ELF Score: A Validated Serum Test Strongly Predictive of Fibrosis in Systemic Sclerosis.

Background/Purpose: The absence of a serum test predictive of activity or severity in Scleroderma (SSc) is a major burden both for clinical intervention studies and for clinical management. Recently, a large multicenter study has identified an algorithm of three serum biomarkers as predictive of severity and clinical outcome in Chronic Liver Disease.The algorithm,known as Enhanced Liver Fibrosis (ELF), includes the measurement of serum concentrations of hyaluronic acid, TIMP-1 and aminoterminal propeptide of procollagen type III. Objectives: To evaluate the predictive value of ELF test as surrogate outcome measure of fibrosis in SSc. Methods: 210 patients with SSc, all satisfying the ACR criteria for the classification of the disease, were enrolled in the study. 260 sera were analysed of which 90 were longitudinal samples from 40 patients. All patients were investigated for clinical and serological subset, disease duration, vascular, skin, joint, tendon, muscle, esophago-gastrointestinal, lung, heart and kidney involvement, disease severity, disease activity and HAQ-DI. ELF score was determined blindly by an independent commercial service (iQur, London, UK). Correlations were calculated using Spearman’s correlation test and an unpaired two-tailed T-test was used to perform comparison between groups. All the variables found to be correlated in univariate analysis were subsequently assessed by step-wise regression analysis. Data were analysed employing SPSS18 software. Results: The mean ELF score in SSc patients sera was 8.751.05 (normal range 5.97). ELF score correlated with many clinical and instrumental measures assessing fibrotic involvement including: mRSS (r 0.32;p 0.0001),DLCO (r0.29; p 0.0001), FVC % (r0.21; p 0.038), Ejection Fraction (r0.21; p0.0009). The other variables found to be correlated with ELF in univariate analysis were Age, ESR, CRP and FVC. Step-wise regression analysis indicated that the single measures independently associated with ELF score were mRSS, DLCO, EF and age whereas ESR, CRP and FVC were not. When compared to complex clinical indexes ELF score was also found to be correlated with EScSG-Activity Index (r 0.20; p 0.0015), total Medger’s disease severity score (r0.35;p0.0001) and total HAQ-DI (r0.32;p0.0001).Most interestingly, in the longitudinal samples ELF test showed a strong sensitivity to change, keeping the correlation with mRSS ( r0.34 p 0.0009), DLCO (r0.029; p0.029 ). In the same cohort of longitudinal samples ELF test was again correlated with total Severity (r0.49; p 0.0001), EScSG-Activity index ( r 0.21;p0.046), and HAQ-DI (r 0.42;p0.0001). Conclusion: The ELF test is a simple serum test that strongly correlates with several measures of fibrosis in SSc. It has a clear face validity for measuring the concentration of molecules involved in extracellular matrix turnover and strongly correlates with fibrotic severity and activity in SSc. The profound sensitivity to change with changes in mRSS, severity, activity and lung function indicate a clear discriminant validity of ELF as biomarker in SSc fibrotic involvement. ELF test should be included in the algorithm of activity index in Scleroderma and considered as outcome in clinical tria