Artificial Intelligence Detected the Relationship Between Nuclear Morphological Features and Molecular Abnormalities of Papillary Thyroid Carcinoma.

Papillary thyroid carcinoma (PTC) is the most common type of thyroid carcinoma and has characteristic nuclear features. Genetic abnormalities of PTC affect recent molecular target therapeutic strategy towards RET-altered cases, and they affect clinical prognosis and progression. However, there has been insufficient objective analysis of the correlation between genetic abnormalities and nuclear features. Using our newly developed methods, we studied the correlation between nuclear morphology and molecular abnormalities of PTC with the aim of predicting genetic abnormalities of PTC. We studied 72 cases of PTC and performed genetic analysis to detect BRAF p.V600E mutation and RET fusions. Nuclear features of PTC, such as nuclear grooves, pseudo-nuclear inclusions, and glassy nuclei, were also automatically detected by deep learning models. After analyzing the correlation between genetic abnormalities and nuclear features of PTC, logistic regression models could be used to predict gene abnormalities. Nuclear features were accurately detected with over 0.90 of AUCs in every class. The ratio of glassy nuclei to nuclear groove and the ratio of pseudo-nuclear inclusion to glassy nuclei were significantly higher in cases that were positive for RET fusions (p = 0.027, p = 0.043, respectively) than in cases that were negative for RET fusions. RET fusions were significantly predicted by glassy nuclei/nuclear grooves, pseudo-nuclear inclusions/glassy nuclei, and age (p = 0.023). Our deep learning models could accurately detect nuclear features. Genetic abnormalities had a correlation with nuclear features of PTC. Furthermore, our artificial intelligence model could significantly predict RET fusions of classic PTC.

Clinicopathological and molecular features of renal cell carcinomas with haemangioblastoma-like features distinct from clear cell renal cell carcinoma.

AIMS: Haemangioblastomas arise in the central nervous system. Rarely, haemangioblastomas may develop in extra-neural sites, such as the kidneys. A few reported cases of renal cell carcinomas (RCCs) with haemangioblastoma-like features have exhibited both clear cell renal cell carcinoma (CCRCC)- and haemangioblastoma-like components. The clinicopathological and molecular characteristics of RCCs with haemangioblastoma-like features were analysed, focusing on VHL alterations, in comparison with CCRCCs partially resembling haemangioblastoma. METHODS AND RESULTS: Four RCCs with haemangioblastoma-like features and five CCRCCs partially resembling haemangioblastoma were included. The RCCs with haemangioblastoma-like features were indolent and lacked adverse prognostic factors. All RCCs with haemangioblastoma-like features had a well-circumscribed appearance and a thick fibromuscular capsule, with fibromuscular bundles extending into the tumour to varying degrees in the three tumours. Each RCC with haemangioblastoma-like features exhibited CCRCC-like areas with indistinct tubular structures and foci of haemangioblastoma-like areas, in which vessels and short spindle cells overwhelmed tumour cells. Whereas haemangioblastoma-like areas in the CCRCCs partially resembling haemangioblastoma exhibited sparse vessels and spindle cells and distinct clear cells. The RCCs with haemangioblastoma-like features exhibited a unique immunohistochemical profile, with positive staining for inhibin-α, S100, carbonic-anhydrase-9, keratin7, and high molecular weight keratin and negative staining for (alpha-methylacyl-CoA racemase) AMACR. RCC with haemangioblastoma-like features did not display any VHL alterations, including VHL mutation, 3p LOH, and methylation of the VHL promoter region, and the two tumours harboured a likely oncogenic missense variant of MTOR (c.7280T>G). CONCLUSION: The histopathological, immunohistochemical, and molecular findings suggest that RCC with haemangioblastoma-like features is a distinct entity from CCRCC.

Pathologic Image Classification of Flat Urothelial Lesions Using Pathologic Criteria-Based Deep Learning.

OBJECTIVES: Pathologic diagnosis of flat urothelial lesions is subject to high interobserver variability. We expected that deep learning could improve the accuracy and consistency of such pathologic diagnosis, although the learning process is a black box. We therefore propose a new approach for pathologic image classification incorporating the diagnostic process of the pathologist into a deep learning method. METHODS: A total of 267 H&E-stained slides of normal urothelium and urothelial lesions from 127 cases were examined. Six independent convolutional neural networks were trained to classify pathologic images according to six pathologic criteria. We then used these networks in the main training for the final diagnosis. RESULTS: Compared with conventional manual analysis, our method significantly improved the classification accuracy of images of flat urothelial lesions. The automated classification showed almost perfect agreement (weighted κ = 0.98) with the consensus reading. In addition, our approach provides the advantages of reliable diagnosis corresponding to histologic interpretation. CONCLUSIONS: We used deep learning to establish an automated subtype classifier for flat urothelial lesions that successfully combines traditional morphologic approaches and complex deep learning to achieve a learning mechanism that seems plausible to the pathologist.

Microtubule-organizing center-mediated structural atypia in low- and high-grade urothelial carcinoma.

Urothelial carcinoma (UC) comprises two subtypes, low grade (LG-UC) and high grade (HG-UC), with different pathological and clinical behavior. LG-UC and HG-UC are classified based on cellular and structural atypia of pathological findings. The mechanisms responsible for maintaining structural atypia, such as the disturbance of nuclear polarity, remain unclear. In this study, we studied microtubule-organizing center (MTOC)-mediated nuclear polarity in UC subtypes. We evaluated six cases with normal urothelium (NU), 10 LG-UC cases, and 10 HG-UC cases by double immunofluorescence staining of γ-tubulin as a marker of MTOC and E-cadherin as a marker of each cell border. The number and position of γ-tubulin dots of expression in more than 100 cells per case were assessed using the spatial relationship with the nucleus and surface-basal axis. We found one γ-tubulin dot in most normal and tumor cells, and more than two γ-tubulin dots in 4.6% of NU cells, 6.1% of LG-UC cells, and 9.8% of HG-UC cells. More than three γ-tubulin dots were found only in 1.2% of HG-UC cells. Surface side positioning of γ-tubulin was found in 77.4% of normal urothelial cells, 63.8% of LG-UC cells, and 39.2% of HG-UC cells, whereas aberrant lateral and basal side positioning of γ-tubulin was found in 22.6% of normal urothelial cells, 36.1% of LG-UC cells, and 60.8% of HG-UC cells. We concluded that numerical and positional aberrations of MTOC in UC cases were strongly correlated with both cellular and structural atypia as well as abnormal cell proliferation.