Objective: To describe the pattern of arthropathy and HLA-DRB1 alleles associated with PMR in order to develop a diagnostic algorithm that could help distinguish PMR and RF-negative (RF -ve) late-onset RA (LO-RA) at presentation.
Methods: This was a prospective study of all patients presenting with PMR or LO-RA over a 10-yr period to one physician. Demographic, clinical and laboratory data were collected at presentation and during a minimum of 5 yrs of follow-up. The accuracy of the initial diagnosis was systematically reviewed.
Results: One hundred and forty-two patients with LO-RA, 147 with PMR and 42 with PMR + TA were studied. Peripheral synovitis was observed in 23% of the PMR patients. In comparison with RF -ve LO-RA, PMR patients were younger (P < 0.001), myalgia more frequent [100 vs 16% (P < 0.001)] and arthritis of PIP, MCP and wrist were less frequent (P < 0.001). The combination of wrist + MCP/PIP or wrist + PIP + MCP were highly suggestive of RF -ve LO-RA (P < 0.001). HLA-DRB1*0101/0102 and *0401 were significantly increased in PMR patients compared with healthy controls. Plasma viscosity and arthritis in the wrist, in combination with at least one MCP or PIP joint at disease onset, were predictive of whether a non-erosive RF -ve patient would ultimately be diagnosed as having RF -ve LO-RA or PMR (+/-/arthritis).
Conclusion: Our longitudinal follow-up data were consistent with RF -ve LO-RA being a separate disease entity to PMR despite some phenotypic and immunogenetic similarities at disease onset. A diagnostic algorithm was derived using baseline clinical features to predict the final diagnosis of RF -ve, non-erosive patients.