Purpose: The molecular classification of breast cancer is crucial for effective treatment. The emergence of digital pathology has ushered in a new era in which weakly supervised learning leveraging whole-slide images has gained prominence in developing deep learning models because this approach alleviates the need for extensive manual annotation. Weakly supervised learning was employed to classify the molecular subtypes of breast cancer. Materials and Methods: Our approach capitalizes on two whole-slide image datasets: one consisting of breast cancer cases from the Korea University Guro Hospital (KG) and the other originating from The Cancer Genomic Atlas dataset (TCGA). Furthermore, we visualized the inferred results using an attention-based heat map and reviewed the histomorphological features of the most attentive patches. Results: The KG+TCGA-trained model achieved an area under the receiver operating characteristics value of 0.749. An inherent challenge lies in the imbalance among subtypes. Additionally, discrepancies between the two datasets resulted in different molecular subtype proportions. To mitigate this imbalance, we merged the two datasets, and the resulting model exhibited improved performance. The attentive patches correlated well with widely recognized histomorphologic features. The triple-negative subtype has a high incidence of high-grade nuclei, tumor necrosis, and intratumoral tumor-infiltrating lymphocytes. The luminal A subtype showed a high incidence of collagen fibers. Conclusion The artificial intelligence (AI) model based on weakly supervised learning showed promising performance. A review of the most attentive patches provided insights into the predictions of the AI model. AI models can become invaluable screening tools that reduce costs and workloads in practice.

Purpose: The molecular classification of breast cancer is crucial for effective treatment. The emergence of digital pathology has ushered in a new era in which weakly supervised learning leveraging whole-slide images has gained prominence in developing deep learning models because this approach alleviates the need for extensive manual annotation. Weakly supervised learning was employed to classify the molecular subtypes of breast cancer. Materials and Methods: Our approach capitalizes on two whole-slide image datasets: one consisting of breast cancer cases from the Korea University Guro Hospital (KG) and the other originating from The Cancer Genomic Atlas dataset (TCGA). Furthermore, we visualized the inferred results using an attention-based heat map and reviewed the histomorphological features of the most attentive patches. Results: The KG+TCGA-trained model achieved an area under the receiver operating characteristics value of 0.749. An inherent challenge lies in the imbalance among subtypes. Additionally, discrepancies between the two datasets resulted in different molecular subtype proportions. To mitigate this imbalance, we merged the two datasets, and the resulting model exhibited improved performance. The attentive patches correlated well with widely recognized histomorphologic features. The triple-negative subtype has a high incidence of high-grade nuclei, tumor necrosis, and intratumoral tumor-infiltrating lymphocytes. The luminal A subtype showed a high incidence of collagen fibers. Conclusion The artificial intelligence (AI) model based on weakly supervised learning showed promising performance. A review of the most attentive patches provided insights into the predictions of the AI model. AI models can become invaluable screening tools that reduce costs and workloads in practice.

Purpose: The molecular classification of breast cancer is crucial for effective treatment. The emergence of digital pathology has ushered in a new era in which weakly supervised learning leveraging whole-slide images has gained prominence in developing deep learning models because this approach alleviates the need for extensive manual annotation. Weakly supervised learning was employed to classify the molecular subtypes of breast cancer. Materials and Methods: Our approach capitalizes on two whole-slide image datasets: one consisting of breast cancer cases from the Korea University Guro Hospital (KG) and the other originating from The Cancer Genomic Atlas dataset (TCGA). Furthermore, we visualized the inferred results using an attention-based heat map and reviewed the histomorphological features of the most attentive patches. Results: The KG+TCGA-trained model achieved an area under the receiver operating characteristics value of 0.749. An inherent challenge lies in the imbalance among subtypes. Additionally, discrepancies between the two datasets resulted in different molecular subtype proportions. To mitigate this imbalance, we merged the two datasets, and the resulting model exhibited improved performance. The attentive patches correlated well with widely recognized histomorphologic features. The triple-negative subtype has a high incidence of high-grade nuclei, tumor necrosis, and intratumoral tumor-infiltrating lymphocytes. The luminal A subtype showed a high incidence of collagen fibers. Conclusion The artificial intelligence (AI) model based on weakly supervised learning showed promising performance. A review of the most attentive patches provided insights into the predictions of the AI model. AI models can become invaluable screening tools that reduce costs and workloads in practice.

This review explores spatial mapping technologies in cancer research, highlighting their crucial role in understanding the complexities of the tumor microenvironment (TME). The TME, which is an intricate ecosystem of diverse cell types, has a significant impact on tumor dynamics and treatment outcomes. This review closely examines cutting-edge spatial mapping technologies, categorizing them into capture-, imaging-, and antibody-based approaches. Each technology was scrutinized for its advantages and disadvantages, factoring in aspects such as spatial profiling area, multiplexing capabilities, and resolution. Additionally, we draw attention to the nuanced choices researchers face, with capture-based methods lending themselves to hypothesis generation, and imaging/antibody-based methods that fit neatly into hypothesis testing. Looking ahead, we anticipate a scenario in which multi-omics data are seamlessly integrated, artificial intelligence enhances data analysis, and spatiotemporal profiling opens up new dimensions.

Gastrointestinal stromal tumors (GISTs) are rare digestive system malignancies with extragastrointestinal stromal tumors (EGISTs) being even less. Diagnosing GISTs usually requires the identification of c-kit (CD117) expression by immunohistochemical staining. A 53-year-old woman complaining of dyspepsia was referred for the evaluation of a 1.5-cm extrinsic compression at the greater curvature of the proximal antrum. EUS revealed a multiseptated mass with positive Doppler findings. Abdominal CT showed that she harbored a large, 20-cm mass in her abdominal cavity, most likely arising from the right ovary. Surgery revealed a hypervascular tumor arising from the mesentery and attached to the gastric lesser curvature. Pathological examination revealed negativity for c-kit, but positivity for the protein “Discovered on GIST-1” (DOG1), confirming the EGIST diagnosis. Herein, we report this rare case of a c-kit-negative EGIST originating in the mesentery, which was diagnosed based on staining for DOG1.

Gastrointestinal stromal tumors (GISTs) are rare digestive system malignancies with extragastrointestinal stromal tumors (EGISTs) being even less. Diagnosing GISTs usually requires the identification of c-kit (CD117) expression by immunohistochemical staining. A 53-year-old woman complaining of dyspepsia was referred for the evaluation of a 1.5-cm extrinsic compression at the greater curvature of the proximal antrum. EUS revealed a multiseptated mass with positive Doppler findings. Abdominal CT showed that she harbored a large, 20-cm mass in her abdominal cavity, most likely arising from the right ovary. Surgery revealed a hypervascular tumor arising from the mesentery and attached to the gastric lesser curvature. Pathological examination revealed negativity for c-kit, but positivity for the protein “Discovered on GIST-1” (DOG1), confirming the EGIST diagnosis. Herein, we report this rare case of a c-kit-negative EGIST originating in the mesentery, which was diagnosed based on staining for DOG1.

Purpose: The most recent 2017 World Health Organization (WHO) classification of pancreatic neuroendocrine neoplasms (PanNENs) has refined the three-tiered 2010 scheme by separating grade 3 pancreatic neuroendocrine tumors (G3 PanNETs) from poorly differentiated pancreatic neuroendocrine carcinomas (PanNECs). However, differentiating between G3 Pan- NETs and PanNECs is difficult in clinical practice. Materials and Methods: Eighty-two surgically resected PanNENs were collected from 16 institutions and reclassified according to the 2017 WHO classification based on the histological features and proliferation index (mitosis and Ki-67). Immunohistochemical stains for ATRX, DAXX, retinoblastoma, p53, Smad4, p16, and MUC1 were performed for 15 high-grade PanNENs. Results: Re-classification resulted in 20 G1 PanNETs (24%), 47 G2 PanNETs (57%), eight G3 well-differentiated PanNETs (10%), and seven poorly differentiated PanNECs (9%). PanNECs showed more frequent diffuse nuclear atypia, solid growth patterns and apoptosis, less frequent organoid growth and regular vascular patterns, and absence of low-grade PanNET components than PanNETs. The Ki-67 index was significantly higher in PanNEC (58.2%± 15.1%) compared to G3 PanNET (22.6%±6.1%, p < 0.001). Abnormal expression of any two of p53, p16, MUC1, and Smad4 could discriminate PanNECs from G3 PanNETs with 100% specificity and 87.5% sensitivity. Conclusion Histological features supporting the diagnosis of PanNECs over G3 PanNETs were the absence of a low-grade PanNET component in the tumor, the presence of diffuse marked nuclear atypia, solid growth pattern, frequent apoptosis and markedly increased proliferative activity with homogeneous Ki-67 labeling. Immunohistochemical stains for p53, p16, MUC1, and Smad4 may be helpful in distinguishing PanNECs from G3 PanNETs in histologically ambiguous cases, especially in diagnostic practice when only small biopsied tissues are available.

BACKGROUND/AIMS: Gastric schwannomas are rare benign mesenchymal tumors that are difficult to differentiate from other mesenchymal tumors with malignant potential, such as gastrointestinal stromal tumors. This study aimed to evaluate the characteristic findings of gastric schwannomas via endoscopic ultrasonography (EUS). METHODS: We retrospectively reviewed the EUS findings of 27 gastric schwannoma cases that underwent surgical excision at Pusan National University Hospital during 2007 to 2014. RESULTS: Gastric schwannomas were mainly located in the middle third of the stomach with a mean tumor size of 32 mm. All lesions exhibited hypoechoic echogenicity, and 24 lesions (88.9%) exhibited heterogeneous echogenicity. Seventeen lesions (63.0%) exhibited decreased echogenicity compared to the normal proper muscle layer. Distinct borders were observed in 24 lesions (88.9%), lobulated margins were observed in six lesions (22.2%), and marginal haloes were observed in 24 lesions (88.9%). Hyperechogenic spots were observed in 21 lesions (77.8%), calcifications were observed in one lesion (3.7%), and cystic changes were observed in two lesions (7.4%). CONCLUSIONS: During EUS, gastric schwannomas appear as heterogeneously hypoechoic lesions with decreased echogenicity compared to the normal proper muscle layer. These features may be helpful for differentiating gastric schwannomas from other mesenchymal tumors.
Busan ; Endosonography ; Gastrointestinal Stromal Tumors ; Neurilemmoma* ; Retrospective Studies ; Stomach

No abstract available.
Adenocarcinoma* ; Carcinoma, Papillary* ; Neoplasm Metastasis* ; Thyroid Gland*