Background/Purpose: The CRISS is a composite index of response in diffuse cutaneous systemic sclerosis (dcSSC). It is a 2-step process for the probability of improvement ranging from 0.0 (no improvement) to 1.0. Patients with decline in cardio-pulmonary-renal involvement are assigned score 0.0 in step 1. All remainng subjects are scored for improvement in FVC, mRSS, patient (PT GA) and physician globals (MD GA), and HAQ-DI.

Two independent studies including in total 464 patients from 6 different centres have shown that skin, lung and overall fibrosis correlate with three serum biomarkers (Procollagen III N terminal pro-peptides (PIIINP), Tissue inhibitor of Metalloproteinase I (TIMP-1) and Hyaluronic Acid (HA)), either singularly or combined in the ELF Test algorithm. Here we aimed to determine whether CRISS correlated with the serum concentration of one or more of the ELF biomarkers and explore the scope to build a combined score with better performance in measuring the probability of clinical response in dcSSc.

Methods: 31 Consecutive dcSSc patients were included in the study at a single centre. Clinical data and serum were collected at baseline, at 12 and 24 months. CRISS and ELF score were calculated at the same time points. CRISS scores were calculated using the publishedd formula and we assessed correlation coefficients between CRISS scores and biomarkers. Comparison between two groups and correlation were performed using Mann-Whitney and Spearman’s tests, respectively. P<0.05 was considered statistically significant. Statistical analysis was carried out using GraphPad Prism Version 7.

Results: Thirty-one dcSSc patients were enrolled (12 M; mean age=50.3 ±11.5). CRISS at 12 months was 0% in 19 patients, whereas in the remaining 12 patients it ranged from 0.1 to 1.0 (median=0.35). Five out of the 12 CRISS “responders” had a CRISS 24-12 >0.0, the remaining 7 had no further response (CRISS=0). Overall CRISS 24-0 was >0.0 in 16 patients (median=0.065, range=0.01-0.93). No clinical features at baseline were significantly different in the responders vs non responders at 12 or 24 months, including mRSS, FVC%, DLCO%, autoantibody profile, age, disease duration, Medsger Severity Score and CRP ( P>0.05 for all). TIMP-1 concentation at baseline was significantly lower in patients with CRISS>0 vs. CRISS=0 (218.5 vs 265.9 respectively, p= 0.03), similar trend was observed for TIMP-1 at 12 months and CRISS 24 months. Accordingly TIMP-1 at baseline or 12 months statistically significant correlation with CRISS 24 months ( R=-0.389 and -0.374, respectively; p < 0.05 for both).

Conclusion: This is the first evaluation of CRISS in an a observational setting. Although limited by single centre setting and low number, we show that in any 12 months interval a proportion between 38 and 51% of patients show a CRISS >0. Whereas only 6 to 22.5% of patients showed a CRISS>0.10. None of the routinely used clinical parameters can predict CRISS at 12 months whereas TIMP-1 concentration at baseline is significantly lower in the responders. This needs to be explored in additional cohorts and opens the possibility to include serum biomarkers of fibrogenesis within a combined response index for Systemic Sclerosis.

« Back to 2017 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/correlation-of-the-american-college-of-rheumatology-provisional-composite-response-index-in-systemic-sclerosis-with-serum-biomarkers-of-fibrogenesis-in-an-observational-cohort/