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.

Hepatic blood flow regulation but not oxygen extraction capability is impaired in prolonged experimental abdominal sepsis.

Impaired oxygen utilization has been proposed to play a significant role in sepsis-induced liver dysfunction, but its magnitude and temporal course during prolonged resuscitation is controversial. The aim of this study is to evaluate the capability of the liver to increase oxygen extraction in sepsis during repeated acute portal vein blood flow reduction. Twenty anesthetized and mechanically ventilated pigs with hepatic hemodynamic monitoring were randomized to fecal peritonitis or controls (n = 10, each). After 8-h untreated sepsis, the animals were resuscitated for three days. The ability to increase hepatic O2 extraction was evaluated by repeated, acute decreases in hepatic oxygen delivery (Do2) via reduction of portal flow. Blood samples for liver function and liver biopsies were obtained repeatedly. Although liver function tests, ATP content, and Do2 remained unaltered, there were signs of liver injury in blood samples and overt liver cell necrosis in biopsies. With acute portal vein occlusion, hepatic Do2 decreased more in septic animals compared with controls [max. decrease: 1.66 ± 0.68 mL/min/kg in sepsis vs. 1.19 ± 0.42 mL/min/kg in controls; portal venous flow (Qpv) reduction-sepsis interaction: P = 0.028]. Hepatic arterial buffer response (HABR) was impaired but recovered after 3-day resuscitation, whereas hepatic oxygen extraction increased similarly during the procedures in both groups (max. increase: 0.27 ± 0.13 in sepsis vs. 0.18 ± 0.09 in controls; all P > 0.05). Our data indicate maintained capacity of the liver to acutely increase O2 extraction, whereas blood flow regulation is transiently impaired with the potential to contribute to liver injury in sepsis.NEW & NOTEWORTHY The capacity to acutely increase hepatic O2 extraction with portal flow reduction is maintained in sepsis with accompanying liver injury, but hepatic blood flow regulation is impaired.

Image-based assessment of extracellular mucin-to-tumor area predicts consensus molecular subtypes (CMS) in colorectal cancer.

The backbone of all colorectal cancer classifications including the consensus molecular subtypes (CMS) highlights microsatellite instability (MSI) as a key molecular pathway. Although mucinous histology (generally defined as >50% extracellular mucin-to-tumor area) is a "typical" feature of MSI, it is not limited to this subgroup. Here, we investigate the association of CMS classification and mucin-to-tumor area quantified using a deep learning algorithm, and  the expression of specific mucins in predicting CMS groups and clinical outcome. A weakly supervised segmentation method was developed to quantify extracellular mucin-to-tumor area in H&E images. Performance was compared to two pathologists' scores, then applied to two cohorts: (1) TCGA (n = 871 slides/412 patients) used for mucin-CMS group correlation and (2) Bern (n = 775 slides/517 patients) for histopathological correlations and next-generation Tissue Microarray construction. TCGA and CPTAC (n = 85 patients) were used to further validate mucin detection and CMS classification by gene and protein expression analysis for MUC2, MUC4, MUC5AC and MUC5B. An excellent inter-observer agreement between pathologists' scores and the algorithm was obtained (ICC = 0.92). In TCGA, mucinous tumors were predominantly CMS1 (25.7%), CMS3 (24.6%) and CMS4 (16.2%). Average mucin in CMS2 was 1.8%, indicating negligible amounts. RNA and protein expression of MUC2, MUC4, MUC5AC and MUC5B were low-to-absent in CMS2. MUC5AC protein expression correlated with aggressive tumor features (e.g., distant metastases (p = 0.0334), BRAF mutation (p < 0.0001), mismatch repair-deficiency (p < 0.0001), and unfavorable 5-year overall survival (44% versus 65% for positive/negative staining). MUC2 expression showed the opposite trend, correlating with less lymphatic (p = 0.0096) and venous vessel invasion (p = 0.0023), no impact on survival.The absence of mucin-expressing tumors in CMS2 provides an important phenotype-genotype correlation. Together with MSI, mucinous histology may help predict CMS classification using only histopathology and should be considered in future image classifiers of molecular subtypes.

Systematically higher Ki67 scores on core biopsy samples compared to corresponding resection specimen in breast cancer: a multi-operator and multi-institutional study.

Ki67 has potential clinical importance in breast cancer but has yet to see broad acceptance due to inter-laboratory variability. Here we tested an open source and calibrated automated digital image analysis (DIA) platform to: (i) investigate the comparability of Ki67 measurement across corresponding core biopsy and resection specimen cases, and (ii) assess section to section differences in Ki67 scoring. Two sets of 60 previously stained slides containing 30 core-cut biopsy and 30 corresponding resection specimens from 30 estrogen receptor-positive breast cancer patients were sent to 17 participating labs for automated assessment of average Ki67 expression. The blocks were centrally cut and immunohistochemically (IHC) stained for Ki67 (MIB-1 antibody). The QuPath platform was used to evaluate tumoral Ki67 expression. Calibration of the DIA method was performed as in published studies. A guideline for building an automated Ki67 scoring algorithm was sent to participating labs. Very high correlation and no systematic error (p = 0.08) was found between consecutive Ki67 IHC sections. Ki67 scores were higher for core biopsy slides compared to paired whole sections from resections (p ≤ 0.001; median difference: 5.31%). The systematic discrepancy between core biopsy and corresponding whole sections was likely due to pre-analytical factors (tissue handling, fixation). Therefore, Ki67 IHC should be tested on core biopsy samples to best reflect the biological status of the tumor.

Taking tumour budding to the next frontier - a post International Tumour Budding Consensus Conference (ITBCC) 2016 review.

Tumour budding in colorectal cancer, defined as single tumour cells or small clusters containing four or fewer tumour cells, is a robust and independent biomarker of aggressive tumour biology. On the basis of published data in the literature, the evidence is certainly in favour of reporting tumour budding in routine practice. One important aspect of implementing tumour budding has been to establish a standardised and evidence-based scoring method, as was recommended by the International Tumour Budding Consensus Conference (ITBCC) in 2016. Further developments have aimed at establishing methods for automated tumour budding assessment. A digital approach to scoring tumour buds has great potential to assist in performing an objective budding count but, like the manual consensus method, must be validated and standardised. The aim of the present review is to present general considerations behind the ITBCC scoring method, and a broad overview of the current situation and challenges regarding automated tumour budding detection methods.

Tumour budding and CD8+ T cells: 'attackers' and 'defenders' in rectal cancer with and without neoadjuvant chemoradiotherapy.

AIM: Tumour budding ('attacker') and CD8+ T cells ('defender') are recognised as important parameters for risk stratification in colon cancers and, combined, may have an even stronger clinical impact. Here, we determine the value of tumour budding and CD8+ in rectal cancer patients treated with/without neoadjuvant therapy. METHODS AND RESULTS: Using digital scans of all tumour slides/case, we analysed CD8+ T cell counts in two patient cohorts: 45 neoadjuvantly treated and 47 primarily surgically treated (totalling n = 543 slides) after double-staining of the surgical resection specimen for pan-cytokeratin and CD8+ . Tumour buds in hot-spots were manually counted (area = 0.785 mm2 ) and CD8+ T cell counts were analysed separately both in tumour budding hot-spots and the densest CD8+ regions throughout the tumour. In neoadjuvantly treated patients, only tumour budding and not CD8+ T cells was associated with tumour features, including more advanced ypT (P = 0.0062), venous invasion (P = 0.002), lymphatic invasion (P = 0.0003) and perineural invasion (P = 0.0017), as well as higher American Joint Committee on Cancer (AJCC) tumour regression score (P = 0.0035), indicating less tumour response. Overall survival was also worse in patients with high-grade budding in univariate analysis only. In contrast, all three variables, namely tumour budding (P = 0.0347), CD8+ T cells in budding hot-spots (P = 0.0382) and CD8+ T cells in the densest areas (P = 0.0117) were also associated with worse (budding) and better (CD8) survival time in the multivariate setting. CONCLUSION: In rectal cancer, tumour budding has clinical relevance in both primarily surgically treated patients and in those with neoadjuvantly treated patients, where it characterises highly aggressive residual disease. CD8+ T cell counts appear not to have prognostic relevance in the neoadjuvant context.

Tumour budding/T cell infiltrates in colorectal cancer: proposal of a novel combined score.

AIMS: The tumour-node-metastasis classification system is used for prognostication purposes and to guide patient management. However, in colorectal cancer (CRC), additional markers are needed to stratify prognostic subgroups. Two promising markers have emerged from large bodies of research: tumour budding and T cell host response (CD3, CD8 and CD45RO infiltrates). However, attempts to combine these two parameters have been sparse. The aim of this study was to perform an assessment of potential protagonists that could be used in a combined score (budding/T cell score, BTS). METHODS AND RESULTS: This descriptive, retrospective study was performed on a multipunch tissue microarray containing material from 345 patients with stages I-IV CRC. Areas from tumour centre, front and microenvironment were stained for pancytokeratin/CD3, pancytokeratin/CD8 and pancytokeratin/CD45RO. Tumour buds were scored manually and T cell infiltrates digitally using open-source software. Tumour buds, T cell counts and combined BTS were associated with clinicopathological features and overall survival (OS). A higher combined BTS score (buds/CD8, tumour centre) performed better than budding or CD8/CD3 alone in predicting nodal metastases (P < 0.0001, OR = 1.466, 95% CI = 1.115-1.928). Only higher BTS (buds/CD3) were significantly associated with poorer OS on multivariate analysis (P = 0.012, hazard ratio = 1.218, 95% confidence interval = 1.044-1.419). CONCLUSIONS: Although CD8+ /CD3+ T cells are predictive of tumour biology in CRC, we found a combined BTS to be stronger in predicting survival and certain features with high clinical relevance, such as nodal metastases, in comparison to budding or T cells alone. Further studies combining T cell infiltrates and tumour budding are necessary to optimise risk assessment of CRC.

[Non-cirrhotic portal hypertension - The point of view of the pathologis]. / Nicht-zirrhotische portale Hypertension aus der Sicht des Pathologen.

Non-cirrhotic portal hypertension - The point of view of the pathologist Abstract. Non-cirrhotic portal hypertension (NCPH) is a heterogeneous group of liver disorders leading to portal hypertension. Many disorders are associated with NCPH. Based on the site of resistance to blood flow, they are classified as pre-hepatic, hepatic, and post-hepatic. Hepatic causes are further subdivided into pre-sinusoidal, sinusoidal and post-sinusoidal. The diagnosis of idiopathic non-cirrhotic portal hypertension (INCPH) can be made if all these disorders have been excluded and consequently no clear liver disease has been identified. INCPH is clinically characterized by features of PHT, moderate to massive splenomegaly, with or without hypersplenism, and preserved liver functions. This review is focused on pathological features of INCPH, which is thought to be caused largely by parenchymal vascular obstruction. INCPH must be considered in every patient presenting with clinical signs of portal hypertension where cirrhosis appears to be absent and the cause of portal hypertension is not clear. In most instances, patients present with bleeding esophageal varices. In order to exclude severe fibrosis or cirrhosis, liver histology remains essential in the diagnosis of INCPH.

Therapeutic targeting of tumor-associated macrophages in pancreatic neuroendocrine tumors.

Pancreatic neuroendocrine tumors (PNETs) represent a heterogeneous group of neuroendocrine neoplasms with varying biological behavior and response to treatment. Although targeted therapies have been shown to improve the survival for patients at advanced stage, resistance to current therapies frequently occurs during the course of therapy. Previous reports indicate that the infiltration of tumor-associated macrophages (TAMs) in PNETs might correlate with tumor progression and metastasis formation. We aimed to evaluate the prognostic and functional impact of TAMs in human PNETs in vitro and in vivo and to investigate the effect of therapeutic targeting TAMs in a genetic PNET mouse model. TAM expression pattern was assessed immunohistochemically in human PNET tissue sections and a tissue-micro-array of PNET tumors with different functionality, stage, and grading. The effect of liposomal clodronate on TAM cell viability was analyzed in myeloid cell lines and isolated murine bone macrophages (mBMM). In vivo, RIP1Tag2 mice developing insulinomas were treated with liposomal clodronate or PBS-Liposomes. Tumor progression, angiogenesis and immune cell infiltration were assessed by immunohistochemistry. In human, insulinomas TAM density was correlated with invasiveness and malignant behavior. Moreover, TAM infiltration in liver metastases was significantly increased compared to primary tumors. In vitro, Liposomal clodronate selectively inhibited the viability of myeloid cells and murine bone macrophages, leaving PNET tumor cell lines largely unaffected. In vivo, repeated application of liposomal clodronate to RIP1Tag2 mice significantly diminished the malignant transformation of insulinomas, which was accompanied by a reduced infiltration of F4/80-positive TAM cells and simultaneously by a decreased microvessel density, suggesting a pronounced effect of clodronate-induced myeloid depletion on tumor angiogenesis. Concomitant treatment with the antiangiogenic TKI sunitinib, however, did not show any synergistic effects with liposomal clodronate. TAMs are crucial for malignant transformation in human PNET and correlate with metastatic behavior. Pharmacological targeting of TAMs via liposomal clodronate disrupts tumor progression in the RIP1Tag2 neuroendocrine tumor model and was associated with reduced tumor angiogenesis. Based on these results, using liposomal clodronate to target proangiogenic myeloid cells could be employed as novel therapeutic avenue in highly angiogenic tumors such as PNET.

Glucagon-like-peptide-1 receptor expression in normal and diseased human thyroid and pancreas.

Glucagon-like-peptide-1 (GLP1) analogs may induce thyroid or pancreatic diseases in animals, raising questions about their use in diabetic patients. There is, however, controversy regarding expression of GLP1 receptors (GLP1R) in human normal and diseased thyroid and pancreas. Here, 221 human thyroid and pancreas samples were analyzed for GLP1R immunohistochemistry and compared with quantitative in vitro GLP1R autoradiography. Neither normal nor hyperplastic human thyroids containing parafollicular C cells express GLP1R with either method. Papillary thyroid cancer do not, and medullary thyroid carcinomas rarely express GLP1R. Insulin- and somatostatin-producing cells in the normal pancreas express a high density of GLP1R, whereas acinar cells express them in low amounts. Ductal epithelial cells do not express GLP1R. All benign insulinomas express high densities of GLP1R, whereas malignant insulinomas rarely express them. All ductal pancreatic carcinomas are GLP1R negative, whereas 6/20 PanIN 1/2 and 0/12 PanIN 3 express GLP1R. Therefore, normal thyroid, including normal and hyperplastic C cells, or papillary thyroid cancer are not targets for GLP1 analogs in humans. Conversely, all pancreatic insulin- and somatostatin-producing cells are physiological GLP1 targets, as well as most acini. As normal ductal epithelial cells or PanIN 3 or ductal pancreatic carcinomas do not express GLP1R, it seems unlikely that GLP1R is related to neoplastic transformation in pancreas. GLP1R-positive medullary thyroid carcinomas and all benign insulinomas are candidates for in vivo GLP1R targeting.

Prognostic and predictive roles of MGMT protein expression and promoter methylation in sporadic pancreatic neuroendocrine neoplasms.

BACKGROUND/AIMS: O(6)-methylguanine-methyltransferase (MGMT) is an important enzyme of DNA repair. MGMT promoter methylation is detectable in a subset of pancreatic neuroendocrine neoplasms (pNEN). A subset of pNEN responds to the alkylating agent temozolomide (TMZ). We wanted to correlate MGMT promoter methylation with MGMT protein loss in pNEN, correlate the findings with clinico-pathological data and determine the role of MGMT to predict response to TMZ chemotherapy. METHODS: We analysed a well-characterized collective of 141 resected pNEN with median follow-up of 83 months for MGMT protein expression and promoter methylation using methylation-specific PCR (MSP). A second collective of 10 metastasized, pretreated and progressive patients receiving TMZ was used to examine the predictive role of MGMT by determining protein expression and promoter methylation using primer extension-based quantitative PCR. RESULTS: In both collectives there was no correlation between MGMT protein expression and promoter methylation. Loss of MGMT protein was associated with an adverse outcome, this prognostic value, however, was not independent from grade and stage in multivariate analysis. Promoter hypermethylation was significantly associated with response to TMZ. CONCLUSION: Loss of MGMT protein expression is associated with adverse outcome in a surgical series of pNET. MGMT promoter methylation could be a predictive marker for TMZ chemotherapy in pNEN, but further, favourably prospective studies will be needed to confirm this result and before this observation can influence clinical routine.

A Composite Decision Rule of CD8+ T-cell Density in Tumor Biopsies Predicts Efficacy in Early-stage, Immunotherapy Trials.

PURPOSE: To examine whether CD8+ T-cell numbers in paired tumor biopsies in early-stage clinical trials can be used as an early indicator of clinical benefit for cancer immunotherapies. EXPERIMENTAL DESIGN: Paraffin sections of tumor biopsies were stained immunohistochemically for CD8+ T cells, which were digitally enumerated. The tumor biopsies were from cancer patients in early-phase trials testing novel immunotherapeutic agents. Paired biopsies taken before the start of treatment and on-treatment were compared. A total of 155 patients were used as the training set and an additional 221 patients were used as the validation set. RESULTS: Using the Cox proportional hazard model, a ≥0.9- increase in fold change (FC) on a ln scale in CD8+ T cells (corresponding to a 2.5-fold increase on the linear scale), from baseline, demonstrated a greater association with prolonged progression-free survival and allowed improved differentiation between groups above and below the threshold. Similarly, a ≥6.2 threshold in geometric mean of the on-treatment density (OTD) of T cells, which approximately corresponds to 500 cells/mm2, correlated with longer PFS. The combination of both criteria (FC and OTD) provided the best discrimination between clinically nonactive and active compounds. CONCLUSIONS: We propose that a composite score of CD8+ T-cell density in paired biopsies taken before and on-treatment may be a new biomarker to inform on clinical outcomes in early immunotherapy clinical trials.

Real-life data from standardized preanalytical coding (SPREC) in tissue biobanking and its dual use for sample characterization and process optimization.

The standardized preanalytical code (SPREC) aggregates warm ischemia (WIT), cold ischemia (CIT), and fixation times (FIT) in a precise format. Despite its growing importance underpinned by the European in vitro diagnostics regulation or broad preanalytical programs by the National Institutes of Health, little is known about its empirical occurrence in biobanked surgical specimen. In several steps, the Tissue Bank Bern achieved a fully informative SPREC code with insights from 10,555 CIT, 4,740 WIT, and 3,121 FIT values. During process optimization according to LEAN six sigma principles, we identified a dual role of the SPREC code as a sample characteristic and a traceable process parameter. With this preanalytical study, we outlined real-life data in a variety of organs with specific differences in WIT, CIT, and FIT values. Furthermore, our FIT data indicate the potential to adapt the SPREC fixation toward concrete paraffin-embedding time points and to extend its categories beyond 72 h due to weekend delays. Additionally, we identified dependencies of preanalytical variables from workload, daytime, and clinics that were actionable with LEAN process management. Thus, streamlined biobanking workflows during the day were significantly resilient to workload peaks, diminishing the turnaround times of native tissue processing (i.e. CIT) from 74.6 to 46.1 min under heavily stressed conditions. In conclusion, there are surgery-specific preanalytics that are surgico-pathologically limited even under process optimization, which might affect biomarker transfer from one entity to another. Beyond sample characteristics, SPREC coding is highly beneficial for tissue banks and Institutes of Pathology to track WIT, CIT, and FIT for process optimization and monitoring measurements.

Hepatocellular neoplasms with loss of liver fatty acid binding protein: Clinicopathologic features and molecular profiling.

HNF1A-inactivated hepatocellular adenomas (H-HCA) show steatosis, no atypia and loss of liver fatty acid binding protein (LFABP). LFABP loss also occurs in hepatocellular carcinoma (HCC). This study examines 68 LFABP-negative tumors: 33 typical H-HCA, 10 atypical hepatocellular neoplasms (AHN), 7 well-differentiated (WD) HCC, 18 moderately or poorly differentiated (MD/PD) HCC. Capture based sequencing was performed in 13 cases (8 AHN, 5 WD-HCC). Patients with HCA, AHN and WD-HCC were nearly all women. AHN and WD-HCC resembled H-HCA but had higher degree of atypia and/or reticulin loss. Variant features like inconspicuous fat (59% vs. 12%, p = 0.03), predominance of eosinophilic cells (59% vs. 21%, p = 0.01) and pseudoacini were more common in AHN and WD-HCC. Myxoid change and prominent lipofuscin were more common in WD-HCC (29% each) than H-HCA and AHN combined (2% and 7% respectively). Compared to WD-HCC, LFABP-negative MD/PD HCC were more commonly associated with male gender, viral hepatitis and cirrhosis. Biallelic HNF1A alterations were seen in all 13 (100%) sequenced cases. Additional mutations and/or copy number alterations were observed in 38% of AHN and 100% of WD-HCC. Diffuse glutamine synthetase (GS) staining was seen in 13% of cases, with no nuclear ß-catenin or Wnt signaling alterations. In conclusion, variant features such as lack of fat, peliosis, myxoid change, pseudoacini and abundant lipofuscin are more common in AHN and/or WD-HCC. LFABP-negative MD/PD HCC have different clinicopathologic features compared to WD-HCC. The significance of diffuse GS in a subset of these cases is unclear.

Impact of scanner variability on lymph node segmentation in computational pathology.

Computer-aided diagnostics in histopathology are based on the digitization of glass slides. However, heterogeneity between the images generated by different slide scanners can unfavorably affect the performance of computational algorithms. Here, we evaluate the impact of scanner variability on lymph node segmentation due to its clinical importance in colorectal cancer diagnosis. 100 slides containing 276 lymph nodes were digitized using 4 different slide scanners, and 50 of the lymph nodes containing metastatic cancer cells. These 400 scans were subsequently annotated by 2 experienced pathologists to precisely label lymph node boundary. Three different segmentation methods were then applied and compared: Hematoxylin-channel-based thresholding (HCT), Hematoxylin-based active contours (HAC), and a convolution neural network (U-Net). Evaluation of U-Net trained from both a single scanner and an ensemble of all scanners was completed. Mosaic images based on representative tiles from a scanner were used as a reference image to normalize the new data from different test scanners to evaluate the performance of a pre-trained model. Fine-tuning was carried out by using weights of a model trained on one scanner to initialize model weights for other scanners. To evaluate the domain generalization, domain adversarial learning and stain mix-up augmentation were also implemented. Results show that fine-tuning and domain adversarial learning decreased the impact of scanner variability and greatly improved segmentation across scanners. Overall, U-Net with stain mix-up (Matthews correlation coefficient (MCC) = 0.87), domain adversarial learning (MCC = 0.86), and HAC (MCC = 0.87) were shown to outperform HCT (MCC = 0.81) for segmentation of lymph nodes when compared against the ground truth. The findings of this study should be considered for future algorithms applied in diagnostic routines.

Classification of colorectal tissue images from high throughput tissue microarrays by ensemble deep learning methods.

Tissue microarray (TMA) core images are a treasure trove for artificial intelligence applications. However, a common problem of TMAs is multiple sectioning, which can change the content of the intended tissue core and requires re-labelling. Here, we investigate different ensemble methods for colorectal tissue classification using high-throughput TMAs. Hematoxylin and Eosin (H&E) core images of 0.6 mm or 1.0 mm diameter from three international cohorts were extracted from 54 digital slides (n = 15,150 cores). After TMA core extraction and color enhancement, five different flows of independent and ensemble deep learning were applied. Training and testing data with 2144 and 13,006 cores included three classes: tumor, normal or "other" tissue. Ground-truth data were collected from 30 ngTMA slides (n = 8689 cores). A test augmentation is applied to reduce the uncertain prediction. Predictive accuracy of the best method, namely Soft Voting Ensemble of one VGG and one CapsNet models was 0.982, 0.947 and 0.939 for normal, "other" and tumor, which outperformed to independent or ensemble learning with one base-estimator. Our high-accuracy algorithm for colorectal tissue classification in high-throughput TMAs is amenable to images from different institutions, core sizes and stain intensity. It helps to reduce error in TMA core evaluations with previously given labels.

Frequency and Significance of Pathologic Pulmonary Findings in Postmortem Examinations-A Single Center Experience before COVID-19.

Coronavirus disease 2019 (COVID-19) has shown the importance of postmortem investigation of deceased patients. For a correct interpretation of the pulmonary findings in this new era, it is, however, crucial to be familiar with pathologic pulmonary conditions observed in postmortem investigations in general. Adequate postmortem histopathological evaluation of the lungs may be affected by suboptimal gross work up, autolysis or poor fixation. Using a standardized preparation approach which consisted in instillation of 4% buffered formaldehyde through the large bronchi for proper fixation and preparing large frontal tissue sections of 1-2 cm thickness after at least 24 h fixation, we comprehensively analyzed postmortem pulmonary findings from consecutive adult autopsies of a two-year period before the occurrence of COVID-19 (2016-2017). In total, significant pathological findings were observed in 97/189 patients (51%), with 28 patients showing more than one pathologic condition. Acute pneumonia was diagnosed 33/128 times (26%), embolism 24 times (19%), primary pulmonary neoplasms 18 times (14%), organizing pneumonia and other fibrosing conditions 14 times (11%), pulmonary metastases 13 times (10%), diffuse alveolar damage 12 times (9%), severe emphysema 9 times (7%) and other pathologies, e.g., amyloidosis 5/128 times (4%). Pulmonary/cardiopulmonary disease was the cause of death in 60 patients (32%). Clinical and pathological diagnoses regarding lung findings correlated completely in 75 patients (40%). Autopsy led to confirmation of a clinically suspected pulmonary diagnosis in 57 patients (39%) and clarification of an unclear clinical lung finding in 16 patients (8%). Major discrepant findings regarding the lungs (N = 31; 16%) comprised cases with clinical suspicions that could not be confirmed or new findings not diagnosed intra vitam. A significant proportion of acute pneumonias (N = 8; 24% of all cases with this diagnosis; p = 0.011) was not diagnosed clinically. We confirmed the frequent occurrence of pulmonary pathologies in autopsies, including inflammatory and neoplastic lesions as the most frequent pathological findings. Acute pneumonia was an important cause for discrepancy between clinical and postmortem diagnostics.

RHAMM in liver metastases of stage IV colorectal cancer with mismatch-repair proficient status correlates with tumor budding, cytotoxic T-cells and PD-1/PD-L1.

BACKGROUND: During the last decades, the management for metastatic colorectal cancer patients has improved due to novel therapeutic approaches. A mismatch-repair deficient status seems to favour a better response to checkpoint inhibitor therapy, but the question arises whether a specific subgroup of stage IV patients with mismatch-repair (MMR) proficient status should also be considered. RHAMM (Receptor for Hyaluronic Acid Mediated Motility/HAMMR/CD168) is characterized by tumor progression and immunogenicity. Therefore, the aim of this study is to determine whether RHAMM within the CRLM of MMR-proficient patients correlate with a more immunological microenvironment, represented by cytotoxic T-cells, PD-1 and PD-1. METHODS: Two patient cohorts of liver metastases from MMR colorectal cancers were included into the study (n = 81 and 76) using ngTMA® technology and immunohistochemically analyzed for RHAMM, cytotoxic T-cells (CD8+), PD-1/PD-L1, intrametastatic budding (IMB) and perimetastatic budding (PMB). RESULTS: RHAMM-positive IMB was linked to a higher PD-L1 expression (r = 0.32; p = 0.233 and r = 0.28; p = 0.044) in the center and periphery of the metastasis and RHAMM-positive PMB was associated with a higher expression of PD-1 (r = 0.33; p = 0.0297), and especially PD-L1 (r = 0.604; p < 0.0001 and r = 0.43; p = 0.003) in the center and periphery of the metastasis. IMB and PMB were additionally associated with a higher count of CD8+ T-cells (p < 0.0001; r = 0.58; p < 0.0001; r = 0.53). CONCLUSIONS: The RHAMM status can be assessed in IMB/PMB either in biopsies or in resections of colorectal cancer liver metastases. A positive RHAMM status in IMB and/or PMB may be a potential indicator for a checkpoint inhibitor therapy for stage IV colorectal cancer patients with MMR proficient status.

CUX1-Transcriptional Master Regulator of Tumor Progression in Pancreatic Neuroendocrine Tumors.

Recently, we identified the homeodomain transcription factor Cut homeobox 1 (CUX1) as mediator of tumour de-differentiation and metastatic behaviour in human insulinoma patients. In insulinomas, CUX1 enhanced tumour progression by stimulating proliferation and angiogenesis in vitro and in vivo. In patients with non-functional pancreatic neuroendocrine tumours (PanNET), however, the impact of CUX1 remains to be elucidated. Here, we analysed CUX1 expression in two large independent cohorts (n = 43 and n = 141 tissues) of non-functional treatment-naïve and pre-treated PanNET patients, as well as in the RIP1Tag2 mouse model of pancreatic neuroendocrine tumours. To further assess the functional role of CUX1, expression profiling of DNA damage-, proliferation- and apoptosis-associated genes was performed in CUX1-overexpressing Bon-1 cells. Validation of differentially regulated genes was performed in Bon-1 and QGP1 cells with knock-down and overexpression strategies. CUX1 expression assessed by a predefined immunoreactivity score (IRS) was significantly associated with shorter progression-free survival (PFS) of pre-treated PanNET patients (23 vs. 8 months; p = 0.005). In treatment-naïve patients, CUX1 was negatively correlated with grading and recurrence-free survival (mRFS of 39 versus 8 months; p = 0.022). In both groups, high CUX1 levels indicated a metastatic phenotype. Functionally, CUX1 upregulated expression of caspases and death associated protein kinase 1 (DAPK1), known as mediators of tumour progression and resistance to cytotoxic drugs. This was also confirmed in both cell lines and human tissues. In the RIP1Tag2 mouse model, CUX1 expression was associated with advanced tumour stage and resistance to apoptosis. In summary, we identified the transcription factor CUX1 as mediator of tumour progression in non-functional PanNET in vitro and in vivo, indicating that the CUX1-dependent signalling network is a promising target for future therapeutic intervention.

Are tumour grade and tumour budding equivalent in colorectal cancer? A retrospective analysis of 771 patients.

BACKGROUND: Tumour grade is traditionally considered in the management of patients with colorectal cancer. However, a large body of literature suggests that a related feature, namely tumour budding, may have a more important clinical impact. The aim of our study is to determine the correlation between tumour grade and tumour budding and their impact on patient outcome. METHODS: A retrospective collective of 771 patients with colorectal cancer were included in the study. Clinicopathological information included tumour grade (World Health Organisation 2010; G1, G2 and G3) and tumour budding evaluated as BD1, BD2 and BD3 and representing 0-4 buds, 5-9 buds and 10 or more buds per 0.785 mm2, respectively. RESULTS: Tumour grade and tumour budding were correlated (p < 0.0001, percent concordance: 33.8%). Of the BD1 cases, 18.1% were of G3. Only two BD3 cases were G1. Both high tumour grade and tumour budding were associated with higher pT, lymph node metastasis, distant metastasis and lymphatic and venous vessel invasion (p < 0.01, all), but only tumour grade was additionally associated with right-sided tumour location and mucinous histology. Higher tumour budding led to worse overall (p = 0.0286) and disease-free survival (p = 0.001), but tumour grade did not. Budding was independent of tumour grade in multivariate analysis. DISCUSSION: Tumour grade and tumour budding are distinct features, as recognised by their different clinicopathological associations, reflecting different underlying biological processes. Nonetheless, tumour budding seems to outperform tumour grade in terms of predicting disease-free survival.