Genomics and emerging biomarkers for immunotherapy of colorectal cancer

The underlying molecular mechanisms and prognostic factors of RNA binding protein in colorectal cancer: a study based on multiple online databases

BACKGROUND

RNA binding protein (RBP) is an active factor involved in the occurrence and development of colorectal cancer (CRC). Therefore, the potential mechanism of RBP in CRC needs to be clarified by dry-lab analyses or wet-lab experiments.

METHODS

The differential RBP gene obtained from the GEPIA 2 (Gene Expression Profiling Interactive Analysis 2) were performed functional enrichment analysis. Then, the alternative splicing (AS) events related to survival were acquired by univariate regression analysis, and the correlation between RBP and AS was analyzed by R software. The online databases were conducted to analyze the mutation and methylation of RBPs in CRC. Moreover, 5 key RBP signatures were obtained through univariate and multivariate Cox regression analysis and established as RBP prognosis model. Subsequently, the above model was verified through another randomized group of TCGA CRC cohorts. Finally, multiple online databases and qRT-PCR analysis were carried to further confirm the expression of the above 5 RBP signatures in CRC.

RESULTS

Through a comprehensive bioinformatics analysis, it was revealed that RBPs had genetic and epigenetic changes in CRC. We obtained 300 differentially expressed RBPs in CRC samples. The functional analysis suggested that they mainly participated in spliceosome. Then, a regulatory network for RBP was established to participate in AS and DDX39B was detected to act as a potentially essential factor in the regulation of AS in CRC. Our analysis discovered that 11 differentially expressed RBPs with a mutation frequency higher than 5%. Furthermore, we found that 10 differentially expressed RBPs had methylation sites related to the prognosis of CRC, and a prognostic model was constructed by the 5 RBP signatures. In another randomized group of TCGA CRC cohorts, the prognostic performance of the 5 RBP signatures was verified.

CONCLUSION

The potential mechanisms that regulate the aberrant expression of RBPs in the development of CRC was explored, a network that regulated AS was established, and the RBP-related prognosis model was constructed and verified, which could improve the individualized prognosis prediction of CRC.

An immune-related model based on INHBA, JAG2 and CCL19 to predict the prognoses of colon cancer patients

Background

Colorectal cancer (CRC) is the leading cause of cancer deaths and most common malignant tumors worldwide. Immune-related genes (IRGs) can predict prognoses of patients and the effects of immunotherapy. A series of colon cancer (CCa) samples from The Cancer Genome Atlas (TCGA) were analyzed to provide a new perspective into this field.

Methods

Differential IRGs and IRGs with significant clinical outcomes (sIRGs) were calculated by the limma algorithm and univariate COX regression analysis. The potential molecular mechanisms of IRGs were detected by PPI, KEGG and GO analysis. Immune-related risk score model (IRRSM) was established based on multivariate COX regression analysis. Based on the median risk score of IRRSM, the high-risk group and low-risk group were distinguished. The expression levels of IHNBA and JAG2 and relationships between IHNBA and clinical features were verified by RT-qPCR.

Results

6 differential sIRGs of patients with CCa were selected by univariate COX regression analysis. Based on the sIRGs (INHBA, JAG2 and CCL19), the IRRSM was established to predict survival probability of CCa patients and to explore the potential correlations with clinical features. Furthermore, IRRSM reflected the infiltration status of 22 types of immune cells. The expression levels of IHNBA and JAG2 were higher in CCa tissues than that in adjacent normal tissues. The expression levels of IHNBA and JAG2 were increased in advanced T stages.

Conclusion

Our results illustrated that some sIRGs showed the latent value of predicting the prognoses of CCa patients and the clinical features. This study could provide a new insight for immune research and treatment strategies in CCa patients.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12935-021-02000-z.

Artificial intelligence-based pathology for gastrointestinal and hepatobiliary cancers

Artificial intelligence (AI) can extract complex information from visual data. Histopathology images of gastrointestinal (GI) and liver cancer contain a very high amount of information which human observers can only partially make sense of. Complementing human observers, AI allows an in-depth analysis of digitised histological slides of GI and liver cancer and offers a wide range of clinically relevant applications. First, AI can automatically detect tumour tissue, easing the exponentially increasing workload on pathologists. In addition, and possibly exceeding pathologist's capacities, AI can capture prognostically relevant tissue features and thus predict clinical outcome across GI and liver cancer types. Finally, AI has demonstrated its capacity to infer molecular and genetic alterations of cancer tissues from histological digital slides. These are likely only the first of many AI applications that will have important clinical implications. Thus, pathologists and clinicians alike should be aware of the principles of AI-based pathology and its ability to solve clinically relevant problems, along with its limitations and biases.

Towards the Interpretability of Machine Learning Predictions for Medical Applications Targeting Personalised Therapies: A Cancer Case Survey

Artificial Intelligence is providing astonishing results, with medicine being one of its favourite playgrounds. Machine Learning and, in particular, Deep Neural Networks are behind this revolution. Among the most challenging targets of interest in medicine are cancer diagnosis and therapies but, to start this revolution, software tools need to be adapted to cover the new requirements. In this sense, learning tools are becoming a commodity but, to be able to assist doctors on a daily basis, it is essential to fully understand how models can be interpreted. In this survey, we analyse current machine learning models and other in-silico tools as applied to medicine-specifically, to cancer research-and we discuss their interpretability, performance and the input data they are fed with. Artificial neural networks (ANN), logistic regression (LR) and support vector machines (SVM) have been observed to be the preferred models. In addition, convolutional neural networks (CNNs), supported by the rapid development of graphic processing units (GPUs) and high-performance computing (HPC) infrastructures, are gaining importance when image processing is feasible. However, the interpretability of machine learning predictions so that doctors can understand them, trust them and gain useful insights for the clinical practice is still rarely considered, which is a factor that needs to be improved to enhance doctors' predictive capacity and achieve individualised therapies in the near future.

Colorectal Cancer and Immunity: From the Wet Lab to Individuals

Simple Summary

Tackling the current dilemma of colorectal cancer resistance to immunotherapy is puzzling and requires novel therapeutic strategies to emerge. However, characterizing the intricate interactions between cancer and immune cells remains difficult because of the complexity and heterogeneity of both compartments. Developing rationales is intellectually feasible but testing them can be experimentally challenging and requires the development of innovative procedures and protocols. In this review, we delineated useful in vitro and in vivo models used for research in the field of immunotherapy that are or could be applied to colorectal cancer management and lead to major breakthroughs in the coming years.

Abstract

Immunotherapy is a very promising field of research and application for treating cancers, in particular for those that are resistant to chemotherapeutics. Immunotherapy aims at enhancing immune cell activation to increase tumor cells recognition and killing. However, some specific cancer types, such as colorectal cancer (CRC), are less responsive than others to the current immunotherapies. Intrinsic resistance can be mediated by the development of an immuno-suppressive environment in CRC. The mutational status of cancer cells also plays a role in this process. CRC can indeed be distinguished in two main subtypes. Microsatellite instable (MSI) tumors show a hyper-mutable phenotype caused by the deficiency of the DNA mismatch repair machinery (MMR) while microsatellite stable (MSS) tumors show a comparatively more “stable” mutational phenotype. Several studies demonstrated that MSI CRC generally display good prognoses for patients and immunotherapy is considered as a therapeutic option for this type of tumors. On the contrary, MSS metastatic CRC usually presents a worse prognosis and is not responsive to immunotherapy. According to this, developing new and innovative models for studying CRC response towards immune targeted therapies has become essential in the last years. Herein, we review the in vitro and in vivo models used for research in the field of immunotherapy applied to colorectal cancer.

TTN/OBSCN 'Double-Hit' predicts favourable prognosis, 'immune-hot' subtype and potentially better immunotherapeutic efficacy in colorectal cancer

Colorectal cancer (CRC) remains a leading cause of cancer-related deaths worldwide. Although treatment strategies for solid tumours have been revolutionized by immunotherapy, only a small subset of CRC patients benefit. Using two-independent cohorts, we found the common frequently mutated genes TTN and OBSCN had the significant correlation with higher tumour mutation burden (TMB) and favourable overall survival. TTN and OBSCN also displayed significant commutation phenomenon. Therefore, based on the status of TTN and OBSCN, we stratified patients into 'Double-WT' phenotype, 'Single-Hit' phenotype and 'Double-Hit' phenotype. Importantly, the 'Double-Hit' phenotype had favourable prognosis, low malignant events propensity, and highest TMB, immune cells infiltration abundance, POLE mutation rate, microsatellite instability ratio, as well as immune checkpoints expression compared with the other two phenotypes. These results indicated that the 'Double-Hit' phenotype suggested 'immune-hot' tumours and potentially better immunotherapeutic efficacy. Bioinformatic algorithm assessment of immunotherapy responses also confirmed this conclusion, and the 'Double-Hit' phenotype was found to be a better predictor of immunotherapy than PD-L1, PD-1, CTLA-4, TMB and microsatellite status. This study revealed CRC patients with TTN/OBSCN 'Double-Hit' was significantly associated favourable prognosis, 'immune-hot' subtype and potentially better immunotherapeutic efficacy.

T-cell immunoglobulin and ITIM domain, as a potential immune checkpoint target for immunotherapy of colorectal cancer

The importance of the tumor microenvironment in cancer progression has been well studied for many years. Immune checkpoint inhibitors (ICIs) are regarded as potential strategies in enhancing the immune responses in patients with cancer, particularly colorectal cancer (CRC). Notably, CRCs are extraordinarily heterogeneous and mostly are microsatellite-stable (MSS) or cold tumors, which means that the immune response is not usually as strong as that of foreign cells. T-cell immunoglobulin and ITIM domain (TIGIT) is a new immune checkpoint receptor overexpressed inside the CRC tumor-immune microenvironments. Moreover, several studies have shown that TIGIT in combination with other ICIs and/or conventional treatments, can lead to a robust anti-tumor response in CRC. This review looks deep inside TIGIT expression patterns, their various functions, and possible immunotherapy strategies to increase survival rates and decrease immune-related adverse events.

Next Generation Imaging Techniques to Define Immune Topographies in Solid Tumors

In recent years, cancer immunotherapy experienced remarkable developments and it is nowadays considered a promising therapeutic frontier against many types of cancer, especially hematological malignancies. However, in most types of solid tumors, immunotherapy efficacy is modest, partly because of the limited accessibility of lymphocytes to the tumor core. This immune exclusion is mediated by a variety of physical, functional and dynamic barriers, which play a role in shaping the immune infiltrate in the tumor microenvironment. At present there is no unified and integrated understanding about the role played by different postulated models of immune exclusion in human solid tumors. Systematically mapping immune landscapes or "topographies" in cancers of different histology is of pivotal importance to characterize spatial and temporal distribution of lymphocytes in the tumor microenvironment, providing insights into mechanisms of immune exclusion. Spatially mapping immune cells also provides quantitative information, which could be informative in clinical settings, for example for the discovery of new biomarkers that could guide the design of patient-specific immunotherapies. In this review, we aim to summarize current standard and next generation approaches to define Cancer Immune Topographies based on published studies and propose future perspectives.

Tumor microenvironment derived signature predicting relapse-free survival in I-III cancer and preliminary experiment verification

The recurrence in colon cancer contributed to great difficulties in diagnostic and therapeutic treatment. Tumor microenvironment (TME) gains increasing attention recently. After univariate Cox analysis on relapse-free survival (RFS) and ESTIMATE analysis, WGCNA was further conducted to determine the TME and relapse-related genes in I-III colon cancer. Functional enrichment analyses were conducted. Furthermore, seven genes were screened to build a prognostic signature via LASSO and multivariate Cox analysis. Univariate followed multivariate Cox analysis all showed that the risk group calculated by the signature as a significant predictors. The ROC curves showed great prognostic in the internal training group, internal verification group, and independent external verification group. In the training group, the AUC at 1, 3, and 5 years was 0.737, 0.79, and 0.756. In addition, correlation analysis presented that the signature and genes involved in were significantly associated with the TME. Moreover, 3 of 7 genes (FAM78A, SGIP1, and MMP9) were validated to be associated with PDL1 through qRT-PCR.

Emerging Role of Immunotherapy for Colorectal Cancer with Liver Metastasis

Colorectal cancer (CRC) is the third most common malignant tumor in the world and the second leading cause of cancer-related deaths, with the liver as the most common site of distant metastasis. The prognosis of CRC with liver metastasis is poor, and most patients cannot undergo surgery. In addition, conventional antitumor approaches such as chemotherapy, radiotherapy, targeted therapy, and surgery result in unsatisfactory outcomes. In recent years, immunotherapy has shown good prospects in the treatment of assorted tumors by enhancing the host's antitumor immune function, and it may become a new effective treatment for liver metastasis of CRC. However, challenges remain in applying immunotherapy to CRC with liver metastasis. This review examines how the microenvironment and immunosuppressive landscape of the liver favor tumor progression. It also highlights the latest research advances in immunotherapy for colorectal liver metastasis and identifies immunotherapy as a treatment regimen with a promising future in clinical applications.

Simultaneous Tumor and Stroma Targeting by Oncolytic Viruses

Current cancer therapeutics often insufficiently eradicate malignant cells due to the surrounding dense tumor stroma. This multi-componential tissue consists of mainly cancer-associated fibroblasts, the (compact) extracellular matrix, tumor vasculature, and tumor-associated macrophages, which all exert crucial roles in maintaining a pro-tumoral niche. Their continuous complex interactions with tumor cells promote tumor progression and metastasis, emphasizing the challenges in tumor therapy development. Over the last decade, advances in oncolytic virotherapy have shown that oncolytic viruses (OVs) are a promising multi-faceted therapeutic platform for simultaneous tumor and stroma targeting. In addition to promoting tumor cell oncolysis and systemic anti-tumor immunity, accumulating data suggest that OVs can also directly target stromal components, facilitating OV replication and spread, as well as promoting anti-tumor activity. This review provides a comprehensive overview of the interactions between native and genetically modified OVs and the different targetable tumor stromal components, and outlines strategies to improve stroma targeting by OVs.

Antiangiogenic immunotherapy suppresses desmoplastic and chemoresistant intestinal tumors in mice

Mutations in APC promote colorectal cancer (CRC) progression through uncontrolled WNT signaling. Patients with desmoplastic CRC have a significantly worse prognosis and do not benefit from chemotherapy, but the mechanisms underlying the differential responses of APC-mutant CRCs to chemotherapy are not well understood. We report that expression of the transcription factor prospero homeobox 1 (PROX1) was reduced in desmoplastic APC-mutant human CRCs. In genetic Apc-mutant mouse models, loss of Prox1 promoted the growth of desmoplastic, angiogenic, and immunologically silent tumors through derepression of Mmp14. Although chemotherapy inhibited Prox1-proficient tumors, it promoted further stromal activation, angiogenesis, and invasion in Prox1-deficient tumors. Blockade of vascular endothelial growth factor A (VEGFA) and angiopoietin-2 (ANGPT2) combined with CD40 agonistic antibodies promoted antiangiogenic and immunostimulatory reprogramming of Prox1-deficient tumors, destroyed tumor fibrosis, and unleashed T cell-mediated killing of cancer cells. These results pinpoint the mechanistic basis of chemotherapy-induced hyperprogression and illustrate a therapeutic strategy for chemoresistant and desmoplastic CRCs.

Analysis of Spatial Organization of Suppressive Myeloid Cells and Effector T Cells in Colorectal Cancer-A Potential Tool for Discovering Prognostic Biomarkers in Clinical Research

The development and progression of solid tumors such as colorectal cancer (CRC) are known to be affected by the immune system and cell types such as T cells, natural killer (NK) cells, and natural killer T (NKT) cells are emerging as interesting targets for immunotherapy and clinical biomarker research. In addition, CD3⁺ and CD8⁺ T cell distribution in tumors has shown positive prognostic value in stage I–III CRC. Recent developments in digital computational pathology support not only classical cell density based tumor characterization, but also a more comprehensive analysis of the spatial cell organization in the tumor immune microenvironment (TiME). Leveraging that methodology in the current study, we tried to address the question of how the distribution of myeloid derived suppressor cells in TiME of primary CRC affects the function and location of cytotoxic T cells. We applied multicolored immunohistochemistry to identify monocytic (CD11b⁺CD14⁺) and granulocytic (CD11b⁺CD15⁺) myeloid cell populations together with proliferating and non-proliferating cytotoxic T cells (CD8⁺Ki67^+/–). Through automated object detection and image registration using HALO software (IndicaLabs), we applied dedicated spatial statistics to measure the extent of overlap between the areas occupied by myeloid and T cells. With this approach, we observed distinct spatial organizational patterns of immune cells in tumors obtained from 74 treatment-naive CRC patients. Detailed analysis of inter-cell distances and myeloid-T cell spatial overlap combined with integrated gene expression data allowed to stratify patients irrespective of their mismatch repair (MMR) status or consensus molecular subgroups (CMS) classification. In addition, generation of cell distance-derived gene signatures and their mapping to the TCGA data set revealed associations between spatial immune cell distribution in TiME and certain subsets of CD8⁺ and CD4⁺ T cells. The presented study sheds a new light on myeloid and T cell interactions in TiME in CRC patients. Our results show that CRC tumors present distinct distribution patterns of not only T effector cells but also tumor resident myeloid cells, thus stressing the necessity of more comprehensive characterization of TiME in order to better predict cancer prognosis. This research emphasizes the importance of a multimodal approach by combining computational pathology with its detailed spatial statistics and gene expression profiling. Finally, our study presents a novel approach to cancer patients’ characterization that can potentially be used to develop new immunotherapy strategies, not based on classical biomarkers related to CRC biology.

Omics technologies for improved diagnosis and treatment of colorectal cancer: Technical advancement and major perspectives

Colorectal cancer (CRC) ranks third among the most commonly occurring cancers worldwide, and it causes half a million deaths annually. Alongside mechanistic study for CRC detection and treatment by conventional techniques, new technologies have been developed to study CRC. These technologies include genomics, transcriptomics, proteomics, and metabolomics which elucidate DNA markers, RNA transcripts, protein and, metabolites produced inside the colon and rectum part of the gut. All these approaches form the omics arena, which presents a remarkable opportunity for the discovery of novel prognostic, diagnostic and therapeutic biomarkers and also delineate the underlying mechanism of CRC causation, which may further help in devising treatment strategies. This review also mentions the latest developments in metagenomics and culturomics as emerging evidence suggests that metagenomics of gut microbiota has profound implications in the causation, prognosis, and treatment of CRC. A majority of bacteria cannot be studied as they remain unculturable, so culturomics has also been strengthened to develop culture conditions suitable for the growth of unculturable bacteria and identify unknown bacteria. The overall purpose of this review is to succinctly evaluate the application of omics technologies in colorectal cancer research for improving the diagnosis and treatment strategies.

Development of a prognostic model based on an immunogenomic landscape analysis of colorectal cancer

Screening and therapeutic programs for colorectal cancer (CRC) are invasive or not effective and unable to meet patient needs. Major advances in immunogenomics may change this status but need more exploration. Differentially expressed genes and immune-related genes (IRGs) were identified by computational methods. A prognostic model was established and validated based on survival-related IRGs via stepwise multivariate Cox regression analysis. Nine IRGs were selected and identified as survival-related genes. A 7-gene prognostic model could offer a preliminary and valid determination of risk in CRC patients. The area under the curve of the receiver operating characteristic was 0.672. The 7-gene prognostic model might be used as a novel prognostic tool in CRC patients.

Clinical-Grade Detection of Microsatellite Instability in Colorectal Tumors by Deep Learning

BACKGROUND & AIMS

Microsatellite instability (MSI) and mismatch-repair deficiency (dMMR) in colorectal tumors are used to select treatment for patients. Deep learning can detect MSI and dMMR in tumor samples on routine histology slides faster and less expensively than molecular assays. However, clinical application of this technology requires high performance and multisite validation, which have not yet been performed.

METHODS

We collected H&E-stained slides and findings from molecular analyses for MSI and dMMR from 8836 colorectal tumors (of all stages) included in the MSIDETECT consortium study, from Germany, the Netherlands, the United Kingdom, and the United States. Specimens with dMMR were identified by immunohistochemistry analyses of tissue microarrays for loss of MLH1, MSH2, MSH6, and/or PMS2. Specimens with MSI were identified by genetic analyses. We trained a deep-learning detector to identify samples with MSI from these slides; performance was assessed by cross-validation (N = 6406 specimens) and validated in an external cohort (n = 771 specimens). Prespecified endpoints were area under the receiver operating characteristic (AUROC) curve and area under the precision-recall curve (AUPRC).

RESULTS

The deep-learning detector identified specimens with dMMR or MSI with a mean AUROC curve of 0.92 (lower bound, 0.91; upper bound, 0.93) and an AUPRC of 0.63 (range, 0.59-0.65), or 67% specificity and 95% sensitivity, in the cross-validation development cohort. In the validation cohort, the classifier identified samples with dMMR with an AUROC of 0.95 (range, 0.92-0.96) without image preprocessing and an AUROC of 0.96 (range, 0.93-0.98) after color normalization.

CONCLUSIONS

We developed a deep-learning system that detects colorectal cancer specimens with dMMR or MSI using H&E-stained slides; it detected tissues with dMMR with an AUROC of 0.96 in a large, international validation cohort. This system might be used for high-throughput, low-cost evaluation of colorectal tissue specimens.

Evolutionary dynamics of neoantigens in growing tumors

Cancers accumulate mutations that lead to neoantigens, novel peptides that elicit an immune response, and consequently undergo evolutionary selection. Here we establish how negative selection shapes the clonality of neoantigens in a growing cancer by constructing a mathematical model of neoantigen evolution. The model predicts that, without immune escape, tumor neoantigens are either clonal or at low frequency; hypermutated tumors can only establish after the evolution of immune escape. Moreover, the site frequency spectrum of somatic variants under negative selection appears more neutral as the strength of negative selection increases, which is consistent with classical neutral theory. These predictions are corroborated by the analysis of neoantigen frequencies and immune escape in exome and RNA sequencing data from 879 colon, stomach and endometrial cancers.

The Unmet Needs of the Diagnosis, Staging, and Treatment of Gastrointestinal Tumors

New scientific insights in cancer biology and immunobiology have changed the clinical practice of medical oncology in recent years. The molecular stratification of solid tumors has led to improved clinical outcomes and is a key part in the diagnostic workup. Beyond mutational spectra (like Rat sarcoma [RAS] mutations or tumor mutational burden), the investigation of the immunological microenvironment has attracted more efforts. Especially as immunotherapies have changed the standard treatment for some solid tumors dramatically and have become an important part of routine oncology, also for gastrointestinal tumors. Still only a subgroup of patients benefits from immunotherapy in gastrointestinal tumors with prominent examples from colorectal, pancreatic or gastric cancer. Not only microsatellite instability as a marker for response to immunotherapy has shown its utility, there plenty of other approaches currently being investigated to better stratify and understand the microenvironment. But these insights have not translated into clinical utility. Reasons for this are limited technical capabilities for stratification and for coping with heterogeneity of tumor cells and the microenvironment as such. So the situation for gastrointestinal tumors has shown mainly progress for a subgroup of immunotherapy-receptive tumors (eg, microsatellite instability), but advances for the remaining majority have been in the area of stratification and combinatorial therapies, including approaches without chemotherapy. Molecular stratification (eg, B-Rapidly Accelerated Fibrosarcoma [BRAF] V600E mutation in colorectal cancer or NRG1 fusions in Kirsten-rat sarcoma (KRAS) Wild-Type Pancreatic Cancer) has clearly improved the possibilities for directed therapies, but there is a plethora of clinical situations where further developments are needed to improve patient care. Finding these areas and identifying the technical approach to unravel the complexities is the next decisive step. Here the recent advances are summarized and an outlook on possible diagnostic and treatment options in areas of unmet need is given with the context of new molecular imaging possibilities and cutting edge advances in nuclear medicine.

An immune landscape based prognostic signature predicts the immune status and immunotherapeutic responses of patients with colorectal cancer

AIMS

Colorectal cancer (CRC) is one of the most common cancers with poor prognosis worldwide. The advent of immunotherapy has greatly improved survival in refractory patients of CRC. In this study, we aimed to identify reliable immune classification and biomarkers that predict immunotherapeutic responses in CRC patients.

MATERIALS AND METHODS

Based on transcriptome profiles of two publicly available CRC datasets, we performed single-sample gene set enrichment analysis (ssGSEA) to calculate the relative abundance of 29 immune-related items of each sample. Unsupervised clustering was used to classify CRC patients. Furthermore, an immune prognostic signature was constructed using the least absolute shrinkage and selection operator (LASSO) Cox regression analysis.

KEY FINDINGS

The CRC patients were clustered into high, medium, low immune infiltration subtypes based on the immune landscape. There was significant heterogeneity among the three subtypes. The high immune infiltration group showed higher expression of programmed cell death-ligand 1 and better prognosis than the median and low immune infiltration groups. Furthermore, we constructed a 7-immune-related prognostic gene signature and found that the signature had high predictive value and was superior to other clinicopathological parameters. Finally, the correlation analysis of the signature with immune cell infiltration and immune checkpoint molecules suggested that the signature had the potential to assess the immunotherapeutic responses of CRC patients.

SIGNIFICANCE

The immune landscape and prognostic signature of CRC contribute to a deeper understanding of the tumor microenvironment and guide accurate immunotherapy.

A deep learning model to predict RNA-Seq expression of tumours from whole slide images

Deep learning methods for digital pathology analysis are an effective way to address multiple clinical questions, from diagnosis to prediction of treatment outcomes. These methods have also been used to predict gene mutations from pathology images, but no comprehensive evaluation of their potential for extracting molecular features from histology slides has yet been performed. We show that HE2RNA, a model based on the integration of multiple data modes, can be trained to systematically predict RNA-Seq profiles from whole-slide images alone, without expert annotation. Through its interpretable design, HE2RNA provides virtual spatialization of gene expression, as validated by CD3- and CD20-staining on an independent dataset. The transcriptomic representation learned by HE2RNA can also be transferred on other datasets, even of small size, to increase prediction performance for specific molecular phenotypes. We illustrate the use of this approach in clinical diagnosis purposes such as the identification of tumors with microsatellite instability.

RNA-sequencing of tumour tissue can provide important diagnostic and prognostic information but this is costly and not routinely performed in all clinical settings. Here, the authors show that whole slide histology slides—part of routine care—can be used to predict RNA-sequencing data and thus reduce the need for additional analyses.

A novel prognostic signature of immune-related genes for patients with colorectal cancer

Colorectal cancer (CRC) is one of the most commonly diagnosed cancers with an estimated 1.8 million new cases worldwide and associated with high mortality rates of 881 000 CRC-related deaths in 2018. Screening programs and new therapies have only marginally improved the survival of CRC patients. Immune-related genes (IRGs) have attracted attention in recent years as therapeutic targets. The aim of this study was to identify an immune-related prognostic signature for CRC. To this end, we combined gene expression and clinical data from the CRC data sets of The Cancer Genome Atlas (TCGA) into an integrated immune landscape profile. We identified a total of 476 IRGs that were differentially expressed in CRC vs normal tissues, of which 18 were survival related according to univariate Cox analysis. Stepwise multivariate Cox proportional hazards analysis established an immune-related prognostic signature consisting of SLC10A2, FGF2, CCL28, NDRG1, ESM1, UCN, UTS2 and TRDC. The predictive ability of this signature for 3- and 5-year overall survival was determined using receiver operating characteristics (ROC), and the respective areas under the curve (AUC) were 79.2% and 76.6%. The signature showed moderate predictive accuracy in the validation and GSE38832 data sets as well. Furthermore, the 8-IRG signature correlated significantly with tumour stage, invasion, lymph node metastasis and distant metastasis by univariate Cox analysis, and was established an independent prognostic factor by multivariate Cox regression analysis for CRC. Gene set enrichment analysis (GSEA) revealed a relationship between the IRG prognostic signature and various biological pathways. Focal adhesions and ECM-receptor interactions were positively correlated with the risk scores, while cytosolic DNA sensing and metabolism-related pathways were negatively correlated. Finally, the bioinformatics results were validated by real-time RT-qPCR. In conclusion, we identified and validated a novel, immune-related prognostic signature for patients with CRC, and this signature reflects the dysregulated tumour immune microenvironment and has a potential for better CRC patient management.

Pan-cancer image-based detection of clinically actionable genetic alterations

Molecular alterations in cancer can cause phenotypic changes in tumor cells and their micro-environment. Routine histopathology tissue slides - which are ubiquitously available - can reflect such morphological changes. Here, we show that deep learning can consistently infer a wide range of genetic mutations, molecular tumor subtypes, gene expression signatures and standard pathology biomarkers directly from routine histology. We developed, optimized, validated and publicly released a one-stop-shop workflow and applied it to tissue slides of more than 5000 patients across multiple solid tumors. Our findings show that a single deep learning algorithm can be trained to predict a wide range of molecular alterations from routine, paraffin-embedded histology slides stained with hematoxylin and eosin. These predictions generalize to other populations and are spatially resolved. Our method can be implemented on mobile hardware, potentially enabling point-of-care diagnostics for personalized cancer treatment. More generally, this approach could elucidate and quantify genotype-phenotype links in cancer.

Colorectal Cancer Immune Infiltrates: Significance in Patient Prognosis and Immunotherapeutic Efficacy

Colorectal cancer occurrence and progression involve multiple aspects of host immune deficiencies. In these events, immune cells vary their phenotypes and functions over time, thus enabling the immune microenvironment to be “tumor-inhibiting” as well as “tumor-promoting” as a whole. Because of the association of tumoricidal T cell infiltration with favorable survival in cancer patients, the Immunoscore system was established. Critically, the tumoral Immunoscore serves as an indicator of CRC patient prognosis independent of patient TNM stage and suggests that patients with high Immunoscores in their tumors have prolonged survival in general. Accordingly, stratifications according to tumoral Immunoscores provide new insights into CRC in terms of comparing disease severity, forecasting disease progression, and making treatment decisions. An important application of this system will be to shed light on candidate selection in immunotherapy for CRC, because the T cells responsible for determining the Immunoscore serve as responders to immune checkpoint inhibitors. However, the Immunoscore system merely provides a standard procedure for identifying the tumoral infiltration of cytotoxic and memory T cells, while information concerning the survival and function of these cells is still absent. Moreover, other infiltrates, such as dendritic cells, macrophages, and B cells, can still influence CRC prognosis, implying that those might also influence the therapeutic efficacy of immune checkpoint inhibitors. On these bases, this review is designed to introduce the Immunoscore system by presenting its clinical significance and application in CRC.

Potential Therapeutic Targets of B7 Family in Colorectal Cancer

Programmed cell death protein 1 (PD-1)/programmed death ligand 1 (PD-L1) pathway blockade has impressively benefited cancer patients with a wide spectrum of tumors. However, its efficacy in colorectal cancer (CRC) is modest, and only a small subset of patients benefits from approved checkpoint inhibitors. Newer checkpoints that target additional immunomodulatory pathways are becoming necessary to activate durable antitumor immune responses in patients with CRC. In this review, we evaluated the mRNA expression of all 10 reported B7 family members in human CRC by retrieving and analyzing the TCGA database and reviewed the current understanding of the top three B7 family checkpoint molecules (B7-H3, VISTA, and HHLA2) with the highest mRNA expression, introducing them as putative therapeutic targets in CRC.

B7-H3 inhibits the IFN-γ-dependent cytotoxicity of Vγ9Vδ2 T cells against colon cancer cells

The immunoregulatory protein B7-H3, a member of the B7 family, has been confirmed to be highly expressed in colon cancer. However, the exact influence of B7-H3 on the features and antitumor ability of γδT cells in colon cancer remains unknown. In the present study, we investigated that the proportions of B7-H3⁺ γδT cells were distinctly increased in the peripheral blood and tumor tissues of colon cancer patients. B7-H3 blockade or knockdown promoted proliferation, inhibited cell apoptosis and induced the expression of activation markers (CD25 and CD69) on Vδ2 T cells. In contrast, treatment with the B7-H3 agonist 4H7 had the opposite effect. Furthermore, B7-H3 suppressed IFN-γ expression by inhibiting T-bet in Vδ2 T cells. Moreover, B7-H3 mediated the inhibition of Vδ2 T cell cytotoxicity via the downregulation of IFN-γ and perforin/granzyme B expression. More importantly, blocking the B7-H3 function significantly enhanced the cytotoxicity of Vδ2 T cells against colon cancer cells in vivo. Therefore, the inhibition or blockade of B7-H3 is a potential immunotherapeutic approach for colon cancer.

The role of the innate immune system in the development and treatment of hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is the most common primary liver cancer. Most patients present with advanced or metastatic HCC at diagnosis and face a dismal prognosis. Tyrosine kinases are the gold standard treatment for this disease but yield limited survival benefits. Immune checkpoint inhibitors that augment adaptive immunity have been tested in HCC. Complex interactions between tumor cells, lymphocytes and the tumor environment determine the efficacy of such immunotherapies. Innate immune mechanisms – known drivers of liver disease progression in pre-HCC conditions such as fibrosis or cirrhosis – may either support or counteract tumor-related immune activation. In this review, we will highlight current concepts of the role of the innate immune system in hepatocarcinogenesis and discuss their relevance for translation into clinics.

The Association Between Inflammation, Epithelial Mesenchymal Transition and Stemness in Colorectal Carcinoma

Background

Inflammation plays an important albeit dual role in carcinogenesis. Survival studies have highlighted the prognostic significance of peritumorous inflammation. Currently, the theoretical background allows inflammation, epithelial mesenchymal transition (EMT) and the closely associated stem cell differentiation in colorectal carcinoma (CRC) to be linked. However, there is scarce direct morphological evidence.

Purpose and methods

The aim of our study was to investigate the role of inflammation in cancer growth and invasion by analyzing the association between inflammation and known morphological prognostic features of colorectal cancer, EMT, stemness and mismatch repair (MMR) protein expression. The study was designed as a retrospective morphological and immunohistochemical assessment of 553 consecutive cases of surgically treated primary CRC.

Results

There were statistically significant associations between high-grade inflammation and lower pT (p = 0.002), absence of lymph node metastases (p < 0.001) and less frequent lymphatic (p = 0.003), venous (p = 0.017), arterial (p = 0.012), perineural (p = 0.001) and intraneural (p = 0.01) invasion. In contrast, Crohn's like reaction (CLR) by density of lymphoid follicles in the invasive front lacked significant differences in regard to pT, pN, tumor invasion into surrounding structures (blood or lymphatic vessels, nerves), grade or necrosis (all p > 0.05). The expression of E-cadherin, CD44 and MMR proteins yielded no statistically significant associations with peritumorous inflammation by Klintrup-Mäkinen score or the density of lymphoid follicles. Nevertheless, E-cadherin levels were significantly associated with the density of eosinophils (p = 0.007).

Conclusion

High-grade peritumorous inflammation is associated with beneficial morphologic CRC features, including less frequent manifestations of invasion, and is not secondary to tissue damage and necrosis. CLR is not associated with cancer spread by pTN; this finding indirectly suggests an independent role of CLR in carcinogenesis. Further, inflammation by Klintrup-Mäkinen grade and CLR is not dependent on epithelial-mesenchymal transition and stem cell differentiation. Our study highlights the complex associations between inflammation, tumor morphology, EMT, stemness and MMR protein expression in human CRC tissues.

Perspectives on Treatment of Metastatic Colorectal Cancer with Immune Checkpoint Inhibitor Therapy

Despite lengthening survival, death rates from metastatic colorectal cancer (CRC) remain unacceptably high, with a bright spot being the demonstration of durable responses in patients with CRC who have mismatch repair-deficient (dMMR) and/or microsatellite instability-high (MSI-H) tumors and are treated with immune checkpoint inhibitor therapy. Nivolumab and pembrolizumab, as well as nivolumab in combination with low-dose ipilimumab-all checkpoint inhibitors-are currently approved by the U.S. Food and Drug Administration (FDA) for patients with MSI-H/dMMR metastatic CRC that progressed following treatment with a fluoropyrimidine, oxaliplatin, and irinotecan. Nonetheless, there are a number of questions and considerations in the use of these checkpoint inhibitor therapies. Using a question-and-answer format, this review summarizes the scientific rationale for immune checkpoint inhibitor therapy in CRC, including the effects of tumor factors such as genetic aberrations and mutational load on the immune response, particularly in patients with MSI-H/dMMR disease. We discuss response patterns, response criteria, and immune-related adverse events using findings from published efficacy and safety data of immune checkpoint inhibitor therapy in metastatic CRC. We also discuss issues surrounding treatment sequencing, incorporating approved checkpoint inhibitors into the current treatment paradigm, and the multiple investigational strategies that may optimize immunotherapy for advanced CRC in the future, including novel combination therapies. IMPLICATIONS FOR PRACTICE: Colorectal cancer (CRC) is the third most common cancer in the U.S. Despite advances in chemotherapy, survival remains poor for patients with metastatic CRC. Certain immunotherapy agents have demonstrated long-lasting responses in previously treated patients with immune-responsive microsatellite instability-high/mismatch repair-deficient metastatic CRC, leading to U.S. Food and Drug Administration approval of the immune checkpoint inhibitors nivolumab (with or without low-dose ipilimumab) and pembrolizumab in this population. Combination therapy (e.g., nivolumab with low-dose ipilimumab) has demonstrated numerically higher response rates and improved long-term clinical benefit relative to anti-programmed death-1 monotherapy. Ongoing trials are evaluating immunotherapy in the broader CRC population and novel combinations to optimize immunotherapy for advanced CRC.

The Paradox of Cancer Immune Exclusion: Immune Oncology Next Frontier

Checkpoint inhibitor therapy (CIT) has revolutionized cancer treatment but it has also reached a standstill when an absent dialog between cancer and immune cells makes it irrelevant. This occurs with high prevalence in the context of "immune silent" and, even perhaps, "immune-excluded" tumors. The latter are characterized by T cells restricted to the periphery of cancer nests. Since in either case T cells do not come in direct contact with most cancer cells, CIT rests immaterial. Adoptive cell therapy (ACT), may also be affected by limited access to antigen-bearing cancer cells. While lack of immunogenicity intuitively explains the immune silent phenotype, immune exclusion is perplexing. The presence of T cells at the periphery suggests that chemo-attraction recruits them and an immunogenic stimulus promotes their persistence. However, what stops the T cells from infiltrating the tumors' nests and reaching the germinal center (GC)? Possibly, a concentric gradient of increased chemo-repulsion or decreased chemo-attraction demarcates an abrupt "do not trespass" warning. Various hypotheses suggest physical or functional barriers but no definitive consensus exists over the weight that each plays in human cancers. On one hand, it could be hypothesized that the intrinsic biology of cancer cells may degenerate from a "cancer stem cell" (CSC)-like phenotype in the GC toward a progressively more immunogenic phenotype prone to immunogenic cell death (ICD) at the periphery. On the other hand, the intrinsic biology of the cancer cells may not change but it is the disorderly architecture of the tumor microenvironment (TME) that alters in a centripetal direction cancer cell metabolism, both directly and indirectly, the function of surrounding stromal cells. In this chapter, we examine whether the paradoxical exclusion of T cells from tumors may serve as a model to understand the requirements for tumor immune infiltration and, correspondingly, we put forth strategies to restore the dialog between immune cells and cancer to enhance the effectiveness of immune oncology (IO) approaches.

Dickkopf-related protein 1, a new biomarker for local immune status and poor prognosis among patients with colorectal liver Oligometastases: a retrospective study

Background

It was reported that tumor-expressed dickkopf-related (DKK) proteins affect micro-environment. However, the influence of DKK1 on colorectal cancer (CRC) liver oligometastases (CRCLOM) remains unclear.

Methods

CRC cases after resection of liver oligometastases were enrolled in Sun Yat-Sen University Cancer Center with intact clinical data. Serum DKK1 was detected by ELISA assay. Immunofluorescent staining examination for CD3 and CD8 in slices were also conducted.

Results

Among 65 patients included, the recurrence-free survival (RFS) and overall survival (OS) were significantly better in the low serum DKK1 group (RFS: P = 0.021; OS: P = 0.043). DKK1 was overexpressed in stage IV CRC patients in TCGA data. The number of CD8+ tumor-infiltrating lymphocytes (TILs) in invasive margin of CRC liver oligometastases was significantly higher in low serum DKK1 group (P = 0.042).

Conclusion

Elevated serum DKK1 level was associated with poorer RFS and OS, and less CD8+ TILs in invasive margin in CRC liver oligometastases. DKK1 might serve as a supplementalprognostic factor for clinical risk score and a potential target for immunotherapy.

Macrophage-derived CCL5 facilitates immune escape of colorectal cancer cells via the p65/STAT3-CSN5-PD-L1 pathway

Infiltrated macrophages are an important constituent of the tumor microenvironment and play roles in tumor initiation and progression by promoting immune evasion. However, the molecular mechanism by which macrophage-derived cytokines foster immune escape of colorectal cancer (CRC) is unclear. Here, we demonstrated that macrophage infiltration induced by lipopolysaccharide (LPS) or a high-cholesterol diet (HCD) significantly promoted CRC growth. Similarly, LPS and poly (I:C) remarkably increased the volume of CT26 cell allograft tumors. C-C motif chemokine ligand 5 (CCL5), which is secreted by macrophages, inhibited T-cell-mediated killing of HT29 cells and promoted immune escape by stabilizing PD-L1 in vitro and in vivo. Mechanistically, CCL5 resulted in formation of nuclear factor kappa-B p65/STAT3 complexes, which bound to the COP9 signalosome 5 (CSN5) promoter, leading to its upregulation. Moreover, CSN5 modulated the deubiquitination and stability of PD-L1. High expression of CSN5 in CRC was associated with significantly shorter survival. Furthermore, compound-15 was identified as an inhibitor of CSN5, and destabilized PD-L1 to alleviate the tumor burden. Our results suggest that the novel CCL5-p65/STAT3-CSN5-PD-L1 signaling axis is significantly activated by LPS or HCD-driven macrophage infiltration in an animal model of CRC, which likely has therapeutic and prognostic implications for human cancers.

Impact of Prior Bevacizumab Treatment on VEGF-A and PlGF Levels and Outcome Following Second-Line Aflibercept Treatment: Biomarker Post Hoc Analysis of the VELOUR Trial

PURPOSE

Aflibercept is a targeted anti-VEGF therapy used to treat patients with metastatic colorectal cancer (mCRC) following progression on oxaliplatin-based regimens. This post hoc study evaluated the effect of prior bevacizumab treatment and growth factor levels on patient outcomes associated with aflibercept in the VELOUR phase III trial.

EXPERIMENTAL DESIGN

Baseline biomarker plasma concentrations were measured using a bead-based multiplex assay. Patients were grouped according to prior bevacizumab treatment, second-line treatment, and serum biomarker concentrations, and analyzed for overall survival (OS) and progression-free survival (PFS).

RESULTS

Plasma samples were available for 553 patients (placebo n = 265; aflibercept n = 288), of which 169 had received prior bevacizumab. Nine biomarkers implicated in angiogenesis or bevacizumab resistance correlated with prior bevacizumab therapy. VEGF-A and placental growth factor (PlGF) were the most significantly increased in patients who had received prior bevacizumab compared with those who had not received prior bevacizumab. In the placebo group, patients with high VEGF-A (>144 pg/mL) levels at baseline had worse OS and PFS compared with patients with lower levels at baseline (9.6 vs. 12.9 months). This was also seen in patients who received placebo and had high baseline PlGF (>8 pg/mL; 9.7 vs. 11.7 months). In the aflibercept group, prolonged OS and PFS were observed regardless of baseline VEGF-A or PlGF levels.

CONCLUSIONS

High VEGF-A and PlGF serum levels may underlie development of resistance to bevacizumab in patients with mCRC. Aflibercept retains its activity regardless of baseline VEGF-A and PlGF levels and may be an effective second-line treatment for patients with bevacizumab-induced resistance.

Tim-3 suppresses the killing effect of Vγ9Vδ2 T cells on colon cancer cells by reducing perforin and granzyme B expression

Gamma delta (γδ) T cell-based tumor immunotherapy has been one of the most promising cancer immunotherapeutic strategies. However, the key regulators of the Vγ9Vδ2 T cell-mediated antitumor response remain unclear. Recently, mounting reports have indicated that Tim-3 performs critical roles in the regulation of the activities of immune cells, including Vγ9Vδ2 T cells. However, the roles of Tim-3 in Vγ9Vδ2 T cell-mediated killing of colon cancer cells and the underlying mechanism remain largely unknown. Here, the proportion of Tim-3+ γδ T cells was significantly increased in both the peripheral blood and colon cancer tissue of patients and was significantly associated with TNM staging and tumor volume. Additionally, the activation of Tim-3 signaling significantly inhibited the killing efficiency of Vγ9Vδ2 T cells against colon cancer cells. In addition, Tim-3 signaling reduced the expression of perforin and granzyme B in Vγ9Vδ2 T cells. Blocking the perforin/granzyme B pathway also decreased the cytotoxicity of Vγ9Vδ2 T cells to colon cancer cells. Moreover, Tim-3 signaling reduced the perforin and granzyme B expression of Vγ9Vδ2 T cells in an ERK1/2 signaling pathway-dependent manner. This knowledge reveals that Tim-3 may be a promising therapeutic target to improve Vγ9Vδ2 T cell-based adoptive immunotherapy for colon cancer.

miR-375-3p suppresses tumorigenesis and partially reverses chemoresistance by targeting YAP1 and SP1 in colorectal cancer cells

Clinically, one of the principal factors in the failure of advanced colorectal cancer (CRC) treatment is chemoresistance to 5-fluorouracil (5FU)-based chemotherapy. Although microRNA-375-3p (miR-375) is considered a tumor suppressor in multiple cancers, the mechanism of miR-375 in the regulation of drug resistance in CRC remains unclear. In this study, we investigated the chemosensitivity of miR-375 to 5FU in CRC from biological and clinical aspects. We found that miR-375 was significantly downregulated in CRC tissues and cell lines, and low miR-375 expression was strongly correlated with poor overall survival in CRC patients. Overexpression of miR-375 sensitized CRC cells to a broad spectrum of chemotherapeutic drugs in vitro and in vivo. Further mechanistic analysis demonstrated that miR-375 enhanced CRC cell sensitivity to 5FU by directly targeting YAP1 and SP1. MiR-375 downregulated YAP1, resulting in reduced expression of the Hippo-YAP1 pathway downstream genes CTGF, cyclin D1 and BIRC5 (also known as survivin). Overall, miR-375 was confirmed as a prospective molecular biomarker in the chemoresistance and prognosis of CRC patients, and the synergy between miR-375 and chemotherapeutic drugs could be a promising therapeutic strategy for CRC patients, especially with chemoresistance.

Primary Tumor Sidedness Predicts Bevacizumab Benefit in Metastatic Colorectal Cancer Patients

The emerging debate between primary tumor location and clinical outcome of bevacizumab treated metastatic colorectal cancer (mCRC) continues. The aim of the present study is to investigate the association between the primary tumor location and clinical outcome of 115 mCRC patients receiving bevacizumab based treatment. A meta-analysis including 21 studies was carried out to confirm the conclusion. In our prospective study, we found that right-sided mCRC commonly occurred in older cases (p = 0.03) with multiple-site metastasis (p = 0.03). Progression-free survival (PFS) of the left-sided patients undergoing bevacizumab plus a FOLFIRI regimen was superior to the right-sided cases (p = 0.03, crude HR = 0.31, 95%CI = 0.11-0.87; adjusted HR = 0.21, 95%CI = 0.06-0.66). The meta-analysis confirmed that efficacy of bevacizumab-based treatment in left-sided mCRC patients was better than the right-sided cases in the overall population (P h = 0.24, combined OR = 1.36, 95%CI = 1.07-1.72), RAS/BRAF wild-type (P h = 0.19, combined OR = 1.66, 95%CI = 1.17-2.34), clinical trial (P h = 0.23, combined OR = 1.42, 95%CI = 1.07-1.88), Caucasian population (P h = 0.18, combined OR = 1.37, 95%CI = 1.02-1.85) and first-line (P h = 0.19, combined OR = 1.48, 95%CI = 1.13-1.96) subgroups. Improved survival of bevacizumab plus chemotherapy treated left-sided mCRC patients was observed in the overall population [P h < 0.01, combined MSR = 1.09, 95%CI = 1.00-1.18 for PFS; P h < 0.01, combined MSR = 1.24, 95%CI = 1.13-1.36 for overall survival (OS)], especially in the RAS/BRAF wild-type (P h = 0.09, combined MSR = 1.10, 95%CI = 1.03-1.19 for PFS; P h = 0.02, combined MSR = 1.34, 95%CI = 1.21-1.49 for OS). These findings indicate that primary tumor sidedness can predict clinical outcome of bevacizumab-treated RAS/BRAF wild-type mCRC patients and the left-sided patients may benefit more from bevacizumab plus FOLFIRI.

Impact of Consensus Molecular Subtype on Survival in Patients With Metastatic Colorectal Cancer: Results From CALGB/SWOG 80405 (Alliance)

PURPOSE

To determine the predictive and prognostic value of the consensus molecular subtypes (CMSs) of colorectal cancer (CRC) that represent a merging of gene expression-based features largely in primary tumors from six independent classification systems and provide a framework for capturing the intrinsic heterogeneity of CRC in patients enrolled in CALGB/SWOG 80405.

PATIENTS AND METHODS

CALGB/SWOG 80405 is a phase III trial that compared the addition of bevacizumab or cetuximab to infusional fluorouracil, leucovorin, and oxaliplatin or fluorouracil, leucovorin, and irinotecan as first-line treatment of advanced CRC. We characterized the CMS classification using a novel NanoString gene expression panel on primary CRCs from 581 patients enrolled in this study to assess the prognostic and predictive value of CMSs in these patients.

RESULTS

The CMSs are highly prognostic for overall survival (OS; P < .001) and progression-free survival (PFS; P < .001). Furthermore, CMSs were predictive for both OS (P for interaction < .001) and PFS (P for interaction = .0032). In the CMS1 cohort, patients treated with bevacizumab had a significantly longer OS than those treated with cetuximab (P < .001). In the CMS2 cohort, patients treated with cetuximab had a significantly longer OS than patients treated with bevacizumab (P = .0046).

CONCLUSION

These findings highlight the possible clinical utility of CMSs and suggests that refinement of the CMS classification may provide a path toward identifying patients with metastatic CRC who are most likely to benefit from specific targeted therapy as part of the initial treatment.

Optimizing oncolytic virotherapy in cancer treatment

In the wake of the success of modern immunotherapy, oncolytic viruses (OVs) are currently seen as a potential therapeutic option for patients with cancer who do not respond or fail to achieve durable responses following treatment with immune checkpoint inhibitors. OVs offer a multifaceted therapeutic platform because they preferentially replicate in tumour cells, can be engineered to express transgenes that augment their cytotoxic and immunostimulatory activities, and modulate the tumour microenvironment to optimize immune-mediated tumour eradication, both at locoregional and systemic sites of disease. Lysis of tumour cells releases tumour-specific antigens that trigger both the innate and adaptive immune systems. OVs also represent attractive combination partners with other systemically delivered agents by virtue of their highly favourable safety profiles. Rational combinations of OVs with different immune modifiers and/or antitumour agents, based on mechanisms of tumour resistance to immune-mediated attack, may benefit the large, currently underserved, population of patients who respond poorly to immune checkpoint inhibition.

Deep learning can predict microsatellite instability directly from histology in gastrointestinal cancer

Microsatellite instability determines whether patients with gastrointestinal cancer respond exceptionally well to immunotherapy. However, in clinical practice, not every patient is tested for MSI, because this requires additional genetic or immunohistochemical tests. Here we show that deep residual learning can predict MSI directly from H&E histology, which is ubiquitously available. This approach has the potential to provide immunotherapy to a much broader subset of patients with gastrointestinal cancer.

The Interplay of Autophagy and Tumor Microenvironment in Colorectal Cancer-Ways of Enhancing Immunotherapy Action

Autophagy as a primary homeostatic and catabolic process is responsible for the degradation and recycling of proteins and cellular components. The mechanism of autophagy has a crucial role in several cellular functions and its dysregulation is associated with tumorigenesis, tumor-stroma interactions, and resistance to cancer therapy. A growing body of evidence suggests that autophagy is also a key regulator of the tumor microenvironment and cellular immune response in different types of cancer, including colorectal cancer (CRC). Furthermore, autophagy is responsible for initiating the immune response especially when it precedes cell death. However, the role of autophagy in CRC and the tumor microenvironment remains controversial. In this review, we identify the role of autophagy in tumor microenvironment regulation and the specific mechanism by which autophagy is implicated in immune responses during CRC tumorigenesis and the context of anticancer therapy.

Harnessing the innate immune system and local immunological microenvironment to treat colorectal cancer

Significant progress in the development of new immunotherapies has led to successful clinical trials for malignant melanoma and non-small cell lung cancer; however, for the majority of solid tumours of the gastrointestinal tract, little or no progress has been seen. The efficacy of immunotherapies is limited by the complexities of a diverse set of immune cells, and interactions between the tumour cells and all other cells in the local microenvironment of solid tumours. A large fraction of immune cells present in and around solid tumours derive from the innate arm of the immune system and using these cells against tumours offers an alternative immunotherapeutic option, especially as current strategies largely harness the adaptive arm of the immune system. This option is currently being investigated and attempts at using the innate immune system for gastrointestinal cancers are showing initial results. Several important factors, including cytokines, chemotherapeutics and the microbiome, influence the plasticity and functionality of innate (myeloid) cells in the microenvironment, and this complexity of regulation has limited translation into successful trials so far. In this review, current concepts of the immunobiology of the innate arm in the tumour microenvironment are presented in the context of clinical translation.

Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study

BACKGROUND

For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images.

METHODS AND FINDINGS

We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I-IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a "deep stroma score," which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27-3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I-IV CRC patients from the "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14-2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5-3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34-2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows.

CONCLUSIONS

In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.

Topography of cancer-associated immune cells in human solid tumors

Lymphoid and myeloid cells are abundant in the tumor microenvironment, can be quantified by immunohistochemistry and shape the disease course of human solid tumors. Yet, there is no comprehensive understanding of spatial immune infiltration patterns (‘topography’) across cancer entities and across various immune cell types. In this study, we systematically measure the topography of multiple immune cell types in 965 histological tissue slides from N = 177 patients in a pan-cancer cohort. We provide a definition of inflamed (‘hot’), non-inflamed (‘cold’) and immune excluded patterns and investigate how these patterns differ between immune cell types and between cancer types. In an independent cohort of N = 287 colorectal cancer patients, we show that hot, cold and excluded topographies for effector lymphocytes (CD8) and tumor-associated macrophages (CD163) alone are not prognostic, but that a bivariate classification system can stratify patients. Our study adds evidence to consider immune topographies as biomarkers for patients with solid tumors.

High-Throughput Screening of Combinatorial Immunotherapies with Patient-Specific In Silico Models of Metastatic Colorectal Cancer

Solid tumors are rich ecosystems of numerous different cell types whose interactions lead to immune escape and resistance to immunotherapy in virtually all patients with metastatic cancer. Here, we have developed a 3D model of human solid tumor tissue that includes tumor cells, fibroblasts, and myeloid and lymphoid immune cells and can represent over a million cells over clinically relevant timeframes. This model accurately reproduced key features of the tissue architecture of human colorectal cancer and could be informed by individual patient data, yielding in silico tumor explants. Stratification of growth kinetics of these explants corresponded to significantly different overall survival in a cohort of patients with metastatic colorectal cancer. We used the model to simulate the effect of chemotherapy, immunotherapies, and cell migration inhibitors alone and in combination. We classified tumors according to tumor and host characteristics, showing that optimal treatment strategies markedly differed between these classes. This platform can complement other patient-specific ex vivo models and can be used for high-throughput screening of combinatorial immunotherapies.Significance: This patient-informed in silico tumor growth model allows testing of different cancer treatment strategies and immunotherapies on a cell/tissue level in a clinically relevant scenario. Cancer Res; 78(17); 5155-63. ©2018 AACR.