Improving fine needle aspiration to predict the tumor biological aggressiveness in pancreatic neuroendocrine tumors using Ki-67 proliferation index, phosphorylated histone H3 (PHH3), and BCL-2.

INTRODUCTION: Surgery is the only known cure for sporadic pancreatic neuroendocrine tumors (PNETs). Therefore, the prediction of the PNETs biological aggressiveness evaluated on endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) has a significant impact on clinical management. The proliferation rate of Ki-67 in PNETs can help to predict the biological aggressiveness of the tumor. In addition, there is a relatively new proliferation marker called phosphorylated histone H3 (PHH3) that can identify and quantify dividing cells in tissue samples, which is a marker highly specific to mitotic figures. Other markers such as BCL-2 also contribute to tumorigenesis and may be involved in the differentiation of neuroendocrine cells. MATERIALS AND METHODS: A retrospective observational study was performed on patients undergoing surveillance for PNETs from January 2010 to May 2021. Data collection included the patients' age, sex, tumor location, tumor size in the surgical specimen, and tumor grade in FNA. The 2019 World Health Organization (WHO) classification guideline was followed to diagnose PNETs, including grade and stage. Immunohistochemical stainings for Ki-67, PHH3 and BCL-2 in PNETs were performed. RESULTS: After excluding cell blocks containing fewer than 100 tumor cells, 44 patients with EUS-FNA and surgical resection specimens were included in this study. There were 19 cases of G1 PNETs, 20 cases of G2 PNETs, and 5 cases of G3 PNETs. The grade assigned based on the Ki-67 index was higher and more sensitive than that based on the mitotic count using H&E slides in some cases of G2 and G3 PNETs. However, there was no significant difference between the mitotic count using PHH3-positive tumor cells and the Ki-67 index to grade PNETs. All grade 1 tumors (19 cases) on surgical resection specimens were correctly graded on FNA (100 % concordance rate). Within the 20 G2 PNETs, 15 cases of grade 2 on surgical resection specimens were graded correctly on FNA based on the Ki-67 index only. Five cases of grade 2 PNETs on surgical resection specimens were graded as grade 1 on FNA when using only the Ki-67 index. Three of five grade 3 tumors on surgical resection specimens were graded as grade 2 on FNA based on the Ki-67 index only. Using only FNA Ki-67 to predict PNET tumor grade, the concordance (accuracy) rate was 81.8 % in total. However, all these eight cases (5 cases of G2 PNETs and 3 cases of G3 PNETs) were graded correctly by using the Ki-67 index plus mitotic rate (using PHH3 IHC stains). Four of 18 (22.2 %) patients with PNETs were positive for BCL-2 stain. In these 4 cases positive for BCL-2 stains, 3 cases were G2 PNETs and one case was G3 PNETs. CONCLUSION: Grade and the proliferative rate in EUS-FNA can be used to predict the tumor grade in surgical resection specimens. However, when using only FNA Ki-67 to predict PNET tumor grade, about 18 % of cases were downgraded by one level. To solve the problem, immunohistochemical staining for BCL-2 and especially PHH3 would be helpful. Our results demonstrated that the mitotic count using PHH3 IHC stains not only improved the accuracy and precision of PNET grading in the surgical resection specimens, but also could reliably be used in routine scoring of mitotic figures of FNA specimens.

Squamous Cell Carcinoma of Skin Cancer Margin Classification From Digital Histopathology Images Using Deep Learning.

OBJECTIVES: Now a days, squamous cell carcinoma (SCC) margin assessment is done by examining histopathology images and inspection of whole slide images (WSI) using a conventional microscope. This is time-consuming, tedious, and depends on experts' experience which may lead to misdiagnosis and mistreatment plans. This study aims to develop a system for the automatic diagnosis of skin cancer margin for squamous cell carcinoma from histopathology microscopic images by applying deep learning techniques. METHODS: The system was trained, validated, and tested using histopathology images of SCC cancer locally acquired from Jimma Medical Center Pathology Department from seven different skin sites using an Olympus digital microscope. All images were preprocessed and trained with transfer learning pre-trained models by fine-tuning the hyper-parameter of the selected models. RESULTS: The overall best training accuracy of the models become 95.3%, 97.1%, 89.8%, and 89.9% on EffecientNetB0, MobileNetv2, ResNet50, VGG16 respectively. In addition to this, the best validation accuracy of the models was 94.7%, 91.8%, 87.8%, and 86.7% respectively. The best testing accuracy of the models at the same epoch was 95.2%, 91.5%, 87%, and 85.5% respectively. From these models, EfficientNetB0 showed the best average training and testing accuracy than the other models. CONCLUSIONS: The system assists the pathologist during the margin assessment of SCC by decreasing the diagnosis time from an average of 25 minutes to less than a minute.

The concordance between wide-area transepithelial sampling with computer-assisted 3-dimensional analysis (WATS-3D) and standard endoscope biopsy in the detection of Barrett's esophagus and esophageal dysplasia.

INTRODUCTION: Barrett's esophagus (BE) is a premalignant condition that leads to susceptibility to developing adenocarcinoma. The most common endoscopic surveillance technique is forceps biopsy, which involves sampling the specimen every 1 to 2 cm along the length of the lesion. This technique has a low sensitivity and often leaves the majority of the esophageal mucosa untested. Recently, the use of wide-area transepithelial sampling with computer-assisted 3-dimensional analysis (WATS-3D) has received much attention. However, there is little known about this novel technique, and this research aims to add to our knowledge of WATS-3D by comparing it to traditional forceps biopsy. MATERIALS AND METHODS: A retrospective observational study was performed. All existing GI biopsy cases diagnosed with WATS-3D were identified from the institutional pathology databases of NYU Langone Hospital - Long Island from 2019 to 2021. Data collection included patients' age, sex, and dysplasia results. Existing pathology reports and CDx diagnostics were reviewed. All the existing slides of the biopsy cases were pulled out and reviewed. Dysplasia was classified as no dysplasia, indefinite for dysplasia, lowgrade dysplasia, and high-grade dysplasia. RESULTS: A total of 109 cases were included in this study. There are 59 cases diagnosed as BE with forceps biopsy, 72 cases by WATS-3D, and 77 cases by WATS-3D combined with forceps biopsy. The sensitivity of detecting BE was significantly increased by WATS-3D and further by WATS-3D combined with forceps biopsy. In 59 cases diagnosed as BE with forceps biopsy, 50 cases were classified as no dysplasia, 3 cases were indefinite for dysplasia, 5 cases were low-grade dysplasia, and 1 case was high-grade dysplasia. In 72 cases diagnosed as BE by WATS-3D, 64 cases were classified as no dysplasia, 7 cases were indefinite for dysplasia, 1 case was high-grade dysplasia, and no cases with low-grade dysplasia. In 77 cases diagnosed as BE by WATS-3D combined with forceps biopsy, 63 cases were classified as no dysplasia, 8 cases were indefinite for dysplasia, 5 cases with low-grade dysplasia, and 1 case was highgrade dysplasia. The maximal longitudinal extent of the esophageal mucosal changes strongly correlated with the severity of BE. CONCLUSION: Compared to traditional forceps biopsy, WATS-3D was more sensitive in finding intestinal metaplasia. However, WATS-3D could not clearly discriminate low-grade dysplasia from indefinite for dysplasia and tended to classify low-grade dysplasia as indefinite for dysplasia. The addition of WATS-3D to forceps biopsy resulted in an increase in diagnostic yield and thus an increase in the quality of patient care.