The natural history and prognosis of rheumatoid arthritis: association of radiographic outcome with process variables, joint motion and immune proteins

Abstract

SummaryObjective: The purposes of the present study were: 1) to investigate how the long‐term course of outcome and inflammatory variables could be described in individual patients and suitably summarized in groups of patients; 2) to investigate the associations between outcome and inflammatory variables on the basis of the defined summary measures; and 3) to investigate new prognostic aspects of RA by means of frozen sera and DNA specimens.Patients and methods: During the period 1966–78, 685 Danish Caucasian patients with RA, classified according to the 1958 American Rheumatism Association (ARA) criteria, were admitted to the Department of Rheumatology of Aarhus University Hospital. For scientific purposes all patients went through the same examination programme, including biochemical variables, clinical evaluation of 68 diarthrodial joints, and radiographic evaluation of 46 diarthrodial joints. Since 1987, data from these patients have been organized in a database. The data are arranged according to onset of disease. This thesis is based on about 600 000 data‐points from 257 patients.Results: The thesis is based on six studies. The first study shows that early symptomatic improvement of RA during gold treatment was stable over several years, but when evaluated radiographically, the condition continued to deteriorate. In the second study, six main types of radiographic progression were identified: (a) a rare type with no radiographic progression at all (<1%); (b) a type with a slow or moderate onset, but an increasing progression rate (exponential growth type) (9%); (c) a linear type (30%); (d) a type with a moderate to fast onset, and a stable progression rate (the square root type) (11%); (e) a type with a fast onset, but a later decreasing progression rate (the first order kinetics type) (30%) and (f) a type characterized by slow onset, then acceleration and later deceleration (the sigmoid type) (20%). The fact that there was a systematic progression was used to define a system of radiographic events, which could be used as outcome measures in prediction models of the long‐term course of RA. The third study shows that low serum levels of the complement‐activating serum lectin, mannan (mannose) binding protein (lectin) (MBP=MBL), are associated with a higher erythrocyte sedimentation rate (ESR) (p=0.006), joint swelling score (JS score) (p=0.019), limitation of joint motion score (LM score) (p=0.027), and annual increase in radiographic destruction score (R score) (p=0.053). The fourth study demonstrated a highly significant association between summary measures of inflammatory variables and radiographic outcome, as defined in the second study, indicating that the degree of inflammation is important for the development of destructive joint damage in RA. The fifth study showed that MBL‐insufficient patients (two defective structural MBL alleles, or one defective allele combined with a low‐expression variant of the normal allele) had a relative risk of a severe radiographic event of 3.1 compared with the MBL competent group (p<0.0001). The sixth study showed that the relative risk (RR) of early interleukin (IL)‐1α auto‐antibodies (aAb) positive patients developing serious radiographic joint destruction was significantly lower than for IL‐1α aAb‐negative patients, RR=0.29 (p=0.04). In rheumatoid factor (RF) positive patients RR was only 0.18 (p=0.02). Patients who seroconverted >2 years after the onset of RA showed the most aggressive development of joint erosion, with RR of serious radiographic joint destruction of 2.56 (p=0.048). Other factors investigated in subgroups of the patients were HLA‐DR4, chemokine receptor 5 (CCR 5) genotypes, IL‐6 aAb, vascular endothelial growth factor (VEGF) aAb, and interferon (IFN)‐γ aAb. About 80% of the patients were HLA‐DR4 positive, indicating the importance of HLA‐DR4 as a predisposing factor for RA. There was no association between IL‐6 aAb and radiographic outcome, or CCR5 genotypes and radiographic outcome. VEGF aAb and IFN‐γ aAb were quantitatively unimportant.Conclusion: In spite of a general improvement in single measures of inflammatory variables, and a general deterioration in radiographic outcome of RA, there is a highly significant association between summary measures of inflammatory variables and radiographic outcome. The progression of radiographic damage in RA follows mathematical patterns. A new method of evaluating the long‐term radiographic outcome by means of Kaplan–Meier plots is demonstrated. It is shown that MBL and IL‐1α aAb are predictors of the prognosis of RA and may play important roles in the pathogenesis of RA.