Clinical, radiographic and HLA associations as markers for different patterns of psoriatic arthritis.

OBJECTIVE: The aim of this study was to examine whether the five clinical forms of psoriatic arthritis (PsA) identified by Moll and Wright (Semin Arthritis Rheum 1973;3:55-78) could be clearly distinguished, especially as the disease evolved over time, to analyse whether radiographic features or HLA associations could define subsets with greater precision and to identify predictors of disease outcome.
METHODS: Seventy-three patients (37 males and 36 females) were followed for a median time of 8 yr (range 1-16 yr). A standard clinical protocol was used to assess patients at each visit and two clinical scores. based on the joint areas involved, were defined to evaluate the mode of onset and the evolution of arthritis. X-ray films of the hands, feet and sacroiliac joints were taken and the patients were divided into two categories according to the presence or absence of erosions and an X-ray erosion score was also used. Three classification methods were used to define the different clinical subsets. HLA-A, B and DR antigens were tested by standard microlymphocytotoxicity assays. A multiple linear regression model was used in the statistical analysis.
RESULTS: The five classical clinical subsets defined by Moll and Wright did not remain since distinct peripheral arthritis patterns tended to evolve over time. Only two discrete groups were identified, axial disease (AD) (sacroilitis with or without peripheral arthritis) in 29% of cases and peripheral disease (PD) without sacroilitis in 71%. AD was positively associated with the duration of arthritis (P < 0.04), presence of mutilation (P < 0.02) and the joint area score over disease evolution (JASE) (P < 0.02). There were erosions in 71% of the patients. Erosions correlated with the presence of mutilation (P < 0.007) and with the JASE (P < 0.0005). HLA-B27 was found in 43% of patients with AD, but only in 11% of PD patients (P < 0.01). No other clear HLA correlations were found.
CONCLUSIONS: Despite the relatively small number of patients, this longitudinal study suggests that only two clinical subsets can be clearly defined in PsA, AD and PD; these are primarily determined on clinical grounds although HLA-B27 is strongly associated with AD. The evolution of PD pattern with time means that narrower peripheral arthritis subsets are of little clinical use.