AI-based histopathology image analysis reveals a distinct subset of endometrial cancers.

Endometrial cancer (EC) has four molecular subtypes with strong prognostic value and therapeutic implications. The most common subtype (NSMP; No Specific Molecular Profile) is assigned after exclusion of the defining features of the other three molecular subtypes and includes patients with heterogeneous clinical outcomes. In this study, we employ artificial intelligence (AI)-powered histopathology image analysis to differentiate between p53abn and NSMP EC subtypes and consequently identify a sub-group of NSMP EC patients that has markedly inferior progression-free and disease-specific survival (termed 'p53abn-like NSMP'), in a discovery cohort of 368 patients and two independent validation cohorts of 290 and 614 from other centers. Shallow whole genome sequencing reveals a higher burden of copy number abnormalities in the 'p53abn-like NSMP' group compared to NSMP, suggesting that this group is biologically distinct compared to other NSMP ECs. Our work demonstrates the power of AI to detect prognostically different and otherwise unrecognizable subsets of EC where conventional and standard molecular or pathologic criteria fall short, refining image-based tumor classification. This study's findings are applicable exclusively to females.

Practical Application of Deep Learning in Diagnostic Neuropathology-Reimagining a Histological Asset in the Era of Precision Medicine.

In the past few decades, neuropathology has experienced several paradigm shifts with the introduction of new technologies. Deep learning, a rapidly progressing subfield of machine learning, seems to be the next innovation to alter the diagnostic workflow. In this review, we will explore the recent changes in the field of neuropathology and how this has led to an increased focus on molecular features in diagnosis and prognosis. Then, we will examine the work carried out to train deep learning models for various diagnostic tasks in neuropathology, as well as the machine learning frameworks they used. Focus will be given to both the challenges and successes highlighted therein, as well as what these trends may tell us about future roadblocks in the widespread adoption of this new technology. Finally, we will touch on recent trends in deep learning, as applied to digital pathology more generally, and what this may tell us about the future of deep learning applications in neuropathology.

Prostate Cancer Risk Stratification by Digital Histopathology and Deep Learning.

PURPOSE: Prostate cancer (PCa) represents a highly heterogeneous disease that requires tools to assess oncologic risk and guide patient management and treatment planning. Current models are based on various clinical and pathologic parameters including Gleason grading, which suffers from a high interobserver variability. In this study, we determine whether objective machine learning (ML)-driven histopathology image analysis would aid us in better risk stratification of PCa. MATERIALS AND METHODS: We propose a deep learning, histopathology image-based risk stratification model that combines clinicopathologic data along with hematoxylin and eosin- and Ki-67-stained histopathology images. We train and test our model, using a five-fold cross-validation strategy, on a data set from 502 treatment-naïve PCa patients who underwent radical prostatectomy (RP) between 2000 and 2012. RESULTS: We used the concordance index as a measure to evaluate the performance of various risk stratification models. Our risk stratification model on the basis of convolutional neural networks demonstrated superior performance compared with Gleason grading and the Cancer of the Prostate Risk Assessment Post-Surgical risk stratification models. Using our model, 3.9% of the low-risk patients were correctly reclassified to be high-risk and 21.3% of the high-risk patients were correctly reclassified as low-risk. CONCLUSION: These findings highlight the importance of ML as an objective tool for histopathology image assessment and patient risk stratification. With further validation on large cohorts, the digital pathology risk classification we propose may be helpful in guiding administration of adjuvant therapy including radiotherapy after RP.

Multi-scale relational graph convolutional network for multiple instance learning in histopathology images.

Graph convolutional neural networks have shown significant potential in natural and histopathology images. However, their use has only been studied in a single magnification or multi-magnification with either homogeneous graphs or only different node types. In order to leverage the multi-magnification information and improve message passing with graph convolutional networks, we handle different embedding spaces at each magnification by introducing the Multi-Scale Relational Graph Convolutional Network (MS-RGCN) as a multiple instance learning method. We model histopathology image patches and their relation with neighboring patches and patches at other scales (i.e., magnifications) as a graph. We define separate message-passing neural networks based on node and edge types to pass the information between different magnification embedding spaces. We experiment on prostate cancer histopathology images to predict the grade groups based on the extracted features from patches. We also compare our MS-RGCN with multiple state-of-the-art methods with evaluations on several source and held-out datasets. Our method outperforms the state-of-the-art on all of the datasets and image types consisting of tissue microarrays, whole-mount slide regions, and whole-slide images. Through an ablation study, we test and show the value of the pertinent design features of the MS-RGCN.

VOLTA: an enVironment-aware cOntrastive ceLl represenTation leArning for histopathology.

In clinical oncology, many diagnostic tasks rely on the identification of cells in histopathology images. While supervised machine learning techniques necessitate the need for labels, providing manual cell annotations is time-consuming. In this paper, we propose a self-supervised framework (enVironment-aware cOntrastive cell represenTation learning: VOLTA) for cell representation learning in histopathology images using a technique that accounts for the cell's mutual relationship with its environment. We subject our model to extensive experiments on data collected from multiple institutions comprising over 800,000 cells and six cancer types. To showcase the potential of our proposed framework, we apply VOLTA to ovarian and endometrial cancers and demonstrate that our cell representations can be utilized to identify the known histotypes of ovarian cancer and provide insights that link histopathology and molecular subtypes of endometrial cancer. Unlike supervised models, we provide a framework that can empower discoveries without any annotation data, even in situations where sample sizes are limited.

Harmonized molecular classification; assessment of a single-test ProMisE NGS tool.

OBJECTIVES: Despite recommendations for integrating molecular classification of endometrial cancers (EC) into pathology reporting and clinical management, uptake is inconsistent. To assign ProMisE subtype, all molecular components must be available (POLE mutation status, mismatch repair (MMR) and p53 immunohistochemistry (IHC)) and often these are assessed at different stages of care and/or at different centres resulting in delays in treatment. We assessed a single-test DNA-based targeted next generation sequencing (NGS) molecular classifier (ProMisE NGS), comparing concordance and prognostic value to the original ProMisE classifier. METHODS: DNA was extracted from formalin-fixed paraffin embedded (FFPE) ECs that had previously undergone ProMisE molecular classification (POLE sequencing, IHC for p53 and MMR). DNA was sequenced using the clinically validated Imagia Canexia Health Find It™ amplicon-based NGS gene panel assay to assess for pathogenic POLE mutations (unchanged from original ProMisE), TP53 mutations (in lieu of p53 IHC), and microsatellite instability (MSI) (in lieu of MMR IHC),with the same order of segregation as original ProMisE used for subtype assignment. Molecular subtype assignment of both classifiers was compared by concordance metrics and Kaplan-Meier survival statistics. RESULTS: The new DNA-based NGS molecular classifier (ProMisE NGS) was used to determine the molecular subtype in 164 ECs previously classified with ProMisE. 159/164 cases were concordant with a kappa statistic of 0.96 and an overall accuracy of 0.97. Prognostic differences in progression-free, disease-specific and overall survival between the four molecular subtypes were observed for the new NGS classifier, recapitulating the survival curves of the original ProMisE classifier. ProMisE NGS was 100% concordant between matched biopsy and hysterectomy samples. CONCLUSION: ProMisE NGS is feasible on standard FFPE material, demonstrates high concordance with the original ProMisE classifier and maintains prognostic value in EC. This test has the potential to facilitate implementation of molecular classification of EC at the time of first diagnosis.

Deep learning-based histotype diagnosis of ovarian carcinoma whole-slide pathology images.

Ovarian carcinoma has the highest mortality of all female reproductive cancers and current treatment has become histotype-specific. Pathologists diagnose five common histotypes by microscopic examination, however, histotype determination is not straightforward, with only moderate interobserver agreement between general pathologists (Cohen's kappa 0.54-0.67). We hypothesized that machine learning (ML)-based image classification models may be able to recognize ovarian carcinoma histotype sufficiently well that they could aid pathologists in diagnosis. We trained four different artificial intelligence (AI) algorithms based on deep convolutional neural networks to automatically classify hematoxylin and eosin-stained whole slide images. Performance was assessed through cross-validation on the training set (948 slides corresponding to 485 patients), and on an independent test set of 60 patients from another institution. The best-performing model achieved a diagnostic concordance of 81.38% (Cohen's kappa of 0.7378) in our training set, and 80.97% concordance (Cohen's kappa 0.7547) on the external dataset. Eight cases misclassified by ML in the external set were reviewed by two subspecialty pathologists blinded to the diagnoses, molecular and immunophenotype data, and ML-based predictions. Interestingly, in 4 of 8 cases from the external dataset, the expert review pathologists rendered diagnoses, based on blind review of the whole section slides classified by AI, that were in agreement with AI rather than the integrated reference diagnosis. The performance characteristics of our classifiers indicate potential for improved diagnostic performance if used as an adjunct to conventional histopathology.

Gastrointestinal symptoms and the severity of COVID-19: Disorders of gut-brain interaction are an outcome.

BACKGROUND: Many of the studies on COVID-19 severity and its associated symptoms focus on hospitalized patients. The aim of this study was to investigate the relationship between acute GI symptoms and COVID-19 severity in a clustering-based approach and to determine the risks and epidemiological features of post-COVID-19 Disorders of Gut-Brain Interaction (DGBI) by including both hospitalized and ambulatory patients. METHODS: The study utilized a two-phase Internet-based survey on: (1) COVID-19 patients' demographics, comorbidities, symptoms, complications, and hospitalizations and (2) post-COVID-19 DGBI diagnosed according to Rome IV criteria in association with anxiety (GAD-7) and depression (PHQ-9). Statistical analyses included univariate and multivariate tests. RESULTS: Five distinct clusters of symptomatic subjects were identified based on the presence of GI symptoms, loss of smell, and chest pain, among 1114 participants who tested positive for SARS-CoV-2. GI symptoms were found to be independent risk factors for severe COVID-19; however, they did not always coincide with other severity-related factors such as age >65 years, diabetes mellitus, and Vitamin D deficiency. Of the 164 subjects with a positive test who participated in Phase-2, 108 (66%) fulfilled the criteria for at least one DGBI. The majority (n = 81; 75%) were new-onset DGBI post-COVID-19. Overall, 86% of subjects with one or more post-COVID-19 DGBI had at least one GI symptom during the acute phase of COVID-19, while 14% did not. Depression (65%), but not anxiety (48%), was significantly more common in those with post-COVID-19 DGBI. CONCLUSION: GI symptoms are associated with a severe COVID-19 among survivors. Long-haulers may develop post-COVID-19 DGBI. Psychiatric disorders are common in post-COVID-19 DGBI.

Tumor-associated antigen PRAME exhibits dualistic functions that are targetable in diffuse large B cell lymphoma.

PRAME is a prominent member of the cancer testis antigen family of proteins, which triggers autologous T cell-mediated immune responses. Integrative genomic analysis in diffuse large B cell lymphoma (DLBCL) uncovered recurrent and highly focal deletions of 22q11.22, including the PRAME gene, which were associated with poor outcome. PRAME-deleted tumors showed cytotoxic T cell immune escape and were associated with cold tumor microenvironments. In addition, PRAME downmodulation was strongly associated with somatic EZH2 Y641 mutations in DLBCL. In turn, PRC2-regulated genes were repressed in isogenic PRAME-KO lymphoma cell lines, and PRAME was found to directly interact with EZH2 as a negative regulator. EZH2 inhibition with EPZ-6438 abrogated these extrinsic and intrinsic effects, leading to PRAME expression and microenvironment restoration in vivo. Our data highlight multiple functions of PRAME during lymphomagenesis and provide a preclinical rationale for synergistic therapies combining epigenetic reprogramming with PRAME-targeted therapies.

The utility of color normalization for AI-based diagnosis of hematoxylin and eosin-stained pathology images.

The color variation of hematoxylin and eosin (H&E)-stained tissues has presented a challenge for applications of artificial intelligence (AI) in digital pathology. Many color normalization algorithms have been developed in recent years in order to reduce the color variation between H&E images. However, previous efforts in benchmarking these algorithms have produced conflicting results and none have sufficiently assessed the efficacy of the various color normalization methods for improving diagnostic performance of AI systems. In this study, we systematically investigated eight color normalization algorithms for AI-based classification of H&E-stained histopathology slides, in the context of using images both from one center and from multiple centers. Our results show that color normalization does not consistently improve classification performance when both training and testing data are from a single center. However, using four multi-center datasets of two cancer types (ovarian and pleural) and objective functions, we show that color normalization can significantly improve the classification accuracy of images from external datasets (ovarian cancer: 0.25 AUC increase, p = 1.6 e-05; pleural cancer: 0.21 AUC increase, p = 1.4 e-10). Furthermore, we introduce a novel augmentation strategy by mixing color-normalized images using three easily accessible algorithms that consistently improves the diagnosis of test images from external centers, even when the individual normalization methods had varied results. We anticipate our study to be a starting point for reliable use of color normalization to improve AI-based, digital pathology-empowered diagnosis of cancers sourced from multiple centers. © 2021 The Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Deep-learning based classification distinguishes sarcomatoid malignant mesotheliomas from benign spindle cell mesothelial proliferations.

Sarcomatoid mesothelioma is an aggressive malignancy that can be challenging to distinguish from benign spindle cell mesothelial proliferations based on biopsy, and this distinction is crucial to patient treatment and prognosis. A novel deep learning based classifier may be able to aid pathologists in making this critical diagnostic distinction. SpindleMesoNET was trained on cases of malignant sarcomatoid mesothelioma and benign spindle cell mesothelial proliferations. Performance was assessed through cross-validation on the training set, on an independent set of challenging cases referred for expert opinion ('referral' test set), and on an externally stained set from outside institutions ('externally stained' test set). SpindleMesoNET predicted the benign or malignant status of cases with AUC's of 0.932, 0.925, and 0.989 on the cross-validation, referral and external test sets, respectively. The accuracy of SpindleMesoNET on the referral set cases (92.5%) was comparable to the average accuracy of 3 experienced pathologists on the same slide set (91.7%). We conclude that SpindleMesoNET can accurately distinguish sarcomatoid mesothelioma from benign spindle cell mesothelial proliferations. A deep learning system of this type holds potential for future use as an ancillary test in diagnostic pathology.

Adult-type granulosa cell tumor of the ovary: a FOXL2-centric disease.

Adult-type granulosa cell tumors (aGCTs) account for 90% of malignant ovarian sex cord-stromal tumors and 2-5% of all ovarian cancers. These tumors are usually diagnosed at an early stage and are treated with surgery. However, one-third of patients relapse between 4 and 8 years after initial diagnosis, and there are currently no effective treatments other than surgery for these relapsed patients. As the majority of aGCTs (>95%) harbor a somatic mutation in FOXL2 (c.C402G; p.C134W), the aim of this study was to identify genetic mutations besides FOXL2 C402G in aGCTs that could explain the clinical diversity of this disease. Whole-genome sequencing of 10 aGCTs and their matched normal blood was performed to identify somatic mutations. From this analysis, a custom amplicon-based panel was designed to sequence 39 genes of interest in a validation cohort of 83 aGCTs collected internationally. KMT2D inactivating mutations were present in 10 of 93 aGCTs (10.8%), and the frequency of these mutations was similar between primary and recurrent aGCTs. Inactivating mutations, including a splice site mutation in candidate tumor suppressor WNK2 and nonsense mutations in PIK3R1 and NLRC5, were identified at a low frequency in our cohort. Missense mutations were identified in cell cycle-related genes TP53, CDKN2D, and CDK1. From these data, we conclude that aGCTs are comparatively a homogeneous group of tumors that arise from a limited set of genetic events and are characterized by the FOXL2 C402G mutation. Secondary mutations occur in a subset of patients but do not explain the diverse clinical behavior of this disease. As the FOXL2 C402G mutation remains the main driver of this disease, progress in the development of therapeutics for aGCT would likely come from understanding the functional consequences of the FOXL2 C402G mutation.

Endometrial Cancer Molecular Risk Stratification is Equally Prognostic for Endometrioid Ovarian Carcinoma.

PURPOSE: Endometrioid ovarian carcinoma (ENOC) is generally associated with a more favorable prognosis compared with other ovarian carcinomas. Nonetheless, current patient treatment continues to follow a "one-size-fits-all" approach. Even though tumor staging offers stratification, personalized treatments remain elusive. As ENOC shares many clinical and molecular features with its endometrial counterpart, we sought to investigate The Cancer Genome Atlas-inspired endometrial carcinoma (EC) molecular subtyping in a cohort of ENOC. EXPERIMENTAL DESIGN: IHC and mutation biomarkers were used to segregate 511 ENOC tumors into four EC-inspired molecular subtypes: low-risk POLE mutant (POLEmut), moderate-risk mismatch repair deficient (MMRd), high-risk p53 abnormal (p53abn), and moderate-risk with no specific molecular profile (NSMP). Survival analysis with established clinicopathologic and subtype-specific features was performed. RESULTS: A total of 3.5% of cases were POLEmut, 13.7% MMRd, 9.6% p53abn, and 73.2% NSMP, each showing distinct outcomes (P < 0.001) and survival similar to observations in EC. Median OS was 18.1 years in NSMP, 12.3 years in MMRd, 4.7 years in p53abn, and not reached for POLEmut cases. Subtypes were independent of stage, grade, and residual disease in multivariate analysis. CONCLUSIONS: EC-inspired molecular classification provides independent prognostic information in ENOC. Our findings support investigating molecular subtype-specific management recommendations for patients with ENOC; for example, subtypes may provide guidance when fertility-sparing treatment is desired. Similarities between ENOC and EC suggest that patients with ENOC may benefit from management strategies applied to EC and the opportunity to study those in umbrella trials.

Synthesis of diagnostic quality cancer pathology images by generative adversarial networks.

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.

TMEM30A loss-of-function mutations drive lymphomagenesis and confer therapeutically exploitable vulnerability in B-cell lymphoma.

Transmembrane protein 30A (TMEM30A) maintains the asymmetric distribution of phosphatidylserine, an integral component of the cell membrane and 'eat-me' signal recognized by macrophages. Integrative genomic and transcriptomic analysis of diffuse large B-cell lymphoma (DLBCL) from the British Columbia population-based registry uncovered recurrent biallelic TMEM30A loss-of-function mutations, which were associated with a favorable outcome and uniquely observed in DLBCL. Using TMEM30A-knockout systems, increased accumulation of chemotherapy drugs was observed in TMEM30A-knockout cell lines and TMEM30A-mutated primary cells, explaining the improved treatment outcome. Furthermore, we found increased tumor-associated macrophages and an enhanced effect of anti-CD47 blockade limiting tumor growth in TMEM30A-knockout models. By contrast, we show that TMEM30A loss-of-function increases B-cell signaling following antigen stimulation-a mechanism conferring selective advantage during B-cell lymphoma development. Our data highlight a multifaceted role for TMEM30A in B-cell lymphomagenesis, and characterize intrinsic and extrinsic vulnerabilities of cancer cells that can be therapeutically exploited.

The coming 15 years in gynaecological pathology: digitisation, artificial intelligence, and new technologies.

Surgical pathology forms the cornerstone of modern oncological medicine, owing to the wealth of clinically relevant information that can be obtained from tissue morphology. Although several ancillary testing modalities have been added to surgical pathology, the way in which we view and interpret tissue morphology has remained largely unchanged since the inception of our profession. In this review, we discuss new technological advances that promise to transform the way in which we access tissue morphology and how we use it to guide patient care.

Molecular profiling and molecular classification of endometrioid ovarian carcinomas.

OBJECTIVE: Endometrioid ovarian carcinomas (EOCs) comprise 5-10% of all ovarian cancers and commonly co-occur with synchronous endometrioid endometrial cancer (EEC). We sought to examine the molecular characteristics of pure EOCs in patients without concomitant EEC. METHODS: EOCs and matched normal samples were subjected to massively parallel sequencing targeting 341-468 cancer-related genes (n = 8) or whole-genome sequencing (n = 28). Mutational frequencies of EOCs were compared to those of high-grade serous ovarian cancers (HGSOCs; n = 224) and EECs (n = 186) from The Cancer Genome Atlas, and synchronous EOCs (n = 23). RESULTS: EOCs were heterogeneous, frequently harboring KRAS, PIK3CA, PTEN, CTNNB1, ARID1A and TP53 mutations. EOCs were distinct from HGSOCs at the mutational level, less frequently harboring TP53 but more frequently displaying KRAS, PIK3CA, PIK3R1, PTEN and CTNNB1 mutations. Compared to synchronous EOCs and pure EECs, pure EOCs less frequently harbored PTEN, PIK3R1 and ARID1A mutations. Akin to EECs, EOCs could be stratified into the four molecular subtypes: 3% POLE (ultramutated), 19% MSI (hypermutated), 17% copy-number high (serous-like) and 61% copy-number low (endometrioid). In addition to microsatellite instability, a subset of EOCs harbored potentially targetable mutations, including AKT1 and ERBB2 hotspot mutations. EOCs of MSI (hypermutated) subtype uniformly displayed a good outcome. CONCLUSIONS: EOCs are heterogeneous at the genomic level and harbor targetable genetic alterations. Despite the similarities in the repertoire of somatic mutations between pure EOCs, synchronous EOCs and EECs, the frequencies of mutations affecting known driver genes differ. Further studies are required to define the impact of the molecular subtypes on the outcome and treatment of EOC patients.

Pharmacological systems analysis defines EIF4A3 functions in cell-cycle and RNA stress granule formation.

The RNA helicase EIF4A3 regulates the exon junction complex and nonsense-mediated mRNA decay functions in RNA transcript processing. However, a transcriptome-wide network definition of these functions has been lacking, in part due to the lack of suitable pharmacological inhibitors. Here we employ short-duration graded EIF4A3 inhibition using small molecule allosteric inhibitors to define the transcriptome-wide dependencies of EIF4A3. We thus define conserved cellular functions, such as cell cycle control, that are EIF4A3 dependent. We show that EIF4A3-dependent splicing reactions have a distinct genome-wide pattern of associated RNA-binding protein motifs. We also uncover an unanticipated role of EIF4A3 in the biology of RNA stress granules, which sequester and silence the translation of most mRNAs under stress conditions and are implicated in cell survival and tumour progression. We show that stress granule induction and maintenance is suppressed on the inhibition of EIF4A3, in part through EIF4A3-associated regulation of G3BP1 and TIA1 scaffold protein expression.

Targeting Hypoxia-Induced Carbonic Anhydrase IX Enhances Immune-Checkpoint Blockade Locally and Systemically.

Treatment strategies involving immune-checkpoint blockade (ICB) have significantly improved survival for a subset of patients across a broad spectrum of advanced solid cancers. Despite this, considerable room for improving response rates remains. The tumor microenvironment (TME) is a hurdle to immune function, as the altered metabolism-related acidic microenvironment of solid tumors decreases immune activity. Here, we determined that expression of the hypoxia-induced, cell-surface pH regulatory enzyme carbonic anhydrase IX (CAIX) is associated with worse overall survival in a cohort of 449 patients with melanoma. We found that targeting CAIX with the small-molecule SLC-0111 reduced glycolytic metabolism of tumor cells and extracellular acidification, resulting in increased immune cell killing. SLC-0111 treatment in combination with immune-checkpoint inhibitors led to the sensitization of tumors to ICB, which led to an enhanced Th1 response, decreased tumor growth, and reduced metastasis. We identified that increased expression of CA9 is associated with a reduced Th1 response in metastatic melanoma and basal-like breast cancer TCGA cohorts. These data suggest that targeting CAIX in the TME in combination with ICB is a potential therapeutic strategy for enhancing response and survival in patients with hypoxic solid malignancies.