Spatial heterogeneity of immune regulators drives dynamic changes of local immune responses, affecting disease outcomes in pancreatic cancer.

PURPOSE: Pancreatic ductal adenocarcinoma (PDAC) is considered a low immunogenic tumor with "cold" tumor microenvironment (TME) and is mostly unresponsive to immune checkpoint blockade therapies. Here we decipher the impact of intratumoral heterogeneity of immune determinants on antitumor response. EXPERIMENTAL DESIGN: We performed spatial proteomic and transcriptomic analyses and multiplexed immunofluorescence on multiple tumor regions, including tumor center (TC) and invasive front (IF), from 220 PDAC-patients, classified according to their transcriptomic immune signaling into high-immunogenic (HI-PDACs, n=54) and low-immunogenic tumors (LI-PDACs, n=166). Spatial compartments (tumor: Pancytokeratin+/CD45- and leukocytes: Pancytokeratin-/CD45+) were defined by fluorescent imaging. RESULTS: HI-PDACs exhibited higher densities of cytotoxic T lymphocytes with upregulation of T-cell priming-associated immune determinants, including CD40, ITGAM, GITR, CXCL10, GZMB, IFNG and HLA-DR, which was significantly more prominent at the IF than the TC. In contrast, LI-PDACs exhibited immune evasive TMEs with downregulation of immune determinants and a negative gradient from TC to IF. Patients with HI-PDACs had significantly better outcomes; however, they showed more frequently exhausted immune phenotypes. CONCLUSIONS: Our results indicate strategic differences in the regulation of immune determinants, which lead to different levels of effectiveness of antitumor responses between high- and low-immunogenic tumors and dynamic spatial changes, which affect the evolution of immune evasion and patient outcomes. This supports coevolution of tumor and immune cells and may help define therapeutic vulnerabilities to improve antitumor immunity and harness the responsiveness to immune checkpoint inhibitors in PDAC patients.

Immunohistochemical analysis of tumor budding in stage II colon cancer: exploring zero budding as a prognostic marker.

Tumor budding, a biomarker traditionally evaluated using hematoxylin and eosin (H&E) staining, has gained recognition as a prognostic biomarker for stage II colon cancer. Nevertheless, while H&E staining offers valuable insights, its limitations prompt the utilization of pan-cytokeratin immunohistochemistry (IHC). Consequently, this study seeks to evaluate the prognostic significance of tumor budding using IHC in a contemporary cohort of stage II colon cancer patients, aiming to deepen our understanding of this critical facet in cancer prognosis. We conducted a retrospective, population-based cohort study including 493 patients with stage II colon cancer and evaluated tumor budding using IHC, following the H&E-based guidelines proposed by the International Tumor Budding Consensus Conference Group. Correlation between H&E-based and IHC-based tumor budding was assessed using a four-tiered scoring system that included a zero budding (Bd0) category. Survival analyses explored the prognostic significance of tumor budding assessed by IHC and H&E. As expected, IHC-based tumor budding evaluation yielded significantly higher bud counts compared to H&E (p < 0.01). Interestingly, 21 patients were identified with no tumor budding using IHC. This was associated with significantly improved recurrence-free survival (HR = 5.19, p = 0.02) and overall survival (HR = 4.47, p = 0.04) in a multivariate analysis when compared to tumors with budding. The Bd0 category demonstrated a 100% predictive value for the absence of recurrence. In conclusion, IHC-based tumor budding evaluation in stage II colon cancer provides additional prognostic information. The absence of tumor budding is associated with a favorable prognosis and may serve as a potential marker for identifying patients with no risk of recurrence.

The prognostic value of tumor budding in a thoroughly characterized stage II colon cancer population in the context of a national screening program.

Tumor budding as a prognostic marker in colorectal cancer has not previously been investigated in a cohort of screened stage II colon cancer patients. We assessed the prognostic significance of tumor budding in a thoroughly characterized stage II colon cancer population comprising surgically resected patients in the Region of Southern Denmark from 2014 to 2016. Tumors were re-staged according to the 8th edition of UICC TNM Classification, undergoing detailed histopathological evaluation and tumor budding assessment following guidelines from the International Tumor Budding Consensus Conference. Prognostic evaluation utilized Kaplan-Meier curves, log-rank tests, and Cox proportional hazard models for time to recurrence (TTR), recurrence-free survival (RFS), and overall survival (OS). Out of 497 patients, 20% were diagnosed through the national colorectal cancer screening program. High-grade tumor budding (Bd3) was found in 19% of tumors and was associated with glandular subtype, perineural invasion, mismatch repair proficient tumors, and tumor recurrence (p < 0.001, p < 0.001, p = 0.045, and p = 0.007 respectively). In multivariable Cox regression, high-grade budding was a significant prognostic factor for TTR compared to low-grade (Bd3 HR 2.617; p = 0.007). An association between tumor budding groups and RFS was observed, and the difference was significant in univariable analysis for high-grade compared to low-grade tumor budding (Bd3 HR 1.461; p = 0.041). No significant differences were observed between tumor budding groups and OS. High-grade tumor budding is a predictor of recurrence in a screened population of patients with stage II colon cancer and should be considered a high-risk factor in a shared decision-making process when stratifying patients to adjuvant chemotherapy.

An interpretable machine learning system for colorectal cancer diagnosis from pathology slides.

Considering the profound transformation affecting pathology practice, we aimed to develop a scalable artificial intelligence (AI) system to diagnose colorectal cancer from whole-slide images (WSI). For this, we propose a deep learning (DL) system that learns from weak labels, a sampling strategy that reduces the number of training samples by a factor of six without compromising performance, an approach to leverage a small subset of fully annotated samples, and a prototype with explainable predictions, active learning features and parallelisation. Noting some problems in the literature, this study is conducted with one of the largest WSI colorectal samples dataset with approximately 10,500 WSIs. Of these samples, 900 are testing samples. Furthermore, the robustness of the proposed method is assessed with two additional external datasets (TCGA and PAIP) and a dataset of samples collected directly from the proposed prototype. Our proposed method predicts, for the patch-based tiles, a class based on the severity of the dysplasia and uses that information to classify the whole slide. It is trained with an interpretable mixed-supervision scheme to leverage the domain knowledge introduced by pathologists through spatial annotations. The mixed-supervision scheme allowed for an intelligent sampling strategy effectively evaluated in several different scenarios without compromising the performance. On the internal dataset, the method shows an accuracy of 93.44% and a sensitivity between positive (low-grade and high-grade dysplasia) and non-neoplastic samples of 0.996. On the external test samples varied with TCGA being the most challenging dataset with an overall accuracy of 84.91% and a sensitivity of 0.996.

Lymph node metastases and recurrence in pT1 colorectal cancer: Prediction with the International Budding Consortium Score-A retrospective, multi-centric study.

BACKGROUND: The 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). OBJECTIVE: The aim of the study was to propose a predictive and reliable score to optimise the clinical management of endoscopically resected pT1 CRC patients. METHODS: This 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. RESULTS: The 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. CONCLUSION: The 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.

Engineering the future of 3D pathology.

In recent years, technological advances in tissue preparation, high-throughput volumetric microscopy, and computational infrastructure have enabled rapid developments in nondestructive 3D pathology, in which high-resolution histologic datasets are obtained from thick tissue specimens, such as whole biopsies, without the need for physical sectioning onto glass slides. While 3D pathology generates massive datasets that are attractive for automated computational analysis, there is also a desire to use 3D pathology to improve the visual assessment of tissue histology. In this perspective, we discuss and provide examples of potential advantages of 3D pathology for the visual assessment of clinical specimens and the challenges of dealing with large 3D datasets (of individual or multiple specimens) that pathologists have not been trained to interpret. We discuss the need for artificial intelligence triaging algorithms and explainable analysis methods to assist pathologists or other domain experts in the interpretation of these novel, often complex, large datasets.

The Complexity of Shapes: How the Circularity of Tumor Nodules Affects Prognosis in Colorectal Cancer.

The current stratification of tumor nodules in colorectal cancer (CRC) staging is subjective and leads to high interobserver variability. In this study, the objective assessment of the shape of lymph node metastases (LNMs), extranodal extension (ENE), and tumor deposits (TDs) was correlated with outcomes. A test cohort and a validation cohort were included from 2 different institutions. The test cohort consisted of 190 cases of stage III CRC. Slides with LNMs and TDs were annotated and processed using a segmentation algorithm to determine their shape. The complexity ratio was calculated for every shape and correlated with outcomes. A cohort of 160 stage III CRC cases was used to validate findings. TDs showed significantly more complex shapes than LNMs with ENE, which were more complex than LNMs without ENE (P < .001). In the test cohort, patients with the highest sum of complexity ratios had significantly lower disease-free survival (P < .01). When only the nodule with the highest complexity was considered, this effect was even stronger (P < .001). This maximum complexity ratio per patient was identified as an independent prognostic factor in the multivariate analysis (hazard ratio, 2.47; P < .05). The trends in the validation cohort confirmed the results. More complex nodules in stage III CRC were correlated with significantly worse disease-free survival, even if only based on the most complex nodule. These results suggest that more complex nodules reflect more invasive tumor biology. As most of the more complex nodules were diagnosed as TDs, we suggest providing a more prominent role for TDs in the nodal stage and include an objective complexity measure in their definition.

Swiss digital pathology recommendations: results from a Delphi process conducted by the Swiss Digital Pathology Consortium of the Swiss Society of Pathology.

Integration of digital pathology (DP) into clinical diagnostic workflows is increasingly receiving attention as new hardware and software become available. To facilitate the adoption of DP, the Swiss Digital Pathology Consortium (SDiPath) organized a Delphi process to produce a series of recommendations for DP integration within Swiss clinical environments. This process saw the creation of 4 working groups, focusing on the various components of a DP system (1) scanners, quality assurance and validation of scans, (2) integration of Whole Slide Image (WSI)-scanners and DP systems into the Pathology Laboratory Information System, (3) digital workflow-compliance with general quality guidelines, and (4) image analysis (IA)/artificial intelligence (AI), with topic experts for each recruited for discussion and statement generation. The work product of the Delphi process is 83 consensus statements presented here, forming the basis for "SDiPath Recommendations for Digital Pathology". They represent an up-to-date resource for national and international hospitals, researchers, device manufacturers, algorithm developers, and all supporting fields, with the intent of providing expectations and best practices to help ensure safe and efficient DP usage.

National digital pathology projects in Switzerland: A 2023 update.

The Swiss Digital Pathology Consortium (SDiPath) was founded in 2018 as a working group of the Swiss Society for Pathology with the aim of networking, training, and promoting digital pathology (DP) at a national level. Since then, two national surveys have been carried out on the level of knowledge, dissemination, use, and needs in DP, which have resulted in clear fields of action. In addition to organizing symposia and workshops, national guidelines were drawn up and an initiative for a national DP platform actively codesigned. With the growing use of digital image processing and artificial intelligence tools, continuous monitoring, evaluation, and exchange of experiences will be pursued, along with best practices.

Digital image analysis and artificial intelligence in pathology diagnostics-the Swiss view.

Digital pathology (DP) is increasingly entering routine clinical pathology diagnostics. As digitization of the routine caseload advances, implementation of digital image analysis algorithms and artificial intelligence tools becomes not only attainable, but also desirable in daily sign out. The Swiss Digital Pathology Consortium (SDiPath) has initiated a Delphi process to generate best-practice recommendations for various phases of the process of digitization in pathology for the local Swiss environment, encompassing the following four topics: i) scanners, quality assurance, and validation of scans; ii) integration of scanners and systems into the pathology laboratory information system; iii) the digital workflow; and iv) digital image analysis (DIA)/artificial intelligence (AI). The current article focuses on the DIA-/AI-related recommendations generated and agreed upon by the working group and further verified by the Delphi process among the members of SDiPath. Importantly, they include the view and the currently perceived needs of practicing pathologists from multiple academic and cantonal hospitals as well as private practices.

WebSTR: A Population-wide Database of Short Tandem Repeat Variation in Humans.

Short tandem repeats (STRs) are consecutive repetitions of one to six nucleotide motifs. They are hypervariable due to the high prevalence of repeat unit insertions or deletions primarily caused by polymerase slippage during replication. Genetic variation at STRs has been shown to influence a range of traits in humans, including gene expression, cancer risk, and autism. Until recently STRs have been poorly studied since they pose significant challenges to bioinformatics analyses. Moreover, genome-wide analysis of STR variation in population-scale cohorts requires large amounts of data and computational resources. However, the recent advent of genome-wide analysis tools has resulted in multiple large genome-wide datasets of STR variation spanning nearly two million genomic loci in thousands of individuals from diverse populations. Here we present WebSTR, a database of genetic variation and other characteristics of genome-wide STRs across human populations. WebSTR is based on reference panels of more than 1.7 million human STRs created with state of the art repeat annotation methods and can easily be extended to include additional cohorts or species. It currently contains data based on STR genotypes for individuals from the 1000 Genomes Project, H3Africa, the Genotype-Tissue Expression (GTEx) Project and colorectal cancer patients from the TCGA dataset. WebSTR is implemented as a relational database with programmatic access available through an API and a web portal for browsing data. The web portal is publicly available at https://webstr.ucsd.edu.

Pathologist Computer-Aided Diagnostic Scoring of Tumor Cell Fraction: A Swiss National Study.

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.

Multi-Institutional Evaluation of Pathologists' Assessment Compared to Immunoscore.

BACKGROUND: The Immunoscore (IS) is a quantitative digital pathology assay that evaluates the immune response in cancer patients. This study reports on the reproducibility of pathologists' visual assessment of CD3+- and CD8+-stained colon tumors, compared to IS quantification. METHODS: An international group of expert pathologists evaluated 540 images from 270 randomly selected colon cancer (CC) cases. Concordance between pathologists' T-score, corresponding hematoxylin-eosin (H&E) slides, and the digital IS was evaluated for two- and three-category IS. RESULTS: Non-concordant T-scores were reported in more than 92% of cases. Disagreement between semi-quantitative visual assessment of T-score and the reference IS was observed in 91% and 96% of cases before and after training, respectively. Statistical analyses showed that the concordance index between pathologists and the digital IS was weak in two- and three-category IS, respectively. After training, 42% of cases had a change in T-score, but no improvement was observed with a Kappa of 0.465 and 0.374. For the 20% of patients around the cut points, no concordance was observed between pathologists and digital pathology analysis in both two- and three-category IS, before or after training (all Kappa < 0.12). CONCLUSIONS: The standardized IS assay outperformed expert pathologists' T-score evaluation in the clinical setting. This study demonstrates that digital pathology, in particular digital IS, represents a novel generation of immune pathology tools for reproducible and quantitative assessment of tumor-infiltrated immune cell subtypes.

Pseudobudding: ruptured glands do not represent true tumor buds.

Tumor budding (TB) is a strong biomarker of poor prognosis in colorectal cancer and other solid cancers. TB is defined as isolated single cancer cells or clusters of up to four cancer cells at the invasive tumor front. In areas with a large inflammatory response at the invasive front, single cells and cell clusters surrounding fragmented glands are observed appearing like TB. Occurrence of these small groups is referred to as pseudobudding (PsB), which arises due to external influences such as inflammation and glandular disruption. Using a combination of orthogonal approaches, we show that there are clear biological differences between TB and PsB. TB is representative of active invasion by presenting features of epithelial-mesenchymal transition and exhibiting increased deposition of extracellular matrix within the surrounding tumor microenvironment (TME), whereas PsB represents a reactive response to heavy inflammation where increased levels of granulocytes within the surrounding TME are observed. Our study provides evidence that areas with a strong inflammatory reaction should be avoided in the routine diagnostic assessment of TB. © 2023 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Fully Automated Tumor Bud Assessment in Hematoxylin and Eosin-Stained Whole Slide Images of Colorectal Cancer.

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.

Spatial distribution of CD3- and CD8-positive lymphocytes as pretest for POLE wild-type in molecular subgroups of endometrial carcinoma.

Introduction: Over the years, the molecular classification of endometrial carcinoma has evolved significantly. Both POLEmut and MMRdef cases share tumor biological similarities like high tumor mutational burden and induce strong lymphatic reactions. While therefore use case scenarios for pretesting with tumor-infiltrating lymphocytes to replace molecular analysis did not show promising results, such testing may be warranted in cases where an inverse prediction, such as that of POLEwt, is being considered. For that reason we used a spatial digital pathology method to quantitatively examine CD3+ and CD8+ immune infiltrates in comparison to conventional histopathological parameters, prognostics and as potential pretest before molecular analysis. Methods: We applied a four-color multiplex immunofluorescence assay for pan-cytokeratin, CD3, CD8, and DAPI on 252 endometrial carcinomas as testing and compared it to further 213 cases as validation cohort from a similar multiplexing assay. We quantitatively assessed immune infiltrates in microscopic distances within the carcinoma, in a close distance of 50 microns, and in more distant areas. Results: Regarding prognostics, high CD3+ and CD8+ densities in intra-tumoral and close subregions pointed toward a favorable outcome. However, TCGA subtyping outperforms prognostication of CD3 and CD8 based parameters. Different CD3+ and CD8+ densities were significantly associated with the TCGA subgroups, but not consistently for histopathological parameter. In the testing cohort, intra-tumoral densities of less than 50 intra-tumoral CD8+ cells/mm2 were the most suitable parameter to assume a POLEwt, irrespective of an MMRdef, NSMP or p53abn background. An application to the validation cohort corroborates these findings with an overall sensitivity of 95.5%. Discussion: Molecular confirmation of POLEmut cases remains the gold standard. Even if CD3+ and CD8+ cell densities appeared less prognostic than TCGA, low intra-tumoral CD8+ values predict a POLE wild-type at substantial percentage rates, but not vice versa. This inverse correlation might be useful to increase pretest probabilities in consecutive POLE testing. Molecular subtyping is currently not conducted in one-third of cases deemed low-risk based on conventional clinical and histopathological parameters. However, this percentage could potentially be increased to two-thirds by excluding sequencing of predicted POLE wild-type cases, which could be determined through precise quantification of intra-tumoral CD8+ cells.

Semi-Supervised Learning to Automate Tumor Bud Detection in Cytokeratin-Stained Whole-Slide Images of Colorectal Cancer.

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.

Spatially restricted tumour-associated and host-associated immune drivers correlate with the recurrence sites of pancreatic cancer.

OBJECTIVE: Most patients with pancreatic ductal adenocarcinoma (PDAC) will experience recurrence after resection. Here, we investigate spatially organised immune determinants of PDAC recurrence. DESIGN: PDACs (n=284; discovery cohort) were classified according to recurrence site as liver (n=93/33%), lung (n=49/17%), local (n=31/11%), peritoneal (n=38/13%) and no-recurrence (n=73/26%). Spatial compartments were identified by fluorescent imaging as: pancytokeratin (PanCK)+CD45- (tumour cells); CD45+PanCK- (leucocytes) and PanCK-CD45- (stromal cells), followed by transcriptomic (72 genes) and proteomic analysis (51 proteins) for immune pathway targets. Results from next-generation sequencing (n=194) were integrated. Finally, 10 tumours from each group underwent immunophenotypic analysis by multiplex immunofluorescence. A validation cohort (n=109) was examined in parallel. RESULTS: No-recurrent PDACs show high immunogenicity, adaptive immune responses and are rich in pro-inflammatory chemokines, granzyme B and alpha-smooth muscle actin+ fibroblasts. PDACs with liver and/or peritoneal recurrences display low immunogenicity, stemness phenotype and innate immune responses, whereas those with peritoneal metastases are additionally rich in FAP+ fibroblasts. PDACs with local and/or lung recurrences display interferon-gamma signalling and mixed adaptive and innate immune responses, but with different leading immune cell population. Tumours with local recurrences overexpress dendritic cell markers whereas those with lung recurrences neutrophilic markers. Except the exclusive presence of RNF43 mutations in the no-recurrence group, no genetic differences were seen. The no-recurrence group exhibited the best, whereas liver and peritoneal recurrences the poorest prognosis. CONCLUSIONS: Our findings demonstrate distinct inflammatory/stromal responses in each recurrence group, which might affect dissemination patterns and patient outcomes. These findings may help to inform personalised adjuvant/neoadjuvant and surveillance strategies in PDAC, including immunotherapeutic modalities.

Computer-Assisted Diagnosis of Lymph Node Metastases in Colorectal Cancers Using Transfer Learning With an Ensemble Model.

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.