Genomics and emerging biomarkers for immunotherapy of colorectal cancer

Integrated analysis reveals a novel 5-fluorouracil resistance-based prognostic signature with promising implications for predicting the efficacy of chemotherapy and immunotherapy in patients with colorectal cancer

BACKGROUND

5-Fluorouracil (5-FU) has been used as a standard first-line treatment for colorectal cancer (CRC) patients. Although 5-FU-based chemotherapy and immune checkpoint blockade (ICB) have achieved success in treating CRC, drug resistance and low response rates remain substantial limitations. Thus, it is necessary to construct a 5-FU resistance-related signature (5-FRSig) to predict patient prognosis and identify ideal patients for chemotherapy and immunotherapy.

METHODS

Using bulk and single-cell RNA sequencing data, we established and validated a novel 5-FRSig model using stepwise regression and multiple CRC cohorts and evaluated its associations with the prognosis, clinical features, immune status, immunotherapy, neoadjuvant therapy, and drug sensitivity of CRC patients through various bioinformatics algorithms. Unsupervised consensus clustering was performed to categorize the 5-FU resistance-related molecular subtypes of CRC. The expression levels of 5-FRSig, immune checkpoints, and immunoregulators were determined using quantitative real-time polymerase chain reaction (RT‒qPCR). Potential small-molecule agents were identified via Connectivity Map (CMap) and molecular docking.

RESULTS

The 5-FRSig and cluster were confirmed as independent prognostic factors in CRC, as patients in the low-risk group and Cluster 1 had a better prognosis. Notably, 5-FRSig was significantly associated with 5-FU sensitivity, chemotherapy response, immune cell infiltration, immunoreactivity phenotype, immunotherapy efficiency, and drug selection. We predicted 10 potential compounds that bind to the core targets of 5-FRSig with the highest affinity.

CONCLUSION

We developed a valid 5-FRSig to predict the prognosis, chemotherapeutic response, and immune status of CRC patients, thus optimizing the therapeutic benefits of chemotherapy combined with immunotherapy, which can facilitate the development of personalized treatments and novel molecular targeted therapies for patients with CRC.

PD-1/PD-L1 inhibitors for early and middle stage microsatellite high-instability and stable colorectal cancer: a review

BACKGROUND

Programmed cell death receptor 1 (PD-1) and programmed cell death ligand 1 (PD-L1) are important immune checkpoint molecules that contribute to tumor immune evasion. However, the main treatment modalities for patients with early and intermediate stage colorectal cancer (CRC) are surgery, and the role of PD-1/PD-L1 inhibitors in these patients is not yet clear. Therefore, this study aims to review the treatment progress of PD-1/PD-L1 inhibitors for early- and intermediate-stage microsatellite high-instability (MSI-H) and stable (MSS) colorectal cancer, in order to provide more options for patients with early- and intermediate-stage colorectal cancer.

MATERIALS AND METHODS

A scoping review of clinical trial registries ( Clinicaltrials.gov and EU clinical trial registers) and PubMed/Medline database of trials on PD-1/PD-L1 Inhibitors for early and middle-stage MSI-H and MSS CRC was done up to March 2024.

RESULTS

A total of 19 trials related to early to mid-stage MSH-I or MSS CRC were included. Among them, 6 trials are in recruiting status, 3 trials are in active, not recruiting status, 3 trials are completed, 1 trial is terminated, and 1 trial is unknown. Of these, 9 trials involve MSI-H type CRC, and 10 trials involve MSS type CRC. Preclinical phase I/II trials are predominant, with only 3 clinical phase III trials. In trials related to MSI-H type CRC, 4 studies involve PD-1/PD-L1 inhibitors combined with neoadjuvant therapy, and 5 studies involve combination therapy. In trials related to MSS type CRC, 3 studies involve PD-1/PD-L1 inhibitors combined with targeted therapy, 2 studies involve PD-1/PD-L1 inhibitors combined with chemotherapy, 1 study involves PD-1/PD-L1 inhibitor combined immunotherapy, 1 study involves PD-1/PD-L1 inhibitors combined with bacterial therapy, and 3 studies involve PD-1/PD-L1 inhibitors combined with comprehensive therapy. As for primary outcome measures, 4 trials select pathological complete response rates, 3 trials select progression-free survival rate, 3 trials select objective response rate, 3 trials select overall survival rate, 4 trials select disease-free survival rate, 1 trial selects clinical complete response rate, and 1 trial selects percentage of participants with a dose-limiting toxicity.

CONCLUSION

For early- and middle-stage MSI-H and MSS CRC, PD-1/PD-L1 inhibitors have shown some therapeutic efficacy, as evidenced by phase I/II studies. However, contemporary trial designs exhibit heterogeneity, with relatively few inclusion criteria, the use of various drug combinations and regimens, and significant variations in reported endpoints. Nevertheless, more double-arm, multicenter, randomized controlled trials are still needed to confirm the efficacy of immunotherapy.

A radiomics and genomics-derived model for predicting metastasis and prognosis in colorectal cancer

Approximately 50% of colorectal cancer (CRC) patients would develop metastasis with poor prognosis, therefore, it is necessary to effectively predict metastasis in clinical treatment. In this study, we aimed to establish a machine-learning model for predicting metastasis in CRC patients by considering radiomics and transcriptomics simultaneously. Here, 1023 patients with CRC from three centers were collected and divided into five queues (Dazhou Central Hospital n = 517, Nanchong Central Hospital n = 120 and the Cancer Genome Atlas (TCGA) n = 386). A total of 854 radiomics features were extracted from tumor lesions on CT images, and 217 differentially expressed genes were obtained from non-metastasis and metastasis tumor tissues using RNA sequencing. Based on radiotranscriptomic (RT) analysis, a novel RT model was developed and verified through genetic algorithms (GA). Interleukin (IL)-26, a biomarker in RT model, was verified for its biological function in CRC metastasis. Furthermore, 15 radiomics variables were screened through stepwise regression, which was highly correlated with the IL26 expression level. Finally, a radiomics model (RA) was established by combining GA and stepwise regression analysis with radiomics features. The RA model exhibited favorable discriminatory ability and accuracy for metastasis prediction in two independent verification cohorts. We designed multicenter, multi-scale cohorts to construct and verify novel combined radiomics and genomics models for predicting metastasis in CRC. Overall, RT model and RA model might help clinicians in directing personalized diagnosis and therapeutic regimen selection for patients with CRC.

Machine learning reveals diverse cell death patterns in lung adenocarcinoma prognosis and therapy

Cancer cell growth, metastasis, and drug resistance pose significant challenges in the management of lung adenocarcinoma (LUAD). However, there is a deficiency in optimal predictive models capable of accurately forecasting patient prognoses and guiding the selection of targeted treatments. Programmed cell death (PCD) pathways play a pivotal role in the development and progression of various cancers, offering potential as prognostic indicators and drug sensitivity markers for LUAD patients. The development and validation of predictive models were conducted by integrating 13 PCD patterns with comprehensive analysis of bulk RNA, single-cell RNA transcriptomics, and pertinent clinicopathological details derived from TCGA-LUAD and six GEO datasets. Utilizing the machine learning algorithms, we identified ten critical differentially expressed genes associated with PCD in LUAD, namely CHEK2, KRT18, RRM2, GAPDH, MMP1, CHRNA5, TMPRSS4, ITGB4, CD79A, and CTLA4. Subsequently, we conducted a programmed cell death index (PCDI) based on these genes across the aforementioned cohorts and integrated this index with relevant clinical features to develop several prognostic nomograms. Furthermore, we observed a significant correlation between the PCDI and immune features in LUAD, including immune cell infiltration and the expression of immune checkpoint molecules. Additionally, we found that patients with a high PCDI score may exhibit resistance to immunotherapy and standard adjuvant chemotherapy regimens; however, they may benefit from other FDA-supported drugs such as docetaxel and dasatinib. In conclusion, the PCDI holds potential as a prognostic signature and can facilitate personalized treatment for LUAD patients.

Regression-based Deep-Learning predicts molecular biomarkers from pathology slides

Deep Learning (DL) can predict biomarkers from cancer histopathology. Several clinically approved applications use this technology. Most approaches, however, predict categorical labels, whereas biomarkers are often continuous measurements. We hypothesize that regression-based DL outperforms classification-based DL. Therefore, we develop and evaluate a self-supervised attention-based weakly supervised regression method that predicts continuous biomarkers directly from 11,671 images of patients across nine cancer types. We test our method for multiple clinically and biologically relevant biomarkers: homologous recombination deficiency score, a clinically used pan-cancer biomarker, as well as markers of key biological processes in the tumor microenvironment. Using regression significantly enhances the accuracy of biomarker prediction, while also improving the predictions' correspondence to regions of known clinical relevance over classification. In a large cohort of colorectal cancer patients, regression-based prediction scores provide a higher prognostic value than classification-based scores. Our open-source regression approach offers a promising alternative for continuous biomarker analysis in computational pathology.

ENOX2 inhibition enhances infiltration of effector memory T-cell and mediates response to chemotherapy in immune-quiescent nasopharyngeal carcinoma

INTRODUCTION

The immunosuppressive tumor microenvironment is a major barrier for chemotherapy. Different chemosensitization approaches to reinstate immunological surveillance for cancers that are immune quiescent at the outset, have thus been devised. Cancer-specific ENOX2 expression is correlated with abnormal cell growth and has been proposed as a cellular target for anti-cancer activity. However, the potential effects of ENOX2 on the interaction between immune system and tumor cells remain elusive.

OBJECTIVES

To understand the mechanisms by which tumor-intrinsic ENOX2-mediated alterations in anti-tumor activity of T-cells and response to chemotherapy.

METHODS

In situ multiplexed immunohistochemistry with single cell and bulk RNA sequencing data from nasopharyngeal carcinoma (NPC) human tissues were used to define tumor phenotypes. Two NPC cell lines, with distinct ENOX2 expression, were used in a co-culture platform to study tumor-immune interactions between cancer cells/spheroids and T-cells. The effect of cisplatin treatment with ENOX2 inhibition by idronoxil (IDX) were tested in vitro and in vivo. Multi-parametric flow cytometry was used to characterize T-cell infiltrates in an NPC tumor humanized mouse model treated with combined treatment.

RESULTS

NPC predominantly displayed an immune-excluded profile. This "cold-phenotype" was shown to exhibit higher ENOX2 expression and was associate with poorer progression-free survival (PFS). The therapeutic combination of IDX with cisplatin was effective in promoting CD8+ effector memory T cell (Tem) differentiation and mobilization. This Tem signature was highly cytotoxic, with Tem-mediated preferential lysis of higher ENOX2-expressing NPC cells. A combination-treated humanized mouse model showing dramatic shrinkage in tumors, were intra-tumoral Tem-enriched.

CONCLUSION

Tumor-intrinsic ENOX2 expression is associated with tumor phenotype and PFS in NPC. Targeting ENOX2 with IDX and cisplatin impose qualitative control of T-cell response by preferentially increasing immune cells infiltration, Tem differentiation and tumor suppression. We suggest that ENOX2 inhibition may be a promising therapeutic strategy to enhance the effects of chemotherapy.

The current status and prospect of immunotherapy in colorectal cancer

Metastatic colorectal cancer (mCRC) is a heterogeneous disease. We reviewed the current clinical trials on immunotherapy in metastatic colorectal cancer with high microsatellite instability and microsatellite stability. Owing to the advances in immunotherapy, its use has gradually expanded from second- and third-line therapies to first-line, early neoadjuvant, and adjuvant therapies. Based on current research results, immunotherapy has shown very good results in dMMR/MSI-H patients, whether it is neoadjuvant therapy for operable patients or first-line or multi-line therapy for advanced patients. KEYNOTE 016 study also showed that patients with MSS were basically ineffective in single immunotherapy. Moreover, immunotherapy for colorectal cancer may also require identification of new biomarkers.

Tumor-suppressive zinc finger protein 24 (ZNF24) sensitizes colorectal cancer cells to 5-fluorouracil by inhibiting the Wnt pathway and activating the p53 signaling

Carcinogenesis and colorectal cancer (CRC) development are associated with dysregulation of various pathways, including Wnt and p53. 5-fluorouracil (5-FU) is a common chemotherapeutic agent for CRC treatment, but its efficacy is restricted by drug resistance. Doxycycline is an orally active tetracycline antibiotic known for its antimicrobial and anticancer cell proliferation activities. This study intends to delineate the potential role of bioinformatically predicted ZNF24 in the 5-FU resistance of CRC cells. The expression of ZNF24 was measured in clinically collected CRC tissues and cells. Afterward, ectopic ZNF24 expression was induced by DOX to evaluate the viability, colony-forming ability and sphere-forming ability of CRC cells. It was found that ZNF24 was validated to be poorly expressed in CRC tissues, and ectopic expression of ZNF24 was revealed to restrict the malignant phenotypes of CRC cells. In addition, restored ZNF24 attenuated 5-FU resistance of CRC cells by inhibiting the Wnt pathway and activating p53 signaling. Furthermore, an inhibitor of Wnt production 2 (IWP-2) treatment was an alternative to ZNF24 up-regulation in sensitizing CRC cells to 5-FU treatment. In conclusion, our results indicate that ZNF24 inhibits 5-FU resistance of CRC cells by suppressing the Wnt pathway and activating p53 signaling, which offers a potential strategy for managing chemoresistance in CRC.

Genetic profile of Chinese patients with small bowel cancer categorized by anatomic location

BACKGROUND

Small bowel cancer (SBC) is a very rare solid malignancy. Consequently, compared with other malignant gastrointestinal tumors, our knowledge regarding SBC, specifically its molecular attributes, remains limited. Herein, we aim to provide an overview of the gene characteristics of Chinese patients with SBC, We particularly focus on elucidating the genetic intricacies that differentiate SBC patients whose primary tumors originate in distinct anatomical regions within the small bowel.

METHODS

During the period ranging from February 2018 to December 2022, a total of 298 tumor samples were consecutively collected from Chinese patients diagnosed with small bowel cancer.. Next-generation sequencing (NGS) was performed to detect gene mutation, assess microsatellite instability (MSI), and evaluate tumor mutational burden (TMB). Additionally,, IHC was used to analyze the level of PD-L1 expression within the samples.

RESULTS

The outcomes of the next-generation sequencing (NGS) unveiled the predominant gene mutations observed in Chinese patients with small bowel cancer (SBC). The top ten gene mutations identified were as follows: TP53 (53%), KRAS (51%), APC (31%), SMAD4 (19%), VEGFA (15%), CDKN2A (15%), RAC1 (15%), LRP1B (14%), MGMT (14%, CD74 (13%). Subsequent analysis revealed disparities in the gene landscape between the cohort in this study and that of the Memorial Sloan Kettering Cancer Center (MSKCC), Notably, distinguishable mutational frequencies were identified in several genes, including ERBB2, FBXW7, PIK3CA, etc. which exhibited contrasting presence in both this cohort and the MSKCC cohort.. Furthermore, we noticed variations in the frequency of gene mutations among SBC patients depending on the specific anatomical site where the tumors originated within the small bowel. In addition, the distribution of patients with high microsatellite instability (MSI-H) and tumor mutational burden (TMB) levels varied among SBC patients with tumors originating from the duodenum, jejunum, and ileum.

CONCLUSION

Chinese patients with small bowel cancer exhibited a distinct genetic profile in comparison to other populations, highlighting a unique genetic landscape. Furthermore, noticeable disparities in the genetic landscape were observed between patients with cancer situated in the duodenum and those with cancer affecting other regions of the small bowel, this suggests that these patients should be treated differently.

Advances in Polymeric Micelles: Responsive and Targeting Approaches for Cancer Immunotherapy in the Tumor Microenvironment

In recent years, to treat a diverse array of cancer forms, considerable advancements have been achieved in the field of cancer immunotherapies. However, these therapies encounter multiple challenges in clinical practice, such as high immune-mediated toxicity, insufficient accumulation in cancer tissues, and undesired off-target reactions. To tackle these limitations and enhance bioavailability, polymer micelles present potential solutions by enabling precise drug delivery to the target site, thus amplifying the effectiveness of immunotherapy. This review article offers an extensive survey of recent progress in cancer immunotherapy strategies utilizing micelles. These strategies include responsive and remodeling approaches to the tumor microenvironment (TME), modulation of immunosuppressive cells within the TME, enhancement of immune checkpoint inhibitors, utilization of cancer vaccine platforms, modulation of antigen presentation, manipulation of engineered T cells, and targeting other components of the TME. Subsequently, we delve into the present state and constraints linked to the clinical utilization of polymeric micelles. Collectively, polymer micelles demonstrate excellent prospects in tumor immunotherapy by effectively addressing the challenges associated with conventional cancer immunotherapies.

Nonoperative Management of dMMR/MSI-H Colorectal Cancer following Neoadjuvant Immunotherapy: A Narrative Review

Immunotherapy with PD-1 blockade has achieved a great success in colorectal cancers (CRCs) with high microsatellite instability (MSI-H) and deficient mismatch repair (dMMR), and has become the first-line therapy in metastatic setting. Studies of neoadjuvant immunotherapy also report exciting results, showing high rates of clinical complete response (cCR) and pathological complete response. The high efficacy and long duration of response of immunotherapy has prompt attempts to adopt watch-and-wait strategy for patients achieving cCR following the treatment. Thankfully, the watch-and-wait approach has been proposed for nearly 20 years for patients undergoing chemoradiotherapy and has gained ground among patients as well as clinicians. In this narrative review, we combed through the available information on immunotherapy for CRC and on the watch-and-wait strategy in chemoradiotherapy, and looked forward to a future where neoadjuvant immunotherapy as a curative therapy would play a big part in the treatment of MSI-H/dMMR CRC.

Advances in immunotyping of colorectal cancer

Immunotherapy has transformed treatment for various types of malignancy. However, the benefit of immunotherapy is limited to a minority of patients with mismatch-repair-deficient (dMMR) and microsatellite instability-high (MSI-H) (dMMR-MSI-H) colorectal cancer (CRC). Understanding the complexity and heterogeneity of the tumor immune microenvironment (TIME) and identifying immune-related CRC subtypes will improve antitumor immunotherapy. Here, we review the current status of immunotherapy and typing schemes for CRC. Immune subtypes have been identified based on TIME and prognostic gene signatures that can both partially explain clinical responses to immune checkpoint inhibitors and the prognosis of patients with CRC. Identifying immune subtypes will improve understanding of complex CRC tumor heterogeneity and refine current immunotherapeutic strategies.

F-Box and Leucine-Rich Repeat Protein 7 Is a Prognostic Biomarker and Is Correlated with the Immunosuppressive Microenvironment in Colorectal Cancer

Background: Colorectal cancer (CRC) is a common malignancy of the digestive system, but its specific mechanisms of occurrence and development remain incompletely understood. F-Box and leucine-rich repeat protein 7 (FBXL7) is a subunit of the Skp-cullin-F-box ubiquitin ligase, involved in cell cycle regulation, endothelial cell damage, and inflammatory immunological responses. However, the role of FBXL7 in CRC remains unknown. In this study, we investigated the clinical significance and potential mechanism of FBXL7 expression in CRC progression. Methods: We utilized data from The Cancer Genome Atlas (TCGA) and the University of California Santa Cruz Xena (UCSC Xena) database for bioinformatic analyses. Clinical CRC samples were used to confirm FBXL7 expression. Gene set enrichment analysis (GSEA) and various databases, such as TCGA, UCSC Xena, cBioPortal, University of ALabama at Birmingham CANcer data analysis portal, MethSurv, Tumor Immune Estimation Resource (TIMER), TIMER2.0, Tumor-Immune System Interaction Database, and Tumor Immune Dysfunction and Exclusion Database (TIDB), were used to investigate the role of FBXL7 in CRC. Statistical analysis was performed using R (v.3.6.3) or GraphPad Prism 8.0. Results: Our findings revealed the predictive significance of FBXL7 in CRC patients. FBXL7 expression was associated with tumor stage, lymph node stage, pathological stage, perineural invasion, and lymphatic invasion. GSEA analysis identified associations between FBXL7 and extracellular matrix organization, as well as immune-related pathways. Immunological analysis revealed a correlation between high FBXL7 expression and the development of an immunosuppressive microenvironment. Conclusion: Identifying FBXL7 as a novel biomarker for CRC could shed light on the promotion of CRC development by the immune environment.

Construction of a novel cancer-associated fibroblast-related signature to predict clinical outcome and immune response in colon adenocarcinoma

The interaction between the tumour and the surrounding microenvironment determines the malignant biological behaviour of the tumour. Cancer-associated fibroblasts (CAFs) coordinate crosstalk between cancer cells in the tumour immune microenvironment (TIME) and are extensively involved in tumour malignant behaviours, such as immune evasion, invasion and drug resistance. Here, we performed differential and prognostic analyses of genes associated with CAFs and constructed CAF-related signatures (CAFRs) to predict clinical outcomes in individuals with colon adenocarcinoma (COAD) based on machine learning algorithms. The CAFRs were further validated in an external independent cohort, GSE17538. Additionally, Cox regression, receiver operating characteristic (ROC) and clinical correlation analysis were utilised to systematically assess the CAFRs. Moreover, CIBERSORT, single sample Gene Set Enrichment Analysis (ssGSEA) and Estimation of Stromal and Immune cells in MAlignant Tumor tissues using Expression data (ESTIMATE) analysis were utilised to characterise the TIME in patients with COAD. Microsatellite instability (MSI) and tumour mutation burden were also analysed. Furthermore, Gene Set Variation Analysis (GSVA), Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene Ontology (GO) elucidated the biological functions and signalling pathways involved in the CAFRs. Consensus clustering analysis was used for the immunological analysis of patients with COAD. Finally, the pRRophic algorithm was used for sensitivity analysis of common drugs. The CAFRs constructed herein can better predict the prognosis in COAD. The cluster analysis based on the CAFRs can effectively differentiate between immune 'hot' and 'cold' tumours, determine the beneficiaries of immune checkpoint inhibitors (ICIs) and provide insight into individualised treatment for COAD.

Case Report: Genetic and immune microenvironmental characteristics of a rectal cancer patient with MSS/PD-L1-negative recurrent hepatopulmonary metastasis who achieved complete remission after treatment with PD-1 inhibitor

Currently, microsatellite high instability (MSI-H)/mismatch repair protein deletion (dMMR) has become a crucial biomarker for utilizing immune checkpoint inhibitors in patients with advanced colorectal cancer (mCRC). However, the proportion of MSI-H/dMMR in advanced patients is only about 5% and mCRC patients with microsatellite stability (MSS)/proficient mismatch repair (pMMR) exhibit poor responses to immunotherapy. Although diverse immune combination therapy regimens have been examined in patients with advanced colorectal cancer who demonstrate MSS/pMMR, these approaches have not yielded favorable efficacy and only a limited proportion of patients have benefited, especially for advanced colorectal cancer patients with liver metastases. Therefore, the mechanism of benefit and potential biomarkers of immunotherapy in patients with MSS/pMMR mCRC deserve more in-depth exploration. Here, we present a case study of a rectal cancer patient with MSS and PD-L1-negative recurrent hepatopulmonary metastases who attained complete remission (CR) and sustained benefits with immunotherapy after systemic therapy had failed. The analysis of the patient's genetic and immune microenvironmental characteristics revealed that mutations in DNA damage repair (DDR) pathway genes and the existence of abundant tumor-infiltrating lymphocytes could contribute to his potential benefit.

Recent application of artificial intelligence on histopathologic image-based prediction of gene mutation in solid cancers

PURPOSE

Evaluation of genetic mutations in cancers is important because distinct mutational profiles help determine individualized drug therapy. However, molecular analyses are not routinely performed in all cancers because they are expensive, time-consuming and not universally available. Artificial intelligence (AI) has shown the potential to determine a wide range of genetic mutations on histologic image analysis. Here, we assessed the status of mutation prediction AI models on histologic images by a systematic review.

METHODS

A literature search using the MEDLINE, Embase and Cochrane databases was conducted in August 2021. The articles were shortlisted by titles and abstracts. After a full-text review, publication trends, study characteristic analysis and comparison of performance metrics were performed.

RESULTS

Twenty-four studies were found mostly from developed countries, and their number is increasing. The major targets were gastrointestinal, genitourinary, gynecological, lung and head and neck cancers. Most studies used the Cancer Genome Atlas, with a few using an in-house dataset. The area under the curve of some of the cancer driver gene mutations in particular organs was satisfactory, such as 0.92 of BRAF in thyroid cancers and 0.79 of EGFR in lung cancers, whereas the average of all gene mutations was 0.64, which is still suboptimal.

CONCLUSION

AI has the potential to predict gene mutations on histologic images with appropriate caution. Further validation with larger datasets is still required before AI models can be used in clinical practice to predict gene mutations.

Development of an immune-related gene prognostic risk model and identification of an immune infiltration signature in the tumor microenvironment of colon cancer

BACKGROUND

Colon cancer is a common and highly malignant tumor. Its incidence is increasing rapidly with poor prognosis. At present, immunotherapy is a rapidly developing treatment for colon cancer. The aim of this study was to construct a prognostic risk model based on immune genes for early diagnosis and accurate prognostic prediction of colon cancer.

METHODS

Transcriptome data and clinical data were downloaded from the cancer Genome Atlas database. Immunity genes were obtained from ImmPort database. The differentially expressed transcription factors (TFs) were obtained from Cistrome database. Differentially expressed (DE) immune genes were identified in 473 cases of colon cancer and 41 cases of normal adjacent tissues. An immune-related prognostic model of colon cancer was established and its clinical applicability was verified. Among 318 tumor-related transcription factors, differentially expressed transcription factors were finally obtained, and a regulatory network was constructed according to the up-down regulatory relationship.

RESULTS

A total of 477 DE immune genes (180 up-regulated and 297 down-regulated) were detected. We developed and validated twelve immune gene models for colon cancer, including SLC10A2, FABP4, FGF2, CCL28, IGKV1-6, IGLV6-57, ESM1, UCN, UTS2, VIP, IL1RL2, NGFR. The model was proved to be an independent prognostic variable with good prognostic ability. A total of 68 DE TFs (40 up-regulated and 23 down-regulated) were obtained. The regulation network between TF and immune genes was plotted by using TF as source node and immune genes as target node. In addition, Macrophage, Myeloid Dendritic cell and CD4⁺ T cell increased with the increase of risk score.

CONCLUSION

We developed and validated twelve immune gene models for colon cancer, including SLC10A2, FABP4, FGF2, CCL28, IGKV1-6, IGLV6-57, ESM1, UCN, UTS2, VIP, IL1RL2, NGFR. This model can be used as a tool variable to predict the prognosis of colon cancer.

Deep learning image analysis quantifies tumor heterogeneity and identifies microsatellite instability in colon cancer

BACKGROUND AND OBJECTIVES

Deep learning utilizing convolutional neural networks (CNNs) applied to hematoxylin & eosin (H&E)-stained slides numerically encodes histomorphological tumor features. Tumor heterogeneity is an emerging biomarker in colon cancer that is, captured by these features, whereas microsatellite instability (MSI) is an established biomarker traditionally assessed by immunohistochemistry or polymerase chain reaction.

METHODS

H&E-stained slides from The Cancer Genome Atlas (TCGA) colon cohort are passed through the CNN. Resulting imaging features are used to cluster morphologically similar slide regions. Tile-level pairwise similarities are calculated and used to generate a tumor heterogeneity score (THS). Patient-level THS is then correlated with TCGA-reported biomarkers, including MSI-status.

RESULTS

H&E-stained images from 313 patients generated 534 771 tiles. Deep learning automatically identified and annotated cells by type and clustered morphologically similar slide regions. MSI-high tumors demonstrated significantly higher THS than MSS/MSI-low (p < 0.001). THS was higher in MLH1-silent versus non-silent tumors (p < 0.001). The sequencing derived MSIsensor score also correlated with THS (r = 0.51, p < 0.0001).

CONCLUSIONS

Deep learning provides spatially resolved visualization of imaging-derived biomarkers and automated quantification of tumor heterogeneity. Our novel THS correlates with MSI-status, indicating that with expanded training sets, translational tools could be developed that predict MSI-status using H&E-stained images alone.

A glimpse of research cores and frontiers on the relationship between long noncoding RNAs (lncRNAs) and colorectal cancer (CRC) using the VOSviewer tool

As lncRNAs are essential participants in colorectal carcinogenesis. This study aimed to use the VOSviewer tool to access the research cores and frontiers on the relationship between lncRNAs and CRC. Our findings showed that the mechanism of lncRNA in the occurrence and development of CRC was the core theme of the field. (1) Immunotherapy and immune microenvironment of CRC and lncRNAs, (2) CRC and lncRNAs in exosomes and (3) CRC and lncRNA-targeted therapy might represent three research frontiers. A comprehensive understanding of their existing mechanisms and the search for new regulatory paradigms are the core topics of future research. This knowledge will also help us select appropriate targeting methods and select appropriate preclinical models to promote clinical translation and ultimately achieve precise treatment of CRC.

Predicting colorectal cancer microsatellite instability with a self-attention-enabled convolutional neural network

This study develops a method combining a convolutional neural network model, INSIGHT, with a self-attention model, WiseMSI, to predict microsatellite instability (MSI) based on the tiles in colorectal cancer patients from a multicenter Chinese cohort. After INSIGHT differentiates tumor tiles from normal tissue tiles in a whole slide image, features of tumor tiles are extracted with a ResNet model pre-trained on ImageNet. Attention-based pooling is adopted to aggregate tile-level features into slide-level representation. INSIGHT has an area under the curve (AUC) of 0.985 for tumor patch classification. The Spearman correlation coefficient of tumor cell fraction given by expert pathologist and INSIGHT is 0.7909. WiseMSI achieves a specificity of 94.7% (95% confidence interval [CI] 93.7%-95.7%), a sensitivity of 84.7% (95% CI 82.6%-86.9%), and an AUC of 0.954 (95% CI 0.948-0.960). Comparative analysis shows that this method has better performance than the other five classic deep learning methods.

Clinical significance of neoadjuvant chemotherapy for locally advanced colorectal cancer patients with deficient mismatch repair: possibly residual value in the era of immunotherapy

Background

Deficient mismatch repair (dMMR) or microsatellite instability is one of the well-established molecular biomarkers in colorectal cancer (CRC). The efficiency of neoadjuvant chemotherapy (NAC) in locally advanced colorectal cancer (LACC) patients with dMMR is unclear.

Objectives

We assessed the tumor response and clinical outcome in LACC patients with dMMR received NAC.

Design

Retrospective, single-center analysis.

Methods

From 2013 to 2018, a total of 577 LACC patients with dMMR who underwent radical surgery were identified. Among them, 109 patients who received adjuvant chemotherapy were further screened out for analysis. According to whether receiving NAC or not, 109 patients were divided into two groups with the purpose of retrospectively analyzing their characteristics, treatment, and survival results, especially the 5-year disease-free survival (DFS) and 5-year overall survival.

Results

Baseline characteristics were matched between the two groups. One of 40 patients in NAC group recurred, while 13 of 69 patients in non-NAC group recurred. Univariate and multivariate analyses showed that NAC (hazard ratio: 0.115; 95% confidence interval: 0.015-0.897; p = 0.039) was independent influence factor for DFS. In NAC group, there were 13/40 (32.5%) patients for tumor regression grade 1 and 27/40 (67.5%) patients converted clinical positive N-stage into negative N-stage.

Conclusion

In this study, NAC was associated with better tumor downstaging and longer 5-year DFS in LACC patients with dMMR. Consequently, NAC might be an additional treatment choice when it comes to such patients in the future.

Oncotherapeutic Strategies in Early Onset Colorectal Cancer

Early onset colorectal cancer (EOCRC), defined as colorectal cancers in patients aged less than 50 years, is becoming an increasingly common issue, globally. Since 1994, the incidence of this condition has been rising by 2% annually. Approximately one in five patients under 50 years of age diagnosed with colorectal cancer have an underlying genetic predisposition syndrome. The detection of cancer among the other 80% of patients poses a considerable task, as there is no family history to advocate for commencing early screening in this group. Patients with EOCRC have distinct social, spiritual, fertility, and financial needs from their older counterparts that need to be addressed. This review discusses the risk factors associated with the development of EOCRC and current best practice for the management of this disease.

Preoperative Prediction of Microsatellite Instability in Rectal Cancer Using Five Machine Learning Algorithms Based on Multiparametric MRI Radiomics

Objectives: To establish and verify radiomics models based on multiparametric MRI for preoperatively identifying the microsatellite instability (MSI) status of rectal cancer (RC) by comparing different machine learning algorithms. Methods: This retrospective study enrolled 383 (training set, 268; test set, 115) RC patients between January 2017 and June 2022. A total of 4148 radiomics features were extracted from multiparametric MRI, including T₂-weighted imaging, T₁-weighted imaging, apparent diffusion coefficient, and contrast-enhanced T₁-weighted imaging. The analysis of variance, correlation test, univariate logistic analysis, and a gradient-boosting decision tree were used for the dimension reduction. Logistic regression, Bayes, support vector machine (SVM), K-nearest neighbor (KNN), and tree machine learning algorithms were used to build different radiomics models. The relative standard deviation (RSD) and bootstrap method were used to quantify the stability of these five algorithms. Then, predictive performances of different models were assessed using area under curves (AUCs). The performance of the best radiomics model was evaluated using calibration and discrimination. Results: Among these 383 patients, the prevalence of MSI was 14.62% (56/383). The RSD value of logistic regression algorithm was the lowest (4.64%), followed by Bayes (5.44%) and KNN (5.45%), which was significantly better than that of SVM (19.11%) and tree (11.94%) algorithms. The radiomics model based on logistic regression algorithm performed best, with AUCs of 0.827 and 0.739 in the training and test sets, respectively. Conclusions: We developed a radiomics model based on the logistic regression algorithm, which could potentially be used to facilitate the individualized prediction of MSI status in RC patients.

An immune, stroma, and epithelial-mesenchymal transition-related signature for predicting recurrence and chemotherapy benefit in stage II-III colorectal cancer

BACKGROUND

Debates exist on the treatment decision of the stage II/III colorectal cancer (CRC) due to the insufficiency of the current TNM stage-based risk stratification system. Epithelial-mesenchymal transition (EMT) and tumor microenvironment (TME) have both been linked to CRC progression in recent studies. We propose to improve the prognosis prediction of CRC by integrating TME and EMT.

METHODS

In total, 2382 CRC patients from seven datasets and one in-house cohort were collected, and 1640 stage II/III CRC patients with complete survival information and gene expression profiles were retained and divided into a training cohort and three independent validation cohorts. Integrated analysis of 398 immune, stroma, and epithelial-mesenchymal transition (ISE)-related genes identified an ISE signature independently associated with the recurrence of CRC. The underlying biological mechanism of the ISE signature and its influence on adjuvant chemotherapy was further explored.

RESULTS

We constructed a 26-gene signature which was significantly associated with poor outcome in Training cohort (p < 0.001, HR [95%CI] = 4.42 [3.25-6.01]) and three independent validation cohorts (Validation cohort-1: p < 0.01, HR [95%CI] = 1.70 [1.15-2.51]; Validation cohort-2: p < 0.001, HR [95% CI] = 2.30 [1.67-3.16]; Validation cohort-3: p < 0.01, HR [95% CI] = 2.42 [1.25-4.70]). After adjusting for known clinicopathological factors, multivariate cox analysis confirmed the ISE signature's independent prognostic value. Subgroup analysis found that stage III patients with low ISE score might benefit from adjuvant chemotherapy (p < 0.001, HR [95%CI] = 0.15 [0.04-0.55]). Hypergeometric test and enrichment analysis revealed that low-risk group was enriched in thr immune pathway while high-risk group was associated with the EMT pathway and CMS4 subtype.

CONCLUSION

We proposed an ISE signature for robustly predicting the recurrence of stage II/III CRC and help treatment decision by identifying patients who will not benefit from current standard treatment.

Artificial intelligence applications in predicting the behavior of gastrointestinal cancers in pathology

Recent research has provided a wealth of data supporting the application of artificial intelligence (AI)-based applications in routine pathology practice. Indeed, it is clear that these methods can significantly support an accurate and rapid diagnosis by eliminating errors, increasing reliability, and improving workflow. In addition, the effectiveness of AI in the pathological evaluation of prognostic parameters associated with behavior, course, and treatment in many types of tumors has also been noted. Regarding gastrointestinal system (GIS) cancers, the contribution of AI methods to pathological diagnosis has been investigated in many studies. On the other hand, studies focusing on AI applications in evaluating parameters to determine tumor behavior are relatively few. For this purpose, the potential of AI models has been studied over a broad spectrum, from tumor subtyping to the identification of new digital biomarkers. The capacity of AI to infer genetic alterations of cancer tissues from digital slides has been demonstrated. Although current data suggest the merit of AI-based approaches in assessing tumor behavior in GIS cancers, a wide range of challenges still need to be solved, from laboratory infrastructure to improving the robustness of algorithms, before incorporating AI applications into real-life GIS pathology practice. This review aims to present data from AI applications in evaluating pathological parameters related to the behavior of GIS cancer with an overview of the opportunities and challenges encountered in implementing AI in pathology.

MAST: a hybrid Multi-Agent Spatio-Temporal model of tumor microenvironment informed using a data-driven approach

Motivation

Recently, several computational modeling approaches, such as agent-based models, have been applied to study the interaction dynamics between immune and tumor cells in human cancer. However, each tumor is characterized by a specific and unique tumor microenvironment, emphasizing the need for specialized and personalized studies of each cancer scenario.

Results

We present MAST, a hybrid Multi-Agent Spatio-Temporal model which can be informed using a data-driven approach to simulate unique tumor subtypes and tumor-immune dynamics starting from high-throughput sequencing data. It captures essential components of the tumor microenvironment by coupling a discrete agent-based model with a continuous partial differential equations-based model.The application to real data of human colorectal cancer tissue investigating the spatio-temporal evolution and emergent properties of four simulated human colorectal cancer subtypes, along with their agreement with current biological knowledge of tumors and clinical outcome endpoints in a patient cohort, endorse the validity of our approach.

Availability and implementation

MAST, implemented in Python language, is freely available with an open-source license through GitLab (https://gitlab.com/sysbiobig/mast), and a Docker image is provided to ease its deployment. The submitted software version and test data are available in Zenodo at https://dx.doi.org/10.5281/zenodo.7267745.

Supplementary information

Supplementary data are available at Bioinformatics Advances online.

Nuclear translocation of Gasdermin D sensitizes colorectal cancer to chemotherapy in a pyroptosis-independent manner

Gasdermin D (GSDMD) has recently been identified as a cytoplasmic effector protein that plays a central role in pyroptosis of immune cells. However, GSDMD is a universally expressed protein, and its function beyond pyroptosis, especially in cancer cells, has not been well characterized. Here, we report that predominant localization of GSDMD in the nucleoplasm in vivo indicates favorable clinical outcomes in colorectal cancer, while a lack of nuclear localization of GSDMD is associated with poor outcomes. Nuclear GSDMD, rather than cytoplasmic GSDMD, inhibits cell growth and promotes apoptosis in colorectal cancer. Hypoxia in the tumor microenvironment accounts for mild or moderate nuclear translocation of GSDMD in vivo. Under the stimulation of chemotherapy drugs, nuclear GSDMD promotes apoptosis via regulation of its subcellular distribution rather than pyroptosis-related cleavage. After nuclear translocation, GSDMD interacts with PARP-1 to dramatically inhibit its DNA damage repair-related function by functioning like the PARP inhibitor olaparib, thus forming a "hypoxia/chemotherapy-GSDMD nuclear translocation-PARP-1 blockade-DNA damage and apoptosis" axis. This study redefines the pyroptosis-independent function of GSDMD and suggests that the subcellular localization of GSDMD may serve as a molecular indicator of clinical outcomes and a promising therapeutic target in colorectal cancer.

Identification and validation of immunotherapy for four novel clusters of colorectal cancer based on the tumor microenvironment

The incidence and mortality of colorectal cancer (CRC) are increasing year by year. The accurate classification of CRC can realize the purpose of personalized and precise treatment for patients. The tumor microenvironment (TME) plays an important role in the malignant progression and immunotherapy of CRC. An in-depth understanding of the clusters based on the TME is of great significance for the discovery of new therapeutic targets for CRC. We extracted data on CRC, including gene expression profile, DNA methylation array, somatic mutations, clinicopathological information, and copy number variation (CNV), from The Cancer Genome Atlas (TCGA), Gene Expression Omnibus (GEO) (four datasets—GSE14333, GSE17538, GSE38832, and GSE39582), cBioPortal, and FireBrowse. The MCPcounter was utilized to quantify the abundance of 10 TME cells for CRC samples. Cluster repetitive analysis was based on the Hcluster function of the Pheatmap package in R. The ESTIMATE package was applied to compute immune and stromal scores for CRC patients. PCA analysis was used to remove batch effects among different datasets and transform genome-wide DNA methylation profiling into methylation of tumor-infiltrating lymphocyte (MeTIL). We evaluated the mutation differences of the clusters using MOVICS, DeconstructSigs, and GISTIC packages. As for therapy, TIDE and SubMap analyses were carried out to forecast the immunotherapy response of the clusters, and chemotherapeutic sensibility was estimated based on the pRRophetic package. All results were verified in the TCGA and GEO data. Four immune clusters (ImmClust-CS1, ImmClust-CS2, ImmClust-CS3, and ImmClust-CS4) were identified for CRC. The four ImmClusts exhibited distinct TME compositions, cancer-associated fibroblasts (CAFs), functional orientation, and immune checkpoints. The highest immune, stromal, and MeTIL scores were observed in CS2, in contrast to the lowest scores in CS4. CS1 may respond to immunotherapy, while CS2 may respond to immunotherapy after anti-CAFs. Among the four ImmClusts, the top 15 markers with the highest mutation frequency were acquired, and CS1 had significantly lower CNA on the focal level than other subtypes. In addition, CS1 and CS2 patients had more stable chromosomes than CS3 and CS4. The most sensitive chemotherapeutic agents in these four ImmClusts were also found. IHC results revealed that CD29 stained significantly darker in the cancer samples, indicating that their CD29 was highly expressed in colon cancer. This work revealed the novel clusters based on TME for CRC, which would guide in predicting the prognosis, biological features, and appropriate treatment for patients with CRC.

Liposomal oxaliplatin prodrugs loaded with metformin potentiate immunotherapy for colorectal cancer

Tumor hypoxia is confirmed to be associated with the formation of tumor immunosuppression, a general feature of solid tumors, and thus attenuates the effectiveness of various cancer therapies in clinic. We herein develop a tumor microenvironment (TME) modulating liposomal nanomedicine by encapsulating metformin with amphiphilic oxaliplatin prodrug constructed liposomes to potentiate cancer immunotherapy. While metformin could regulate metabolisms of tumor cells to reduce their oxygen consumption and relieve tumor hypoxia, oxaliplatin is a chemotherapy drug that induces immunogenic cell death (ICD). The obtained met-oxa(IV)-liposome upon intravenous injection effectively attenuates tumor hypoxia and induce ICD of cancer cells, thereby collectively suppresses the growth of murine colorectal tumors by eliciting potent antitumor immunity and reversing the immunosuppressive TME. As the result, the treatment with met-oxa(IV)-liposome effectively potentiates the immune checkpoint blockade (ICB) therapy against murine colorectal tumors. This liposomal nanomedicine is highlighted to be a TME modulating liposomal nanomedicine with high potency in suppressing tumor growth, particularly promising in synergizing with ICB therapy by boosting antitumor immune responses.

Comparative analysis of high- and low-level deep learning approaches in microsatellite instability prediction

Deep learning-based approaches in histopathology can be largely divided into two categories: a high-level approach using an end-to-end model and a low-level approach using feature extractors. Although the advantages and disadvantages of both approaches are empirically well known, there exists no scientific basis for choosing a specific approach in research, and direct comparative analysis of the two approaches has rarely been performed. Using the Cancer Genomic Atlas (TCGA)-based dataset, we compared these two different approaches in microsatellite instability (MSI) prediction and analyzed morphological image features associated with MSI. Our high-level approach was based solely on EfficientNet, while our low-level approach relied on LightGBM and multiple deep learning models trained on publicly available multiclass tissue, nuclei, and gland datasets. We compared their performance and important image features. Our high-level approach showed superior performance compared to our low-level approach. In both approaches, debris, lymphocytes, and necrotic cells were revealed as important features of MSI, which is consistent with clinical knowledge. Then, during qualitative analysis, we discovered the weaknesses of our low-level approach and demonstrated that its performance can be improved by using different image features in a complementary way. We performed our study using open-access data, and we believe this study can serve as a useful basis for discovering imaging biomarkers for clinical application.

Fruquintinib in Combination With PD-1 Inhibitors in Patients With Refractory Non-MSI-H/pMMR Metastatic Colorectal Cancer: A Real-World Study in China

Background

Fruquintinib, a vascular endothelial growth factor receptor inhibitor, is a new anticancer drug independently developed in China to treat refractory metastatic colorectal cancer (mCRC). In Japan, regorafenib combined with nivolumab has been demonstrated to be promising in patients with refractory mCRC. Here, in a real-world study, we were aimed to evaluate the efficacy of fruquintinib with various programmed death-1 (PD-1) inhibitors after standard treatment in Chinese non-microsatellite instability-high (MSI-H)/mismatch repair proficient mCRC patients.

Methods

A total of 45 patients with refractory mCRC were involved in the study. They received fruquintinib (3 or 5 mg, orally administered once a day for 3 weeks followed by 1 week off in 4-week cycles) and a PD-1 inhibitor(200 mg pembrolizumab, 3 mg/kg nivolumab, 200 mg sintilimab or camrelizumab, intravenously administered on D1 once every 3 weeks). Progression-free survival (PFS), overall survival (OS), disease control rate (DCR), and objective response rate (ORR) were reviewed and evaluated.

Results

Among the 45 patients, the median age was 54 years (29-85). The ORR was 11.1% (5/45), DCR 62.2% (28/45), median PFS equal 3.8 months, and median OS was 14.9 months. The response duration was 3.4 months. PFS between left and right primary tumors and PFS with or without lung metastases were both not significantly different (p &gt; 0.05), which was inconsistent with the result of REGONIVO study. The multivariate analysis indicated no association of OS benefit in the specified subgroups. No adverse-effect-related deaths were reported.

Conclusions

Fruquintinib, in combination with anti-PD-1, was observed to have clinical activity in a small population of patients with heavily pretreated mCRC in our center. Further studies are needed to verify this outcome in a large population.

SCG2: A Prognostic Marker That Pinpoints Chemotherapy and Immunotherapy in Colorectal Cancer

Background

Fluorouracil (FU)-based chemotherapy regimens are indispensable in the comprehensive treatment of colorectal cancer (CRC). However, the heterogeneity of treated individuals and the severe adverse effects of chemotherapy results in limited overall benefit.

Methods

Firstly, Weighted gene co-expression network analysis (WGCNA) identified modules tightly associated with chemotherapy response. Then, the in-house cohort and prognostic cohorts from TCGA and GEO were subjected to Cox proportional hazards model and survival analysis to ascertain the predictable function of SCG2 on the prognosis of CRC patients. Finally, we performed In vitro experiments, functional analysis, somatic mutation, and copy number variation research to explore the biological characteristics of SCG2.

Results

We identified red and green as the modules most associated with chemotherapy response, in which SCG2 was considered a risky factor with higher expression predicting poorer prognosis. SCG2 expression in the APC non-mutation group was remarkably higher than in the mutation group. The mutation frequencies of amplified genes differed significantly between different SCG2 expression subgroups. Besides, CRC cell lines with SCG2 knockdown have reduced invasive, proliferative, and proliferative capacity. We discovered that the SCG2 high expression subgroup was the immune hot type and considered more suitable for immunotherapy.

Conclusion

This study demonstrates the clinical significance and biological characteristics of SCG2, which could serve as a promising biomarker to identify patients who may benefit from chemotherapy and immunotherapy.

Attention-based multiple instance learning with self-supervision to predict microsatellite instability in colorectal cancer from histology whole-slide images

Microsatellite instability (MSI) is a clinically important characteristic of colorectal cancer. Standard diagnosis of MSI is performed via genetic analyses, however these tests are not always included in routine care. Histopathology whole-slide images (WSIs) are the gold-standard for colorectal cancer diagnosis and are routinely collected. This study develops a model to predict MSI directly from WSIs. Making use of both weakly- and self-supervised deep learning techniques, the proposed model shows improved performance over conventional deep learning models. Additionally, the proposed framework allows for visual interpretation of model decisions. These results are validated in internal and external testing datasets.

ALOX12: A Novel Insight in Bevacizumab Response, Immunotherapy Effect, and Prognosis of Colorectal Cancer

Colorectal cancer is a highly malignant cancer with poor prognosis and mortality rates. As the first biological agent approved for metastatic colorectal cancer (mCRC), bevacizumab was confirmed to exhibit good performance when combined with chemotherapy and immunotherapy. However, the efficacy of both bevacizumab and immunotherapy is highly heterogeneous across CRC patients with different stages. Thus, exploring a novel biomarker to comprehensively assess the prognosis and bevacizumab and immunotherapy response of CRC is of great significance. In our study, weighted gene co-expression network analysis (WGCNA) and the receiver operating characteristic (ROC) curves were employed to identify bevacizumab-related genes. After verification in four public cohorts and our internal cohort, ALOX12 was identified as a key gene related to bevacizumab response. Prognostic analysis and in vitro experiments further demonstrated that ALOX12 was closely associated with the prognosis, tumor proliferation, invasion, and metastasis. Multi-omics data analysis based on mutation and copy number variation (CNV) revealed that RYR3 drove the expression of ALOX12 and the deletion of 17p12 inhibited ALOX12 expression, respectively. Moreover, we interrogated the relationship between ALOX12 and immune cells and checkpoints. The results exhibited that high ALOX12 expression predicted a higher immune infiltration and better immunotherapy response, which was further validated in Tumor Immune Dysfunction and Exclusion (TIDE) and Subclass Mapping (SubMap) methods. Above all, our study provides a stable biomarker for clinical protocol optimization, prognostic assessment, precise treatment, and individualized treatment of CRC.

Development and validation of a radiomics-based nomogram for the preoperative prediction of microsatellite instability in colorectal cancer

BACKGROUND

Preoperative prediction of microsatellite instability (MSI) status in colorectal cancer (CRC) patients is of great significance for clinicians to perform further treatment strategies and prognostic evaluation. Our aims were to develop and validate a non-invasive, cost-effective reproducible and individualized clinic-radiomics nomogram method for preoperative MSI status prediction based on contrast-enhanced CT (CECT)images.

METHODS

A total of 76 MSI CRC patients and 200 microsatellite stability (MSS) CRC patients with pathologically confirmed (194 in the training set and 82 in the validation set) were identified and enrolled in our retrospective study. We included six significant clinical risk factors and four qualitative imaging data extracted from CECT images to build the clinics model. We applied the intra-and inter-class correlation coefficient (ICC), minimal-redundancy-maximal-relevance (mRMR) and the least absolute shrinkage and selection operator (LASSO) for feature reduction and selection. The selected independent prediction clinical risk factors, qualitative imaging data and radiomics features were performed to develop a predictive nomogram model for MSI status on the basis of multivariable logistic regression by tenfold cross-validation. The area under the receiver operating characteristic (ROC) curve (AUC), calibration plots and Hosmer-Lemeshow test were performed to assess the nomogram model. Finally, decision curve analysis (DCA) was performed to determine the clinical utility of the nomogram model by quantifying the net benefits of threshold probabilities.

RESULTS

Twelve top-ranked radiomics features, three clinical risk factors (location, WBC and histological grade) and CT-reported IFS were finally selected to construct the radiomics, clinics and combined clinic-radiomics nomogram model. The clinic-radiomics nomogram model with the highest AUC value of 0.87 (95% CI, 0.81-0.93) and 0.90 (95% CI, 0.83-0.96), as well as good calibration and clinical utility observed using the calibration plots and DCA in the training and validation sets respectively, was regarded as the candidate model for identification of MSI status in CRC patients.

CONCLUSION

The proposed clinic-radiomics nomogram model with a combination of clinical risk factors, qualitative imaging data and radiomics features can potentially be effective in the individualized preoperative prediction of MSI status in CRC patients and may help performing further treatment strategies.

Utility of Continuous Disease Subtyping Systems for Improved Evaluation of Etiologic Heterogeneity

Molecular pathologic diagnosis is important in clinical (oncology) practice. Integration of molecular pathology into epidemiological methods (i.e., molecular pathological epidemiology) allows for investigating the distinct etiology of disease subtypes based on biomarker analyses, thereby contributing to precision medicine and prevention. However, existing approaches for investigating etiological heterogeneity deal with categorical subtypes. We aimed to fully leverage continuous measures available in most biomarker readouts (gene/protein expression levels, signaling pathway activation, immune cell counts, microbiome/microbial abundance in tumor microenvironment, etc.). We present a cause-specific Cox proportional hazards regression model for evaluating how the exposure-disease subtype association changes across continuous subtyping biomarker levels. Utilizing two longitudinal observational prospective cohort studies, we investigated how the association of alcohol intake (a risk factor) with colorectal cancer incidence differed across the continuous values of tumor epigenetic DNA methylation at long interspersed nucleotide element-1 (LINE-1). The heterogeneous alcohol effect was modeled using different functions of the LINE-1 marker to demonstrate the method's flexibility. This real-world proof-of-principle computational application demonstrates how the new method enables visualizing the trend of the exposure effect over continuous marker levels. The utilization of continuous biomarker data without categorization for investigating etiological heterogeneity can advance our understanding of biological and pathogenic mechanisms.

Artificial intelligence for detection of microsatellite instability in colorectal cancer-a multicentric analysis of a pre-screening tool for clinical application

BACKGROUND

Microsatellite instability (MSI)/mismatch repair deficiency (dMMR) is a key genetic feature which should be tested in every patient with colorectal cancer (CRC) according to medical guidelines. Artificial intelligence (AI) methods can detect MSI/dMMR directly in routine pathology slides, but the test performance has not been systematically investigated with predefined test thresholds.

METHOD

We trained and validated AI-based MSI/dMMR detectors and evaluated predefined performance metrics using nine patient cohorts of 8343 patients across different countries and ethnicities.

RESULTS

Classifiers achieved clinical-grade performance, yielding an area under the receiver operating curve (AUROC) of up to 0.96 without using any manual annotations. Subsequently, we show that the AI system can be applied as a rule-out test: by using cohort-specific thresholds, on average 52.73% of tumors in each surgical cohort [total number of MSI/dMMR = 1020, microsatellite stable (MSS)/ proficient mismatch repair (pMMR) = 7323 patients] could be identified as MSS/pMMR with a fixed sensitivity at 95%. In an additional cohort of N = 1530 (MSI/dMMR = 211, MSS/pMMR = 1319) endoscopy biopsy samples, the system achieved an AUROC of 0.89, and the cohort-specific threshold ruled out 44.12% of tumors with a fixed sensitivity at 95%. As a more robust alternative to cohort-specific thresholds, we showed that with a fixed threshold of 0.25 for all the cohorts, we can rule-out 25.51% in surgical specimens and 6.10% in biopsies.

INTERPRETATION

When applied in a clinical setting, this means that the AI system can rule out MSI/dMMR in a quarter (with global thresholds) or half of all CRC patients (with local fine-tuning), thereby reducing cost and turnaround time for molecular profiling.

Deconstructing Immune Cell Infiltration in Human Colorectal Cancer: A Systematic Spatiotemporal Evaluation

Cancer-related immunity has been identified as playing a key role in the outcome of colorectal cancer (CRC); however, the exact mechanisms are only partially understood. In this study, we evaluated a total of 242 surgical specimen of CRC patients using tissue microarrays and immunohistochemistry to evaluate tumor infiltrating immune cells (CD3, CD4, CD8, CD20, CD23, CD45 and CD56) and immune checkpoint markers (CTLA-4, PD-L1, PD-1) in systematically selected tumor regions and their corresponding lymph nodes, as well as in liver metastases. Additionally, an immune panel gene expression assay was performed on 12 primary tumors and 12 consecutive liver metastases. A higher number of natural killer cells and more mature B cells along with PD-1⁺ expressing cells were observed in the main tumor area as compared to metastases. A higher number of metastatic lymph nodes were associated with significantly lower B cell counts. With more advanced lymph node metastatic status, higher leukocyte—particularly T cell numbers—were observed. Eleven differentially expressed immune-related genes were found between primary tumors and liver metastases. Also, alterations of the innate immune response and the tumor necrosis factor superfamily pathways had been identified.

Multi-OMICs data analysis identifies molecular features correlating with tumor immunity in colon cancer

BACKGROUND

There is a current need for new markers with higher sensitivity and specificity to predict immune status and optimize immunotherapy use in colon cancer.

OBJECTIVE

We aimed to investigate the multi-OMICs features associated with colon cancer immunity and response to immunotherapy.

METHODS

We evaluated the association of multi-OMICs data from three colon cancer datasets (TCGA, CPTAC2, and Samstein) with antitumor immune signatures (CD8+ T cell infiltration, immune cytolytic activity, and PD-L1 expression). Using the log-rank test and hierarchical clustering, we explored the association of various OMICs features with survival and immune status in colon cancer.

RESULTS

Two gene mutations (TERT and ERBB4) correlated with antitumor cytolytic activity found also correlated with improved survival in immunotherapy-treated colon cancers. Moreover, the expression of numerous genes was associated with antitumor immunity, including GBP1, GBP4, GBP5, NKG7, APOL3, IDO1, CCL5, and CXCL9. We clustered colon cancer samples into four immuno-distinct clusters based on the expression levels of 82 genes. We have also identified two proteins (PREX1 and RAD50), ten miRNAs (hsa-miR-140, 146, 150, 155, 342, 59, 342, 511, 592 and 1977), and five oncogenic pathways (CYCLIN, BCAT, CAMP, RB, NRL, EIF4E, and VEGF signaling pathways) significantly correlated with antitumor immune signatures.

CONCLUSION

These molecular features are potential markers of tumor immune status and response to immunotherapy.

Multi-Platform-Based Analysis Characterizes Molecular Alterations of the Nucleus in Human Colorectal Cancer

Background: The disturbed molecular alterations of nucleus may promote the development of colorectal cancer (CRC). A multi-platform-based analysis of nucleus of CRC patients helps us to better understand the underlying mechanism of CRC and screen out the potential drug targets for clinical treatment. However, such studies on nucleus in human CRC are still lacking.

Methods: We collected the cancerous and para-cancerous tissues from eight CRC patients and performed a multiplex analysis of the molecular changes of the nucleus, including structural variations (SVs), DNA methylation, chromatin accessibility, proteome and phosphorproteome.

Results: In our study, we revealed a significant molecular change of nucleus of CRC patients using our original proteomic and phosphorylomic datasets. Subsequently, we characterized the molecular alterations of nucleus of CRC patients at multiple dimensionalities, including DNA, mRNA, protein and epigenetic modification. Next, we found that the great molecular changes of nucleus might affect the biological processes named endocytosis and ubiquitin-mediated proteolysis. Besides, we identified DYNC1LI2 and TPR as the potentially hub proteins within the network of nuclear genes in CRC cells. Furthermore, we identified 1905 CRC-specific SVs, and proclaimed 17 CRC-specific SVs were probably associated with the disturbance of immune microenvironment of CRC patients. We also revealed that the SVs of CXCL5, CXCL10 and CXCL11 might be the core SVs among all the immune-relevant SVs. Finally, we identified seven genes as the upstream transcriptional factors potentially regulating the expression of nuclear genes, such as YY1 and JUN, using a multi-omics approach.

Conclusion: Here, we characterized the molecular changes of nucleus of CRC patients, disclosed the potentially core nuclear genes within the network, and identified the probable upstream regulator of nucleus. The findings of this study are helpful to understand the pathogenic molecular changes of nucleus in CRC patients and provide a functional context for drug development in future.

Gene Expression Profile Reveals a Prognostic Signature of Non–MSI-H/pMMR Colorectal Cancer

Studies have demonstrated that non–MSI-H/pMMR colorectal cancer (CRC) has a worse prognosis and relapse rate than microsatellite instability-high (MSI-H)/mismatch repair deficient (dMMR) CRC. Hence, searching for a novel tool to advance the prognostic management of non–MSI-H/pMMR CRC is vital. In this study, using three independent public cohorts and a clinical in-house cohort, we developed and validated a microsatellite stable–associated signature (MSSAS). The initial signature establishment was performed in GSE39582 (n = 454). This was followed by independent validation of this signature in The Cancer Genome Atlas–CRC (n = 312), GSE39084 (n = 54), and in-house cohort (n = 146). As a result, MSSAS was proven to be an independent risk factor for overall survival and relapse-free survival in non–MSI-H/pMMR CRC. Receiver operating characteristic analysis showed that MSSAS had a stable and accurate performance in all cohorts for 1, 3, and 5 years, respectively. Further analysis suggested that MSSAS performed better than age, gender, and the T, N, M, and AJCC stages, adjuvant chemotherapy, tumor mutation burden, neoantigen, and TP53, KRAS, BRAF, and PIK3CA mutations. The clinical validation was executed to further ensure the robustness and clinical feasibility of this signature. In conclusion, MSSAS might be a robust and promising biomarker for advancing clinical management of non–MSI-H/pMMR CRC.

Construction and validation of an immune-related genes prognostic index (IRGPI) model in colon cancer

BACKGROUND

Though immunotherapy has become one of the standard therapies for colon cancer, the overall effective rate of immunotherapy is very low. Constructing an immune-related genes prognostic index (IRGPI) model may help to predict the response to immunotherapy and clinical outcomes.

METHODS

Differentially expressed immune-related genes (DEIRGs) between normal tissues and colon cancer tissues were identified and used to construct the co-expression network. Genes in the module with the most significant differences were further analyzed. Independent prognostic immune-related genes (IRGs) were identified by univariate and multivariate cox regression analysis. Independent prognostic IRGs were used to construct the IRGPI model using the multivariate cox proportional hazards regression model, and the IRGPI model was validated by independent dataset. ROC curves were plotted and AUCs were calculated to estimate the predictive power of the IRGPI model to prognosis. Gene set enrichment analysis (GSEA) was performed to screen the enriched KEGG pathways in the high-risk and low-risk phenotype. Correlations between IRGPI and clinical characteristic, immune checkpoint expression, TMB, immune cell infiltration, immune function, immune dysfunction, immune exclusion, immune subtype were analyzed.

RESULTS

Totally 680 DEIRGs were identified. Three independent IRGs,NR5A2, PPARGC1A and LGALS4, were independently related to survival. NR5A2, PPARGC1A and LGALS4 were used to establish the IRGPI model. Survival analysis showed that patients with high-risk showed worse survival than patients in the low-risk group. The AUC of the IRGPI model for 1-year, 3-year and 5-year were 0.584, 0.608 and 0.697, respectively. Univariate analysis and multivariate cox regression analysis indicated that IRGPI were independent prognostic factors for survival. Stratified survival analysis showed that patients with IRGPI low-risk and low TMB had the best survival, which suggested that combination of TMB and IRGPI can better predict clinical outcome. Immune cell infiltration, immune function, immune checkpoint expression and immune exclusion were different between IRGPI high-risk and low-risk patients.

CONCLUSION

An immune-related genes prognostic index (IRGPI) was constructed and validated in the current study and the IRGPI maybe a potential biomarker for evaluating response to immunotherapy and clinical outcome for colon cancer patients.

Integrative analysis of the molecular mechanisms, immunological features and immunotherapy response of ferroptosis regulators across 33 cancer types

Ferroptosis is a recently described mode of cell death caused by the accumulation of intracellular iron and lipid reactive oxygen species (ROS), which play critical roles in tumorigenesis and cancer progression. However, the underlying molecular mechanisms and promising biomarkers of ferroptosis among cancers remain to be elucidated. In this study, 30 ferroptosis regulators in ferroptosis-related signaling pathways were identified and analyzed in 33 cancer types. We found transcriptomic aberrations and evaluated the prognostic value of ferroptosis regulators across 33 cancer types. Then, we predicted and validated potential transcription factors (including E2F7, KLF5 and FOXM1) and therapeutic drugs (such as cyclophosphamide, vinblastine, and gefitinib) that target ferroptosis regulators in cancer. Moreover, we explored the molecular mechanisms of ferroptosis and found that signaling pathways such as the IL-1 and IL-2 pathways are closely associated with ferroptosis. Additionally, we found that ferroptosis regulators have a close relationship with immunity-related parameters, including the immune score, immune cell infiltration level, and immune checkpoint protein level. Finally, we determined a ferroptosis score using GSVA method. We found that the ferroptosis score effectively predicted ferroptotic cell death in tumor samples. And ferroptosis score is served as an independent prognostic indicator for the incidence and recurrence of cancers. More importantly, patients with high ferroptosis scores received greater benefit from immunotherapy. We aslo created an online webserver based on the nomogram prognostic model to predict the survival in immunotherapy cohort. The reason for this outcome is partially the result of patients with a high ferroptosis rate also having high immune scores, HLA-related gene expression and immune checkpoint protein expression, such as PDL2 and TIM3. Moreover, patients with high ferroptosis scores exhibited CD8 T cell and TIL infiltration and immune-related signaling pathway enrichment. In summary, we systematically summarize the molecular characteristics, clinical relevance and immune features of ferroptosis across cancers and show that the ferroptosis score can be used as a prognostic factor and for the evaluation of immunotherapy effects.

Artificial intelligence in pathological evaluation of gastrointestinal cancers

The integration of artificial intelligence (AI) has shown promising benefits in many fields of diagnostic histopathology, including for gastrointestinal cancers (GCs), such as tumor identification, classification, and prognosis prediction. In parallel, recent evidence suggests that AI may help reduce the workload in gastrointestinal pathology by automatically detecting tumor tissues and evaluating prognostic parameters. In addition, AI seems to be an attractive tool for biomarker/genetic alteration prediction in GC, as it can contain a massive amount of information from visual data that is complex and partially understandable by pathologists. From this point of view, it is suggested that advances in AI could lead to revolutionary changes in many fields of pathology. Unfortunately, these findings do not exclude the possibility that there are still many hurdles to overcome before AI applications can be safely and effectively applied in actual pathology practice. These include a broad spectrum of challenges from needs identification to cost-effectiveness. Therefore, unlike other disciplines of medicine, no histopathology-based AI application, including in GC, has ever been approved either by a regulatory authority or approved for public reimbursement. The purpose of this review is to present data related to the applications of AI in pathology practice in GC and present the challenges that need to be overcome for their implementation.

Evaluating Cancer-Related Biomarkers Based on Pathological Images: A Systematic Review

Many diseases are accompanied by changes in certain biochemical indicators called biomarkers in cells or tissues. A variety of biomarkers, including proteins, nucleic acids, antibodies, and peptides, have been identified. Tumor biomarkers have been widely used in cancer risk assessment, early screening, diagnosis, prognosis, treatment, and progression monitoring. For example, the number of circulating tumor cell (CTC) is a prognostic indicator of breast cancer overall survival, and tumor mutation burden (TMB) can be used to predict the efficacy of immune checkpoint inhibitors. Currently, clinical methods such as polymerase chain reaction (PCR) and next generation sequencing (NGS) are mainly adopted to evaluate these biomarkers, which are time-consuming and expansive. Pathological image analysis is an essential tool in medical research, disease diagnosis and treatment, functioning by extracting important physiological and pathological information or knowledge from medical images. Recently, deep learning-based analysis on pathological images and morphology to predict tumor biomarkers has attracted great attention from both medical image and machine learning communities, as this combination not only reduces the burden on pathologists but also saves high costs and time. Therefore, it is necessary to summarize the current process of processing pathological images and key steps and methods used in each process, including: (1) pre-processing of pathological images, (2) image segmentation, (3) feature extraction, and (4) feature model construction. This will help people choose better and more appropriate medical image processing methods when predicting tumor biomarkers.

A New Risk Score Based on Eight Hepatocellular Carcinoma- Immune Gene Expression Can Predict the Prognosis of the Patients

Background

Hepatocellular carcinoma (HCC) is one of the malignant tumors with high morbidity and mortality worldwide. Immunotherapy has emerged as an increasingly important cancer treatment modality. However, the potential relationship between immune genes and HCC still needs to be explored. The purpose of this study is to construct a new prognostic risk signature to predict the prognosis of HCC patients based on the expression of immune-related genes (IRGs) and explore its potential mechanism.

Methods

We analyzed the gene expression data of 332 HCC patient samples and 46 adjacent normal tissues samples (Solid Tissue Normal including cirrhotic tissue) in The Cancer Genome Atlas (TCGA) database and clinical characteristics. We analyzed the gene expression data, identified differentially expressed IRGs in HCC tissues, filtered IRGs with prognostic value to construct an IRG signature, and classified patients into high and low gene expression groups based on the expression of IRGs in their tumor tissues. We also investigated the potential molecular mechanisms of IRGs through a bioinformatics approach using Protein-Protein Interaction (PPI) network, Kyoto Encyclopedia of Genes and Genomes (KEGG) database analysis and Gene Ontology (GO) database analysis. Differentially expressed IRGs associated with significant clinical outcomes (SIRGs) were identified by univariate Cox regression analysis. An immune-related risk score model (IRRSM) was established based on Lasso Cox regression analysis and multivariate Cox regression analysis. Based on the IRRSM, the immune score of the patients was calculated, and the patients were divided into high-risk and low-risk patients according to the median score, and the differences in survival between the two groups were compared. Then, the correlation analysis between the IRRSM and clinical characteristics was performed, and the IRRSM was validated using the International Cancer Genome Consortium (ICGC) database.

Results

The IRRSM was eventually constructed and confirmed to be an independent prognostic model for HCC patients. The IRRSM was shown to be positively correlated with the infiltration of four types of immune cells.

Conclusion

Our results showed that some SIRGs have potential value for predicting the prognosis and clinical outcomes of HCC patients. IRGs affect the prognosis of HCC patients by regulating the tumor immune microenvironment (TIME). This study provides a new insight for immune research and treatment strategies in HCC patients.

SLAMF8 expression predicts the efficacy of anti-PD1 immunotherapy in gastrointestinal cancers

Objectives

Epstein–Barr virus (EBV) infection is associated with a better response to anti‐PD1 immunotherapy. We hypothesised that genetic alterations induced by EBV infection are responsible for the activation of key immune responses and hence are predictive of anti‐PD1 efficacy.

Methods

With transcriptome data of gastric cancer (GC), we explored differentially expressed genes (DEGs) specific for EBV infection and performed coexpression network analysis using the DEGs to identify the consistent coexpression genes (CCGs) between EBV‐positive and EBV‐negative GC tissues. We selected the tag genes of the CCGs and validated them using RNA sequencing and immunohistochemistry. We established murine models and collected tissues from clinical patients to test the value of SLAMF8 in predicting anti‐PD1 treatment. The location and expression of SLAMF8 were characterised by multiplex immunofluorescence and quantitative PCR. Moreover, exogenous overexpression and RNA‐sequencing analysis were used to test the potential function of SLAMF8.

Results

We identified 290 CCGs and validated the tag gene SLAMF8 in transcriptome data of gastrointestinal cancer (GI). We observed that the T‐cell activation pathway was significantly enriched in high‐expression SLAMF8 GI cancers. Higher SLAMF8 expression was positively associated with CD8 expression and a better response to anti‐PD1 treatment. We further observed dynamically increased expression of SLAMF8 in murine models relatively sensitive to anti‐PD1 treatment. SLAMF8 was mainly expressed on the surface of macrophages. Exogenous overexpression of SLAMF8 in macrophages resulted in enrichment of positive regulation of multiple immune‐related pathways.

Conclusion

Higher SLAMF8 expression may predict better anti‐PD1 immunotherapy efficacy in GI cancer.

Cost-effectiveness of treatment optimisation with biomarkers for immunotherapy in solid tumours: a systematic review protocol

INTRODUCTION

The combination of biomarkers and drugs is the subject of growing interest both from regulators, physicians and companies. This study protocol of a systematic review is aimed to describe available literature evidences about the cost-effectiveness, cost-utility or net-monetary benefit of the use of biomarkers in solid tumour as tools for customising immunotherapy to identify what further research needs.

METHODS AND ANALYSIS

A systematic review of the literature will be carried out according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement guidelines. PubMed and Embase will be queried from June 2010 to June 2021. The PICOS model will be applied: target population (P) will be patients with solid tumours treated with immune checkpoint inhibitors (ICIs); the interventions (I) will be test of the immune checkpoint predictive biomarkers; the comparator (C) will be any other targeted or non-targeted therapy; outcomes (O) evaluated will be health economic and clinical implications assessed in terms of incremental cost-effectiveness ratio, net health benefit, net monetary benefit, life years gained, quality of life, etc; study (S) considered will be economic evaluations reporting cost-effectiveness analysis, cost-utility analysis, net-monetary benefit. The quality of the evidence will be graded according to Grading of Recommendations Assessment, Development and Evaluation.

ETHICS AND DISSEMINATION

This systematic review will assess the cost-effectiveness implications of using biomarkers in the immunotherapy with ICIs, which may help to understand whether this approach is widespread in real clinical practice. This research is exempt from ethics approval because the work is carried out on published documents. We will disseminate this protocol in a related peer-reviewed journal.

PROSPERO REGISTRATION NUMBER

CRD42020201549.

Deep Learning of Histopathological Features for the Prediction of Tumour Molecular Genetics

Advanced diagnostics are enabling cancer treatments to become increasingly tailored to the individual through developments in immunotherapies and targeted therapies. However, long turnaround times and high costs of molecular testing hinder the widespread implementation of targeted cancer treatments. Meanwhile, gold-standard histopathological assessment carried out by a trained pathologist is widely regarded as routine and mandatory in most cancers. Recently, methods have been developed to mine hidden information from histopathological slides using deep learning applied to scanned and digitized slides; deep learning comprises a collection of computational methods which learn patterns in data in order to make predictions. Such methods have been reported to be successful in a variety of cancers for predicting the presence of biomarkers such as driver mutations, tumour mutational burden, and microsatellite instability. This information could prove valuable to pathologists and oncologists in clinical decision making for cancer treatment and triage for in-depth sequencing. In addition to identifying molecular features, deep learning has been applied to predict prognosis and treatment response in certain cancers. Despite reported successes, many challenges remain before the clinical implementation of such diagnostic strategies in the clinical setting is possible. This review aims to outline recent developments in the field of deep learning for predicting molecular genetics from histopathological slides, as well as to highlight limitations and pitfalls of working with histopathology slides in deep learning.

Artificial Intelligence for solid tumor diagnosis in digital pathology

Tumor diagnosis relies on the visual examination of histological slides by pathologists through a microscope eyepiece. Digital pathology, the digitalization of histological slides at high magnification with slides scanners, has raised the opportunity to extract quantitative information thanks to image analysis. In the last decade, medical image analysis has made exceptional progress due to the development of artificial intelligence (AI) algorithms. AI has been successfully used in the field of medical imaging and more recently in digital pathology. The feasibility and usefulness of AI assisted pathology tasks have been demonstrated in the very last years and we can expect those developments to be applied on routine histopathology in the future. In this review, we will describe and illustrate this technique and present the most recent applications in the field of tumor histopathology.