A screening tool to detect interstitial lung disease in systemic sclerosis: the ILD-RISC score

Authors:
Open Access
Publisher website
Google Scholar
Abstract
Objectives: As a screening tool for the presence of systemic sclerosis-associated interstitial lung disease (SSc-ILD) on high-resolution computed tomography (HRCT) is missing, we aimed to develop the ILD-RISC score, a risk algorithm to guide physicians in ordering HRCTs, with specific focus on follow-up visits. Methods: The nominal group technique was used to select items for the multivariable logistic regression with backward selection. The ILD-RISC score was developed from baseline visits of the derivation cohort. After identifying a cut-off favoring sensitivity >85% from the ROC curve, it was validated in a separate cross-sectional cohort, and then applied longitudinally in a specific SSc cohort with negative baseline HRCT. Results: In the derivation cohort (533 patients), 13 variables associated with the presence of SSc-ILD on HRCT were tested. The ILD-RISC score, including FVC%, DLCO/SB%, digital ulcers ever, age and SSc autoantibodies, showed an area under the curve of 79.1% (75.3–83.0%) for the presence of SSc-ILD on HRCT. An ILD-RISC score ≥0.3 had sensitivity 85.6% and specificity 53.6%, as confirmed in the validation cohort (247 patients). Among 819 patients with negative baseline HRCT, 170 developed SSc-ILD: a low ILD-RISC score was detected in almost 50% of the follow-up visits, supporting the sparing of HRCTs. Conclusion: The ILD-RISC score was developed and validated to predict the presence of SSc-ILD. Thus, the ILD-RISC score may help to decide when to order HRCTs at follow-up, allowing the reduction of costs and radiation exposure, but also at the time of SSc diagnosis when resources are limited.
Funding Information
  • Gruppo Italiano Lotta alla Sclerodermia
  • Novartis Foundation for Medical-biological Research
  • Boehringer Ingelheim
  • AstraZeneca
  • Scleroderma Research Foundation (US8247389, EP2331143)

This publication has 30 references indexed in Scilit: