Change in calcinosis over 1 year using the scleroderma clinical trials consortium radiologic scoring system for calcinosis of the hands in patients with systemic sclerosis

Introduction

Calcinosis cutis is a debilitating complication of systemic sclerosis (SSc). We previously developed a radiographic scoring system to assess severity of calcinosis affecting the hands in patients with SSc. We sought to further validate our radiographic scoring system to assess for change over 1 year and to identify factors associated with improvement or progression.

Materials and methods

Baseline and 1-year antero-posterior hand radiographs were obtained in 39 SSc patients with calcinosis prospectively enrolled at 6 centers within the US, Canada, Mexico and Australia. Two readers (one radiologist and one rheumatologist) scored all radiographs using the calcinosis scoring system and a 5-point Likert scale (1 = A lot better, 2 = A little better, 3=No change, 4 = A little worse, 5 = A lot worse) on follow-up. By maximizing the Kappa coefficient of agreement between grouped Likert scale (better/no change/worse) and the percentage of change of calcinosis in the radiographic scoring system, we defined progressive calcinosis as >25% increase in score from baseline at 1-year, stable calcinosis as change in score between -25% to 25%, and improvement of calcinosis as decrease in score by >25%. Nineteen SSc patients from an independent cohort were used for validation.

Results

Inter-rater reliability of the calcinosis scoring system was high with intra-class correlation coefficient of 0.93 (0.89–0.95). The median percentage of change from baseline to 1 year was 12.8% (range -89.3 to 290.2%). Sixteen patients (41%) experienced progression of calcinosis over 1 year; 18 (46%) remained stable; and 5 (13%) had improvement. Patients with progressive calcinosis had lower T-score on bone densitometry (-3.3 vs -1.7, p = 0.044) and higher prevalence of loss of digital pulp on physical exam (56% vs 22%, p = 0.027), with a trend towards lower baseline modified Rodnan skin score (mRSS) (3.8 vs. 5.9, p = 0.057), than patients who did not progress. Patients who experienced improvement in calcinosis had lower prevalence of digital pitting scars (20% vs 71%, p = 0.047) than patients whose calcinosis did not improve. In multivariable analysis, loss of digital pulp remained a predictor of calcinosis progression (OR 5.8, p = 0.023, CI 1.27 - 26.36). In the validation cohort, 2 (11%) patients improved, 10 (53%) remained stable, and 7 (37%) progressed.

Conclusions

We confirmed the excellent inter-rater reliability of our radiographic calcinosis scoring system and demonstrated its usefulness to detect change over time. Approximately 40% of patients experienced progression of calcinosis over 1 year. Loss of digital pulp was predictive of progressive calcinosis providing further evidence that digital ischemia contributes to the progression of calcinosis.