Deep learning predicts chromosomal instability from histopathology images

Chromosomal Instability and Clonal Heterogeneity in Breast Cancer: From Mechanisms to Clinical Applications

BACKGROUND

Chromosomal instability (CIN) and clonal heterogeneity (CH) are fundamental hallmarks of breast cancer that drive tumor evolution, disease progression, and therapeutic resistance. Understanding the mechanisms underlying these phenomena is essential for improving cancer diagnosis, prognosis, and treatment strategies.

METHODS

In this review, we provide a comprehensive overview of the biological processes contributing to CIN and CH, highlighting their molecular determinants and clinical relevance.

RESULTS

We discuss the latest advances in detection methods, including single-cell sequencing and other high-resolution techniques, which have enhanced our ability to characterize intratumoral heterogeneity. Additionally, we explore how CIN and CH influence treatment responses, their potential as therapeutic targets, and their role in shaping the tumor immune microenvironment, which has implications for immunotherapy effectiveness.

CONCLUSIONS

By integrating recent findings, this review underscores the impact of CIN and CH on breast cancer progression and their translational implications for precision medicine.

Targeting chromosomally unstable tumors with a selective KIF18A inhibitor

Chromosome instability is a prevalent vulnerability of cancer cells that has yet to be fully exploited therapeutically. To identify genes uniquely essential to chromosomally unstable cells, we mined the Cancer Dependency Map for genes essential in tumor cells with high levels of copy number aberrations. We identify and validate KIF18A, a mitotic kinesin, as a vulnerability of chromosomally unstable cancer cells. Knockdown of KIF18A leads to mitotic defects and reduction of tumor growth. Screening of a chemical library for inhibitors of KIF18A enzymatic activity identified a hit that was optimized to yield VLS-1272, which is orally bioavailable, potent, ATP non-competitive, microtubule-dependent, and highly selective for KIF18A versus other kinesins. Inhibition of KIF18A's ATPase activity prevents KIF18A translocation across the mitotic spindle, resulting in chromosome congression defects, mitotic cell accumulation, and cell death. Profiling VLS-1272 across >100 cancer cell lines demonstrates that the specificity towards cancer cells with chromosome instability differentiates KIF18A inhibition from other clinically tested anti-mitotic drugs. Treatment of tumor xenografts with VLS-1272 results in mitotic defects leading to substantial, dose-dependent inhibition of tumor growth. The strong biological rationale, robust preclinical data, and optimized compound properties enable the clinical development of a KIF18A inhibitor in cancers with high chromosomal instability.

Detecting cell types and densities in the tumor microenvironment improves prognostic risk assessment for breast cancer

A comprehensive evaluation of the relationship between the densities of various cell types in the breast cancer tumor microenvironment and patient prognosis is currently lacking. Additionally, the absence of a large patch-level whole slide imaging (WSI) dataset of breast cancer with annotated cell types hinders the ability of artificial intelligence to evaluate cell density in breast cancer WSI. We first employed Lasso-Cox regression to build a breast cancer prognosis assessment model based on cell density in a population study. Pathology experts manually annotated a dataset containing over 70,000 patches and used transfer learning based on ResNet152 to develop an artificial intelligence model for identifying different cell types in these patches. The results showed that significant prognostic differences were observed among breast cancer patients stratified by cell density score (P = 0.0018), with the cell density score identified as an independent prognostic factor for breast cancer patients (P < 0.05). In the validation cohort, the predictive performance for overall survival (OS) was satisfactory, with area under the curve (AUC) values of 0.893 (OS) at 1-year, 0.823 (OS) at 3-year, and 0.861 (OS) at 5-year intervals. We trained a robust model based on ResNet152, achieving over 99% classification accuracy for different cell types in patches. These achievements offer new public resources and tools for personalized treatment and prognosis assessment.

Publicly available datasets of breast histopathology H&E whole-slide images: A scoping review

Advancements in digital pathology and computing resources have made a significant impact in the field of computational pathology for breast cancer diagnosis and treatment. However, access to high-quality labeled histopathological images of breast cancer is a big challenge that limits the development of accurate and robust deep learning models. In this scoping review, we identified the publicly available datasets of breast H&E-stained whole-slide images (WSIs) that can be used to develop deep learning algorithms. We systematically searched 9 scientific literature databases and 9 research data repositories and found 17 publicly available datasets containing 10 385 H&E WSIs of breast cancer. Moreover, we reported image metadata and characteristics for each dataset to assist researchers in selecting proper datasets for specific tasks in breast cancer computational pathology. In addition, we compiled 2 lists of breast H&E patches and private datasets as supplementary resources for researchers. Notably, only 28% of the included articles utilized multiple datasets, and only 14% used an external validation set, suggesting that the performance of other developed models may be susceptible to overestimation. The TCGA-BRCA was used in 52% of the selected studies. This dataset has a considerable selection bias that can impact the robustness and generalizability of the trained algorithms. There is also a lack of consistent metadata reporting of breast WSI datasets that can be an issue in developing accurate deep learning models, indicating the necessity of establishing explicit guidelines for documenting breast WSI dataset characteristics and metadata.

Integrating histopathology and transcriptomics for spatial tumor microenvironment profiling in a melanoma case study

Local structures formed by cells in the tumor microenvironment (TME) play an important role in tumor development and treatment response. This study introduces SPoTLIghT, a computational framework providing a quantitative description of the tumor architecture from hematoxylin and eosin (H&E) slides. We trained a weakly supervised machine learning model on melanoma patients linking tile-level imaging features extracted from H&E slides to sample-level cell type quantifications derived from RNA-sequencing data. Using this model, SPoTLIghT provides spatial cellular maps for any H&E image, and converts them in graphs to derive 96 interpretable features capturing TME cellular organization. We show how SPoTLIghT's spatial features can distinguish microenvironment subtypes and reveal nuanced immune infiltration structures not apparent in molecular data alone. Finally, we use SPoTLIghT to effectively predict patients' prognosis in an independent melanoma cohort. SPoTLIghT enhances computational histopathology providing a quantitative and interpretable characterization of the spatial contexture of tumors.

Deep learning-based histological predictions of chromosomal instability in colorectal cancer

Colorectal cancer (CRC) is a lethal malignancy and a leading cause of cancer-related mortality worldwide. Chromosomal instability (CIN) is a key driver of genomic instability in CRC and is characterized by aneuploidy and somatic copy-number alterations. This study aimed to predict CIN in CRC using histological data from whole slide images (WSIs). CRC samples from TCGA were analyzed, with tumor regions segmented into tiles and nuclei for feature extraction using convolutional neural network (CNN) and morphologic analysis. Binary classification models were developed to distinguish high and low aneuploidy scores (AS) based on slide-level features. The analysis included 313 patients with 315 WSIs, resulting in over 350,000 tumor tiles and nearly 2.7 million tumor cell nuclei. The ResNet18-SSL model, pre-trained on histopathological images, demonstrated superior accuracy in tile-based AS prediction, while DenseNet121 excelled in nucleus-based prediction. Combining CNN-based and morphological features enhanced the classification accuracy of nucleus-based predictions. Additionally, significant correlations were observed between morphological features and copy-number signatures. Unsupervised clustering of nuclear features revealed that distinct groups are significantly correlated with CIN and TP53 mutations. This study underscores the potential of histological features from WSIs to predict CIN in CRC samples. Nuclear feature analysis, combined with deep-learning techniques, offers a robust method for CIN prediction, highlighting the importance of further research into the relationships between histological and molecular phenotypes.

Targeting chromosomal instability in patients with cancer

Chromosomal instability (CIN) is a hallmark of cancer and a driver of metastatic dissemination, therapeutic resistance, and immune evasion. CIN is present in 60-80% of human cancers and poses a formidable therapeutic challenge as evidenced by the lack of clinically approved drugs that directly target CIN. This limitation in part reflects a lack of well-defined druggable targets as well as a dearth of tractable biomarkers enabling direct assessment and quantification of CIN in patients with cancer. Over the past decade, however, our understanding of the cellular mechanisms and consequences of CIN has greatly expanded, revealing novel therapeutic strategies for the treatment of chromosomally unstable tumours as well as new methods of assessing the dynamic nature of chromosome segregation errors that define CIN. In this Review, we describe advances that have shaped our understanding of CIN from a translational perspective, highlighting both challenges and opportunities in the development of therapeutic interventions for patients with chromosomally unstable cancers.

Integrating Machine Learning with Multi-Omics Technologies in Geroscience: Towards Personalized Medicine

Aging is a fundamental biological process characterized by a progressive decline in physiological functions and an increased susceptibility to diseases. Understanding aging at the molecular level is crucial for developing interventions that could delay or reverse its effects. This review explores the integration of machine learning (ML) with multi-omics technologies-including genomics, transcriptomics, epigenomics, proteomics, and metabolomics-in studying the molecular hallmarks of aging to develop personalized medicine interventions. These hallmarks include genomic instability, telomere attrition, epigenetic alterations, loss of proteostasis, disabled macroautophagy, deregulated nutrient sensing, mitochondrial dysfunction, cellular senescence, stem cell exhaustion, altered intercellular communication, chronic inflammation, and dysbiosis. Using ML to analyze big and complex datasets helps uncover detailed molecular interactions and pathways that play a role in aging. The advances of ML can facilitate the discovery of biomarkers and therapeutic targets, offering insights into personalized anti-aging strategies. With these developments, the future points toward a better understanding of the aging process, aiming ultimately to promote healthy aging and extend life expectancy.

The yin and yang of chromosomal instability in prostate cancer

Metastatic prostate cancer remains an incurable lethal disease. Studies indicate that prostate cancer accumulates genomic changes during disease progression and displays the highest levels of chromosomal instability (CIN) across all types of metastatic tumours. CIN, which refers to ongoing chromosomal DNA gain or loss during mitosis, and derived aneuploidy, are known to be associated with increased tumour heterogeneity, metastasis and therapy resistance in many tumour types. Paradoxically, high CIN levels are also proposed to be detrimental to tumour cell survival, suggesting that cancer cells must develop adaptive mechanisms to ensure their survival. In the context of prostate cancer, studies indicate that CIN has a key role in disease progression and might also offer a therapeutic vulnerability that can be pharmacologically targeted. Thus, a comprehensive evaluation of the causes and consequences of CIN in prostate cancer, its contribution to aggressive advanced disease and a better understanding of the acquired CIN tolerance mechanisms can translate into new tumour classifications, biomarker development and therapeutic strategies.

Joint global and local interpretation method for CIN status classification in breast cancer

Breast cancer is among the cancer types with the highest numbers of new cases. The study of this disease from a microscopic perspective has been a prominent research topic. Previous studies have shown that microRNAs (miRNAs) are closely linked to chromosomal instability (CIN). Correctly predicting CIN status from miRNAs can help to improve the survival of breast cancer patients. In this study, a joint global and local interpretation method called GL_XGBoost is proposed for predicting CIN status in breast cancer. GL_XGBoost integrates the eXtreme Gradient Boosting (XGBoost) and SHapley Additive exPlanation (SHAP) methods. XGBoost is used to predict CIN status from miRNA data, whereas SHAP is used to select miRNA features that have strong relationships with CIN. Furthermore, SHAP's rich visualization strategies enhance the interpretability of the entire model at the global and local levels. The performance of GL_XGBoost is validated on the TCGA-BRCA dataset, and it is shown to have an accuracy of 78.57% and an area under the curve value of 0.87. Rich visual analysis is used to explain the relationships between miRNAs and CIN status from different perspectives. Our study demonstrates an intuitive way of exploring the relationship between CIN and cancer from a microscopic perspective.

Chromosomal instability in aneuploid acute lymphoblastic leukemia associates with disease progression

Chromosomal instability (CIN) lies at the core of cancer development leading to aneuploidy, chromosomal copy-number heterogeneity (chr-CNH) and ultimately, unfavorable clinical outcomes. Despite its ubiquity in cancer, the presence of CIN in childhood B-cell acute lymphoblastic leukemia (cB-ALL), the most frequent pediatric cancer showing high frequencies of aneuploidy, remains unknown. Here, we elucidate the presence of CIN in aneuploid cB-ALL subtypes using single-cell whole-genome sequencing of primary cB-ALL samples and by generating and functionally characterizing patient-derived xenograft models (cB-ALL-PDX). We report higher rates of CIN across aneuploid than in euploid cB-ALL that strongly correlate with intraclonal chr-CNH and overall survival in mice. This association was further supported by in silico mathematical modeling. Moreover, mass-spectrometry analyses of cB-ALL-PDX revealed a "CIN signature" enriched in mitotic-spindle regulatory pathways, which was confirmed by RNA-sequencing of a large cohort of cB-ALL samples. The link between the presence of CIN in aneuploid cB-ALL and disease progression opens new possibilities for patient stratification and offers a promising new avenue as a therapeutic target in cB-ALL treatment.

One label is all you need: Interpretable AI-enhanced histopathology for oncology

Artificial Intelligence (AI)-enhanced histopathology presents unprecedented opportunities to benefit oncology through interpretable methods that require only one overall label per hematoxylin and eosin (H&E) slide with no tissue-level annotations. We present a structured review of these methods organized by their degree of verifiability and by commonly recurring application areas in oncological characterization. First, we discuss morphological markers (tumor presence/absence, metastases, subtypes, grades) in which AI-identified regions of interest (ROIs) within whole slide images (WSIs) verifiably overlap with pathologist-identified ROIs. Second, we discuss molecular markers (gene expression, molecular subtyping) that are not verified via H&E but rather based on overlap with positive regions on adjacent tissue. Third, we discuss genetic markers (mutations, mutational burden, microsatellite instability, chromosomal instability) that current technologies cannot verify if AI methods spatially resolve specific genetic alterations. Fourth, we discuss the direct prediction of survival to which AI-identified histopathological features quantitatively correlate but are nonetheless not mechanistically verifiable. Finally, we discuss in detail several opportunities and challenges for these one-label-per-slide methods within oncology. Opportunities include reducing the cost of research and clinical care, reducing the workload of clinicians, personalized medicine, and unlocking the full potential of histopathology through new imaging-based biomarkers. Current challenges include explainability and interpretability, validation via adjacent tissue sections, reproducibility, data availability, computational needs, data requirements, domain adaptability, external validation, dataset imbalances, and finally commercialization and clinical potential. Ultimately, the relative ease and minimum upfront cost with which relevant data can be collected in addition to the plethora of available AI methods for outcome-driven analysis will surmount these current limitations and achieve the innumerable opportunities associated with AI-driven histopathology for the benefit of oncology.

Artificial Intelligence in Breast Cancer Diagnosis and Personalized Medicine

Breast cancer is a significant cause of cancer-related mortality in women worldwide. Early and precise diagnosis is crucial, and clinical outcomes can be markedly enhanced. The rise of artificial intelligence (AI) has ushered in a new era, notably in image analysis, paving the way for major advancements in breast cancer diagnosis and individualized treatment regimens. In the diagnostic workflow for patients with breast cancer, the role of AI encompasses screening, diagnosis, staging, biomarker evaluation, prognostication, and therapeutic response prediction. Although its potential is immense, its complete integration into clinical practice is challenging. Particularly, these challenges include the imperatives for extensive clinical validation, model generalizability, navigating the "black-box" conundrum, and pragmatic considerations of embedding AI into everyday clinical environments. In this review, we comprehensively explored the diverse applications of AI in breast cancer care, underlining its transformative promise and existing impediments. In radiology, we specifically address AI in mammography, tomosynthesis, risk prediction models, and supplementary imaging methods, including magnetic resonance imaging and ultrasound. In pathology, our focus is on AI applications for pathologic diagnosis, evaluation of biomarkers, and predictions related to genetic alterations, treatment response, and prognosis in the context of breast cancer diagnosis and treatment. Our discussion underscores the transformative potential of AI in breast cancer management and emphasizes the importance of focused research to realize the full spectrum of benefits of AI in patient care.

Applications of Artificial Intelligence in Breast Pathology

CONTEXT.—

Increasing implementation of whole slide imaging together with digital workflow and advances in computing capacity enable the use of artificial intelligence (AI) in pathology, including breast pathology. Breast pathologists often face a significant workload, with diagnosis complexity, tedious repetitive tasks, and semiquantitative evaluation of biomarkers. Recent advances in developing AI algorithms have provided promising approaches to meet the demand in breast pathology.

OBJECTIVE.—

To provide an updated review of AI in breast pathology. We examined the success and challenges of current and potential AI applications in diagnosing and grading breast carcinomas and other pathologic changes, detecting lymph node metastasis, quantifying breast cancer biomarkers, predicting prognosis and therapy response, and predicting potential molecular changes.

DATA SOURCES.—

We obtained data and information by searching and reviewing literature on AI in breast pathology from PubMed and based our own experience.

CONCLUSIONS.—

With the increasing application in breast pathology, AI not only assists in pathology diagnosis to improve accuracy and reduce pathologists' workload, but also provides new information in predicting prognosis and therapy response.

Spatial cellular architecture predicts prognosis in glioblastoma

Intra-tumoral heterogeneity and cell-state plasticity are key drivers for the therapeutic resistance of glioblastoma. Here, we investigate the association between spatial cellular organization and glioblastoma prognosis. Leveraging single-cell RNA-seq and spatial transcriptomics data, we develop a deep learning model to predict transcriptional subtypes of glioblastoma cells from histology images. Employing this model, we phenotypically analyze 40 million tissue spots from 410 patients and identify consistent associations between tumor architecture and prognosis across two independent cohorts. Patients with poor prognosis exhibit higher proportions of tumor cells expressing a hypoxia-induced transcriptional program. Furthermore, a clustering pattern of astrocyte-like tumor cells is associated with worse prognosis, while dispersion and connection of the astrocytes with other transcriptional subtypes correlate with decreased risk. To validate these results, we develop a separate deep learning model that utilizes histology images to predict prognosis. Applying this model to spatial transcriptomics data reveal survival-associated regional gene expression programs. Overall, our study presents a scalable approach to unravel the transcriptional heterogeneity of glioblastoma and establishes a critical connection between spatial cellular architecture and clinical outcomes.

A correlation graph attention network for classifying chromosomal instabilities from histopathology whole-slide images

The chromosome instability (CIN) is one of the hallmarks of cancer and is closely related to tumor metastasis. However, the sheer size and resolution of histopathology whole-slide images (WSIs) already challenges the capabilities of computational pathology. In this study, we propose a correlation graph attention network (MLP-GAT) that can construct graphs for classifying multi-type CINs from the WSIs of breast cancer. We construct a WSIs dataset of breast cancer from the Cancer Genome Atlas Breast Invasive Carcinoma (TCGA-BRCA). Extensive experiments show that MLP-GAT far outperforms accepted state-of-the-art methods and demonstrate the advantages of the constructed graph networks for analyzing WSI data. The visualization shows the difference among the tiles in a WSI. Furthermore, the generalization performance of the proposed method was verified on the stomach cancer. This study provides guidance for studying the relationship between CIN and cancer from the perspective of image phenotype.

Biological insights and novel biomarker discovery through deep learning approaches in breast cancer histopathology

Breast cancer remains a highly prevalent disease with considerable inter- and intra-tumoral heterogeneity complicating prognostication and treatment decisions. The utilization and depth of genomic, transcriptomic and proteomic data for cancer has exploded over recent times and the addition of spatial context to this information, by understanding the correlating morphologic and spatial patterns of cells in tissue samples, has created an exciting frontier of research, histo-genomics. At the same time, deep learning (DL), a class of machine learning algorithms employing artificial neural networks, has rapidly progressed in the last decade with a confluence of technical developments - including the advent of modern graphic processing units (GPU), allowing efficient implementation of increasingly complex architectures at scale; advances in the theoretical and practical design of network architectures; and access to larger datasets for training - all leading to sweeping advances in image classification and object detection. In this review, we examine recent developments in the application of DL in breast cancer histology with particular emphasis of those producing biologic insights or novel biomarkers, spanning the extraction of genomic information to the use of stroma to predict cancer recurrence, with the aim of suggesting avenues for further advancing this exciting field.

Single-cell Heterogeneity-aware Transformer-guided Multiple Instance Learning for Cancer Aneuploidy Prediction from Whole Slide Histopathology Images

Aneuploidy is a hallmark of aggressive malignancies associated with therapeutic resistance and poor survival. Measuring aneuploidy requires expensive specialized techniques that are not clinically applicable. Deep learning analysis of routine histopathology slides has revealed associations with genetic mutations. However, existing studies focus on image patches or tiles, and there is no prior work that predicts aneuploidy using single-cell analysis. Here, we present a single-cell heterogeneity-aware and transformer-guided deep learning framework to predict aneuploidy from whole slide histopathology images. First, we perform nuclei segmentation and classification to obtain individual cancer cells, which are clustered into multiple subtypes. The cell subtype distributions are computed to measure cancer cell heterogeneity. Additionally, morphological features of different cell subtypes are extracted. Further, we leverage a multiple instance learning module with Transformer, which encourages the network to focus on the most informative cancer cells. Lastly, a hybrid network is built to unify cell heterogeneity, morphology, and deep features for aneuploidy prediction. We train and validate our method on two public datasets from TCGA: lung adenocarcinoma (LUAD) and head and neck squamous cell carcinoma (HNSC), with 339 and 245 patients. Our model achieves promising performance with AUC of 0.818 (95% CI: 0.718-0.919) and 0.827 (95% CI: 0.704-0.949) on the LUAD and HNSC test sets, respectively. Through extensive ablation and comparison studies, we demonstrate the effectiveness of each component of the model and superior performance over alternative networks. In conclusion, we present a novel deep learning approach to predict aneuploidy from histopathology images, which could inform personalized cancer treatment.

Deep learning in digital pathology for personalized treatment plans of cancer patients

Over the past decade, many new cancer treatments have been developed and made available to patients. However, in most cases, these treatments only benefit a specific subgroup of patients, making the selection of treatment for a specific patient an essential but challenging task for oncologists. Although some biomarkers were found to associate with treatment response, manual assessment is time-consuming and subjective. With the rapid developments and expanded implementation of artificial intelligence (AI) in digital pathology, many biomarkers can be quantified automatically from histopathology images. This approach allows for a more efficient and objective assessment of biomarkers, aiding oncologists in formulating personalized treatment plans for cancer patients. This review presents an overview and summary of the recent studies on biomarker quantification and treatment response prediction using hematoxylin-eosin (H&E) stained pathology images. These studies have shown that an AI-based digital pathology approach can be practical and will become increasingly important in improving the selection of cancer treatments for patients.

The progress in our understanding of CIN in breast cancer research

Chromosomal instability (CIN) is an important marker of cancer, which is closely related to tumorigenesis, disease progression, treatment efficacy, and patient prognosis. However, due to the limitations of the currently available detection methods, its exact clinical significance remains unknown. Previous studies have demonstrated that 89% of invasive breast cancer cases possess CIN, suggesting that it has potential application in breast cancer diagnosis and treatment. In this review, we describe the two main types of CIN and discuss the associated detection methods. Subsequently, we highlight the impact of CIN in breast cancer development and progression and describe how it can influence treatment and prognosis. The goal of this review is to provide a reference on its mechanism for researchers and clinicians.

Artificial intelligence in cancer research and precision medicine: Applications, limitations and priorities to drive transformation in the delivery of equitable and unbiased care

Artificial intelligence (AI) has experienced explosive growth in oncology and related specialties in recent years. The improved expertise in data capture, the increased capacity for data aggregation and analytic power, along with decreasing costs of genome sequencing and related biologic "omics", set the foundation and need for novel tools that can meaningfully process these data from multiple sources and of varying types. These advances provide value across biomedical discovery, diagnosis, prognosis, treatment, and prevention, in a multimodal fashion. However, while big data and AI tools have already revolutionized many fields, medicine has partially lagged due to its complexity and multi-dimensionality, leading to technical challenges in developing and validating solutions that generalize to diverse populations. Indeed, inner biases and miseducation of algorithms, in view of their implementation in daily clinical practice, are increasingly relevant concerns; critically, it is possible for AI to mirror the unconscious biases of the humans who generated these algorithms. Therefore, to avoid worsening existing health disparities, it is critical to employ a thoughtful, transparent, and inclusive approach that involves addressing bias in algorithm design and implementation along the cancer care continuum. In this review, a broad landscape of major applications of AI in cancer care is provided, with a focus on cancer research and precision medicine. Major challenges posed by the implementation of AI in the clinical setting will be discussed. Potentially feasible solutions for mitigating bias are provided, in the light of promoting cancer health equity.

Chromosomal Instability as Enabling Feature and Central Hallmark of Breast Cancer

Chromosomal instability (CIN) has become a topic of great interest in recent years, not only for its implications in cancer diagnosis and prognosis but also for its role as an enabling feature and central hallmark of cancer. CIN describes cell-to-cell variation in the number or structure of chromosomes in a tumor population. Although extensive research in recent decades has identified some associations between CIN with response to therapy, specific associations with other hallmarks of cancer have not been fully evidenced. Such associations place CIN as an enabling feature of the other hallmarks of cancer and highlight the importance of deepening its knowledge to improve the outcome in cancer. In addition, studies conducted to date have shown paradoxical findings about the implications of CIN for therapeutic response, with some studies showing associations between high CIN and better therapeutic response, and others showing the opposite: associations between high CIN and therapeutic resistance. This evidences the complex relationships between CIN with the prognosis and response to treatment in cancer. Considering the above, this review focuses on recent studies on the role of CIN in cancer, the cellular mechanisms leading to CIN, its relationship with other hallmarks of cancer, and the emerging therapeutic approaches that are being developed to target such instability, with a primary focus on breast cancer. Further understanding of the complexity of CIN and its association with other hallmarks of cancer could provide a better understanding of the cellular and molecular mechanisms involved in prognosis and response to treatment in cancer and potentially lead to new drug targets.

Machine learning can aid in prediction of IDH mutation from H&E-stained histology slides in infiltrating gliomas

While Machine Learning (ML) models have been increasingly applied to a range of histopathology tasks, there has been little emphasis on characterizing these models and contrasting them with human experts. We present a detailed empirical analysis comparing expert neuropathologists and ML models at predicting IDH mutation status in H&E-stained histology slides of infiltrating gliomas, both independently and synergistically. We find that errors made by neuropathologists and ML models trained using the TCGA dataset are distinct, representing modest agreement between predictions (human-vs.-human κ = 0.656; human-vs.-ML model κ = 0.598). While no ML model surpassed human performance on an independent institutional test dataset (human AUC = 0.901, max ML AUC = 0.881), a hybrid model aggregating human and ML predictions demonstrates predictive performance comparable to the consensus of two expert neuropathologists (hybrid classifier AUC = 0.921 vs. two-neuropathologist consensus AUC = 0.920). We also show that models trained at different levels of magnification exhibit different types of errors, supporting the value of aggregation across spatial scales in the ML approach. Finally, we present a detailed interpretation of our multi-scale ML ensemble model which reveals that predictions are driven by human-identifiable features at the patch-level.

Deep Learning-Based Prediction of Molecular Tumor Biomarkers from H&E: A Practical Review

Molecular and genomic properties are critical in selecting cancer treatments to target individual tumors, particularly for immunotherapy. However, the methods to assess such properties are expensive, time-consuming, and often not routinely performed. Applying machine learning to H&E images can provide a more cost-effective screening method. Dozens of studies over the last few years have demonstrated that a variety of molecular biomarkers can be predicted from H&E alone using the advancements of deep learning: molecular alterations, genomic subtypes, protein biomarkers, and even the presence of viruses. This article reviews the diverse applications across cancer types and the methodology to train and validate these models on whole slide images. From bottom-up to pathologist-driven to hybrid approaches, the leading trends include a variety of weakly supervised deep learning-based approaches, as well as mechanisms for training strongly supervised models in select situations. While results of these algorithms look promising, some challenges still persist, including small training sets, rigorous validation, and model explainability. Biomarker prediction models may yield a screening method to determine when to run molecular tests or an alternative when molecular tests are not possible. They also create new opportunities in quantifying intratumoral heterogeneity and predicting patient outcomes.

Deep learning model to predict Epstein-Barr virus associated gastric cancer in histology

The detection of Epstein-Barr virus (EBV) in gastric cancer patients is crucial for clinical decision making, as it is related with specific treatment responses and prognoses. Despite its importance, the limited medical resources preclude universal EBV testing. Herein, we propose a deep learning-based EBV prediction method from H&E-stained whole-slide images (WSI). Our model was developed using 319 H&E stained WSI (26 EBV positive; TCGA dataset) from the Cancer Genome Atlas, and 108 WSI (8 EBV positive; ISH dataset) from an independent institution. Our deep learning model, EBVNet consists of two sequential components: a tumor classifier and an EBV classifier. We visualized the learned representation by the classifiers using UMAP. We externally validated the model using 60 additional WSI (7 being EBV positive; HGH dataset). We compared the model's performance with those of four pathologists. EBVNet achieved an AUPRC of 0.65, whereas the four pathologists yielded a mean AUPRC of 0.41. Moreover, EBVNet achieved an negative predictive value, sensitivity, specificity, precision, and F1-score of 0.98, 0.86, 0.92, 0.60, and 0.71, respectively. Our proposed model is expected to contribute to prescreen patients for confirmatory testing, potentially to save test-related cost and labor.

Chromosomal Instability, Selection and Competition: Factors That Shape the Level of Karyotype Intra-Tumor Heterogeneity

Intra-tumor heterogeneity (ITH) is a pan-cancer predictor of survival, with high ITH being correlated to a dismal prognosis. The level of ITH is, hence, a clinically relevant characteristic of a malignancy. ITH of karyotypes is driven by chromosomal instability (CIN). However, not all new karyotypes generated by CIN are viable or competitive, which limits the amount of ITH. Here, we review the cellular processes and ecological properties that determine karyotype ITH. We propose a framework to understand karyotype ITH, in which cells with new karyotypes emerge through CIN, are selected by cell intrinsic and cell extrinsic selective pressures, and propagate through a cancer in competition with other malignant cells. We further discuss how CIN modulates the cell phenotype and immune microenvironment, and the implications this has for the subsequent selection of karyotypes. Together, we aim to provide a comprehensive overview of the biological processes that shape the level of karyotype heterogeneity.

Copy number heterogeneity identifies ER+ breast cancer patients that do not benefit from adjuvant endocrine therapy

BACKGROUND

Endocrine therapy forms the backbone of adjuvant treatment for oestrogen-receptor-positive (ER+) breast cancer. However, it remains unclear whether adjuvant treatment improves survival rates in low-risk patients. Low intra-tumour heterogeneity (ITH) has been shown to confer low risk for recurrent disease. Here, it is studied if chromosomal copy-number ITH (CNH) can identify low-risk ER+, lymph-node-negative breast cancer patients who do not benefit from adjuvant endocrine therapy.

METHODS

Lymph-node-negative ER+ patients from the observational METABRIC dataset were retrospectively analysed (n = 708). CNH was determined from a single bulk copy-number measurement for each patient. Survival rates were compared between patients that did or did not receive adjuvant endocrine therapy for CNH-low, middle and high groups with Cox proportional-hazards models, using propensity-score weights to correct for confounders.

RESULTS

Adjuvant endocrine therapy improved the relapse-free survival (RFS) for CNH-high patients treatment (HR = 0.55), but not for CNH-low patients treatment (HR = 0.88). For CNH-low patients adjuvant endocrine therapy was associated with impaired OS (HR = 1.62).

CONCLUSIONS

This retrospective study of lymph-node-negative, ER+ breast cancer finds that patients identified as low risk using CNH do not benefit from adjuvant endocrine therapy.

Deep-Learning to Predict BRCA Mutation and Survival from Digital H&E Slides of Epithelial Ovarian Cancer

BRCA 1/2 genes mutation status can already determine the therapeutic algorithm of high grade serous ovarian cancer patients. Nevertheless, its assessment is not sufficient to identify all patients with genomic instability, since BRCA 1/2 mutations are only the most well-known mechanisms of homologous recombination deficiency (HR-d) pathway, and patients displaying HR-d behave similarly to BRCA mutated patients. HRd assessment can be challenging and is progressively overcoming BRCA testing not only for prognostic information but more importantly for drugs prescriptions. However, HR testing is not already integrated in clinical practice, it is quite expensive and it is not refundable in many countries. Selecting patients who are more likely to benefit from this assessment (BRCA 1/2 WT patients) at an early stage of the diagnostic process, would allow an optimization of genomic profiling resources. In this study, we sought to explore whether somatic BRCA1/2 genes status can be predicted using computational pathology from standard hematoxylin and eosin histology. In detail, we adopted a publicly available, deep-learning-based weakly supervised method that uses attention-based learning to automatically identify sub regions of high diagnostic value to accurately classify the whole slide (CLAM). The same model was also tested for progression free survival (PFS) prediction. The model was tested on a cohort of 664 (training set: n = 464, testing set: n = 132) ovarian cancer patients, of whom 233 (35.1%) had a somatic BRCA 1/2 mutation. An area under the curve of 0.7 and 0.55 was achieved in the training and testing set respectively. The model was then further refined by manually identifying areas of interest in half of the cases. 198 images were used for training (126/72) and 87 images for validation (55/32). The model reached a zero classification error on the training set, but the performance was 0.59 in terms of validation ROC AUC, with a 0.57 validation accuracy. Finally, when applied to predict PFS, the model achieved an AUC of 0.71, with a negative predictive value of 0.69, and a positive predictive value of 0.75. Based on these analyses, we have planned further steps of development such as proving a reference classification performance, exploring the hyperparameters space for training optimization, eventually tweaking the learning algorithms and the neural networks architecture for better suiting this specific task. These actions may allow the model to improve performances for all the considered outcomes.

Spatial Characterization of Tumor-Infiltrating Lymphocytes and Breast Cancer Progression

Tumor-infiltrating lymphocytes (TILs) have been established as a robust prognostic biomarker in breast cancer, with emerging utility in predicting treatment response in the adjuvant and neoadjuvant settings. In this study, the role of TILs in predicting overall survival and progression-free interval was evaluated in two independent cohorts of breast cancer from the Cancer Genome Atlas (TCGA BRCA) and the Carolina Breast Cancer Study (UNC CBCS). We utilized machine learning and computer vision algorithms to characterize TIL infiltrates in digital whole-slide images (WSIs) of breast cancer stained with hematoxylin and eosin (H&E). Multiple parameters were used to characterize the global abundance and spatial features of TIL infiltrates. Univariate and multivariate analyses show that large aggregates of peritumoral and intratumoral TILs (forests) were associated with longer survival, whereas the absence of intratumoral TILs (deserts) is associated with increased risk of recurrence. Patients with two or more high-risk spatial features were associated with significantly shorter progression-free interval (PFI). This study demonstrates the practical utility of Pathomics in evaluating the clinical significance of the abundance and spatial patterns of distribution of TIL infiltrates as important biomarkers in breast cancer.

Artificial intelligence applied to breast pathology

The convergence of digital pathology and computer vision is increasingly enabling computers to perform tasks performed by humans. As a result, artificial intelligence (AI) is having an astoundingly positive effect on the field of pathology, including breast pathology. Research using machine learning and the development of algorithms that learn patterns from labeled digital data based on "deep learning" neural networks and feature-engineered approaches to analyze histology images have recently provided promising results. Thus far, image analysis and more complex AI-based tools have demonstrated excellent success performing tasks such as the quantification of breast biomarkers and Ki67, mitosis detection, lymph node metastasis recognition, tissue segmentation for diagnosing breast carcinoma, prognostication, computational assessment of tumor-infiltrating lymphocytes, and prediction of molecular expression as well as treatment response and benefit of therapy from routine H&E images. This review critically examines the literature regarding these applications of AI in the area of breast pathology.

Chromosomal instability and aneuploidy as causes of cancer drug resistance

High levels of aneuploidy and chromosomal instability (CIN) are correlated with poor patient outcomes, though the mechanism(s) underlying this relationship have not been established. Recent evidence has demonstrated that chromosome copy number changes can function as point mutation-independent sources of drug resistance in cancer, which may partially explain this clinical association. CIN generates intratumoral heterogeneity in the form of gene dosage alterations, upon which the selective pressures induced by drug treatments can act. Thus, although CIN and aneuploidy impair cell fitness under most conditions, CIN can augment cellular adaptability, establishing CIN as a bet-hedging mechanism in tumor evolution. CIN may also endow cancers with unique vulnerabilities, which could be exploited therapeutically to achieve better patient outcomes.

Chromosomal instability accelerates the evolution of resistance to anti-cancer therapies

Aneuploidy is a ubiquitous feature of human tumors, but the acquisition of aneuploidy typically antagonizes cellular fitness. To investigate how aneuploidy could contribute to tumor growth, we triggered periods of chromosomal instability (CIN) in human cells and then exposed them to different culture environments. We discovered that transient CIN reproducibly accelerates the acquisition of resistance to anti-cancer therapies. Single-cell sequencing revealed that these resistant populations develop recurrent aneuploidies, and independently deriving one chromosome-loss event that was frequently observed in paclitaxel-resistant cells was sufficient to decrease paclitaxel sensitivity. Finally, we demonstrated that intrinsic levels of CIN correlate with poor responses to numerous therapies in human tumors. Our results show that, although CIN generally decreases cancer cell fitness, it also provides phenotypic plasticity to cancer cells that can allow them to adapt to diverse stressful environments. Moreover, our findings suggest that aneuploidy may function as an under-explored cause of therapy failure.

Making Sense of Genetic Information: The Promising Evolution of Clinical Stratification and Precision Oncology Using Machine Learning

Precision medicine is a medical approach to administer patients with a tailored dose of treatment by taking into consideration a person's variability in genes, environment, and lifestyles. The accumulation of omics big sequence data led to the development of various genetic databases on which clinical stratification of high-risk populations may be conducted. In addition, because cancers are generally caused by tumor-specific mutations, large-scale systematic identification of single nucleotide polymorphisms (SNPs) in various tumors has propelled significant progress of tailored treatments of tumors (i.e., precision oncology). Machine learning (ML), a subfield of artificial intelligence in which computers learn through experience, has a great potential to be used in precision oncology chiefly to help physicians make diagnostic decisions based on tumor images. A promising venue of ML in precision oncology is the integration of all available data from images to multi-omics big data for the holistic care of patients and high-risk healthy subjects. In this review, we provide a focused overview of precision oncology and ML with attention to breast cancer and glioma as well as the Bayesian networks that have the flexibility and the ability to work with incomplete information. We also introduce some state-of-the-art attempts to use and incorporate ML and genetic information in precision oncology.