Objective: To compare radiological scoring systems in early RA, and then to determine if the rate of radiological progression in the first year of disease (RRP01) predicts the subsequent rate of progression up to 8 years (RRP1-8).
Methods: Hand and feet radiographs were performed at 0, 1, 2 and 8 years in 65 patients presenting with nonerosive RA and symptoms for less than 3 years. All films were scored by Sharp's method; Sharp, Larsen and carpometacarpal (CMC) ratio methods were compared in 23 patients.
Results: Significant change in score was detected over the first year using Sharp (median of 15.5 vs 7.5, p < 0.00002) and Larsen (30.5 vs 22.5, p < 0.0002), but not CMC ratio (1.13 vs 1.14, p = 0.07). Sharp had greater range and sensitivity of change, and had better inter and intraobserver reproducibility. RRP01 showed a skewed distribution (median 0.7 units/month; range -0.3 to +6.1). Spearman correlation between RRP01 and RRP1-8 was r = 0.57, p < 0.001.
Conclusion: Both Sharp and Larsen indices are sensitive to change in the first year of RA, but Sharp has advantages of greater sensitivity and reproducibility: CMC ratio is not useful. Radiological progression rate in the first year correlates only moderately with subsequent rate of progression up to 8 years. This suggests that radiological progression is nonuniform and that different patterns of progression over time may exist.