Celiac disease is a chronic intestinal disease with an immune response to antigens in wheat gluten. Disease occurs after the target antigen gliadin has been modified by the enzyme tissue transglutaminase, which allows subsequent presentation in the context of specific HLA molecules [19]. It is likely that RA is also the result of post-translational modifications of antigens by enzymatic activity (deiminases), and subsequent immune-mediated destruction of the synovium [9]. Regulation of peptidyl arginine deiminase (PADI) activity appears also to be involved in the maintenance of immunological tolerance to antigens. Thus, citrullination of proteins in RA patients may unmask essential epitopes that bind strongly to HLA class II molecules and result in initiation of an autoimmune response.

Two recent studies [14, 15] address different aspects of the value of anti-CCP antibodies as a risk factor for the development of RA. A longitudinal study [14] in a cohort of 79 RA patients in whom blood donations were available prior to the development of the disease showed that half of the patients produced anti-CCP antibodies and IgM-RF before the onset of RA. At a similar frequency, 10 of 16 patients who developed systemic lupus erythematosus or mixed connective tissue disease were positive for antinuclear antibodies 0.7–4.5 years before the onset of symptoms [16]. In comparison, 82% of IDDM patients have antibodies to glutamic acid decarboxylase 10 years prior to the disease [17]. In the RA study [14], antibody positive patients proved to develop more severe RA. These data indicate that 50% of the RA patients had a detectable abnormality in immune tolerance before the early clinically detectable disease stages.

The autoimmune response in immune-mediated diseases, such as RA appears to develop at different stages in each individual – this has several important implications. Autoantibodies serve as early indicators of a definite break in tolerance and may provide insight into the pathogenesis of RA. This raises the possibility to predict RA development in high-risk populations and may further allow a more precise "window of opportunity" for early and effective treatment interventions. Moreover, modulation of the immune response to a given antigen might alter the future disease course. A further important recent lesson learned is that serologic prediction of the disease can be greatly improved by considering the presence of combinations of autoantibodies [20]. Although a clear pathogenic link between RF and anti-CCP antibody is missing, their co-presence is highly indicative of RA [21]. Likely antibody combinations reflect a spreading of the immune response to include more than one antigenic determinant with an associated increase in the risk of progression to disease [21]. Spreading of this immune response is probably genetically determined [22].

The value of positive autoantibodies in RA can only be considered midst a complex pattern of additional predictive markers. Another recent prospective study [15] evaluated patients with undifferentiated early arthritis. Of the 936 patients, 21.9% had RA at inclusion, 32% after 1 year, 38% after 2 years and 40% after 3 years of follow-up. Importantly, the presence of anti-CCP antibodies was identified as a significant risk factor for RA with an odds ratio (OR) of 37.8 (95% confidence interval [CI] 13.8–111.9) while IgM-RF had an OR of 9.8 (95% CI 4.1–23.4). Major predictive clinical variables were morning stiffness, polyarthritis, symmetric arthritis and erosions on radiographs. Although serologic abnormalities had a high OR, these data emphasize the importance of a clinical pattern that needs to be identified for preclinical conditions with antibody positivity.

Genetic factors, for example HLA-DR haplotypes and PADI gene polymorphisms, represent further intriguing candidates for prediction. Hill et al., identified the specific role of MHC class II molecules in presenting citrullinated peptides to the immune system by studying T-cell responses to citrulline-containing peptides in HLA-DRB1*0401 transgenic (DR4-IE tg) mice [22]. They demonstrated that the conversion of arginine to citrulline at the peptide side-chain position interacting with the shared epitope significantly increased peptide-MHC affinity and led to the activation of CD4⁺ T cells. These data could, therefore, explain how DRB1 alleles with the shared epitope initiate an autoimmune response to citrullinated peptides in RA patients. Further studies [23, 24] focussed on an involvement of peptidylarginine deiminases citrullinating enzymes (encoded by PADI genes). In a Japanese case-control linkage disequilibrium study [23], PADI type 4 was identified as a susceptibility locus for RA; this was not found in a UK population of RA patients [24] indicating the need for further studies. The role of genetics in RA is further supported by the observation that an anti-CCP+ member of a multicase RA family has an estimated 69.4% risk of developing RA within 5 years [15].

A number of questions have been raised by these studies: (1) Since only 50% of RA patients develop autoantibodies before the onset of disease, the question remains whether other unknown humoral disturbances might be present in the remaining patients or if there are different subgroups of patients independent of the presence of RF and/or anti-CCP antibodies before disease develops; (2) The American College of Rheumatology criteria [25] for RA are commonly used for classification based on patient's history, physical examination, laboratory and radiographic findings. Early diagnosis of RA in patients with arthritis of recent onset [26, 27] is often difficult. New aspects of the prognostic value of autoantibodies lead us to question whether our current diagnostic standard in RA needs reevaluation; (3) Although we have progressed remarkably in our therapeutic armamentarium as well as in identifying high-risk subjects for RA through antigen-specific autoantibodies, screening the appropriate population for RA susceptibility remains a challenging task.

Identification of an autoimmune response to specific autoantigens in the pre-RA period raises the prospect of prediction and prevention. This may allow early treatment interventions (secondary prevention) and potentially disease prevention (primary prevention). However, accurate disease prediction is vital for secondary prevention in the context of clinical manifestations. At last, this will require not only a reliable view through the "window of opportunity" but also a subsequent successful step through a "gate of new opportunities".