Background There are muliple potential therapeutic a gents under development for sstemic sclerosis (SSc). Given that SSc is uncommon, there is interest in cohort enrichment to maximize the potential of identifying a drug effect, One area of interest is a modified Rodnan skin score (mRSS) cut-off as an inclusion criterion in early diffuse SSc trials. Autoantibody profiles of SSc patients vary by geographical region. Given that the mRSS differs between autoantibody groups and that many trials are multinational, it is important to understand the differences in mRSS patterns between autoantibodies to choose optimal mRSS cut-off points that facilitates enrichment for patients likely to develop skin progression.
Objectives We sought to determine the baseline mRSS to maximally predict progressive skin involvement over one year in US patients with early diffuse SSc, and assess if this differed by autoantibody.
Methods We examined a US single-center cohort of prospectively followed early diffuse SSc patients seen for an initial visit between 1980 and 2014. To reflect modern early diffuse SSc clinical trial design, early was defined as <3 years of symptoms and diffuse SSc as proximal skin thickening. Eligible patients had ≥2 mRSS within one year of the first SSc center visit, for a minimum of 3 mRSS over one year. The outcome of interest was mRSS progression at any time over that year, defined as an increase of 5 points or more and a 25% increase in the mRSS from baseline. Autoantibodies were confirmed by immunodiffusion for Scl-70, immunofluorescence for anti-centromere and immunoprecipitation for RNA polymerase III (RNAP) and all other SSc-associated antibodies. Descriptive statistics using Receiver Operating Curve (ROC) analysis were used with mRSS progression as the state variable to determine cut-off points for mRSS.
Results Among 317 eligible patients the mean age was 50.8±13.5 years, 74% were female and 92% Caucasian. The median disease duration was 0.90 years (IQR 0.63, 1.46). The mean baseline mRSS was 23±11. 25% (n=79) were scl-70 positive, 56% n=178) RNAP positive, 54 (17%) other SSc-associated antibody and 2% (n=6) missing. The median baseline mRSS (IQR) for RNAP was 24 (17,33), compared to 18 (13,25) for Scl-70 and 20 (13,28) for other antibodies (p=0.0004). In total, 163 patients (51%) developed progressive skin disease over one year. 59% of RNAP positive patients progressed, versus 48% of Scl-70 and 33% of other antibody positive patients (p=0.004). The mean time to peak mRSS was 0.51±0.23 years overall. The baseline mRSS below which there was the maximal likelihood of predicting progressive skin disease, with the greatest area under the curve (AUC) in Scl-70 positive patients was ≤24 (AUC 0.72; 95%CI 0.64, 0.80). This contrasted with RNAP patients, for which the maximal AUC occurred at ≤29 (AUC 0.74;95% CI 0.67, 0.80). Patients with other autoantibodies had a much lower baseline mRSS of ≤17 for optimal prediction (AUC 0.67; 95%CI 0.53, 0.80).
Conclusions The mRSS cut-off that predicted skin progression over one year was different between RNAP, Scl-70 and other SSc antibodies. his supports the concept that cohort enrichment in clinical trials may be enhanced by modifying the mRSS cut-off point for the major autoantibody groups. A lower cut-off point will exclude some US patients who are likely to progress.
Disclosure of Interest R. Domsic Consultant for: Biogen-Idec, M. Lucas: None declared, V. Steen: None declared, R. Lafyatis Grant/research support from: Regeneron, Consultant for: Regeneron, T. Medsger: None declared