without clinical arthritis, and having unclassified arthritis. These recommendations further suggest that “pre-RA” should refer to the period of time or phase, assessed retrospectively, before RA can be classified. Studies examining the pre-RA phase usually compare patients who were recently classified as RA and those not classified as RA but followed over a similar period of time. These studies compare risk factors in each group. The term “individuals at risk” implies a longitudinal study design in which individuals with risk factors for RA are followed to determine the probability of, or time course to, developing classifiable RA given a set of risk factors. This article focuses on individuals at risk who have symptoms and/or one or more other risk factors for RA but who have not developed clinical arthritis it also looks at studies that examined patients in the pre-RA phase.

Despite an increased interest in the preclinical phases of RA, there are few studies designed to examine the symptoms in this period. This article therefore extrapolates findings about symptoms in the preclinical phase from existing studies and summarizes these from studies that included information about clinical symptoms. Included are studies examining (1) screening tools to detect RA, and (2) physical features examined in cohorts of people with undifferentiated arthritis to predict who developed classifiable RA or persistent arthritis. This article also highlights the signs and symptoms being captured prospectively among patients followed in cohort studies of people at risk for RA, including cohorts of people who screened positively for increased risk of RA (eg, FDRs of individuals with RA, or those who carry antibodies associated with RA, or have symptoms without clinical arthritis [eg, arthralgia without synovitis]). Underscoring the need to understand symptoms in the presynovitis phase, a study by van de Stadt and colleagues identified that joint symptoms are important components of predicting future RA in autoantibody-positive individuals initially without arthritis.

Screening and triage tools

Screening tools have been developed to screen populations with varying degrees of risk to detect people with RA. These tools may identify people with musculoskeletal (MSK) symptoms, or may be used to help triage those with symptoms or signs to either general practitioner or rheumatologist care in an attempt to improve early identification of inflammatory arthritis and RA. Because MSK symptoms are common, these tools are designed to identify only those people with true inflammatory disease. A recent systematic review of strategies to promote early diagnosis identified available screening tools to increase detection of individuals with inflammatory arthritis. The symptoms queried by these self-report tools included questions about the presence of joint pain, stiffness, swelling, symmetry, duration of disease, and functional limitation ( Tables 1 and 2 ). Some tools (such as the Connective Screening Questionnaire [CSQ]) were less specific for RA and queried an extensive list of extra-articular features associated with autoimmune diseases in general (eg, mucositis, rash, nodules). Ascertainment of these might identify not only RA but also a variety of connective tissue diseases. The questions that pertain to symptoms, the populations examined with these tools, and their performance characteristics are listed in Table 1 . The clinical signs and symptoms incorporated into these screening tools were selected by a variety of methods, including qualitative studies of patients, and Delphi consensus with input from patients and providers. These tools most often reflect domains that are part of existing classification criteria either for autoimmune diseases in general or for specific diseases. Although these features may be important to help clinicians to identify eventual persistent and classifiable RA, earlier detection of inflammatory arthritis may need to systematically study symptoms or signs not necessarily included in classification criteria because they may have important predictive value. This approach is the basis of many of the newer cohort studies examining individuals at risk of RA, the goals of which include identifying incident inflammatory arthritis in addition to classifiable RA.

Undifferentiated arthritis cohorts

Undifferentiated arthritis cohort studies were developed to detect and treat arthritis early in the disease course to improve outcomes. These cohorts identify individuals who already have clinically detectable inflammatory arthritis (eg, swollen joints) within a specified duration of symptoms (eg, within 1 year) and follow them over time.

Because the wish is to identify the clinical signs and symptoms that manifest in people who ultimately develop persistent disease, one approach is to identify the signs and symptoms used in cohorts of undifferentiated arthritis followed prospectively to derive prediction rules for the development of classifiable RA. The Leiden Prediction Rule is one such rule in which a score including clinical symptoms and signs could predict people who go on to develop classifiable RA by 1987 criteria. This tool was created for use in patients with undifferentiated arthritis. The rule weighted involvement of upper and lower extremities, small joint involvement, symmetry, numbers of tender and swollen joints, and morning stiffness greater than 60 minutes as predictive signs and symptoms for the development of RA. Combining the weights from the clinical variables with other demographic and laboratory variables (age, sex, C-reactive protein level, rheumatoid factor [RF] and anti–cyclic citrullinated protein [anti-CCP] positivity), a threshold of greater than or equal to 9 points was 31% and 99% specific for RA by 1987 criteria. A threshold of greater than or equal to 6 points was 81% sensitive and 79% specific predictive of the development of classifiable RA by American College of Rheumatology (ACR/EULAR) 2010 criteria in later studies performed after the development of the newer classification criteria (see Table 2 ).

Examining undifferentiated arthritis cohorts for clinical features that predicted the development of persistent arthritis can also provide an understanding of the clinical features that occur before the development of classifiable arthritis. A systematic review by Kuriya and colleagues examined the diagnostic and prognostic value of elements of history and physical examination with respect to persistent arthritis published in observational studies of undifferentiated peripheral inflammatory arthritis. The clinical elements in histories that predicted the development of persistent arthritis included symptom duration greater than 12 weeks, early morning stiffness greater than 1 hour, and small joint or wrist swelling. The physical examination features that predicted disease persistence included higher number of swollen joints and the presence of metatarsophalangeal compression pain (see Table 2 ).