Genomic characterization of vulvar squamous cell carcinoma reveals differential gene expression based on clinical outcome.

OBJECTIVE: The greatest challenge in the management of vulvar squamous cell carcinoma (VSCC) is treatment of recurrent disease where options for surgery and radiation have been exhausted, or treatment of disease where distant metastasis is present. Identification of mutations differentially expressed between tumor from patients who died of aggressive disease and tumor from patients with an indolent course could reveal novel prognostic indicators and guide development of therapeutic drugs. METHODS: From 202 consecutive patients with VSCC, patients who recurred and died of disease (group A) were identified and matched by age, tumor size, depth of invasion and nodal status with those whose disease did not recur (group B). Tumors from 21 patients were subjected to whole exome sequencing of DNA and RNA, immunohistochemistry (IHC) antibodies of PD-L1 and P16, and in-situ hybridization (ISH) for high-risk HPV. RESULTS: Analysis of DNA and RNA revealed six genes that were strongly differentially expressed between group A and B: TGM3, ACVR2A, ROS1, NFEL2, CCND1 and BCL6. Clinically relevant DNA mutations were significantly greater in group A versus B: 7 vs 2.3 mutations per patient. The most common genomic alterations were mutations in TP53 and the promoter region of TERT. Other common genomic events include alterations of FAT1, CDKN2A, PIK3CA, CCND1, and LRP1B. All samples were MSI stable and tumor mutational burden (TMB) was similar in groups A and B. Most VSCC specimens (81%) were positive for PD-L1. CONCLUSIONS: ACVR2A and TGM3 are significantly under-expressed in tumors with poor outcome, suggesting they may play a role in tumor suppression. Clinical outcome of VSCC appears independent of MSI, TMB, or PD-L1 status.

SHIFT: speedy histological-to-immunofluorescent translation of a tumor signature enabled by deep learning.

Spatially-resolved molecular profiling by immunostaining tissue sections is a key feature in cancer diagnosis, subtyping, and treatment, where it complements routine histopathological evaluation by clarifying tumor phenotypes. In this work, we present a deep learning-based method called speedy histological-to-immunofluorescent translation (SHIFT) which takes histologic images of hematoxylin and eosin (H&E)-stained tissue as input, then in near-real time returns inferred virtual immunofluorescence (IF) images that estimate the underlying distribution of the tumor cell marker pan-cytokeratin (panCK). To build a dataset suitable for learning this task, we developed a serial staining protocol which allows IF and H&E images from the same tissue to be spatially registered. We show that deep learning-extracted morphological feature representations of histological images can guide representative sample selection, which improved SHIFT generalizability in a small but heterogenous set of human pancreatic cancer samples. With validation in larger cohorts, SHIFT could serve as an efficient preliminary, auxiliary, or substitute for panCK IF by delivering virtual panCK IF images for a fraction of the cost and in a fraction of the time required by traditional IF.

Predicting primary site of secondary liver cancer with a neural estimator of metastatic origin.

Purpose: Pathologists rely on relevant clinical information, visual inspection of stained tissue slide morphology, and sophisticated molecular diagnostics to accurately infer the biological origin of secondary metastatic cancer. While highly effective, this process is expensive in terms of time and clinical resources. We seek to develop and evaluate a computer vision system designed to reasonably infer metastatic origin of secondary liver cancer directly from digitized histopathological whole slide images of liver biopsy. Approach: We illustrate a two-stage deep learning approach to accomplish this task. We first train a model to identify spatially localized regions of cancerous tumor within digitized hematoxylin and eosin (H&E)-stained tissue sections of secondary liver cancer based on a pathologist's annotation of several whole slide images. Then, a second model is trained to generate predictions of the cancers' metastatic origin belonging to one of three distinct clinically relevant classes as confirmed by immunohistochemistry. Results: Our approach achieves a classification accuracy of 90.2% in determining metastatic origin of whole slide images from a held-out test set, which compares favorably to an established clinical benchmark by three board-certified pathologists whose accuracies ranged from 90.2% to 94.1% on the same prediction task. Conclusions: We illustrate the potential impact of deep learning systems to leverage morphological and structural features of H&E-stained tissue sections to guide pathological and clinical determination of the metastatic origin of secondary liver cancers.

Variational Autoencoding Tissue Response to Microenvironment Perturbation.

This work applies deep variational autoencoder learning architecture to study multi-cellular growth characteristics of human mammary epithelial cells in response to diverse microenvironment perturbations. Our approach introduces a novel method of visualizing learned feature spaces of trained variational autoencoding models that enables visualization of principal features in two dimensions. We find that unsupervised learned features more closely associate with expert annotation of cell colony organization than biologically-inspired hand-crafted features, demonstrating the utility of deep learning systems to meaningfully characterize features of multi-cellular growth characteristics in a fully unsupervised and data-driven manner.