A new computed tomography index for quantification of interstitial lung disease in systemic sclerosis is associated with lung function parameters in the short term follow-up

Background: New computer-assisted methods for the objective quantification of interstitial lung disease (ILD) at computed tomography (CT), based on the evaluation of mean lung attenuation (MLA), skewness and kurtosis have been recently investigated in Systemic Sclerosis (SSc). We developed a computerised integrated index (CII) based on a weighted evaluation of MLA, skewness and kurtosis and investigated its reliability for the quantitative assessment of SSc-ILD and its associations with lung function parameters in a cross-sectional study. Objectives: To identify the CII cut off value with the highest sensitivity and specificity for CT-detected ILD and to investigate its impact on lung function parameters over-time of baseline assessed CII. Methods: SSc patients meeting the new ACR/EULAR classification criteria, who had undergone a volumetric lung CT study from July 1 st 2014 to June 30th 2015, had been evaluated at baseline for ILD quantification by Goh et al. method and the previously referred dedicated software and had their CII calculated, were enrolled in a prospective study including complete clinical, serological, and functional assessment at baseline and at 1 year follow-up (FU). Results: Thirty-nine out of 83 (47%) SSc patients (79 females, 4 males; mean age 56.4±11.3 years; median disease duration 12 years; 18 diffuse cutaneous and 65 limited cutaneous SSc) had evidence of ILD as assessed by volumetric CT of the lungs at baseline. CII in patients with ILD was significantly lower than in those without ILD (−0.4929±0.9933 versus 0.4145±0.8059 HU; p<0.0001). ROC analysis revealed that the best discriminating CII value for ILD was 0.1966: sensitivity 0.81 (95% C.I. 0.68 to 0.92); specificity 0.66 (95% C.I. 0.52 to 0.80). Out of the 44 ILD negative patients, 22 (50%) presented a CII value lower than the cut-off, and 13 of them (59%) were found to have a diffusing lung capacity for CO (DLCO)<80% of predicted. At 1 year FU, the CII was significantly correlated with total lung capacity -TLC (r=0.45, p=0.004) and DLCO (r=0.29, p=0.045). Out of the 22 patients with a CII <0.1966 but no ILD at visual evaluation, 11 (50%) developed a FVC decline at 1 year, and 8 (36.7%) a DLCO decline. Conclusions: Here we confirm that quantitative computer-assisted CT of the lungs could be a reliable method for SSc-ILD evaluation and found that it could also be useful in predicting the evolution of lung function in the short-term FU. Disclosure of Interest: None declared DOI: 10.1136/annrheumdis-2018-eular.7559 Citation: Ann Rheum Dis, volume 77, supplement Suppl, year 2018, page A760Session: Scleroderma, myositis and related syndromes