Limitations of tissue microarrays compared with whole tissue sections in survival analysis

Performance Analysis of Leica Biosystems Monoclonal Antibody Programmed Cell Death Ligand 1 Clone 73-10 on Breast, Colorectal, and Hepatocellular Carcinomas

Programmed cell death receptor 1/Programmed cell death ligand 1 (PD-L1) checkpoint pathway is responsible for the control of immune cell responses. Immunotherapy using checkpoint inhibitors, such as anti-PD-L1 therapy, aids disease management and potentiates clinical outcomes. This study aimed to analyze the performance of the Leica Biosystems (LBS) USA FDA class I in vitro diagnostic monoclonal antibody (clone 73-10) to detect PD-L1 expression in breast, colorectal, and hepatocellular carcinomas compared with the class III FDA-approved PD-L1 detecting antibodies [SP263 (Ventana), 22C3 (Dako), and 28-8 (Dako)] using 208 unique tissue microarray-based cases for each tumor type. The interassay concordances between LBS 73-10 clone and other PD-L1 antibodies ranged from 0.59 to 0.95 Cohen kappa coefficient (K) and from 0.66 to 0.90 (K) for cutoff values of 1% and 50% tumor proportion score (TPS), respectively. The 73-10 clones showed inter-pathologist agreements ranging from 0.53 to 1.0 (K) and 0.34 to 0.94 (K) for cutoff values of 1% and 50% TPS, respectively. For the immune cell proportion score (IPS) using a cutoff of 1%, the Kappa coefficient of interassay concordances and inter-pathologist agreements ranged from 0.34 to 0.94. The 73-10 clone assay's sensitivity ranged from 78.3% to 100% (TPS ≥1%), 100% (TPS ≥50%), and 77.4% to 93.5% (IPS ≥1%), while its specificity was 97.9% to 100% (TPS ≥1%), 99.5% to 99.8% (TPS ≥50%), and 97.9% to 100% (IPS ≥1%). This exploratory evaluation of LBS 73-10 monoclonal antibody on a large set of breast, colorectal, and hepatocellular carcinomas showed the assay's technical performance is comparable to the FDA-approved companion/complementary diagnostics PD-L1 detection assays.

Image-based multiplex immune profiling of cancer tissues: translational implications. A report of the International Immuno-oncology Biomarker Working Group on Breast Cancer

Recent advances in the field of immuno-oncology have brought transformative changes in the management of cancer patients. The immune profile of tumours has been found to have key value in predicting disease prognosis and treatment response in various cancers. Multiplex immunohistochemistry and immunofluorescence have emerged as potent tools for the simultaneous detection of multiple protein biomarkers in a single tissue section, thereby expanding opportunities for molecular and immune profiling while preserving tissue samples. By establishing the phenotype of individual tumour cells when distributed within a mixed cell population, the identification of clinically relevant biomarkers with high-throughput multiplex immunophenotyping of tumour samples has great potential to guide appropriate treatment choices. Moreover, the emergence of novel multi-marker imaging approaches can now provide unprecedented insights into the tumour microenvironment, including the potential interplay between various cell types. However, there are significant challenges to widespread integration of these technologies in daily research and clinical practice. This review addresses the challenges and potential solutions within a structured framework of action from a regulatory and clinical trial perspective. New developments within the field of immunophenotyping using multiplexed tissue imaging platforms and associated digital pathology are also described, with a specific focus on translational implications across different subtypes of cancer. © 2024 The Authors. The Journal of Pathology published by John Wiley & Sons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Findings from a Malaysian multicentre study on oropharyngeal squamous cell carcinoma

BACKGROUND

In addition to the conventional aetiologic agents of oropharyngeal squamous cell carcinoma (OPSCC) such as tobacco usage, alcohol consumption and betel quid usage, it has been established that a proportion of OPSCC are driven by persistent oncogenic human papillomavirus (HPV) infections. Currently, there is a lack of data on the burden of HPV- associated OPSCC in Asian countries including Malaysia.

METHODS

A cross-sectional multicentre study with tissue analysis of Malaysian patients diagnosed with primary OPSCC within a five-year period, from 2015 to 2019 between 01/01/2015 to 31/12/2019 was undertaken. Determination of HPV status was carried out using p16INK4a immunohistochemistry on tissue microarrays constructed from archived formalin-fixed paraffin-embedded tissue.

RESULTS

From the cases identified, 184 cases had sufficient tissue material for analysis. Overall, median age at diagnosis was 63.0 years (IQR = 15) and 76.1% of patients were males. In our cohort, 35.3% of patients were Indian, 34.2% were Chinese, 21.2% were Malay and 9.2% were from other ethnicities. The estimated prevalence of HPV-associated OPSCC in our cohort was 31.0% (CI 24.4-38.2%). The median age for the HPV-associated OPSCC sub-group of patients was not significantly lower than the median age of patients with HPV-independent OPSCC. More than half of HPV-associated OPSCC was seen in patients of Chinese ethnicity (54.4%). Patients with HPV-associated OPSCC had a much better overall survival than patients with HPV-independent OPSCC (Log rank test; p < 0.001). Patients with HPV-associated OPSCC with no habit-related risk factors such as smoking, were found to have much better overall survival when compared to all other sub-groups.

CONCLUSIONS

The findings from our study suggests that prevalence of HPV-associated OPSCC in Malaysia, though not as high as some developed countries, is however on an upward trend. HPV-associated OPSCC appears to be more frequently encountered in patients of Chinese ethnicity. Conventional risk-factors associated with OPSCC such as smoking, alcohol consumption and betel quid chewing should still be considered when estimating prognosis of patients with HPV-associated OPSCC.

Stress Keratin 17 Is a Predictive Biomarker Inversely Associated with Response to Immune Check-Point Blockade in Head and Neck Squamous Cell Carcinomas and Beyond

Low response rates in immune check-point blockade (ICB)-treated head and neck squamous cell carcinoma (HNSCC) drive a critical need for robust, clinically validated predictive biomarkers. Our group previously showed that stress keratin 17 (CK17) suppresses macrophage-mediated CXCL9/CXCL10 chemokine signaling involved in attracting activated CD8+ T cells into tumors, correlating with decreased response rate to pembrolizumab-based therapy in a pilot cohort of ICB-treated HNSCC (n = 26). Here, we performed an expanded analysis of the predictive value of CK17 in ICB-treated HNSCC according to the REMARK criteria and investigated the gene expression profiles associated with high CK17 expression. Pretreatment samples from pembrolizumab-treated HNSCC patients were stained via immunohistochemistry using a CK17 monoclonal antibody (n = 48) and subjected to spatial transcriptomic profiling (n = 8). Our findings were validated in an independent retrospective cohort (n = 22). CK17 RNA expression in pembrolizumab-treated patients with various cancer types was investigated for predictive significance. Of the 48 patients (60% male, median age of 61.5 years), 21 (44%) were CK17 high, and 27 (56%) were CK17 low. A total of 17 patients (35%, 77% CK17 low) had disease control, while 31 patients (65%, 45% CK17 low) had progressive disease. High CK17 expression was associated with a lack of disease control (p = 0.037), shorter time to treatment failure (p = 0.025), and progression-free survival (PFS, p = 0.004), but not overall survival (OS, p = 0.06). A high CK17 expression was associated with lack of disease control in an independent validation cohort (p = 0.011). PD-L1 expression did not correlate with CK17 expression or clinical outcome. CK17 RNA expression was predictive of PFS and OS in 552 pembrolizumab-treated cancer patients. Our findings indicate that high CK17 expression may predict resistance to ICB in HNSCC patients and beyond.

Slide-to-Slide Tissue Transfer and Array Assembly From Limited Samples for Comprehensive Molecular Profiling

Tissue microarrays (TMA) have become an important tool in high-throughput molecular profiling of tissue samples in the translational research setting. Unfortunately, high-throughput profiling in small biopsy specimens or rare tumor samples (eg, orphan diseases or unusual tumors) is often precluded owing to limited amounts of tissue. To overcome these challenges, we devised a method that allows tissue transfer and construction of TMAs from individual 2- to 5-μm sections for subsequent molecular profiling. We named the technique slide-to-slide (STS) transfer, and it requires a series of chemical exposures (so-called xylene-methacrylate exchange) in combination with rehydrated lifting, microdissection of donor tissues into multiple small tissue fragments (methacrylate-tissue tiles), and subsequent remounting on separate recipient slides (STS array slide). We developed the STS technique by assessing the efficacy and analytical performance using the following key metrics: (a) dropout rate, (b) transfer efficacy, (c) success rates using different antigen-retrieval methods, (d) success rates of immunohistochemical stains, (e) fluorescent in situ hybridization success rates, and (f) DNA and (g) RNA extraction yields from single slides, which all functioned appropriately. The dropout rate ranged from 0.7% to 6.2%; however, we applied the same STS technique successfully to fill these dropouts ("rescue" transfer). Hematoxylin and eosin assessment of donor slides confirmed a transfer efficacy of >93%, depending on the size of the tissue (range, 76%-100%). Fluorescent in situ hybridization success rates and nucleic acid yields were comparable with those of traditional workflows. In this study, we present a quick, reliable, and cost-effective method that offers the key advantages of TMAs and other molecular techniques-even when tissue is sparse. The perspectives of this technology in biomedical sciences and clinical practice are promising, given that it allows laboratories to create more data with less tissue.

Breast Cancer Dataset, Classification and Detection Using Deep Learning

Incorporating scientific research into clinical practice via clinical informatics, which includes genomics, proteomics, bioinformatics, and biostatistics, improves patients' treatment. Computational pathology is a growing subspecialty with the potential to integrate whole slide images, multi-omics data, and health informatics. Pathology and laboratory medicine are critical to diagnosing cancer. This work will review existing computational and digital pathology methods for breast cancer diagnosis with a special focus on deep learning. The paper starts by reviewing public datasets related to breast cancer diagnosis. Additionally, existing deep learning methods for breast cancer diagnosis are reviewed. The publicly available code repositories are introduced as well. The paper is closed by highlighting challenges and future works for deep learning-based diagnosis.

Commentary on: SMARCB1 as a novel diagnostic and prognostic biomarker for osteosarcoma

In the last couple of decades, biomarkers have been on the rise for diagnostic and predictive value. There has been a rush to identify new markers using new technologies and drug repurposing approaches. SMARCB1 acronym arises from the SWI/SNF (SWItch/Sucrose Non-Fermentable)-related Matrix-associated Actin-dependent Regulator of Chromatin subfamily B member 1 (SMARCB1). It is a molecule, whose role is associated with the sucrose metabolism. SMARCB1 is also called INI1 (Integrase Interactor 1). The molecule was discovered in the mid-1990s. Its role as a loss-of-function marker for malignant rhabdoid tumors (MRT) of renal and extrarenal origin has enormously expanded the spectrum of involved neoplasms since that time. Several tumors have been characterized by genetic aberrations in the SMARCB1 gene. They include reduction in expression, loss of expression, and mosaic expression. Most of the tumors are sarcomas, but a variegated group of tumors with mixed phenotypes has also been delineated. It is well known that the outcome of patients harboring genetic aberrations in the SMARCB1 gene has been poor. Guo et al. reported that reduced SMARCB1 expression occurred in 70% of osteosarcomas. Their data significantly correlated with poor neoadjuvant response. These authors emphasize a shorter progression-free and overall survival of the patients demonstrating an altered expression of this gene. Interestingly, mRNA in silico analysis established that SMARCB1 expression correlates with the response to chemotherapy of osteosarcoma patients, but there was no reliable correlation between SMARCB1 expression level and metastasis, response to neoadjuvant therapy, overall survival, and progression-free survival. The study involved a tissue microarray (TMA) on bone tumors that may limit the full evaluation of the gene expression. Nevertheless, Guo et al.’s study is remarkable. It expands the list of the tumors harboring an altered SMARCB1 gene expression and suggests that this marker should be investigated in every pathology workup for potential predictive value. On the other side, much work needs to be done if we hope that we strive to provide additional therapeutic strategies for osteosarcoma patients with altered SMARCB1 gene expression.

High expression of p62/SQSTM1 predicts shorter survival for patients with pancreatic cancer

BACKGROUND

Accumulation of the signal adaptor protein p62 has been demonstrated in many forms of cancer, including pancreatic ductal adenocarcinoma (PDAC). Although data from experimental studies suggest that p62 accumulation accelerates the development of PDAC, the association between p62 protein expression and survival in PDAC patients is unclear.

METHODS

Thirty-three tumor specimens from PDAC patients treated by primary surgery were obtained. Immunohistochemical expression of p62, microtubule-associated protein 1A/1B-light chain 3 (LC3), and nuclear factor-erythroid factor 2-related factor 2 (NRF2) in tumor tissue was examined for associations with clinicopathological characteristics and disease-specific survival (DSS).

RESULTS

There was no association between p62 expression and any of the clinicopathological variables. However, high p62 protein expression in tumor cells was significantly associated with shorter DSS (7 months vs. 29 months, p = 0.017). The hazard ratio for death in patients with high p62 protein expression in tumor cells was 2.88 (95% confidence interval: 1.17-7.11, p = 0.022). In multivariable analysis, high p62 expression was an independent prognostic factor for shorter DSS (p = 0.020) when follow up time was more than 5 years. LC3 and NRF2 staining was not associated with survival or other clinicopathological parameters.

CONCLUSION

Our results show that high p62 protein expression in tumor cells is associated with shorter survival following pancreatic tumor resection. This association supports a role for p62 as a prognostic marker in patients with PDAC treated by primary surgery.

Prognosis Value of Immunoregulatory Molecules in Oral Cancer Microenvironment: An Immunohistochemical Study

Objectives: To evaluate the relationship of the immune-checkpoint PD-1/PD-L1 with the clinical evolution of OSCC; to assess survival in OSCC based on the characteristics of TME and histologic risk score; to evaluate the clinical and histopathological relationship of OSCC with immunological TME. Material and Methods: A retrospective study was carried out on 65 samples from patients with OSCC on the floor of the mouth or tongue. Clinicopathological variables and the expression of the biomarkers PD-1, PD-L1, FoxP3, CD4, CD8, CSF1R, and p16 were recorded. The relationship of the clinical and histological variables with the expression of the biomarkers and survival was studied. Results: The univariate and multivariate analysis indicated that positive PD-1 expression was an independent protective factor for survival (overall, disease-free, disease-specific survival) and that high PD-L1 also improved survival. Poorly differentiated histological grades and metastasis were associated with a worse prognosis. Conclusions: PD-1 is a protective survival factor that is maintained independently of PD-L1 expression. High values of PD-L1 expression also improve survival. Higher expression of PD-1 is observed in smaller tumors, and higher expression of PD-L1 is more likely in women. No relationship between the tumor microenvironment and histologic risk score was found to influence the survival patterns studied in the OSCC. There is no evidence of a relationship between the histopathological features and the studied markers, although the positive PD-1 and PD-L1 cases have a lower risk of a high WPOI score, and positive PD-1 expression was associated with a lower DOI.

PD-1 Expression and its Correlation With Prognosis in Clear Cell Renal Cell Carcinoma

BACKGROUND/AIM

Programmed death ligand-1 (PD-L1) and programmed death protein 1 (PD-1) expression levels in many tumors and their correlation with prognosis have been actively studied. However, studies on PD-1 expression and its prognostic value in clear cell renal cell carcinoma (ccRCC) are limited and controversial. In this study, we describe the expression of PD-1 and its prognostic significance and association with clinical features in patients with ccRCC.

MATERIALS AND METHODS

We obtained clinicopathological data from 166 patients with ccRCC who were treated at Gyeongsang National University Hospital, Jinju, Korea between January 2000 and December 2009. Tissue microarray blocks were made using representative paraffin blocks of ccRCC specimens. Two pathologists analyzed PD-L1 and PD-1 expression in both tumor and inflammatory cells.

RESULTS

PD-1 expression in tumor-infiltrating inflammatory cells was significantly correlated with unfavorable disease-free survival (DFS) (p<0.001) and disease-specific survival (DSS) (p=0.002) in the univariate analysis. A statistically significant correlation between PD-1 expression and unfavorable DFS (p=0.025) was observed in the multivariate analysis.

CONCLUSION

PD-1 expression in tumor-infiltrating inflammatory cells serves as an independent prognostic factor for unfavorable DSS in patients with ccRCC.

Expression of PAX2 and PAX8 in Wilms Tumor: A Tissue Microarray-based Immunohistochemical Study

Background & Objective:

There is currently inadequate information about the expression of immunohistochemical markers in pediatric tumors. Paired box genes 2 and 8 (PAX2 and PAX8) genes have an essential role in kidney organogenesis. This study aimed to investigate the IHC expression of PAX2 and PAX8 in Wilms tumor. Such study would be helpful in diagnosis and possibly in differentiation of this tumor from other mimics, especially in those of poorly differentiated type in small needle biopsy specimens.

Methods:

We performed a cross-sectional study on 45 Wilms tumor cases referred to Bahrami pediatric hospital between 2005 and 2015. Demographic data were collected from medical documents. Sections from related paraffin blocks were provided by the tissue microarray method, and immunohistochemical (IHC) staining was done for PAX8 and PAX2.

Results:

The mean tumor size was 9.98±4.95 cm. Favorable histology was seen in 84.4% of samples. PAX2 was expressed in 41 cases (91.1%), and PAX8 in 37 patients (82.2%). PAX2 and PAX8 expression was mostly seen in both blastemal and epithelial components (77.8% and 66.6%), respectively. Tumors with favorable and unfavorable histology did not significantly differ in PAX2 and PAX8 expression (P=0.637). We found a statically significant relationship between PAX8 expression and tumor size (P=0.033).

Conclusion:

PAX2 and PAX8 markers might helpful in diagnosis of Wilms tumor and may differentiate it from other histologically similar kidney tumors. PAX8 expression may be associated with larger tumor size. Tumors with favorable and unfavorable histology may not be different in PAX2 and PAX8 expression.

Epithelial-mesenchymal transition and cancer stem cell interactions in breast phyllodes tumours: immunohistochemical evaluation of EZH2, EZR, HMGA2, CD24 and CD44 in correlation with outcome analysis

AIM

Phyllodes tumours (PTs) categorised as benign, borderline and malignant, account for 1% of all breast tumours. Histological assessment does not always predict tumour behaviour, hindering determination of the clinical course and management.Epithelial-mesenchymal transition (EMT) is an important process during embryogenesis. Dysregulation of EMT causes loss of cell polarity, decreased intercellular adhesion, increased motility and invasiveness, promoting tumour progression. Similarly, cancer stem cells (CSCs) promote tumour growth, resistance and recurrence. The aim of this study is to evaluate expression of CSC markers; enhancer of zeste homolog 2 (EZH2), CD24 and CD44 and EMT associated proteins; ezrin (EZR) and high-mobility group AT-hook 2 (HMGA2) in PTs.

METHOD

Uing tissue microarray sections, immunohistochemistry was performed on 360 PTs. Epithelial and stromal expressions of EZH2, EZR, HMGA2, CD24 and CD44 were evaluated to assess their impact on disease progression and behaviour in correlation with clinicopathological parameters.

RESULTS

Stromal expression of EZH2, EZR and HMGA2 was observed in 73 (20.3%), 53 (14.7%) and 28 (7.8%) of tumours, epithelial expression in 121 (35.9%), 3 (0.8%) and 351 (97.5%) tumours, respectively. CD24 and CD44 staining was absent in both components.

CONCLUSION

Expression of biomarkers correlated significantly with aggressive tumour traits such as stromal hypercellularity, atypia, mitoses and permeative tumour borders.Stromal expression of EZH2 and EZR shortened disease-free survival and overall survival; HMGA2 expression did not alter patient survival. EZH2 and EZR may thus be useful in predicting PT behaviour.

Processing and Analysis of RNA-seq Data from Public Resources

Advances in next generation sequencing (NGS) technologies resulted in a broad array of large-scale gene expression studies and an unprecedented volume of whole messenger RNA (mRNA) sequencing data, or the transcriptome (also known as RNA sequencing, or RNA-seq). These include the Genotype Tissue Expression project (GTEx) and The Cancer Genome Atlas (TCGA), among others. Here we cover some of the commonly used datasets, provide an overview on how to begin the analysis pipeline, and how to explore and interpret the data provided by these publicly available resources.

Upregulation of peroxisome proliferator-activated receptor-α and the lipid metabolism pathway promotes carcinogenesis of ampullary cancer

Ampullary cancer is a rare periampullary cancer currently with no targeted therapeutic agent. It is important to develop a deeper understanding of the carcinogenesis of ampullary cancer. We attempted to explore the characteristics of ampullary cancer in our dataset and a public database, followed by a search for potential drugs. We used a bioinformatics pipeline to analyze complementary (c)DNA microarray data of ampullary cancer and surrounding normal duodenal tissues from five patients. A public database from the National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO) was applied for external validation. Bioinformatics tools used included the Gene Set Enrichment Analysis (GSEA), Database for Annotation, Visualization and Integrated Discovery (DAVID), MetaCore, Kyoto Encyclopedia of Genes and Genomes (KEGG), Hallmark, BioCarta, Reactome, and Connectivity Map (CMap). In total, 9097 genes were upregulated in the five ampullary cancer samples compared to normal duodenal tissues. From the MetaCore analysis, genes of peroxisome proliferator-activated receptor alpha (PPARA) and retinoid X receptor (RXR)-regulated lipid metabolism were overexpressed in ampullary cancer tissues. Further a GSEA of the KEGG, Hallmark, Reactome, and Gene Ontology databases revealed that PPARA and lipid metabolism-related genes were enriched in our specimens of ampullary cancer and in the NCBI GSE39409 database. Expressions of PPARA messenger (m)RNA and the PPAR-α protein were higher in clinical samples and cell lines of ampullary cancer. US Food and Drug Administration (FDA)-approved drugs, including alvespimycin, trichostatin A (a histone deacetylase inhibitor), and cytochalasin B, may have novel therapeutic effects in ampullary cancer patients as predicted by the CMap analysis. Trichostatin A was the most potent agent for ampullary cancer with a half maximal inhibitory concentration of < 0.3 μM. According to our results, upregulation of PPARA and lipid metabolism-related genes are potential pathways in the carcinogenesis and development of ampullary cancer. Results from the CMap analysis suggested potential drugs for patients with ampullary cancer.

Deep learning-enabled breast cancer hormonal receptor status determination from base-level H&E stains

For newly diagnosed breast cancer, estrogen receptor status (ERS) is a key molecular marker used for prognosis and treatment decisions. During clinical management, ERS is determined by pathologists from immunohistochemistry (IHC) staining of biopsied tissue for the targeted receptor, which highlights the presence of cellular surface antigens. This is an expensive, time-consuming process which introduces discordance in results due to variability in IHC preparation and pathologist subjectivity. In contrast, hematoxylin and eosin (H&E) staining—which highlights cellular morphology—is quick, less expensive, and less variable in preparation. Here we show that machine learning can determine molecular marker status, as assessed by hormone receptors, directly from cellular morphology. We develop a multiple instance learning-based deep neural network that determines ERS from H&E-stained whole slide images (WSI). Our algorithm—trained strictly with WSI-level annotations—is accurate on a varied, multi-country dataset of 3,474 patients, achieving an area under the curve (AUC) of 0.92 for sensitivity and specificity. Our approach has the potential to augment clinicians’ capabilities in cancer prognosis and theragnosis by harnessing biological signals imperceptible to the human eye.

Determination of estrogen receptor status (ERS) in breast cancer tissue requires immunohistochemistry, which is sensitive to the vagaries of sample processing and the subjectivity of pathologists. Here the authors present a deep learning model that determines ERS from H&E stained tissue, which could improve oncology decisions in under-resourced settings.

Going deeper through the Gleason scoring scale: An automatic end-to-end system for histology prostate grading and cribriform pattern detection

BACKGROUND AND OBJECTIVE

Prostate cancer is one of the most common diseases affecting men worldwide. The Gleason scoring system is the primary diagnostic and prognostic tool for prostate cancer. Furthermore, recent reports indicate that the presence of patterns of the Gleason scale such as the cribriform pattern may also correlate with a worse prognosis compared to other patterns belonging to the Gleason grade 4. Current clinical guidelines have indicated the convenience of highlight its presence during the analysis of biopsies. All these requirements suppose a great workload for the pathologist during the analysis of each sample, which is based on the pathologist's visual analysis of the morphology and organisation of the glands in the tissue, a time-consuming and subjective task. In recent years, with the development of digitisation devices, the use of computer vision techniques for the analysis of biopsies has increased. However, to the best of the authors' knowledge, the development of algorithms to automatically detect individual cribriform patterns belonging to Gleason grade 4 has not yet been studied in the literature. The objective of the work presented in this paper is to develop a deep-learning-based system able to support pathologists in the daily analysis of prostate biopsies. This analysis must include the Gleason grading of local structures, the detection of cribriform patterns, and the Gleason scoring of the whole biopsy.

METHODS

The methodological core of this work is a patch-wise predictive model based on convolutional neural networks able to determine the presence of cancerous patterns based on the Gleason grading system. In particular, we train from scratch a simple self-design architecture with three filters and a top model with global-max pooling. The cribriform pattern is detected by retraining the set of filters of the last convolutional layer in the network. Subsequently, a biopsy-level prediction map is reconstructed by bi-linear interpolation of the patch-level prediction of the Gleason grades. In addition, from the reconstructed prediction map, we compute the percentage of each Gleason grade in the tissue to feed a multi-layer perceptron which provides a biopsy-level score.

RESULTS

In our SICAPv2 database, composed of 182 annotated whole slide images, we obtained a Cohen's quadratic kappa of 0.77 in the test set for the patch-level Gleason grading with the proposed architecture trained from scratch. Our results outperform previous ones reported in the literature. Furthermore, this model reaches the level of fine-tuned state-of-the-art architectures in a patient-based four groups cross validation. In the cribriform pattern detection task, we obtained an area under ROC curve of 0.82. Regarding the biopsy Gleason scoring, we achieved a quadratic Cohen's Kappa of 0.81 in the test subset. Shallow CNN architectures trained from scratch outperform current state-of-the-art methods for Gleason grades classification. Our proposed model is capable of characterising the different Gleason grades in prostate tissue by extracting low-level features through three basic blocks (i.e. convolutional layer + max pooling). The use of global-max pooling to reduce each activation map has shown to be a key factor for reducing complexity in the model and avoiding overfitting. Regarding the Gleason scoring of biopsies, a multi-layer perceptron has shown to better model the decision-making of pathologists than previous simpler models used in the literature.

Better prognosis of gastric cancer patients with high levels of tumor infiltrating lymphocytes is counteracted by PD-1 expression

The prognostic potential of anti-tumor immune responses is becoming increasingly important in adenocarcinoma of the gastroesophageal junction and stomach (AGE/S) especially regarding the use of immune checkpoint inhibitors. This study analyzes for the first time the prognostic impact of tumor-infiltrating lymphocytes (TILs) and checkpoint inhibitors in a large Caucasian cohort in patients with AGE/S. We screened tissue samples from 438 therapy-naïve patients with AGE/S undergoing surgery between 1992 and 2005, examined in a tissue microarray (TMA) and stained against human CD3, CD4, CD8, PD-1, and PD-L1. Out of 438 tissue samples, 210 were eligible for multivariate analysis. This revealed that high infiltration with CD3+, CD4+, or CD8+ TILs was associated with an increased overall survival in AGE/S patients, which could only be confirmed in multivariate analysis for CD3 (HR: 0.326; p = .023). Independent improved survival was limited to gastric cancer patients and to early tumor stages as long as TILs did not express PD-1 (HR: 1.522; p = .021). Subgroup analyses indicate that TIL-dependent anti-tumor immune response is only effective in gastric cancer patients in early stages of disease in PD-1 negative TILs. Combined analysis of PD-1 and CD3 could serve as a prognostic marker for the clinical outcome of gastric cancer patients and could also be of interest for immunotherapy.

Comprehensive evaluation of methods to assess overall and cell-specific immune infiltrates in breast cancer

Background

Breast cancer (BC) immune infiltrates play a critical role in tumor progression and response to treatment. Besides stromal tumor infiltrating lymphocytes (sTILs) which have recently reached level 1B evidence as a prognostic marker in triple negative BC, a plethora of methods to assess immune infiltration exists, and it is unclear how these compare to each other and if they can be used interchangeably.

Methods

Two experienced pathologists scored sTIL, intra-tumoral TIL (itTIL), and 6 immune cell types (CD3⁺, CD4⁺, CD8⁺, CD20⁺, CD68⁺, FOXP3⁺) in the International Cancer Genomics Consortium breast cancer cohort using hematoxylin and eosin-stained (n = 243) and immunohistochemistry-stained tissue microarrays (n = 254) and whole slides (n = 82). The same traits were evaluated using transcriptomic- and methylomic-based deconvolution methods or signatures.

Results

The concordance correlation coefficient (CCC) between pathologists for sTIL was very good (0.84) and for cell-specific immune infiltrates slightly lower (0.63–0.66). Comparison between tissue microarray and whole slide pathology scores revealed systematically higher values in whole slides (ratio 2.60–5.98). The Spearman correlations between microscopic sTIL and transcriptomic- or methylomic-based assessment of immune infiltrates were highly variable (r = 0.01–0.56). Similar observations were made for cell type-specific quantifications (r = 0.001–0.54). We observed a strong inter-method variability between the omics-derived estimations, which is further cell type dependent. Finally, we demonstrated that most methods more accurately identify highly infiltrated (sTIL ≥ 60%; area under the curve, AUC, 0.64–0.99) as compared to lowly infiltrated tumors (sTIL ≤ 10%; AUC 0.52–0.82).

Conclusions

There is a lower inter-pathologist concordance for cell-specific quantification as compared to overall infiltration quantification. Microscopic assessments are underestimated when considering small cores (tissue microarray) instead of whole slides. Results further highlight considerable differences between the microscopic-, transcriptomic-, and methylomic-based methods in the assessment of overall and cell-specific immune infiltration in BC. We therefore call for extreme caution when assessing immune infiltrates using current methods and emphasize the need for standardized immune characterization beyond TIL.

Prognostic value of PD-L1 expression for surgically treated localized renal cell carcinoma: implications for risk stratification and adjuvant therapies

Background:

We aimed to evaluate the prognostic role of programmed-death receptor ligand (PD-L1) in a multinational cohort of patients with localized renal cell carcinoma (RCC).

Methods:

Formalin-fixed paraffin-embedded blocks of 1017 patients from the Latin American Renal Cancer Group were analyzed. Tissue microarrays were immunostained for PD-L1 using a commercially available monoclonal antibody. Expression of PD-L1 in ⩾5% tumor cells was considered positive. PD-1 expression in immune cells was also assessed. All cases were reviewed twice based on antibody expression and compared with a positive control. Cox proportional hazard regression models were used to identify predictors of recurrence-free survival (RFS) and overall survival (OS).

Results:

A total of 738 cases with complete follow up met criteria. Median age was 57 [interquartile range (IQR): 49–64] years, and median follow up was 34 (IQR: 15–62.9) months. Median tumor size was 5 cm (IQR: 3.0–7.5 cm). Approximately 8.2% and 7.6% of tumors were PD-L1 and programmed cell-death 1 (PD-1) positive, respectively. PD-L1 and PD-1 positivity were significantly associated with higher tumor stage (both p < 0.001), and presence of tumor necrosis and lymphovascular multivariable analyses; PD-L1 positivity was found as a predictor of worse RFS [hazard ratio (HR) = 2.08, p = 0.05] and OS (HR = 2.61, p = 0.02).

Conclusions:

PD-L1 positivity was significantly associated with worse outcomes for patients with localized RCC at intermediate follow up. This marker may help stratify patients for stricter surveillance after surgical treatment and provide a basis for checkpoint-inhibitor therapy in the adjuvant setting.

EGFR, but not COX-2, protein in resected pancreatic ductal adenocarcinoma is associated with poor survival

The effects of EGFR and COX-2 protein overexpression on clinical outcomes in pancreatic ductal adenocarcinoma (PDAC) patients remains unclear. Therefore, the aim of the present study was to evaluate the protein expression of epithelial growth factor receptor (EGFR) and cyclooxygenase-2 (COX-2) in tumor cells in surgically resected PDAC, in comparison with clinicopathological characteristics and clinical outcomes. Immunohistochemical staining of formalin-fixed paraffin-embedded tissue derived from surgically resected tumors was performed. Tissue slides were evaluated for membrane wild-type EGFR and cytoplasmic COX-2 staining using a histoscore system. Statistical associations between EGFR and COX-2 staining and clinicopathological characteristics were examined to predict survival. In a cohort of 32 resected PDAC patients, high EGFR protein expression in tumor cells was significantly associated with shorter median overall survival (7.9 vs. 39.2 months, P=0.0038). The corresponding hazard ratio (HR) for patients with high EGFR protein expression in tumor cells was 3.12 [95% confidence interval (CI): 1.39–7.00, P=0.006]. COX-2 protein expression was not associated with survival (22.6 vs. 24.5 months P=0.60; HR 1.22 95% CI: 0.59–2.51, P=0.60). Following multivariate Cox regression analysis, high EGFR protein expression in tumor cells (P=0.043) remained as significant independent prognostic factor for survival. In conclusion, high wild-type EGFR protein expression, but not COX-2 protein expression, in tumor cells is a prognostic factor for reduced overall survival following pancreatic tumor resection, supporting a role for EGFR in identifying resected patients that may benefit from EGFR-targeted therapy.

Prognostic Impact of Tumor-Infiltrating Lymphocytes and Neutrophils on Survival of Patients with Upfront Resection of Pancreatic Cancer

In patients with pancreatic ductal adenocarcinoma (PDAC), the tumor microenvironment consists of cellular and stromal components that influence prognosis. Hence, tumor-infiltrating lymphocytes (TILs) may predict prognosis more precisely than conventional staging systems. Studies on the impact of TILs are heterogeneous and further research is needed. Therefore, this study aims to point out the importance of peritumoral TILs, tumor-infiltrating neutrophils (TINs), and immune subtype classification in PDAC. Material from 57 patients was analyzed with immunohistochemistry performed for CD3, CD8, CD20, CD66b, α-sma, and collagen. Hot spots with peritumoral TILs and TINs were quantified according to the QTiS algorithm and the distance of TILs hot spots to the tumor front was measured. Results were correlated with overall (OS) and progression-free survival (PFS). High densities of peritumoral hot spots with CD3⁺, CD8⁺, and CD20⁺ TILs correlated significantly with improved OS and PFS. Combined immune cell subtypes predicted improved OS and PFS. High infiltration of CD3⁺ TILs predicted progression after 12 months. The location of TILs' hot spots and their distance to the tumor front did not correlate with patient survival. Peritumoral TILs and the composition of the stroma predict OS and PFS in PDAC.

Protein expression patterns in cancer-associated fibroblasts and cells undergoing the epithelial-mesenchymal transition in ovarian cancers

Recent studies have shown that cancer-associated fibroblasts (CAFs) and the epithelial-mesenchymal transition (EMT) contribute to invasive and metastatic abilities of ovarian cancer (OC) cells. In the present study, we attempted to identify the role of CAF- and EMT-related proteins in OCs, including serous carcinoma, mucinous carcinoma, endometrioid carcinoma and clear cell carcinoma using an immunohistochemical approach. The following CAF-related markers were used: CD10, podoplanin, fibroblast activating protein (FAP), platelet derived growth factor receptor (PDGFRα), PDGFRβ, S100A4 and α-smooth muscle actin (α-SMA). In addition, the following EMT-related markers were investigated: Slug, TWIST1 and ZEB1We performed hierarchical cluster analysis to group the samples according to their scoring. Subgroup 1 was characterized by high expression of CD10, podoplanin, α-SMA, Slug and ZEB1, whereas subgroup 2 was closely associated with high expression of podoplanin, PDGFRα, PDGFRβ, α-SMA, and Slug. In addition, marked expression of CD10 was observed in subgroup 3. High expression of α-SMA was a distinctive feature in subgroup 4, and expression of podoplanin and α-SMA characterized subgroup 5. Each subgroup was correlated with a histological type. The fact that different histological types were associated with different subgroups suggests the presence of distinct and heterogeneous subpopulations of CAFs in OC.

KDM3A is not associated with metastasis and prognosis of breast cancer

Lysine demethylase 3A (KDM3A), also known as JMJD1A, has been associated with metastasis and poor prognosis in several cancer types, including renal cell carcinoma, prostate cancer and Ewing sarcoma. However, little is known regarding the clinicopathological significance of KDM3A expression in breast cancer (BCa). To investigate the clinical relevance of KDM3A expression in the setting of BCa, immunohistochemistry was performed on a tissue microarray consisting of 150 commercially available BCa samples. No significant correlation was identified between KDM3A expression and various clinicopathological variables, including clinical stage, pathological grade, tumor size and the expression statuses of human epidermal growth factor receptor 2, estrogen receptor, and progesterone receptor. In addition, no significant association between KDM3A expression and overall prognosis was observed. Taken together, these findings suggest that there is no significant association between KDM3A expression and clinicopathological variables, indicating that KDM3A may not be associated with the malignant behavior of BCa.

Prognostic implications of CD44, NANOG, OCT4, and BMI1 expression in tongue squamous cell carcinoma

BACKGROUND

Tongue squamous cell carcinoma (SCC) contains a cell subpopulation referred to as cancer stem cells (CSCs), which are responsible for tumor growth, metastasis, and resistance to chemotherapy and radiotherapy. The CSC markers have been used to isolate these cells and as biomarkers to predict overall survival.

METHODS

The CSC markers CD44, NANOG, OCT4, and BMI1 were investigated using reverse transcriptase-quantitative polymerase chain reaction (RT-qPCR) and immunohistochemistry and correlated with clinicopathological parameters.

RESULTS

The CD44 overexpression was associated with disease-related death (P = 0.02) and worst prognosis. NANOG was upregulated in nontumoral margins and associated with T1/T2 classification, lymph node metastasis, and worst prognosis. OCT4 was associated with lymph node metastasis and worst overall survival. BMI1 and CD44v3 were overexpressed in tongue SCC. Coexpression of CD44⁺⁺ /NANOG⁺⁺ was associated with worst overall survival when compared with patients with CD44^-/+ /NANOG^-/+ .

CONCLUSION

The CSC markers might play an important role not only in CSC trait acquisition but also in tongue SCC development and progression.

Automated analysis of co-localized protein expression in histologic sections of prostate cancer

An automated approach based on routinely-processed, whole-slide immunohistochemistry (IHC) was implemented to study co-localized protein expression in tissue samples. Expression of two markers was chosen to represent stromal (CD31) and epithelial (Ki-67) compartments in prostate cancer. IHC was performed on whole-slide sections representing low-, intermediate-, and high-grade disease from 15 patients. The automated workflow was developed using a training set of regions-of-interest in sequential tissue sections. Protein expression was studied on digital representations of IHC images across entire slides representing formalin-fixed paraffin embedded blocks. Using the training-set, the known association between Ki-67 and Gleason grade was confirmed. CD31 expression was more heterogeneous across samples and remained invariant with grade in this cohort. Interestingly, the Ki-67/CD31 ratio was significantly increased in high (Gleason ≥ 8) versus low/intermediate (Gleason ≤7) samples when assessed in the training-set and the whole-tissue block images. Further, the feasibility of the automated approach to process Tissue Microarray (TMA) samples in high throughput was evaluated. This work establishes an initial framework for automated analysis of co-localized protein expression and distribution in high-resolution digital microscopy images based on standard IHC techniques. Applied to a larger sample population, the approach may help to elucidate the biologic basis for the Gleason grade, which is the strongest, single factor distinguishing clinically aggressive from indolent prostate cancer.

Ki-67 in endometrial cancer: scoring optimization and prognostic relevance for window studies

Ki-67, a marker of cellular proliferation, is increasingly being used in pre-surgical window studies in endometrial cancer as a primary outcome measure. Unlike in breast cancer, however, there are no guidelines standardizing its measurement and its clinical relevance as a response biomarker is undetermined. It is, therefore, imperative that Ki-67 scoring protocols are optimized and its association with patient survival rigorously evaluated, in order to be able to clinically interpret the results of these studies. Using the International Ki-67 in Breast Cancer Working Group guidelines as a basis, whole slide, hot spot and invasive edge scoring protocols were evaluated using endometrial biopsies and hysterectomy specimens from 179 women. Whole sections and tissue microarrays, manual and semi-automated scoring using Definiens Developer software were additionally compared. Ki-67 scores were related to clinicopathological variables and cancer-specific survival in uni- and multivariate analysis. Against criteria of time efficiency, intra- and inter-observer variability and consistency, semi-automated hot spot scoring was the preferred method. Ki-67 scores positively correlated with grade, stage and depth of myometrial invasion (P-values all <0.03). By univariate analysis, higher Ki-67 scores were associated with a significant reduction in cancer-specific survival (P≤0.05); however, this effect was substantially attenuated in the multivariate model. In conclusion, hot spot scoring of whole sections using Definiens is an optimal method to quantify Ki-67 in endometrial cancer window study specimens. Measured this way, it is a clinically relevant marker, though further work is required to determine whether reductions in Ki-67 in neoadjuvant intervention studies translate into improved patient outcome.

Clinicopathological characteristics of oestrogen receptor negative, progesterone receptor positive breast cancers: re-evaluating subsets within this group

Aims

The presence of oestrogen and progesterone receptors (ER, PR) in breast carcinoma is an important prognostic indicator as well as a predictor of likely response to hormonal treatment. Current ambiguity surrounds ER-negative (–)/PR-positive (+) breast cancer (BC) as to whether this phenotype exists as a distinct entity. The independent predictive value of PR for treatment considerations is also in question, as some investigators believe ER status to be the single most important therapeutic predictive factor in BC. We undertook this study to determine the existence of ER(–)/PR(+) BC and the prognostic effect, if any, of this phenotype.

Methods

We investigated 267 archival documented ER(–)/PR(+) BCs diagnosed between January 1994 and July 2009. Histological slides were retrieved and reviewed. Tissue microarrays were constructed by selecting two 1 mm cores of tumour per case. Repeat immunohistochemistry was performed for confirmation of the ER(–)/PR(+) status. Clinicopathological parameters including age, ethnicity, tumour size, histological grade, histological subtype, associated ductal carcinoma in situ, lymphovascular invasion and lymph node status were evaluated.

Results

On repeat immunohistochemistry, 92 tumours were confirmed as ER(–)/PR(+) BCs. This phenotype accounted for 1.1% of all BC phenotypes and exhibited different clinicopathological features and survival outcome when compared with other phenotypes. ER(–)/PR(+) tumours showed a trend for an early recurrence and poorer overall survival as compared with the patients with ER(+)/PR(+) tumours and similar to ER(–)/PR(–) tumours.

Conclusions

Our findings suggest that ER(–)/PR(+) BCs exist, although rare, with distinct pathological and clinical characteristics from patients with ER(+)/PR(+) BCs.

Evaluation of P-glycoprotein (Pgp) expression in human osteosarcoma by high-throughput tissue microarray

Survival of osteosarcoma patients is currently limited by the development of metastases and multidrug resistance (MDR). A well-established cause of MDR involves overexpression of P-glycoprotein (Pgp) in tumor cells. However, some discrepancies still exist as to the clinical significance of Pgp in osteosarcoma. We sought to elucidate further whether the Pgp expression correlated with clinical behavior in a series of patients with osteosarcoma via high-throughput tissue microarray (TMA) analysis. Immunohistochemical analysis of Pgp expression in a TMA of 114 specimens with a retrospective review of 70 osteosarcoma patients admitted to the Massachusetts General Hospital (MGH) was performed. High Pgp expression was correlated with metastasis development and poor response to pre-operative chemotherapy in osteosarcoma patients. Eighteen of the fifty-seven patients initially admitted with primary osteosarcoma showed high Pgp expression. Among these 18 patients with high Pgp expression, 13 of 18 (72%) patients eventually developed metastases. There was no significant clinical relevance between Pgp expression and osteosarcoma survival. These results support that high expression of Pgp is important, but cannot be assigned as, an individual predictor in the development of human osteosarcoma. © 2016 Orthopaedic Research Society. Published by Wiley Periodicals, Inc. J Orthop Res 34:1606-1612, 2016.

PIK3CD promoted proliferation in diffuse large B cell lymphoma through upregulation of c-myc

Despite PIK3CD has been extensively reported in cancers, however, little evidence has been available regarding its role in the setting of diffuse large B cell lymphoma (DLBCL). In the present study, to investigate the role of PIK3CD in DLBCL, relevant experiments were carried out on both in vivo clinical tissue level and in vitro cell line level. Prognostic and clinicopathological significance were analyzed after immunohistochemical assay of PIK3CD expression on DLBCL tissue microarray. MTT assay and flow cytometry were employed to evaluate the proliferative variation, cell cycle, and apoptosis. Athymic nude mice xenografted with DLBCL cell line were employed to confirm the role of PIK3CD. It was found that there was a significant difference between expression of PIK3CD and international prognosis index (IPI), performance state (PS), and inferior overall prognosis. Furthermore, PIK3CD can promote proliferation and prevent apoptosis in DLBCL cells in vitro through upregulation of c-myc and p-AKT and in contrast downregulation of p21 and p27. In nude mice model, knock-down of PIK3CD was shown to be able to suppress the proliferation of DLBCL but not significantly compared with control group. Taken together, our study showed that PIK3CD can promote proliferation of DLBCL cells both in vitro and in vivo, suggesting that PIK3CD could be druggable in the therapy of DLBCL.

Upregulation of nemo-like kinase is an independent prognostic factor in colorectal cancer

AIM: To investigate the expression and oncogenic role of nemo-like kinase (NLK) in colorectal cancer.

METHODS: Expression of NLK protein was assessed by immunohistochemistry in tissue specimens from 56 cases of normal colorectal mucosa, 51 cases of colorectal adenoma, and 712 cases of colorectal cancer. In addition, NLK expression was knocked down using a lentivirus carrying NLK small hairpin RNA in colorectal cancer cells. Cell viability methylthiazoletetrazolium assays, colony formation assays, flow cytometry cell cycle assays, Transwell migration assays, and gene expression assays were performed to explore its role on proliferation and migration of colorectal cancer.

RESULTS: Expression of NLK protein progressively increased in tissues from the normal mucosa through adenoma to various stages of colorectal cancer. Overexpression of NLK protein was associated with advanced tumor-lymph node-metastasis stages, poor differentiation, lymph node and distant metastases, and a higher recurrence rate of colorectal cancer (P < 0.05). Multivariate analyses showed that NLK expression was an independent prognostic factor to predict overall survival (hazard ratio 2.57, 95% confidence interval: 1.66-3.98; P < 0.001) and disease-free survival (hazard ratio 1.96, 95% confidence interval: 1.40-2.74: P < 0.001) of colorectal cancer patients. Furthermore, knockdown of NLK expression in colorectal cancer cell lines reduced cell viability, colony formation, and migration, and arrested tumor cells at the G0/G1 phase of the cell cycle. At the gene level, knockdown of NLK expression inhibited matrix metalloproteinase-2 expression in colorectal cancer cells.

CONCLUSION: NLK overexpression is an independent prognostic factor in colorectal cancer and knockdown of NLK expression inhibits colorectal cancer progression and metastasis.

Expression of ZNF148 in different developing stages of colorectal cancer and its prognostic value: a large Chinese study based on tissue microarray

BACKGROUND

It has been speculated that zinc finger protein 148 (ZNF148) is a tumor suppressor. However, to the authors' knowledge, little is known about the clinical significance of ZNF148 expression in patients with colorectal cancer (CRC). The objective of the current study was to clarify the association between ZNF148 expression and the postoperative prognosis of patients with CRC.

METHODS

Tissue microarrays containing 56 normal mucosa, 51 adenoma, 742 CRC (TNM stage I-IV), 16 familial adenomatous polyposis, and 21 metastatic CRC specimens were examined immunohistochemically for ZNF148 expression.

RESULTS

Expression of ZNF148 was found to increase consecutively from normal mucosa to stage I CRC, and then decreased consecutively from stage I to stage IV CRC. Lower expression of ZNF148 in tumors was found to be significantly associated with lymph node metastases, advanced TNM disease stage, poor differentiation, higher rate of disease recurrence, worse overall survival (OS), and shorter disease-free survival. High expression of ZNF148 was also associated with improved OS (P = .025) and disease-free survival (P = .042) in patients with stages II to III CRC. On multivariate Cox analysis, lower ZNF148 expression in tumors, advanced TNM stage, colon cancer, and elevated serum carbohydrate antigen 19-9 (CA19-9) were found to be significant factors for a worse OS. In 16 patients with familial adenomatous polyposis, ZNF148 expression was upregulated at steps toward carcinogenesis. In 21 patients with metastatic CRC, although ZNF148 expression was higher in primary tumors compared with adjacent mucosa, its expression in metastatic tumors was significantly lower than that in primary tumors.

CONCLUSIONS

Although ZNF148 expression is related to colorectal carcinogenesis, high ZNF148 expression in patients with CRC appears to be inversely associated with malignant phenotypes and may serve as a significant prognostic factor after surgery for patients with CRC.