Background/Purpose

Predictors of radiographic progression in erosive osteoarthritis (OA) are important in identifying patients with high risk of disease activity and consequently functional loss. Disease duration, number of tender joints and number of joints with palpable effusion at baseline are already identified as clinical predictors of radiographic progression. The aim of this study is to confirm the existing predictors in a prospective cohort. Additionally, potentially other clinical and radiographic predictors will be identified.

Methods

One hundred and twelve patients with erosive OA were selected from an already existing cohort that was recruited from April 2007 through January 2010 at the Ghent University Hospital. X-rays, clinical and demographic data of the 1^st assessment were present. All patients were reassessed between January 2014 and March 2014. All interphalangeal finger joints on both radiographs were scored according to the Verbruggen and Veys method. Radiographic progression was defined as a joint progressing from at least one anatomical phase, excluding the progression from a ‘N’ phase to a ‘S’ phase. A generalized estimating equation (GEE) model with a binary logistic function was used to explore the following potential clinical and radiographic predictors on joint level: disease duration (≤5 years, >5 years), presence of erosive joints in the dominant hand, presence of painful joints, tender joints, or joints with palpable effusion, the presence of a joint in ‘J’ phase and in ‘E’ phase. All variables were dichotomous (present or absent).

Results

Three clinical and two radiographic predictors were retained: a painful joint, a tender joint, a joint with palpable effusion, a joint in ‘J’ phase and a joint in ‘E’ phase. A joint with palpable effusion was the strongest clinical predictor (odds ratio (OR): 2.474) (table 1). A joint in ‘E’ phase was the strongest radiographic predictor (OR: 90.628) (table 2).

Table  SEQ Table * ARABIC 1: Clinical predictors for radiographic progression by GEE modeling

CI: confidence interval.

Table  SEQ Table * ARABIC 2: Radiographic predictors for radiographic progression by GEE modeling

Conclusion

A painful joint, a tender joint, a joint with palpable effusion, ‘J’ phase and ‘E’ phase were identified as predictors of radiographic progression in erosive OA. The strongest clinical and radiographic predictor was a joint with palpable effusion and the presence of an ‘E’ phase respectively. These predictors should be considered when selecting patients for therapeutic trials with potential disease-modifying osteoarthritic drugs.