From body composition to pulmonary fat attenuation volume: evaluation of new CT-based biomarkers to predict the overall survival in patients with non-specific interstitial pneumonia (NSIP).

<bold>Rationale:</bold> Non-specific interstitial pneumonia (NSIP) is an interstitial lung disease that can result in end-stage fibrosis. Body composition analysis (BCA) and CT pulmonary fat attenuation volume (CTpfav) have emerged as predictive tools of overall survival (OS) in IPF. Our study was aimed at investigating the association of BCA derived markers and CTpfav with OS in NSIP patients. <bold>Patients and Methods:</bold> In this monocentric retrospective study, 71 NSIP patients (55 % female) were studied. Fully automated BCA was performed using an open-source Body and Organ Segmentation framework based on a nnU-Net. Features were aggregated and combined into Sarcopenia, Fat, Myosteatosis and Mediastinal Fat index. The Pulmonary Fat index was created by dividing CTpfav by the lunge volume. All indices' expression was compared by categorizing the cohort into patients who died within five years and those who did not. <bold>Results:</bold> Thirty-eight patients (54%) died during follow up. A higher (> median) Sarcopenia and lower (< median) Mediastinal Fat Index were significantly associated with a better survival rate after 48 months (p<0.001). Patients with a higher Pulmonary Fat Index had a significantly lower survival rate compared to those with lower Index (p=0.0034). In the adjusted multivariate Cox regression, the Pulmonary Fat Index was a significant independent predictor of survival (HR=2.37, 95% CI=1.03–5.48, p=0.043). <bold>Conclusion:</bold> Fully automated BCA and CTpfav provide promising imaging markers, like the Pulmonary Fat Index, to predict overall survival in patients with NSIP.