Interstitial lung disease (ILD) is a heterogeneous group of diseases characterized by varying degrees of lung inflammation and/or fibrosis. We investigated biomarkers to infer whether patients with collagen vascular diseases associated ILD (CVD–ILD) and interstitial pneumonia with autoimmune features (IPAF) benefit from immunosuppressive therapy. We retrospectively investigated patients with CVD–ILD, IPAF, and idiopathic pulmonary fibrosis (IPF) between June 2013 and May 2017 at our department. First, we assessed differences in serum and bronchoalveolar lavage fluid (BALF) levels of cytokines between groups. Second, we assessed the associations of patient’s clinical variables with serum and BALF levels of those cytokines that were different between groups. Finally, we assessed the associations of diagnosis and response to immunosuppressive therapy with serum levels of those cytokines that were different between groups. We included 102 patients (51 with IPF, 35 with IPAF, and 16 with CVD–ILD). Serum and BALF levels of CXCL9, CXCL10, and CXCL11 were significantly elevated in patients with IPAF or CVD–ILD compared with those in patients with IPF. BALF levels of CXCL9 and CXCL10 were correlated with the percentages of lymphocytes and macrophages in BALF. Serum levels of CXCL9 and CXCL10 were correlated with BALF levels. Serum levels of CXCL9, CXCL10, and CXCL11 were correlated C-reactive protein, percent predicted forced vital capacity, alveolar-arterial oxygen difference, and the percentages of lymphocytes and macrophages in BALF. Serum levels of CXCL9, CXCL10, and CXCL11 showed moderate accuracy to distinguish patients with CVD–ILD from those with IPAF and IPF. Pre-treatment serum levels of CXCL9 and CXCL11 showed strong positive correlations with the annual forced vital capacity changes in patients with IPAF and CVD–ILD treated with immunosuppressive drugs. Serum CXCL9, CXCL10, and CXCL11 are potential biomarkers for autoimmune inflammation and predictors of the immunosuppressive therapy responses in ILD with background autoimmunity.

Idiopathic pulmonary fibrosis (IPF) is the most frequent and severe form of idiopathic interstitial pneumonias. Although IPF has not been thought to be associated with bacterial communities, recent papers reported the possible role of microbiome composition in IPF. The roles of microbiomes in respiratory functions and as clinical biomarkers for IPF remain unknown. In this study, we aim to identify the relationship between the microbial environment in the lung and clinical findings. Thirty-four subjects diagnosed with IPF were included in this analysis. The 16S rDNA was purified from bronchoalveolar lavage fluid obtained at the time of diagnosis and analyzed using next-generation sequencing techniques to characterize the bacterial communities. Furthermore, microbiomes from mice with bleomycin-induced lung fibrosis were analyzed. The most prevalent lung phyla were Firmicutes, Proteobacteria and Bacteroidetes. Decreased microbial diversity was found in patients with low forced vital capacity (FVC) and early mortality. Additionally, the diversity and relative abundance of Firmicutes, Streptococcaceae, and Veillonellaceae were significantly associated with FVC, 6-min walk distance, and serum surfactant protein D. Bleomycin-induced lung fibrosis resulted in decrease of diversity and alteration of microbiota in PCoA analysis. These results support the observations in human specimens. This study identified relationships between specific taxa in BALF and clinical findings, which were also supported by experiments in a mouse model. Our data suggest the possibility that loss of microbial diversity is associated with disease activities of IPF.

Surfactant proteins SP-A and SP-D are useful biomarkers in diagnosis, monitoring, and prognosis of idiopathic pulmonary fibrosis (IPF). Despite their high structural homology, their serum concentrations often vary in IPF patients. This retrospective study aimed to investigate distinct compartmentalization of SP-A and SP-D in the vasculature and lungs by bronchoalveolar lavage fluid (BALF)/serum analysis, hydrophilicity and immunohistochemistry. We included 36 IPF patients, 18 sarcoidosis (SAR) patients and 20 healthy subjects. Low-speed centrifugal supernatants of BALF (Sup-1) were obtained from each subject. Sera were also collected from each patient. Furthermore, we separated Sup-1 of IPF patients into hydrophilic supernatant (Sup-2) and hydrophobic precipitate (Ppt) by high-speed centrifugation. We measured SP-A and SP-D levels of each sample with the sandwich ELISA technique. We analyzed the change of the BALF/serum level ratios of the two proteins in IPF patients and their hydrophilicity in BALF. The distribution in the IPF lungs was also examined by immunohistochemical staining. In BALF, SP-A levels were comparable between the groups; however, SP-D levels were significantly lower in IPF patients than in others. Although IPF reduced the BALF/serum level ratios of the two proteins, the change in concentration of SP-D was more evident than SP-A. This suggests a higher disease impact for SP-D. Regarding hydrophilicity, although more than half of the SP-D remained in hydrophilic fractions (Sup-2), almost all of the SP-A sedimented in the Ppt with phospholipids. Hydrophilicity suggests that SP-D migrates into the blood more easily than SP-A in IPF lungs. Immunohistochemistry revealed that SP-A was confined to thick mucus-filling alveolar space, whereas SP-D was often intravascular. This data also suggests that SP-D easily leaks into the bloodstream, whereas SP-A remains bound to surfactant lipids in the alveolar space. The current study investigated distinct compartmentalization of SP-A and SP-D in the vasculature and lungs. Our results suggest that serum levels of SP-D could reflect pathological changes of the IPF lungs more incisively than those of SP-A.

Background: Antifibrotic therapies are available to treat chronic fibrosing interstitial lung diseases (CF-ILDs), including idiopathic pulmonary fibrosis. Early use of these treatments is recommended to slow deterioration of respiratory function and to prevent acute exacerbation. However, identifying patients in the early stages of CF-ILD using chest radiographs is challenging. In this study, we developed and tested a deep-learning algorithm to detect CF-ILD using chest radiograph images.Method: From the image archive of Sapporo Medical University Hospital, 653 chest radiographs from 263 patients with CF-ILDs and 506 from 506 patients without CF-ILD were identified; 921 were used for deep learning and 238 were used for algorithm testing. The algorithm was designed to output a numerical score ranging from 0 to 1, representing the probability of CF-ILD. Using the testing dataset, the algorithm's capability to identify CF-ILD was compared with that of doctors. A second dataset, in which CF-ILD was confirmed using computed tomography images, was used to further evaluate the algorithm's performance.Results: The area under the receiver operating characteristic curve, which indicates the algorithm's detection capability, was 0.979. Using a score cut-off of 0.267, the sensitivity and specificity of detection were 0.896 and 1.000, respectively. These data showed that the algorithm's performance was noninferior to that of doctors, including pulmonologists and radiologists; performance was verified using the second dataset.Conclusions: We developed a deep-learning algorithm to detect CF-ILDs using chest radiograph images. The algorithm's detection capability was noninferior to that of doctors.