November 2, 2010

NIH Campus, Building 49, Room 1A51/1A59
Bethesda, Maryland

Robert H. Carter, MD, NIAMS
Susana A. Serrate-Sztein, MD, NIAMS
Judith James, M.D., Ph.D., Oklahoma Medical Research Foundation


The NIAMS organizes four to five roundtables on an annual basis as a part of the Institute’s long-range planning process. Over the long-term, these discussions help shape the Institute’s thinking about areas of importance in our basic, translational, and clinical research portfolios.

The purpose of the November 2, 2010, roundtable was to assess current opportunities, challenges, and future directions in the study of the preclinical disease phase of autoimmune diseases that are of interest to the NIAMS. The discussion was expected to identify research on disease mechanisms that would enable early detection and subphenotyping of autoimmune diseases, and, ultimately, facilitate implementation of specific preventive interventions. This summary reflects the opinions of the participants and the colleagues they polled prior to the meeting.

Key Mechanisms in Preclinical Autoimmunity

A framework for the development of autoimmune diseases has emerged from observational research, beginning with disease predisposition, which can include genetics, epigenetics, race, gender, and socioeconomic status. The condition can proceed sequentially to:

  1. latent events,
  2. pathogenic events,
  3. pre-diagnosis clinical features; and then
  4. clinical disease

Environmental factors and gene-environment interactions have been implicated in certain autoimmune diseases, and there is a high-degree of specificity in their effects. For example, there is a strong association between rheumatoid arthritis (RA) and exposure to cigarette smoke and air pollution. Exposures to these environmental factors are linked to citrullination, a post-translational peptide modification, and autoantibodies are produced to the altered epitopes on the body’s own molecules. Anti-citrullinated peptide antibodies (ACPAs) have been detected in the serum of relatives of RA patients and other susceptible individuals, as early as four years before the occurrence of clinical disease, and their titers increase closer to diagnosis of the disease. Recently developed assays that detect citrullinated proteins (rather than just citrulline), such as enolases, vimentin, and fibrinogen, may provide greater specificity in future preclinical analyses. Porphyromonas gingivalis-associated periodontal disease precedes some cases of RA, which may be more prevalent in Hispanic populations, is another example of gene-environment interactions on pathogenesis. There is speculation that antibodies to citrullinated peptides from this bacterial infection, followed by epitope spreading, may be a first step in ACPA production. However, more research is needed to develop tools for mapping environmental exposures biologically, and to elucidate the route from environmental effects, such as peptide citrullination, to inflamed joints.

Additional factors need to be considered when assessing the impact of environmental exposures on autoimmunity. Although air pollution and cigarette smoking in the United States have decreased over the past 20 years, RA rates have remained stable. Women, who are predominantly affected by the disease, have not quit smoking at the same rate as men. Estrogen has been associated with RA protection, so decreases in the use of hormone replacement therapy and in the estrogen doses in oral contraceptives may contribute to stable RA rates. The impact of loss of protection on disease onset is also suggested by the development of lupus in older Caucasian women, the occurrence of severe kidney damage in some cases with negligible levels of autoantibodies, and in twin studies, in which both siblings have the genetic risk, but the disease occurs in one and not the other. Smoking is not as strongly associated with the incidence of lupus in susceptible individuals. However, oral contraceptive use and the appearance of antibodies to Epstein-Barr virus (EBV) are associated with lupus risk (but not RA). Other infectious agents may be involved in the early stages of autoimmune diseases, such as lung inflammation. Changes in the gut microbiome in mouse models of RA can trigger joint inflammation.

Several autoimmune diseases may share preclinical characteristics; however, they may be distinguished by unique etiologies. The complex network of potential triggers suggests that many genes and factors, each with small levels of risk, contribute to disease incidence. Understanding the temporal pattern of markers may provide important insights into the individual’s susceptibility to disease progression, from a preclinical phase to symptomatic, clinical disease. Preclinical changes in cytokines, such as interferon signatures, are biomarkers of progression, but are also informative descriptors of disease mechanisms.

Isotype switching has been observed in the progression of endemic pemphigus foliaceus in Limao Verde, Brazil, in which immunoglobulin M (IgM), as well as IgE and non-IgG4, antibodies against the disease’s autoantigen, desmoglein 1 (Dsg1) have been identified as preclinical markers of disease. Anti-Dsg1 IgM antibodies were not found in infants, but they appeared later in childhood in 58% of the cohort. Seven percent of these children developed IgG anti-Dsg1 antibodies, and 1% expressed pathogenic IgG4 anti-Dsg1 antibodies prior to disease symptoms.

In contrast to forms of pemphigus, which have specific autoantigens targeted by pathogenic autoantibodies, there may be a common autoantibody for systemic, rheumatic autoimmune diseases for a widespread autoantigen. Subsequent, progressive appearance of disease-specific autoantibodies and epitope spreading may be in response to changes in antigen presentation and affinity. A general anti-nuclear antibody (ANA) serum test is positive long before the onset of clinical symptoms for several autoimmune diseases, including lupus, RA, scleroderma, and dermatomyositis, but specific, disease-associated anti-nucleic acid antibodies are seen at particular periods thereafter. For example, anti-Ro and anti-La antibodies may appear long before lupus diagnosis, and anti-Sm antibodies are seen just before lupus diagnosis.

However, autoantibodies may be indicators, rather than direct causal agents, of disease. About 5% of ANA-positive individuals develop autoimmune disease, and RA occurs in only 6% of ACPA-positive northern Europeans, so factors other than autoantibodies may be involved in the preclinical phase of autoimmune diseases. Most autoantibodies are found closer to the manifestation of clinical disease.

Two models for early preclinical autoimmunity, that may be involved in different diseases, are under consideration:

  1. Immune dysregulation, in the innate and/or adaptive immune system (such as defects in toll-like receptors and changes in T cell activity, respectively), leading to inflammation, release of tissue antigens (including nucleic acids), appearance of autoantibodies, then tissue damage; or,
  2. Genetic predisposition to produce autoantibodies, then release of autoantigen, followed by epitope spreading, leading to tissue damage.

Exogenous factors may have their greatest influence in preclinical phases, whereas endogenous factors may sustain the process in clinical disease. There may also be amplification loops, based on the interaction of antibodies, immune cells (e.g., neutrophils and plasmacytoid dendritic cells), and cytokines (such as interferons), and the formation of immune complexes, that control the rate and magnitude of disease progression and tissue damage.

Opportunities for Future Research

The current base of knowledge provides opportunities for diagnosis, intervention, and prevention during the preclinical phases of autoimmune diseases. Discoveries and approaches from the closely related and more mature field of research on type 1 diabetes, another autoimmune disease, can provide insights into future directions for studying related conditions. For example, the NIH-supported Environmental Determinants of Diabetes in the Young study is focusing on epidemiological approaches, in preparation for a prevention trial. Samples are being collected from children ever three to six months, over several years, to identify genetic, epigenetic, and environmental factors related to disease onset in genetically-susceptible individuals. Researchers have found more biomarkers than autoantibodies. Cyclosporine treatment is being tested for delay of disease onset.

At the inception of studies in diseases, such as RA, lupus, scleroderma, and vasculitis, which have heterogeneous clinical presentations, subphenotyping, with clinical descriptions and associated molecular, cellular, and genetic markers, will be needed. There are several at-risk populations that can be monitored for the development of disease and to identify transition points, with the large arsenal of available biomarkers:

  • Relatives of rheumatic disease patients, such as unaffected twins and first-degree relatives. In particular, pediatric patients can be followed more accurately for environmental exposures, the latency period for progression to disease is shorter, and they generally change location less frequently than older patients, so they are easier to follow clinically for longer periods of time.
  • Endemic disease populations, e.g., pemphigus foliaceus in Brazil.
  • The general population, through military biorepositories and medical records (which collect blood samples and physical histories across the lifespan, from early adulthood), and small population cohorts, who can be monitored for pulmonary inflammation and periodontal disease, and the appearance of swollen joints, which may be signs of mild disease.

The development of autoimmune disease may also be tracked in patients whose immune systems are being reconstituted after bone marrow transplants to treat their autoimmune conditions, because they presumably have the same genetic risk as before the treatment. Juvenile idiopathic arthritis and lupus are marked by periods of remission and disease activity, and changes in interferon signatures and other markers are harbingers of disease flares, so monitoring these cycles may replicate the progression of disease from preclinical phases of autoimmunity. However, more research is needed to discern whether the biomarkers and mechanisms of remission-to-flare represent preclinical progression, or a more advanced disease stage.

Disease modeling, such as a systems biology approach, would create a structure for understanding the transitions between pre-disease phases. By populating the framework with information on normal and aberrant immune function, the impact of perturbations, such as environmental triggers, on disease progression can be evaluated. Monitoring changes in markers continuously would likely yield more useful information than static views, such as:

  • alterations in cytokine expression, which can provide insights into disease mechanisms and disease progression
  • appearance, activation, and responses of immune cells, e.g., T cells, B cells, plasmacytoid dendritic cells, and neutrophils
  • appearance and presence of autoantigens, e.g., nucleic acids
  • appearance, presence, and changing characteristics of autoantibodies, such as isotype switching, epitope spreading, and changes in antigen presentation and affinity
  • appearance, presence, and changing indicators of tissue damage, environmental exposure (e.g., infections, toxins), and aging

Future research would benefit greatly from multi-site and international collaborations, in which clinical samples, technical resources, and methodologies are shared. Current activities include a U.S. registry for pediatric rheumatic diseases, linked to European pediatric clinical research networks, and an international preclinical autoimmune diseases consortium, that conducts multiplex biomarker analysis on all clinical samples. A multidisciplinary team could capture data on numerous parameters, to provide a broader and more informative picture of the preclinical phase of autoimmune diseases. The collection of as much patient data as possible will be important for current studies, and for future studies, as new technologies and analytic methodologies become available, such as markers of environmental exposure.

Candidate interventions, with relatively safe drugs, could be tested in pilot clinical trials with at-risk cohorts, which could include monitoring changes in biomarkers, conducting mechanistic studies with immune cells, and collecting patient-reported outcomes, to provide insights into disease progression. For example, it has been reported that treatment with hydroxychloroquine (an antimalarial drug) delayed the progression of lupus, which provides the rationale for a prospective, blinded trial, testing the capacity of hydroxychloroquine to delay or prevent lupus in high risk populations. Because there are still large gaps in understanding the development of disease in susceptible individuals, with the possibility that they may never experience significant symptoms, there are ethical concerns about informing them of the realistic risk for developing the disease, and exposing them to treatments that may have potentially toxic side effects. Hence, balancing the risk of experimental interventions in preclinical subjects with the risk of disease outcomes, and the potential efficacy of the treatment, will be essential in designing intervention trials in this novel area of research. It will also be important to select cohorts from the highest risk populations with markers that correlate with preclinical phases that could be responsive to treatment, to obtain rapid results.

Meeting Participants

Patient Representative
Director, Office of Communications and Public Liaison

BUYON, Jill P., M.D.
Professor, Department of Medicine
New York University School of Medicine

CARTER, Robert, M.D.
Deputy Director

CROW, Mary (Peggy) K., M.D.
Physician-in-Chief and Chair, Division of Rheumatology
Department of Medicine
Hospital for Special Surgery

DIAZ, Luis A., M.D.
Chair, Department of Dermatology
University of North Carolina School of Medicine

JAMES, Judith A., M.D., Ph.D.
Chair, Arthritis and Clinical Immunology
Oklahoma Medical Research Foundation

JOHNSON, David B., Ph.D.
Program Director
Division of Allergy, Immunology, and Transplantation
National Institute of Allergy and Infectious Diseases, NIH

KARLSON, Elizabeth W., M.D.
Associate Professor, Division of Rheumatology, Immunology and Allergy
Department of Medicine
Brigham and Women’s Hospital

KATZ, Stephen I., M.D., Ph.D.

LA CAVA, Antonio, M.D.
Professor, Division of Rheumatology and Arthritis
Department of Medicine
Geffen School of Medicine
University of California, Los Angeles

MAO, Su-Yau, Ph.D.
Director, Arthritis Biology Program
Division of Skin and Rheumatic Diseases

MERKEL, Peter, M.D., Ph.D., M.P.H.
Director, Vasculitis Center
Boston University School of Medicine

MOEN, Laura, Ph.D.
Director, Division of Extramural Research Activities

OLSEN, Nancy J., M.D.
Chief, Division of Rheumatology
Department of Medicine
Pennsylvania State University Hershey Medical Center

PASCUAL, Virginia, M.D.
Associate Investigator
Baylor Institute of Immunology Research
Baylor University Medical Center

PISETSKY, David S., M.D., Ph.D.
Professor, Division of Rheumatology and Immunology
Department of Medicine
Duke University School of Medicine

ROSEN, Antony, M.D.
Director, Division of Rheumatology
Department of Medicine
Johns Hopkins University School of Medicine

Co-Chief, Division of Pediatric Rheumatology
Department of Pediatrics
Duke University School of Medicine

SCHER, Jose, U., M.D.
Clinical Instructor, Division of Rheumatology
Department of Medicine
New York University School of Medicine

Director, Division of Skin and Rheumatic Diseases

SIEGEL, Richard, M.D., Ph.D.
Chief, Autoimmunity Branch
Intramural Research Program

WANG, Yan, M.D., Ph.D.
Director, Rheumatic Diseases Genetics and Translational Research Program
Division of Skin and Rheumatic Diseases

WITTER, James, M.D., Ph.D.
Director, Rheumatic Diseases Clinical Program
Division of Skin and Rheumatic Diseases

Last Reviewed: