IS IT POSSIBLE TO IDENTIFY A PHENOTYPE OF ILD PATIENTS WITH A HIGHER RISK OF MORTALITY?

<bold>Background:</bold> In patients with interstitial lung diseases (ILD), some clinical outcomes are associated with mortality. However, it is unknown whether combining these outcomes into groups (i.e. clusters) would help in identifying patients at high risk of mortality. <bold>Aim:</bold> To identify clusters of patients with ILD and verify their association with one-year mortality. <bold>Methods:</bold> Patients diagnosed with ILD underwent assessment of pulmonary function (FVC, DLCO), exercise capacity (6MWT), levels of physical activity in daily life ([PADL] time in moderate activity/day; steps/day; time lying down and standing/day), and sensation of dyspnea in daily life (MRC). K-means cluster analysis was used to stratify patients with ILD, according to similarities in the outcomes investigated. The association of clusters with one-year mortality was verified by Kaplan–Meier analysis using the log-rank test. Finally, the chance of death in one year per cluster was performed using odds-ratio (OR) analysis. <bold>Results:</bold> K-means Cluster analysis identified two clusters. Patients from Cluster 1 (CL 1, n= 27, 64±10 years, 18[66%] women, BMI=27±5 kg/m²) had worse outcomes when compared to those from Cluster 2 (CL2, n=26, 58 ±10 years, 15[58%] women, BMI=28±5 kg/m²). During the one-year follow-up, 16 (30%) patients died, 12 in CL1 (44%) and 4 in CL2 (15%) (p=0.002). There was a significant difference between CL1 and CL2 in the Kaplan-Meier curve (log-rank p=0.009). Finally, patients in CL2 had a 65.3% lower chance of death within one year than patients in CL1 (OR:0.34, p<0.01). <bold>Conclusion:</bold> ILD patients with worse lung function, exercise capacity and lower PADL levels have a higher risk of mortality over a period of one year.