Deep Learning Analysis for Predicting Tumor Spread through Air Space in Early-Stage Lung Adenocarcinoma Pathology Images.

The presence of spread through air spaces (STASs) in early-stage lung adenocarcinoma is a significant prognostic factor associated with disease recurrence and poor outcomes. Although current STAS detection methods rely on pathological examinations, the advent of artificial intelligence (AI) offers opportunities for automated histopathological image analysis. This study developed a deep learning (DL) model for STAS prediction and investigated the correlation between the prediction results and patient outcomes. To develop the DL-based STAS prediction model, 1053 digital pathology whole-slide images (WSIs) from the competition dataset were enrolled in the training set, and 227 WSIs from the National Taiwan University Hospital were enrolled for external validation. A YOLOv5-based framework comprising preprocessing, candidate detection, false-positive reduction, and patient-based prediction was proposed for STAS prediction. The model achieved an area under the curve (AUC) of 0.83 in predicting STAS presence, with 72% accuracy, 81% sensitivity, and 63% specificity. Additionally, the DL model demonstrated a prognostic value in disease-free survival compared to that of pathological evaluation. These findings suggest that DL-based STAS prediction could serve as an adjunctive screening tool and facilitate clinical decision-making in patients with early-stage lung adenocarcinoma.

Computed Tomography-Guided Localization and Extended Segmentectomy for Non-Small Cell Lung Cancer.

Background: Lung cancer is one of the most devastating cancers. Low-dose computed tomography (LDCT) can detect lung cancer at an early stage of the disease when a minimally invasive surgical procedure using video-assisted thoracoscopic surgery is the best strategy. Herein, we discuss the treatment of deep lung tumors between segments or lesions located near the margin of a segment. Patients and Methods: This was a retrospective study conducted from January 2013 to January 2020 using the National Taiwan University Hospital data bank. We included early-stage non-small cell lung cancer (NSCLC) patients who underwent lung surgery and screened out those who received CT-guided localization for extended segmentectomy. Outcome measurements were safety margin, complication rate, and postoperative course. Results: During the study period, 68 patients with early-stage NSCLC received CT-guided localization followed by extended segmentectomy. The mean surgery time was 92.1 ± 30.3 min, and the mean blood loss was 32.8 mL. Mean drainage time was 2.3 ± 1 days, and the total hospital stay was 4.9 ± 1.1 days. Pathological reports showed tumor-free resection margins >2 cm. Sixty-one patients had adenocarcinoma at stage IA and two patients at stage IB. One patient had squamous cell carcinoma at stage IA. Conclusion: CT-guided localization followed by extended segmentectomy allows lung volume preservation with clean safety margins and good clinical outcomes.

Primary Breast Neuroendocrine Tumors: An Analysis of the National Cancer Database.

BACKGROUND: Primary breast neuroendocrine tumors (BNETs) represent < 1% of breast cancers. Diagnosing BNETs can be challenging, and a limited amount of cohort data currently exists in literature. We aimed to describe primary BNET characteristics, treatment modalities, and survival outcomes through the National Cancer Database (NCDB). METHODS: A retrospective cohort analysis was performed using the NCDB from 2004 to 2017. BNET cases were compared with patients with invasive ductal carcinoma (IDC). A matched IDC cohort was created by matching patient age, race, and disease stage. Kaplan-Meier analysis was performed, and hazard ratios (HR) were calculated through the bootstrap sampling method. RESULTS: A total of 1389 BNET and 1,967,401 IDC cases were identified. When compared with IDC patients, BNET patients were older, had more comorbidities, and were more often male (p < 0.01). BNETs were larger, higher grade, and more frequently hormone receptor negative (p < 0.01). While BNET patients were treated with surgery and radiotherapy (p < 0.01) less often compared with IDC patients, they presented at later disease stage (p < 0.001) and received systemic treatment more frequently (53.5% vs. 40%, p < 0.01). Patients with BNET had increased mortality compared with the matched IDC cohort: stage 1 HR 1.8, stage 2 HR 2.0, stage 3 HR 1.8, and stage 4 HR 1.5 (p < 0.001 for all). CONCLUSION: Patients with BNET tend to present at higher clinical stages, are more frequently hormone receptor negative, and have inferior overall survival compared with patients with IDC. Further treatment strategies and studies are needed to elucidate optimal therapies to maximize patient outcomes.