Microsatellite instability (MSI) is an important biomarker in colorectal cancer, influencing both patient prognosis and treatment decisions. Current approaches for MSI prediction from hematoxylin and eosin--stained whole-slide images (WSI) rely on end-to-end deep learning ("black-box") models with limited interpretability, often relying on heatmaps for visualization. However, experienced pathologists can intuitively identify MSI through specific histologic features and have developed manual classification systems such as MS-Path for Lynch syndrome screening. We present a novel hybrid approach that combines computational and pathologist expertise to create an explainable and verifiable method for MSI prediction in colorectal cancer, applicable to resection and biopsy WSI. Our proposed method uses nuclei and tissue segmentation models to automatically quantify MSI-associated histologic features outlined in the Bethesda guidelines, including intraepithelial lymphocytes, grade of differentiation, mucinous components, and tertiary lymphoid structures. After validation on annotated data sets, these features are integrated with clinical data and used in logistic regression and random forest models to predict MSI status. We validated our approach using 3256 WSI from 2267 patients across 7 cohorts from 5 centers. The method achieved an area under the curve of up to 0.88 across all resection cohorts, and 0.90 on biopsies, performing on par with published black-box deep learning models. Importantly, the learned variable importances strongly correlated with manual scoring systems and aligned with manual pathologist assessments. We observed significant intrapatient heterogeneity in predicted scores, emphasizing the importance of whole-case analysis. Our approach also shows potential as a screening tool that could exclude 41% of patients from gold-standard MSI testing while maintaining 95% sensitivity. This study demonstrates that classifiers based on clinical and validated histologic information can predict MSI status as effectively as black-box models while providing complete interpretability. Our method offers an alternative pathway for understandable, explainable, and trustworthy biomarker prediction in computational pathology.

BACKGROUND: Pathology reporting of colorectal cancer (CRC) follows the International Collaboration on Cancer Reporting (ICCR) guidelines which define a set of 25 elements (such as tumor grade, TNM stage and microsatellite instability) to be assessed for diagnosis. With the aim to further develop the CRC diagnostic routine, multiple computational tools have been developed in the last ten years. Despite the excellent sensitivity and potential advantages, including reduced inter-observer variability, many tools do not reach clinical deployment. This suggests that there are critical challenges to address when developing these algorithms given the aim to reliably and automatically assess ICCR elements. METHODS: To summarize existing efforts in deep- and machine learning for ICCR CRC elements and highlight existing gaps between development and clinical deployment, this systematic review collected studies on computational tools for colorectal cancer histopathology analysis published between 2015 and 2024. RESULTS: In total, 4863 studies were retrieved for the analysis, of which 66 remained after screening. Most of the reviewed studies focus on a subset of just three ICCR elements, namely mismatch repair status, BRAFV600E mutation testing, and lymph node status. Notably, many of the studies did not include clinically relevant and validated results, which puts into question their reliability for routine diagnosis. CONCLUSION: These results show the wide gap between research and clinical practice in pathology with the example of CRC diagnosis. There is an unmet need for publicly available datasets addressing a variety of topics, and a stronger focus on clinically important tasks. This review is critical to help the community align their contributions in computation pathology with the clinic and ultimately increase the translational potential of the developed tools. ### Competing Interest Statement Inti Zlobec acts as scientific advisor for Aiforia. ### Funding Statement This work was made possible via funding from several sources. E.B. and A.L.F. were funded by the Swiss National Science Foundation (CRSII5\_193832). A.K. was funded by the Center for Artificial Intelligence in Medicine (CAIM), University of Bern. M.G. and J.F.C. was funded by the Swiss cancer league (KFS-5786-02-2023-R). J.H. was funded by the Institute of Tissue Medicine and Pathology, University of Bern. R.M. was funded by Swiss National Science Foundation (31003A\_166578/1) and the Swiss Government Excellence Scholarship (ESKAS, nr. 2021.0019 / Kosovo / OP). J.G.B. was funded by the Swiss National Science Foundation (10.000.619). ### Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes All data produced in the present study are available upon reasonable request to the authors.

BackgroundThe International Collaboration on Cancer Reporting proposes histological tumour type, lymphovascular invasion, tumour grade, perineural invasion, extent, and dimensions of invasion as risk factors for lymph node metastases and tumour progression in completely endoscopically resected pT1 colorectal cancer (CRC).ObjectiveThe aim of the study was to propose a predictive and reliable score to optimise the clinical management of endoscopically resected pT1 CRC patients.MethodsThis multi-centric, retrospective International Budding Consortium (IBC) study included an international pT1 CRC cohort of 565 patients. All cases were reviewed by eight expert gastrointestinal pathologists. All risk factors were reported according to international guidelines. Tumour budding and immune response (CD8+ T-cells) were assessed with automated models using artificial intelligence. We used the information on risk factors and least absolute shrinkage and selection operator logistic regression to develop a prediction model and generate a score to predict the occurrence of lymph node metastasis or cancer recurrence.ResultsThe IBC prediction score included the following parameters: lymphovascular invasion, tumour buds, infiltration depth and tumour grade. The score has an acceptable discrimination power (area under the curve of 0.68 [95% confidence intervals (CI) 0.61-0.75]; 0.64 [95% CI 0.57-0.71] after internal validation). At a cut-off of 6.8 points to discriminate high-and low-risk patients, the score had a sensitivity and specificity of 0.9 [95% CI 0.8-0.95] and 0.26 [95% 0.22, 0.3], respectively.ConclusionThe IBC score is based on well-established risk factors and is a promising tool with clinical utility to support the management of pT1 CRC patients. image

Enabling the examination of cell-cell relationships in tissue, spatially resolved omics technologies have revolutionised our perspectives on cancer biology. Clinically, the development of immune checkpoint inhibitors (ICI) has advanced cancer therapeutics. However, a major challenge of effective implementation is the identification of predictive biomarkers of response. In this review we examine the potential added predictive value of spatial biomarkers of response to ICI beyond current clinical benchmarks.

Somatic variant testing through next-generation sequencing (NGS) is well integrated into Swiss molecular pathology laboratories and has become a standard diagnostic method for numerous indications in cancer patient care. Currently, there is a wide variation in reporting practices within our country, and as patients move between different hospitals, it is increasingly necessary to standardize NGS reports to ease their reinterpretation. Additionally, as many different stakeholders—oncologists, hematologists, geneticists, pathologists, and patients—have access to the NGS report, it needs to contain comprehensive and detailed information in order to answer the questions of experts and avoid misinterpretation by non-experts. In 2017, the Swiss Institute of Bioinformatics conducted a survey to assess the differences in NGS reporting practices across ten pathology institutes in Switzerland. The survey examined 68 reporting items and identified 48 discrepancies. Based on these findings, the Swiss Society of Molecular Pathology initiated a Delphi method to reach a consensus on a set of recommendations for NGS reporting. Reports should include clinical information about the patient and the diagnosis, technical details about the sample and the test performed, and a list of all clinically relevant variants and variants of uncertain significance. In the absence of a consensus on an actionability scheme, the five-class pathogenicity scheme proposed by the ACMG/AMP guideline must be included in the reports. The Swiss Society of Molecular Pathology recognizes the importance of including clinical actionability in the report and calls on the European community of molecular pathologists and oncologists to reach a consensus on this issue.

Tumor cell fraction (TCF) estimation is a common clinical task with well-established large interobserver variability. It thus provides an ideal test bed to evaluate potential impacts of employing a tumor cell fraction computer-aided diagnostic (TCFCAD) tool to support pathologists’ evaluation. During a National Slide Seminar event, pathologists (n = 69) were asked to visually estimate TCF in 10 regions of interest (ROIs) from hematoxylin and eosin colorectal cancer images intentionally curated for diverse tissue compositions, cellularity, and stain intensities. Next, they re-evaluated the same ROIs while being provided a TCFCAD-created overlay highlighting predicted tumor vs nontumor cells, together with the corresponding TCF percentage. Participants also reported confidence levels in their assessments using a 5-tier scale, indicating no confidence to high confidence, respectively. The TCF ground truth (GT) was defined by manual cell-counting by experts. When assisted, interobserver variability significantly decreased, showing estimates converging to the GT. This improvement remained even when TCFCAD predictions deviated slightly from the GT. The standard deviation (SD) of the estimated TCF to the GT across ROIs was 9.9% vs 5.8% with TCFCAD (P < .0001). The intraclass correlation coefficient increased from 0.8 to 0.93 (95% CI, 0.65-0.93 vs 0.86-0.98), and pathologists stated feeling more confident when aided (3.67 ± 0.81 vs 4.17 ± 0.82 with the computer-aided diagnostic [CAD] tool). TCFCAD estimation support demonstrated improved scoring accuracy, interpathologist agreement, and scoring confidence. Interestingly, pathologists also expressed more willingness to use such a CAD tool at the end of the survey, highlighting the importance of training/education to increase adoption of CAD systems.

Tumor budding (TB), the presence of single cells or small clusters of up to 4 tumor cells at the invasive front of colorectal cancer (CRC), is a proven risk factor for adverse outcomes. International definitions are necessary to reduce interobserver variability. According to the current international guidelines, hotspots at the invasive front should be counted in hematoxylin and eosin (H&E)-stained slides. This is time-consuming and prone to interobserver variability; therefore, there is a need for computer-aided diagnosis solutions. In this study, we report an artificial intelligence-based method for detecting TB in H&E-stained whole slide images. We propose a fully automated pipeline to identify the tumor border, detect tumor buds, characterize them based on the number of tumor cells, and produce a TB density map to identify the TB hotspot. The method outputs the TB count in the hotspot as a computational biomarker. We show that the proposed automated TB detection workflow performs on par with a panel of 5 pathologists at detecting tumor buds and that the hotspot-based TB count is an independent prognosticator in both the univariate and the multivariate analysis, validated on a cohort of n = 981 patients with CRC. Computer-aided detection of tumor buds based on deep learning can perform on par with expert pathologists for the detection and quantification of tumor buds in H&E-stained CRC histopathology slides, strongly facilitating the introduction of budding as an independent prognosticator in clinical routine and clinical trials.

Tumor budding is a histopathological biomarker associated with metastases and adverse survival outcomes in colorectal carcinoma (CRC) patients. It is characterized by the presence of single tumor cells or small clusters of cells within the tumor or at the tumor-invasion front. In order to obtain a tumor budding score for a patient, the region with the highest tumor bud density must first be visually identified by a pathologist, after which buds will be counted in the chosen hotspot field. The automation of this process will expectedly increase efficiency and reproducibility. Here, we present a deep learning convolutional neural network model that automates the above procedure. For model training, we used a semi-supervised learning method, to maximize the detection performance despite the limited amount of labeled training data. The model was tested on an independent dataset in which human- and machine-selected hotspots were mapped in relation to each other and manual and machine detected tumor bud numbers in the manually selected fields were compared. We report the results of the proposed method in comparison with visual assessment by pathologists. We show that the automated tumor bud count achieves a prognostic value comparable with visual estimation, while based on an objective and reproducible quantification. We also explore novel metrics to quantify buds such as density and dispersion and report their prognostic value. We have made the model available for research use on the grand-challenge platform.

Enterocolic phlebitis (EP) is a rare cause of bowel ischemia due to isolated venulitis of the bowel wall and mesentery without arterial involvement. EP is often misdiagnosed as inflammatory bowel disease, carcinoma, or diverticulitis due to non-specific symptoms as well as non-specific clinical and radiological findings. While unresponsive to pharmacotherapy, surgical resection of the affected bowel appears to be the only successful therapy with a very low recurrence rate. Etiology of EP remains unknown. We report a case of EP with rare presentation in the left hemicolon and unusual histological findings emphasizing the heterogeneity of this cause of enterocolic ischemia. The review and comparison of the three entities-EP, mesenteric inflammatory veno-occlusive disease (MIVOD), and idiopathic myointimal hyperplasia of mesenteric veins (IMHMV), all describing patterns of bowel ischemia due to isolated pathology of mesenteric veins-reveal that the current terminology is unclear. EP and MIVOD are very similar and may be considered the same disease. IMHMV, though, differs in localization, symptom duration, and histological findings but also shares features with EP and MIVOD. Further studies and harmonized terminology are inevitable for better understanding of the disease, prevention of unnecessary pharmacotherapy, and reduction in time to diagnosis.

Digital pathology has gone through considerable technical advances during the past few years and certain aspects of digital diagnostics have been widely and swiftly adopted in many centers, catalyzed by the COVID-19 pandemic. However, analysis of requirements, careful planning, and structured implementation should to be considered in order to reap the full benefits of a digital workflow. The aim of this review is to provide a practical, concise and hands-on summary of issues relevant to implementing and developing digital diagnostics in the pathology laboratory. These include important initial considerations, possible approaches to overcome common challenges, potential diagnostic pitfalls, validation and regulatory issues and an introduction to the emerging field of image analysis in routine.

Screening of lymph node metastases in colorectal cancer (CRC) can be a cumbersome task, but it is amenable to artificial intelligence (AI)-assisted diagnostic solution. Here, we propose a deep learning-based workflow for the evaluation of CRC lymph node metastases from digitized hematoxylin and eosin-stained sections. A segmentation model was trained on 100 whole-slide images (WSIs). It achieved a Matthews correlation coefficient of 0.86 (±0.154) and an acceptable Hausdorff distance of 135.59 μm (±72.14 μm), indicating a high congruence with the ground truth. For metastasis detection, 2 models (Xception and Vision Transformer) were independently trained first on a patch-based breast cancer lymph node data set and were then fine-tuned using the CRC data set. After fine-tuning, the ensemble model showed significant improvements in the F1 score (0.797-0.949; P <.00001) and the area under the receiver operating characteristic curve (0.959-0.978; P <.00001). Four independent cohorts (3 internal and 1 external) of CRC lymph nodes were used for validation in cascading segmentation and metastasis detection models. Our approach showed excellent performance, with high sensitivity (0.995, 1.0) and specificity (0.967, 1.0) in 2 validation cohorts of adenocarcinoma cases (n = 3836 slides) when comparing slide-level labels with the ground truth (pathologist reports). Similarly, an acceptable performance was achieved in a validation cohort (n = 172 slides) with mucinous and signet-ring cell histology (sensitivity, 0.872; specificity, 0.936). The patch-based classification confidence was aggregated to overlay the potential metastatic regions within each lymph node slide for visualization. We also applied our method to a consecutive case series of lymph nodes obtained over the past 6 months at our institution (n = 217 slides). The overlays of prediction within lymph node regions matched 100% when compared with a microscope evaluation by an expert pathologist. Our results provide the basis for a computer-assisted diagnostic tool for easy and efficient lymph node screening in patients with CRC.

Colon resection is often the treatment of choice for colorectal cancer (CRC) patients. However, especially for minimally invasive cancer, such as pT1, simply removing the polyps may be enough to stop cancer progression. Different histopathological risk factors such as tumor grade and invasion depth currently found the basis for the need for colon resection in pT1 CRC patients. Here, we investigate two additional risk factors, tumor budding and lymphocyte infiltration at the invasive front, which are known to be clinically relevant. We capture the spatial layout of tumor buds and T-cells and use graph-based deep learning to investigate them as potential risk predictors. Our pT1 Hotspot Tumor Budding T-cell Graph (pT1-HBTG) dataset consists of 626 tumor budding hotspots from 575 patients. We propose and compare three different graph structures, as well as combinations of the node labels. The best-performing Graph Neural Network architecture is able to increase specificity by 20% compared to the currently recommended risk stratification based on histopathological risk factors, without losing any sensitivity. We believe that using a graph-based analysis can help to assist pathologists in making risk assessments for pT1 CRC patients, and thus decrease the number of patients undergoing potentially unnecessary surgery. Both the code and dataset are made publicly available.

We present the case of a 31-year-old woman who was referred with a 12-month history of a tumor on the ulnar side of her dominant right hand. The eventual histopathologic diagnosis was an atypical pleomorphous lipomatous tumor, an entity that has only been recently classified in the World Health Organization Classification of Soft Tissue and Bone Tumors.

Supervised learning is constrained by the availability of labeled data, which are especially expensive to acquire in the field of digital pathology. Making use of open-source data for pre-training or using domain adaptation can be a way to overcome this issue. However, pre-trained networks often fail to generalize to new test domains that are not distributed identically due to tissue stainings, types, and textures variations. Additionally, current domain adaptation methods mainly rely on fully-labeled source datasets. In this work, we propose Self-Rule to Multi-Adapt (SRMA), which takes advantage of self-supervised learning to perform domain adaptation, and removes the necessity of fully-labeled source datasets. SRMA can effectively transfer the discriminative knowledge obtained from a few labeled source domain’s data to a new target domain without requiring additional tissue annotations. Our method harnesses both domains’ structures by capturing visual similarity with intra-domain and cross-domain self-supervision. Moreover, we present a generalized formulation of our approach that allows the framework to learn from multiple source domains. We show that our proposed method outperforms baselines for domain adaptation of colorectal tissue type classification in single and multi-source settings, and further validate our approach on an in-house clinical cohort. The code and trained models are available open-source: https://github.com/christianabbet/SRA.