Inage Terunaga

OBJECTIVE: Endobronchial ultrasound-guided transbronchial needle aspiration is a vital tool for mediastinal and hilar lymph node staging in patients with lung cancer. Despite its high diagnostic performance and safety, it has a limited negative predictive value. Our objective was to evaluate the diagnostic performance of deep learning-based prediction of lung cancer lymph node metastases using convolutional neural networks developed from automatically extracted images of endobronchial ultrasound videos without supervision of the lymph node location. METHODS: Patient and lymph node data were collected from a single-center database. The diagnosis of metastasis was confirmed with endobronchial ultrasound-guided transbronchial needle aspiration and/or surgically resected specimens; the diagnosis of normal lymph node was confirmed with surgically resected specimens only. An annotation system facilitated automated image extraction from endobronchial ultrasound videos. Image frames were randomly selected and split into training and validation datasets on a per-patient basis. A deep learning model with convolutional neural networks, SqueezeNet, was used for image classification via transfer learning based on pretraining from ImageNet. Adaptive moment estimation and stochastic gradient descent were applied as optimizers. RESULTS: SqueezeNet, with adaptive moment estimation, achieved a sensitivity, specificity, accuracy, positive predictive value, and negative predictive value of 96.7% each after 300 epochs, whereas SqueezeNet with stochastic gradient descent achieved 91.1% each. However, SqueezeNet with stochastic gradient descent demonstrated more stable performance than with adaptive moment estimation. CONCLUSIONS: Deep learning-based image classification using convolutional neural networks showed promising diagnostic accuracy for lung cancer nodal metastasis. Future clinical trials are warranted to validate the algorithm's efficacy in a prospective, large-cohort study.

BACKGROUND: According to a large-scale clinical trial in Japan, segmentectomy for small peripheral non-small cell lung cancer has an advantage over lobectomy in terms of overall survival, while it could also increase the incidence of local recurrence. In ipsilateral reoperations, intrathoracic adhesions from a previous surgery increase the risk of lung injury and bleeding, which may result in intraoperative and postoperative complications. The ability of oxidized regenerated cellulose (ORC) sheets to prevent postoperative adhesions has been demonstrated in the abdomen, and the same effect is expected in the thoracic region. The purpose of this study is to provide evidence supporting the application of ORC sheets to the parietal pleura of an open chest wounds to prevent postoperative adhesions in the thoracic region. METHODS: This phase II prospective open-label, randomized, parallel-group study will validate adhesion prevention by applying ORC sheets to the parietal pleura of open chest wounds at the time of surgical closure. In the control group, the chest is closed by the usual procedure without ORC sheets. The primary endpoint is the presence rate of pleural adhesion findings on chest echography performed 4-20 weeks postoperatively. Data analysis will be performed in 2025-2026. DISCUSSION: This study will provide evidence to the adhesion prevention effect of ORC sheet in the thoracic region, with the aim of establishing a strategy to prevent postoperative intrapleural adhesions. TRIAL REGISTRATION: This trial has been registered on the Japan Registry of Clinical Trials 1032230271 (https://jrct.niph.go.jp/latest-detail/jRCT1032230271).

BACKGROUND: Home oxygen therapy (HOT) is used to treat chronic respiratory diseases and is sometimes required in patients with lung cancer after radical surgery. We aimed to identify the risk factors for postoperative home-based oxygen therapy in patients with lung cancer. METHODS: Patients who underwent surgery for primary lung cancer at Chiba University Hospital between January 2019 and March 2021 were included. Patients who did not undergo complete resection, died in hospital after surgery, or used oxygen therapy preoperatively were excluded. Eligible patients were divided into HOT and non-HOT groups. They were retrospectively analyzed for risk factors for postoperative HOT using medical records in a multivariate analysis. RESULTS: A total of 410 patients were included in this study, 24 (5.9%) of whom required HOT after surgery. The HOT group comprised significantly more men, heavy smokers, and patients with pulmonary comorbidities, low percent forced expiratory volume, percent forced vital capacity, predicted postoperative forced expiratory volume in 1 s, and postoperative pulmonary complications on univariate analysis. In a multivariate analysis, independent risk factors for postoperative HOT were pulmonary comorbidities [odds ratio (OR): 5.94; 95% confidence interval (CI): 1.64-21.5; P=0.002) and postoperative pulmonary complications (OR: 5.39; 95% CI: 2.14-13.5; P<0.001). The postoperative HOT application rate was calculated according to a formula developed for this purpose. CONCLUSIONS: Comorbid pulmonary diseases and postoperative pulmonary complications were significantly associated with postoperative HOT in patients with lung cancer.