Mesonephric adenocarcinoma (MA) and mesonephric-like adenocarcinoma (MLA) are uncommon neoplasms of the gynecologic tract that have until recently been poorly understood. Although their morphologic, immunohistochemical, and molecular profiles have been recently defined, little is known about their clinical behavior. Small studies have demonstrated inconsistent findings and no large studies have examined the clinical behavior of these adenocarcinomas. In this multi-institutional study, representing the largest and most stringently defined cohort of cases to date, we examined the clinicopathologic features of 99 MAs and MLAs (30 MAs of the uterine cervix, 44 MLAs of the endometrium, and 25 MLAs of the ovary). Only tumors with characteristic mesonephric morphology and either immunohistochemical or molecular support were included. Our results demonstrate that the majority of mesonephric neoplasms presented at an advanced stage (II to IV) (15/25 [60%] MA of the cervix, 25/43 [58%] MLA of the endometrium, and 7/18 [39%] MLA of the ovary). The majority (46/89 [52%] overall, 12/24 [50%] MA of the cervix, 24/41 [59%] MLA of the endometrium, and 10/24 [42%] MLA of the ovary) developed recurrences, most commonly distant (9/12 [75%] MA of the cervix, 22/24 [92%] MLA of the endometrium, and 5/9 [56%] MLA of the ovary). The 5-year disease-specific survival was 74% (n=26) for MA of cervix, 72% (n=43) for MLA of endometrium, and 71% (n=23) for MLA of ovary. Our results confirm that mesonephric neoplasms are a clinically aggressive group of gynecologic carcinomas that typically present at an advanced stage, with a predilection for pulmonary recurrence.

Deep learning‐based computer vision methods have recently made remarkable breakthroughs in the analysis and classification of cancer pathology images. However, there has been relatively little investigation of the utility of deep neural networks to synthesize medical images. In this study, we evaluated the efficacy of generative adversarial networks to synthesize high‐resolution pathology images of 10 histological types of cancer, including five cancer types from The Cancer Genome Atlas and the five major histological subtypes of ovarian carcinoma. The quality of these images was assessed using a comprehensive survey of board‐certified pathologists (n = 9) and pathology trainees (n = 6). Our results show that the real and synthetic images are classified by histotype with comparable accuracies and the synthetic images are visually indistinguishable from real images. Furthermore, we trained deep convolutional neural networks to diagnose the different cancer types and determined that the synthetic images perform as well as additional real images when used to supplement a small training set. These findings have important applications in proficiency testing of medical practitioners and quality assurance in clinical laboratories. Furthermore, training of computer‐aided diagnostic systems can benefit from synthetic images where labeled datasets are limited (e.g. rare cancers). We have created a publicly available website where clinicians and researchers can attempt questions from the image survey (http://gan.aimlab.ca/). © 2020 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

and Classification (IECC) has reorganized the classification of endocervical adenocarcinomas (ECAs), separating them into human papilloma virus (HPV)-associated (HPVA) and HPVA independent (HPVI) categories. In this study, we sought to revalidate the IECC clinical findings in an independent cohort and assess the mutational differences between HPVA and HPVI ECAs using next generation sequencing. Consecutive cases of ECAs were reclassified under the IECC. Clinicopathologic information was collected and tissue was sent for targeted next-generation sequencing in 33 genes. Associations between HPV status, clinicopathologic parameters and mutation status, with survival were evaluated. The series comprised of 85/100 HPVA (63 HPVA-usual type, 4 villoglandular, 3 mucinous intestinal, 15 mucinous not otherwise specified) and 15/100 HPVI (9 gastric, 4 mesonephric, 1 clear cell, 1 not otherwise specified). HPVA ECAs presented at a lower age (P = 0.001), smaller tumor sizes (P = 0.011), less margin positivity (P = 0.027), less Silva pattern C (P = 0.002), and lower FIGO stages (P = 0.020). HPVA had superior survival compared with HPVI ECA [overall survival (P = 0.0026), disease-specific survival (P = 0.0092), and progression-free survival (P = 0.0041)]. Factors that correlated with worse prognosis irrespective of HPV status were FIGO stage, positive margins and lymphovascular invasion (Po0.05). TP53 mutations were detected in a significantly higher proportion of HPVIs than HPVAs (Poo0.001). The study revalidates the IECC system by reaffirming the clinical and prognostic differences between HPVA and HPVI ECAs in an independent dataset.