A noval identification of 4 systemic sclerosis - interstitial lung disease subgroups using principal component analysis-based cluster analysis

ObjectiveInterstitial lung disease (ILD) is a common and serious complication of systemic sclerosis (SSc). It is usually classified by histologic type, but this classification may not fully reflect the clinical phenotypic variation. This study aimed to examine clinical features and aggregate patients with SSc-ILD based on patients' clinical manifestations, High-resolution computed tomography (HRCT) features and specific antibody expression to achieve precise treatment of SSc-ILD with early identification of complications.MethodsThis study included 103 patients with SSc-ILD. A cluster analysis was performed based on five clinical and serological variables to identify subgroups of patients. The survival rates between obtained clusters and risk factors affecting prognosis were also compared.ResultFour clusters were identified in this study: Cluster 1 (n = 23) represented the lymphocytic interstitial pneumonia (LIP) group with LIP as the predominant HRCT characteristic. Cluster 2 (n = 23) was the worst prognosis group, with the highest Warrick score as well as the highest mortality rate. Cluster 3 (n = 20) with all patients having a negative anti-SCL-70 antibody response. Cluster 4 (n = 28) with all patients were positive for the anti-SCL-70 antibody. It was found that albumin was a protective factor for the prognosis of patients with SSC-ILD patients (p = 0.018), whereas age (p = 0.036) and IgM (p = 0.040) were risk factors.ConclusionThe results of our cluster analysis indicated that based solely on histologic typing, may not be capturing the full heterogeneity of SSc-ILD patients. In order to identify homogeneous patient groups with a specific prognosis, HRCT features and antibody profiles should be taken into consideration.