KRASFormer: a fully vision transformer-based framework for predicting KRAS gene mutations in histopathological images of colorectal cancer.

Detecting the Kirsten Rat Sarcoma Virus (KRAS) gene mutation is significant for colorectal cancer (CRC) patients. The KRAS gene encodes a protein involved in the epidermal growth factor receptor (EGFR) signaling pathway, and mutations in this gene can negatively impact the use of monoclonal antibodies in antiEGFR therapy and affect treatment decisions. Currently, commonly used methods like next-generation sequencing (NGS) identify KRAS mutations but are expensive, time-consuming, and may not be suitable for every cancer patient sample. To address these challenges, we have developed KRASFormer, a novel framework that predicts KRAS gene mutations from Haematoxylin and Eosin (H&E) stained WSIs that are widely available for most CRC patients. KRASFormer consists of two stages: the first stage filters out non-tumor regions and selects only tumour cells using a quality screening mechanism, and the second stage predicts the KRAS gene either 'wildtype' or 'mutant' using a Vision Transformer-based XCiT method. The XCiT employs cross-covariance attention to capture clinically meaningful long-range representations of textural patterns in tumour tissue and KRAS mutant cells. We evaluated the performance of the first stage using an independent CRC-5000 dataset, and the second stage included both The Cancer Genome Atlas colon and rectal cancer (TCGA-CRCDX) and in-house cohorts. The results of our experiments showed that the XCiT outperformed existing state-of-the-art methods, achieving AUCs for ROC curves of 0.691 and 0.653 on TCGA-CRC-DX and in-house datasets, respectively. Our findings emphasize three key consequences: the potential of using H&E-stained tissue slide images for predicting KRAS gene mutations as a cost-effective and time-efficient means for guiding treatment choice with CRC patients; the increase in performance metrics of a Transformer-based model; and the value of the collaboration between pathologists and data scientists in deriving a morphologically meaningful model.

True-T - Improving T-cell response quantification with holistic artificial intelligence based prediction in immunohistochemistry images.

The immune response associated with oncogenesis and potential oncological ther- apeutic interventions has dominated the field of cancer research over the last decade. T-cell lymphocytes in the tumor microenvironment are a crucial aspect of cancer's adaptive immunity, and the quantification of T-cells in specific can- cer types has been suggested as a potential diagnostic aid. However, this is cur- rently not part of routine diagnostics. To address this challenge, we present a new method called True-T, which employs artificial intelligence-based techniques to quantify T-cells in colorectal cancer (CRC) using immunohistochemistry (IHC) images. True-T analyses the chromogenic tissue hybridization signal of three widely recognized T-cell markers (CD3, CD4, and CD8). Our method employs a pipeline consisting of three stages: T-cell segmentation, density estimation from the segmented mask, and prediction of individual five-year survival rates. In the first stage, we utilize the U-Net method, where a pre-trained ResNet-34 is em- ployed as an encoder to extract clinically relevant T-cell features. The segmenta- tion model is trained and evaluated individually, demonstrating its generalization in detecting the CD3, CD4, and CD8 biomarkers in IHC images. In the second stage, the density of T-cells is estimated using the predicted mask, which serves as a crucial indicator for patient survival statistics in the third stage. This ap- proach was developed and tested in 1041 patients from four reference diagnostic institutions, ensuring broad applicability. The clinical effectiveness of True-T is demonstrated in stages II-IV CRC by offering valuable prognostic information that surpasses previous quantitative gold standards, opening possibilities for po- tential clinical applications. Finally, to evaluate the robustness and broader ap- plicability of our approach without additional training, we assessed the universal accuracy of the CD3 component of the True-T algorithm across 13 distinct solid tumors.

Translation of tissue-based artificial intelligence into clinical practice: from discovery to adoption.

Digital pathology (DP), or the digitization of pathology images, has transformed oncology research and cancer diagnostics. The application of artificial intelligence (AI) and other forms of machine learning (ML) to these images allows for better interpretation of morphology, improved quantitation of biomarkers, introduction of novel concepts to discovery and diagnostics (such as spatial distribution of cellular elements), and the promise of a new paradigm of cancer biomarkers. The application of AI to tissue analysis can take several conceptual approaches, within the domains of language modelling and image analysis, such as Deep Learning Convolutional Neural Networks, Multiple Instance Learning approaches, or the modelling of risk scores and their application to ML. The use of different approaches solves different problems within pathology workflows, including assistive applications for the detection and grading of tumours, quantification of biomarkers, and the delivery of established and new image-based biomarkers for treatment prediction and prognostic purposes. All these AI formats, applied to digital tissue images, are also beginning to transform our approach to clinical trials. In parallel, the novelty of DP/AI devices and the related computational science pipeline introduces new requirements for manufacturers to build into their design, development, regulatory and post-market processes, which may need to be taken into account when using AI applied to tissues in cancer discovery. Finally, DP/AI represents challenge to the way we accredit new diagnostic tools with clinical applicability, the understanding of which will allow cancer patients to have access to a new generation of complex biomarkers.

Molecular profiling of aromatase inhibitor sensitive and resistant ER+HER2- postmenopausal breast cancers.

Aromatase inhibitors (AIs) reduce recurrences and mortality in postmenopausal patients with oestrogen receptor positive (ER+) breast cancer (BC), but >20% of patients will eventually relapse. Given the limited understanding of intrinsic resistance in these tumours, here we conduct a large-scale molecular analysis to identify features that impact on the response of ER + HER2- BC to AI. We compare the 15% of poorest responders (PRs, n = 177) as measured by proportional Ki67 changes after 2 weeks of neoadjuvant AI to good responders (GRs, n = 190) selected from the top 50% responders in the POETIC trial and matched for baseline Ki67 categories. In this work, low ESR1 levels are associated with poor response, high proliferation, high expression of growth factor pathways and non-luminal subtypes. PRs having high ESR1 expression have similar proportions of luminal subtypes to GRs but lower plasma estradiol levels, lower expression of estrogen response genes, higher levels of tumor infiltrating lymphocytes and immune markers, and more TP53 mutations.

Technical note on the exploration of COVID-19 in autopsy material.

Interrogation of immune response in autopsy material from patients with SARS-CoV-2 is potentially significant. We aim to describe a validated protocol for the exploration of the molecular physiopathology of SARS-CoV-2 pulmonary disease using multiplex immunofluorescence (mIF).The application of validated assays for the detection of SARS-CoV-2 in tissues, originally developed in our laboratory in the context of oncology, was used to map the topography and complexity of the adaptive immune response at protein and mRNA levels.SARS-CoV-2 is detectable in situ by protein or mRNA, with a sensitivity that could be in part related to disease stage. In formalin-fixed, paraffin-embedded pneumonia material, multiplex immunofluorescent panels are robust, reliable and quantifiable and can detect topographic variations in inflammation related to pathological processes.Clinical autopsies have relevance in understanding diseases of unknown/complex pathophysiology. In particular, autopsy materials are suitable for the detection of SARS-CoV-2 and for the topographic description of the complex tissue-based immune response using mIF.

ICOSeg: Real-Time ICOS Protein Expression Segmentation from Immunohistochemistry Slides Using a Lightweight Conv-Transformer Network.

In this article, we propose ICOSeg, a lightweight deep learning model that accurately segments the immune-checkpoint biomarker, Inducible T-cell COStimulator (ICOS) protein in colon cancer from immunohistochemistry (IHC) slide patches. The proposed model relies on the MobileViT network that includes two main components: convolutional neural network (CNN) layers for extracting spatial features; and a transformer block for capturing a global feature representation from IHC patch images. The ICOSeg uses an encoder and decoder sub-network. The encoder extracts the positive cell's salient features (i.e., shape, texture, intensity, and margin), and the decoder reconstructs important features into segmentation maps. To improve the model generalization capabilities, we adopted a channel attention mechanism that added to the bottleneck of the encoder layer. This approach highlighted the most relevant cell structures by discriminating between the targeted cell and background tissues. We performed extensive experiments on our in-house dataset. The experimental results confirm that the proposed model achieves more significant results against state-of-the-art methods, together with an 8× reduction in parameters.

General Roadmap and Core Steps for the Development of AI Tools in Digital Pathology.

Integrating artificial intelligence (AI) tools in the tissue diagnostic workflow will benefit the pathologist and, ultimately, the patient. The generation of such AI tools has two parallel and yet interconnected processes, namely the definition of the pathologist's task to be delivered in silico, and the software development requirements. In this review paper, we demystify this process, from a viewpoint that joins experienced pathologists and data scientists, by proposing a general pathway and describing the core steps to build an AI digital pathology tool. In doing so, we highlight the importance of the collaboration between AI scientists and pathologists, from the initial formulation of the hypothesis to the final, ready-to-use product.

A Means of Assessing Deep Learning-Based Detection of ICOS Protein Expression in Colon Cancer.

Biomarkers identify patient response to therapy. The potential immune-checkpoint biomarker, Inducible T-cell COStimulator (ICOS), expressed on regulating T-cell activation and involved in adaptive immune responses, is of great interest. We have previously shown that open-source software for digital pathology image analysis can be used to detect and quantify ICOS using cell detection algorithms based on traditional image processing techniques. Currently, artificial intelligence (AI) based on deep learning methods is significantly impacting the domain of digital pathology, including the quantification of biomarkers. In this study, we propose a general AI-based workflow for applying deep learning to the problem of cell segmentation/detection in IHC slides as a basis for quantifying nuclear staining biomarkers, such as ICOS. It consists of two main parts: a simplified but robust annotation process, and cell segmentation/detection models. This results in an optimised annotation process with a new user-friendly tool that can interact with1 other open-source software and assists pathologists and scientists in creating and exporting data for deep learning. We present a set of architectures for cell-based segmentation/detection to quantify and analyse the trade-offs between them, proving to be more accurate and less time consuming than traditional methods. This approach can identify the best tool to deliver the prognostic significance of ICOS protein expression.

Identifying mismatch repair-deficient colon cancer: near-perfect concordance between immunohistochemistry and microsatellite instability testing in a large, population-based series.

AIMS: Establishing the mismatch repair (MMR) status of colorectal cancers is important to enable the detection of underlying Lynch syndrome and inform prognosis and therapy. Current testing typically involves either polymerase chain reaction (PCR)-based microsatellite instability (MSI) testing or MMR protein immunohistochemistry (IHC). The aim of this study was to compare these two approaches in a large, population-based cohort of stage 2 and 3 colon cancer cases in Northern Ireland. METHODS AND RESULTS: The study used the Promega pentaplex assay to determine MSI status and a four-antibody MMR IHC panel. IHC was applied to tumour tissue microarrays with triplicate tumour sampling, and assessed manually. Of 593 cases with available MSI and MMR IHC results, 136 (22.9%) were MSI-high (MSI-H) and 135 (22.8%) showed abnormal MMR IHC. Concordance was extremely high, with 97.1% of MSI-H cases showing abnormal MMR IHC, and 97.8% of cases with abnormal IHC showing MSI-H status. Under-representation of tumour epithelial cells in samples from heavily inflamed tumours resulted in misclassification of several cases with abnormal MMR IHC as microsatellite-stable. MMR IHC revealed rare cases with unusual patterns of MMR protein expression, unusual combinations of expression loss, or secondary clonal loss of expression, as further illustrated by repeat immunostaining on whole tissue sections. CONCLUSIONS: MSI PCR testing and MMR IHC can be considered to be equally proficient tests for establishing MMR/MSI status, when there is awareness of the potential pitfalls of either method. The choice of methodology may depend on available services and expertise.

Systematic evaluation of PAXgene® tissue fixation for the histopathological and molecular study of lung cancer.

Whilst adequate for most existing pathological tests, formalin is generally considered a poor DNA preservative and use of alternative fixatives may prove advantageous for molecular testing of tumour material; an increasingly common approach to identify targetable driver mutations in lung cancer patients. We collected paired PAXgene® tissue-fixed and formalin-fixed samples of block-sized tumour and lung parenchyma, Temno-needle core tumour biopsies and fine needle tumour aspirates (FNAs) from non-small cell lung cancer resection specimens. Traditionally processed formalin fixed paraffin wax embedded (FFPE) samples were compared to paired PAXgene® tissue fixed paraffin-embedded (PFPE) samples. We evaluated suitability for common laboratory tests (H&E staining and immunohistochemistry) and performance for downstream molecular investigations relevant to lung cancer, including RT-PCR and next generation DNA sequencing (NGS). Adequate and comparable H&E staining was seen in all sample types and nuclear staining was preferable in PAXgene® fixed Temno tumour biopsies and tumour FNA samples. Immunohistochemical staining was broadly comparable. PFPE samples enabled greater yields of less-fragmented DNA than FFPE comparators. PFPE samples were also superior for PCR and NGS performance, both in terms of quality control metrics and for variant calling. Critically we identified a greater number of genetic variants in the epidermal growth factor receptor gene when using PFPE samples and the Ingenuity® Variant Analysis pipeline. In summary, PFPE samples are adequate for histopathological diagnosis and suitable for the majority of existing laboratory tests. PAXgene® fixation is superior for DNA and RNA integrity, particularly in low-yield samples and facilitates improved NGS performance, including the detection of actionable lung cancer mutations for precision medicine in lung cancer samples.

Practical guide for the comparison of two next-generation sequencing systems for solid tumour analysis in a universal healthcare system.

AIMS: Although there have been excellent reports in the literature of validating next-generation sequencing, comparisons between two systems are not often published due to cost and time. We set out to establish that targetable mutations could be reliably detected with different gene panels and different chemistries using a common bioinformatics pipeline for meaningful comparisons to be made. METHODS: After running selected formalin-fixed, paraffin-embedded samples through QPCR, Sanger sequencing and the 50 gene hotspot v2 panel from Life Technologies to determine standard-of-care variants, we compared the Oncomine panel from Life Technologies performed on a Personal Genome Machine (PGM) and the eight-gene actionable panel from Qiagen performed on a MiSeq platform. We used a common bioinformatics program following the creation of respective VCF files. RESULTS: Both panels were accurate to above 90%, the actionable panel workflow was easier to perform but the lowest effective starting DNA load was obtained on the Oncomine workflow at 4 ng. Such minimal DNA can help with samples where there is limited material such as those for lung cancer molecular studies. We also discuss gene panel content and propose that increasing the gene profile of a panel will not benefit clinical laboratories where standard-of-care testing is all that is required. CONCLUSIONS: Once recognised, it may be cost-effective for such laboratories to begin validation with an appropriate bioinformatics pipeline for targeted multigene hotspot molecular testing.

Critical Appraisal of Programmed Death Ligand 1 Reflex Diagnostic Testing: Current Standards and Future Opportunities.

INTRODUCTION: Patient suitability to anti-programmed death ligand 1 (PD-L1) immune checkpoint inhibition is key to the treatment of NSCLC. We present, applied to PD-L1 testing: a comprehensive cross-validation of two immunohistochemistry (IHC) clones; our descriptive experience in diagnostic reflex testing; the concordance of IHC to in situ RNA (RNA-ISH); and application of digital pathology. METHODS: Eight hundred thirteen NSCLC tumor samples collected from 564 diagnostic samples were analyzed prospectively, and 249 diagnostic samples analyzed retrospectively in tissue microarray format. Validated methods for IHC and RNA-ISH were tested in tissue microarrays and full sections and the QuPath system were used for digital pathology analysis. RESULTS: Antibody concordance of clones SP263 and 22C3 validation was 97% to 98% in squamous cell carcinoma and adenocarcinomas, respectively. Clinical NSCLC cases were reported as PD-L1-negative (48%), 1% to 49% (23%), and more than 50% (29%), with differences associated to tissue-type and EGFR status. Comparison of IHC and RNA-ISH was highly concordant in both subgroups. Comparison of digital assessment versus manual assessment was highly concordant. Discrepancies were mostly around the 1% clinical threshold. Challenging IHC interpretation included 1) calculating the total tumor cell denominator and the nature of PD-L1 expressing cell aggregates in cytology samples; 2) peritumoral expression of positive immune cells; 3) calculation of positive tumor percentages around clinical thresholds; and 4) relevance of the 100 malignant cell rule. CONCLUSIONS: Sample type and EGFR status dictate differences in the expected percentage of PD-L1 expression. Analysis of PD-L1 is challenging, and interpretative guidelines are discussed. PD-L1 evaluations by RNA-ISH and digital pathology appear reliable, particularly in adenocarcinomas.

Molecular profiling of signet ring cell colorectal cancer provides a strong rationale for genomic targeted and immune checkpoint inhibitor therapies.

BACKGROUND: Signet ring cell colorectal cancer (SRCCa) has a bleak prognosis. Employing molecular pathology techniques we investigated the potential of precision medicine in this disease. METHODS: Using test (n=26) and validation (n=18) cohorts, analysis of mutations, DNA methylation and transcriptome was carried out. Microsatellite instability (MSI) status was established and immunohistochemistry (IHC) was used to test for adaptive immunity (CD3) and the immune checkpoint PDL1. RESULTS: DNA methylation data split the cohorts into hypermethylated (n=18, 41%) and hypomethylated groups (n=26, 59%). The hypermethylated group predominant in the proximal colon was enriched for CpG island methylator phenotype (CIMP), BRAF V600E mutation and MSI (P<0.001). These cases also had a high CD3+ immune infiltrate (P<0.001) and expressed PDL1 (P=0.03 in intra-tumoural lymphoid cells). The hypomethylated group predominant in the distal colon did not show any characteristic molecular features. We also detected a common targetable KIT mutation (c.1621A>C) across both groups. No statistically significant difference in outcome was observed between the two groups. CONCLUSIONS: Our data show that SRCCa phenotype comprises two distinct genotypes. The MSI+/CIMP+/BRAF V600E+/CD3+/PDL1+ hypermethylated genotype is an ideal candidate for immune checkpoint inhibitor therapy. In addition, one fourth of SRCCa cases can potentially be targeted by KIT inhibitors.

Erythropoietin drives breast cancer progression by activation of its receptor EPOR.

Breast cancer is a leading cause of cancer-related deaths. Anemia is common in breast cancer patients and can be treated with blood transfusions or with recombinant erythropoietin (EPO) to stimulate red blood cell production. Clinical studies have indicated decreased survival in some groups of cancer patients treated with EPO. Numerous tumor cells express the EPO receptor (EPOR), posing a risk that EPO treatment would enhance tumor growth, but the mechanisms involved in breast tumor progression are poorly understood.Here, we have examined the functional role of the EPO-EPOR axis in pre-clinical models of breast cancer. EPO induced the activation of PI3K/AKT and MAPK pathways in human breast cancer cell lines. EPOR knockdown abrogated human tumor cell growth, induced apoptosis through Bim, reduced invasiveness, and caused downregulation of MYC expression. EPO-induced MYC expression is mediated through the PI3K/AKT and MAPK pathways, and overexpression of MYC partially rescued loss of cell proliferation caused by EPOR downregulation. In a xenotransplantation model, designed to simulate recombinant EPO therapy in breast cancer patients, knockdown of EPOR markedly reduced tumor growth.Thus, our experiments in vitro and in vivo demonstrate that functional EPOR signaling is essential for the tumor-promoting effects of EPO and underline the importance of the EPO-EPOR axis in breast tumor progression.

Tissue-based next generation sequencing: application in a universal healthcare system.

In the context of solid tumours, the evolution of cancer therapies to more targeted and nuanced approaches has led to the impetus for personalised medicine. The targets for these therapies are largely based on the driving genetic mutations of the tumours. To track these multiple driving mutations the use of next generation sequencing (NGS) coupled with a morphomolecular approach to tumours, has the potential to deliver on the promises of personalised medicine. A review of NGS and its application in a universal healthcare (UHC) setting is undertaken as the technology has a wide appeal and utility in diagnostic, clinical trial and research paradigms. Furthermore, we suggest that these can be accommodated with a unified integromic approach. Challenges remain in bringing NGS to routine clinical use and these include validation, handling of the large amounts of information flow and production of a clinically useful report. These challenges are particularly acute in the setting of UHC where tests are not reimbursed and there are finite resources available. It is our opinion that the challenges faced in applying NGS in a UHC setting are surmountable and we outline our approach for its routine application in diagnostic, clinical trial and research paradigms.

ALK Immunohistochemistry in NSCLC: Discordant Staining Can Impact Patient Treatment Regimen.

INTRODUCTION: Diagnostic immunohistochemistry (IHC) is increasingly accepted as a screening method for anaplastic lymphoma receptor tyrosine kinase gene (ALK) rearrangements in NSCLC. We have sought to establish an ongoing robust external quality assessment process to gauge quality of anaplastic lymphoma kinase (ALK) IHC, which can have an impact on interpretation of patient samples. METHODS: Unstained tissue and cell line samples were distributed on a quarterly basis to participating laboratories from 30 countries. Participants stained the slide using their routine diagnostic ALK IHC method and returned the slide along with their in-house control and methodology details. Slides were assessed by a team of pathologists and scientists. RESULTS: Overall, there was a mean pass rate of 83% (range 71%-98%), with 38 variations in staining protocol. Methods included the following: the Roche D5F3 assay (65% of users, pass rate 93%); Novocastra 5A4 (15% of users, pass rate 65%); Cell Signaling Technology D5F3 (7% of users, pass rate 91%), and Dako ALK1 (5% of users, pass rate 50%). Choice of methodology directly affected final interpretation of distributed ALK-positive and ALK-negative NSCLC cases, which were correctly identified by 89% and 88% of participants, respectively. Antibody detection method was a contributing factor in false-negative staining results. The choice of laboratory controls was found to be unsuitable, and as such, in-house control recommendations are also provided. CONCLUSIONS: ALK IHC is a robust screening technique, but there is concern that some diagnostic laboratories are using inadequate staining methods, which has a direct impact on final interpretation. External assessment helps provide laboratories with continued confidence in their ALK IHC testing.

Quantification of HER2 heterogeneity in breast cancer-implications for identification of sub-dominant clones for personalised treatment.

Breast cancer is a heterogeneous disease, at both an inter- and intra-tumoural level. Appreciating heterogeneity through the application of biomarkers and molecular signatures adds complexity to tumour taxonomy but is key to personalising diagnosis, treatment and prognosis. The extent to which heterogeneity exists, and its interpretation remains a challenge to pathologists. Using HER2 as an exemplar, we have developed a simple reproducible heterogeneity index. Cell-to-cell HER2 heterogeneity was extensive in a proportion of both reported 'amplified' and 'non-amplified' cases. The highest levels of heterogeneity objectively identified occurred in borderline categories and higher ratio non-amplified cases. A case with particularly striking heterogeneity was analysed further with an array of biomarkers in order to assign a molecular diagnosis. Broad biological complexity was evident. In essence, interpretation, depending on the area of tumour sampled, could have been one of three distinct phenotypes, each of which would infer different therapeutic interventions. Therefore, we recommend that heterogeneity is assessed and taken into account when determining treatment options.

Validation of immunocytochemistry as a morphomolecular technique.

Immunocytochemistry (ICC) is a long-established means for clinical laboratories to investigate material for which it is difficult to obtain tissue samples. Unlike immunohistochemistry (IHC), the cells do not retain surrounding tissue environment/architecture. This can be of benefit in that fixation is often immediate and rapid, protecting the cells. Although fixation is frequently observed as the main preanalytic variable of test quality, all cytology preanalytic factors should be identified and controlled. In addition, the validation of ICC should take the same rigorous approach that other molecular pathology techniques follow. A three-step validation protocol is offered here. The end result is a comprehensive, morphomolecular approach to ICC, with an emphasis on therapeutic ICC. Cancer Cytopathol 2016;124:540-5. © 2016 American Cancer Society.