Patients and study design
This was a single-center, retrospective, matched case-control study that included patients with non-small cell lung cancer with IIP who underwent pulmonary resection at Juntendo University between January 2009 and June 2018. The data presented here are from an observational database of surgical candidates at Juntendo Hospital.
This study was approved by the ethics committee of Juntendo University School of Medicine (No. 2,016,095) and performed in accordance with the guidelines of the Declaration of Helsinki and its subsequent amendments. Informed consent was obtained in the form of an opt-out method on the website. This study was supported by JSPS KAKENHI (grant number 21K16523).
Control patients
In total, 135 patients were pathologically diagnosed with IIP. Patients were divided into the AE (n = 13, 9.6%) and non-AE (n = 122, 90.4%) groups. Thirteen cases of AE-IIP and 122 cases of non-AE were matched using a propensity score analysis, and 12 cases in each group were compared.
Immunohistochemical analysis
The subpleural lung tissues without cancer were used in this study, which was collected for the diagnosis of ILD and stored in a paraffin block. We evaluated immunohistochemical staining for CD3 and dual staining for CD4 and CD8 and also for CD4 and dual staining for T-bet, GATA3, STAT3, FOXP3, and CD8 in all non-cancer specimens.
After deparaffinization, monoclonal antibodies against CD3, CD4, CD8, T-bet, and GATA3 were added to the autostaining device. Immunohistochemical staining was performed using a commercially available autostaining device (VENTANA BenchMark GX; Roche Diagnostics K.K., Tokyo, Japan), according to the manufacturer’s protocol. Monoclonal antibodies against STAT3 (Lsbio, Seattle, WA, USA) and FOXP3 (Abcam, Waltham, MA, USA) were used for staining. A dual-labeled cell was defined by the presence of cytoplasmic staining for CD4 (red chromogen), CD3, T-bet, GATA3, STAT3, FOXP3, and CD8 (brown chromogen) and nuclear staining for the interrogated transcription factor (blue chromogen). Slides were visualized using an Olympus AX73 (Olympus Corporation, Tokyo, Japan) microscope. The dual-labeled CD3+/CD4+, CD4+/T-bet+, CD4+/GATA3+, CD4+/STAT3+, CD4+/ FOXP3+, and CD8+ cells were counted in over three independent areas, with the greatest abundance of lymphocytes (e-Fig. 1). T helper (Th)1 cells, Th2 cells, Th17 cells, and Tregs were identified as CD4+ T-bet+ cells, CD4+ GATA3+ cells, CD4+ STAT3+ cells, and CD4+ FOXP3+ cells, respectively. Any visible pigment, including weak or pale staining, indicated positive staining. Cell counts were performed using Image J (NIH, Bethesda, ML, USA). MF, who was blinded to the clinical backgrounds and courses of the patients, performed the immunohistochemical evaluation, and the results were later checked against clinical data. If the difference from other data was > 20%, the patient was re-evaluated. The mean number of cells counted per high-power field was used to calculate the percentages of Th1, Th2, Th17, Treg, and CD8+ cells.
Statistical analysis
Clinicopathological features of patients with ILD with or without AE after surgery were investigated. Continuous variables were compared using Student’s t-test, Welch’s t-test, and the Mann − Whitney U test. Categorical variables were compared using the chi-squared test and Fisher’s exact test. We used a propensity score matching approach to control for clinical selection bias in the two groups when comparing the AE and non-AE immune samples. The propensity score was calculated based on the distribution of IIP, UIP pattern, preoperative arterial oxygen partial pressure, %VC, %DLCO, and the surgical procedure as independent variables. Patients were matched 1:1 using nearest-neighbor matching (caliper width:2 times the standard deviation of the propensity score on the logit scale) without replacement.
All statistical analyses were performed using IBM SPSS Statistics for Windows version 25.0 (IBM Corp., Armonk, NY, USA). P-values < 0.05 were considered statistically significant.