An elevated acute-phase response is associated with increased radiologic damage in rheumatoid arthritis (RA), but development of damage in previously normal joints (“new joint involvement”) has not previously been investigated. This study was undertaken to investigate the hypothesis that when there is suppression of disease activity as judged by the C-reactive protein level, new joint involvement is reduced to a greater extent than is progression in already damaged joints (“damaged joint progression”).


Three hundred fifty-nine patients with active RA were studied as part of a 5-year randomized, prospective, open-label study of disease-modifying antirheumatic drug therapy. Time-averaged CRP was calculated from samples obtained every 6 months, and patients were divided into groups with CRP values of <6, 6–<12, 12–<25, and ≥25 mg/liter. Radiographs of the hands and feet were scored by the Larsen method; a damaged joint was defined as one with a score of ≥2.


The rank correlation between time-integrated CRP and increase in Larsen score was 0.50; the correlation increased to 0.59 for patients entering the study with disease duration of ≤2 years. The percentage of new joint involvement over 5 years varied markedly with time-integrated CRP, from 7.3% in the CRP <6 mg/liter group to 39.1% in the CRP ≥25 mg/liter group (5.4-fold increase). The percentage of damaged joint progression increased from 26.1% in the CRP <6 mg/liter group to 41.6% in the CRP ≥25 mg/liter group (1.6-fold increase).


The results of this study provide further confirmation that high CRP levels over time are associated with greater radiologic progression. Although radiologic progression still occurred in both previously normal and damaged joints despite the presence of normal CRP levels, this consisted of proportionately less new joint involvement compared with damaged joint progression. These findings support the idea that disease-suppressive therapy should be instituted at an early stage in patients with RA, before erosive damage has occurred.