Background VEDOSS project relied on a specific set of inclusion criteria identifying patients in very early stage of systemic sclerosis (SSc) to be submitted to a tight follow up and unravel factors linked to the progression of the disease. The components of the ELF score (PIIINP, TIMP-1 and HA) have been shown to correlate with severity of skin and lung fibrosis in SSc.
Objectives Our aim was to analyze the concentrations of ELF components in a subset of patients enrolled in the VEDOSS study and determine their potential diagnostic value.
Methods Sera from 114 patients enrolled into the VEDOSS data base and 67 SSc controls fulfilling ACR/EULAR 2013 criteria (subsets: 33 diffuse; 34 limited) were obtained from 2 centres. Serum concentrations of ELF components were determined on Siemens Advia Centaur platform. Logistic regression analysis of the results was performed employing classification of SSc as state variable.
Results Among the 114 patients enrolled from VEDOSS, 30 subjects had primary Raynaud's phenomenon (PRP); 54 had Raynaud's phenomenon (RP) and at least one of the following “red flags” (risk factors) for SSc: (Puffy fingers, ANA, SSc specific autoantibodies or SSc-specific capillaroscopic pattern) without fulfilling 2013 ACR/EULAR criteria (score <9). 30 VEDOSS patients fulfilled new ACR/EULAR criteria (score >9) despite having no evidence of skin involvement nor interstitial lung disease at lung high resolution CT or pulmonary arterial hypertension, left ventricular systolic impairment or renal involvement.
All serum biomarkers concentrations correlated with age (p<0.05 for all). Logistic regression analysis using the three biomarkers as variables identified a specific model (SSc score) ranging from -3.92 to 5.77. The new score showed good ability to discriminate between patients with SSc (VEDOSS patients already classified as SSc and SSc controls) versus VEDOSS patients not yet classified as SSc (AOUC under ROC curve: 0.853, 95%CI 0.797–0.910). Within the VEDOSS database, SSc score showed a fair ability to discriminate between VEDOSS patients fulfilling SSc classification criteria, despite lack of skin internal organ involvement (lung, heart and kidneys), and those not fulfilling the criteria (AOUC =0.756, 95%CI 0.646–0.865). Indeed, patients fulfilling SSc classification criteria, had average score of 0.44 vs -0.86 of patients not fulfilling the criteria and the score remained significant even after correcting for age (p=0.026). Furthermore, SSc controls had significantly higher SSc score compared to VEDOSS patients already classified as SSc, confirmed after age correction (p<0.000). No difference in SSc score has been observed between PRP and VEDOSS patients not yet classified as SSc.
Conclusions Our data indicate that the SSc score is a simple serum test based model that can be used in patients with RP to aid in the very early diagnosis of SSc. Furthermore, the identification of VEDOSS patients with a high SSc score test and already classified as SSc even in the absence of internal organ involvement can be used in intervention trials aimed to prevent further disease progression.
Disclosure of Interest None declared