Do GPI Antibodies Serve as a Biomarker for Extraarticular Complications of RA
The pathogenic role of circulating antibodies to glucose-6-phosphate isomerase (GPI) in RA is controversial. In an experimental mouse model, the K/BxN mouse, production of GPI antibodies leads to a spontaneous inflammatory arthritis resembling RA. Moreover, transfer of GPI antibodies from arthritic K/BxN mice to wild-type mice induces an inflammatory arthritis in the host mice. Though initial studies in humans confirmed high levels of GPI antibodies in sera, synovial fluid, and synovium of patients with RA (Schaller et al (Nature Immunol 2:746, 2001)), more recent attempts to confirm these results have not been corroborative (Matsumoto et al (Arthritis Rheum 48:944, 2003). Seeking to reconcile this contradiction, van Gaalen et al (Arthritis Rheum 50:395, 2004) queried whether GPI antibodies in humans might serve as a marker for extraarticular complications of RA.
Methods: Antibodies to GPI were assayed by ELISA in 131 patients followed at the Leiden University Medical Center, The Netherlands with RA. Subjects were divided into the following groups:
- early untreated RA of less than 4 weeks duration,
- uncomplicated treated RA of 3 years duration,
- uncomplicated treated RA of 12 years duration, and
- RA with extraarticular complications (rheumatoid nodules, rheumatoid vasculitis, or Feltys syndrome)
In addition, antibodies to GPI were assayed in 28 healthy control female subjects from the infertility clinic at the Leiden University Medical Center.
Results: Clinical characteristics of each group of RA patients, as would be expected, were disparate. Notably, patients in the rheumatoid vasculitis and Feltys groups tended to be older (mean age 67 years), have longer disease durations, and have a higher percentage of men than the early/untreated or uncomplicated/treated groups. Patients with uncomplicated RA of 12 years and patients with rheumatoid nodules tended to be about 10 years younger than those in the early RA groups or those with vasculitis or Feltys syndrome. Only 72% of patients with rheumatoid vasculitis were RF positive.
Patients with early/untreated RA, uncomplicated/treated RA of 3 years duration, and uncomplicated/ treated RA of 12 years duration demonstrated mean levels of GPI antibodies that were similar to non-RA controls. In contrast, in patients with rheumatoid nodules, rheumatoid vasculitis, and Feltys, mean GPI antibody levels were significantly higher than non-RA controls. This association was strongest in the patients with Feltys, in which 12 of 13 patients (92%) had GPI levels above the upper range of normal.
Conclusion: Elevated levels of GPI antibodies in patients with RA are associated with certain extraarticular complications (rheumatoid nodules, rheumatoid vasculitis, and particularly, Feltys syndrome). These antibodies are uncommon in early or uncomplicated disease.
Editorial Comments: Despite small numbers of patients in the groups of patients with extraarticular complications, these results are compelling in drawing an association between GPI antibodies and extraarticular RA. These results provide a potential explanation in previous studies for the inconsistent association of GPI antibodies with RA, since patients in these studies were not sorted by disease duration or presence of extraarticular complications. According to the current study, anti-GPI antibodies would not be useful in the diagnosis of RA but could have some predictive value for extraarticular disease (if their presence could be shown to predate extraarticular disease complications).
The relative absence of GPI antibodies in early RA is in stark contrast to antibodies to cyclic citrullinated peptide (anti-CCP). Anti-CCP antibodies are detected not only in early RA, but can also predate the development of clinical signs and symptoms of RA in many individuals. Anti-CCP antibodies thus have predictive value for the development of RA and, consequently, quite likely play a more direct role in the pathogenesis of RA. The potential pathogenic role of GPI antibodies to RA in general, or to extraarticular disease specifically, is much less clear.