Search

Background: The progress in Colorectal cancer (CRC) screening and management has resulted in an unprecedented caseload for histopathological diagnosis. While artificial intelligence (AI) presents a potential solution, the predominant emphasis on slide-level aggregation performance without thorough verification of cancer in each location, impedes both explainability and transparency. Effectively addressing these challenges is crucial to ensuring the reliability and efficacy of AI in histology applications. Method: In this study, we created an innovative AI algorithm using transfer learning from a polyp segmentation model in endoscopy. The algorithm precisely localized CRC targets within 0.25 mm2 grids from whole slide imaging (WSI). We assessed the CRC detection capabilities at this fine granularity and examined the influence of AI on the diagnostic behavior of pathologists. The evaluation utilized an extensive dataset comprising 858 consecutive patient cases with 1418 WSIs obtained from an external center. Results: Our results underscore a notable sensitivity of 90.25% and specificity of 96.60% at the grid level, accompanied by a commendable area under the curve (AUC) of 0.962. This translates to an impressive 99.39% sensitivity at the slide level, coupled with a negative likelihood ratio of <0.01, signifying the dependability of the AI system to preclude diagnostic considerations. The positive likelihood ratio of 26.54, surpassing 10 at the grid level, underscores the imperative for meticulous scrutiny of any AI-generated highlights. Consequently, all four participating pathologists demonstrated statistically significant diagnostic improvements with AI assistance. Conclusion: Our transfer learning approach has successfully yielded an algorithm that can be validated for CRC histological localizations in whole slide imaging. The outcome advocates for the integration of the AI system into histopathological diagnosis, serving either as a diagnostic exclusion application or a computer-aided detection (CADe) tool. This integration has the potential to alleviate the workload of pathologists and ultimately benefit patients. [ABSTRACT FROM AUTHOR]

Colorectal cancer (CRC) is one of the most common cancers worldwide and demands more effective treatments. We sought to identify tumor selective CRC antigens and their therapeutic potential for cytotoxic T-cell targeting by transcriptomic and immunohistochemical analysis. LY6G6D was identified as a tumor selectively expressed CRC antigen, mainly in the microsatellite stable (MSS) subtype. A specific anti LY6G6D/CD3 T cell engager (TcE) was generated and demonstrated potent tumor cell killing and T cell activation in vitro. Ex vivo treatment of primary patient-derived CRC tumor slice cultures with the LY6G6D/CD3 TcE led to IFNgγ secretion in LY6G6D positive tumor samples. In vivo, LY6G6D/CD3 TcE monotherapy demonstrated tumor regressions in pre-clinical mouse models of engrafted human CRC tumor cells and PBMCs. Lastly, 2D and 3D cocultures of LY6G6D positive and negative cells were used to explore the bystander killing of LY6G6D negative cells after specific activation of T cells by LY6G6D positive cells. LY6G6D/CD3 TcE treatment was shown to lyse target negative cells in the vicinity of target positive cells through a combined effect of IFNgγ, TNFa and Fas/FasL. In summary, LY6G6D was identified as a selectively expressed CRC antigen that can be utilized to potently re-direct and activate cytotoxic T-cells to lyse LY6G6D expressing CRC using a TcE. This effect can be spread to target negative neighboring tumor cells, potentially leading to improved therapeutic efficacy. [ABSTRACT FROM AUTHOR]

Colorectal cancer (CRC) is one of the most common cancers worldwide and demands more effective treatments. We sought to identify tumor selective CRC antigens and their therapeutic potential for cytotoxic T-cell targeting by transcriptomic and immunohistochemical analysis. LY6G6D was identified as a tumor selectively expressed CRC antigen, mainly in the microsatellite stable (MSS) subtype. A specific anti LY6G6D/CD3 T cell engager (TcE) was generated and demonstrated potent tumor cell killing and T cell activation in vitro. Ex vivo treatment of primary patient-derived CRC tumor slice cultures with the LY6G6D/CD3 TcE led to IFNγ secretion in LY6G6D positive tumor samples. In vivo, LY6G6D/CD3 TcE monotherapy demonstrated tumor regressions in pre-clinical mouse models of engrafted human CRC tumor cells and PBMCs. Lastly, 2D and 3D cocultures of LY6G6D positive and negative cells were used to explore the bystander killing of LY6G6D negative cells after specific activation of T cells by LY6G6D positive cells. LY6G6D/CD3 TcE treatment was shown to lyse target negative cells in the vicinity of target positive cells through a combined effect of IFNγ, TNFα and Fas/FasL. In summary, LY6G6D was identified as a selectively expressed CRC antigen that can be utilized to potently re-direct and activate cytotoxic T-cells to lyse LY6G6D expressing CRC using a TcE. This effect can be spread to target negative neighboring tumor cells, potentially leading to improved therapeutic efficacy.

Approximately 25% of colorectal cancer (CRC) patients experience systemic metastases, with the most frequent target organs being the liver and lung. Metabolic reprogramming has been recognized as one of the hallmarks of cancer. Here, metabolic and functional differences between two CRC cells with different metastatic organotropisms (metastatic KM12SM CRC cells to the liver and KM12L4a to the lung when injected in the spleen and in the tail vein of mice) were analysed in comparison to their parental non-metastatic isogenic KM12C cells, for a subsequent investigation of identified metabolic targets in CRC patients. Meta-analysis from proteomic and transcriptomic data deposited in databases, qPCR, WB, in vitro cell-based assays, and in vivo experiments were used to survey for metabolic alterations contributing to their different organotropism and for the subsequent analysis of identified metabolic markers in CRC patients. Although no changes in cell proliferation were observed between metastatic cells, KM12SM cells were highly dependent on oxidative phosphorylation at mitochondria, whereas KM12L4a cells were characterized by being more energetically efficient with lower basal respiration levels and a better redox management. Lipid metabolism-related targets were found altered in both cell lines, including LDLR, CD36, FABP4, SCD, AGPAT1, and FASN, which were also associated with the prognosis of CRC patients. Moreover, CD36 association with lung metastatic tropism of CRC cells was validated in vivo. Altogether, our results suggest that LDLR, CD36, FABP4, SCD, FASN, LPL, and APOA1 metabolic targets are associated with CRC metastatic tropism to the liver or lung. These features exemplify specific metabolic adaptations for invasive cancer cells which stem at the primary tumour.

There is a pressing need for novel immunotherapeutic targets in colorectal cancer (CRC). Cytotoxic T cell infiltration is well established as a key prognostic indicator in CRC, and it is known that these tumor infiltrating lymphocytes (TILs) target and kill tumor cells. However, the specific antigens that drive these CD8+ T cell responses have not been well characterized. Recently, phosphopeptides have emerged as strong candidates for tumor-specific antigens, as dysregulated signaling in cancer leads to increased and aberrant protein phosphorylation. Here, we identify 120 HLA-I phosphopeptides from primary CRC tumors, CRC liver metastases and CRC cell lines using mass spectrometry and assess the tumor-resident immunity against these posttranslationally modified tumor antigens. Several CRC tumor-specific phosphopeptides were presented by multiple patients’ tumors in our cohort (21% to 40%), and many have previously been identified on other malignancies (58% of HLA-A*02 CRC phosphopeptides). These shared antigens derived from mitogenic signaling pathways, including p53, Wnt and MAPK, and are therefore markers of malignancy. The identification of public tumor antigens will allow for the development of broadly applicable targeted therapeutics. Through analysis of TIL cytokine responses to these phosphopeptides, we have established that they are already playing a key role in tumor-resident immunity. Multifunctional CD8+ TILs from primary and metastatic tumors recognized the HLA-I phosphopeptides presented by their originating tumor. Furthermore, TILs taken from other CRC patients’ tumors targeted two of these phosphopeptides. In another cohort of CRC patients, the same HLA-I phosphopeptides induced higher peripheral T cell responses than they did in healthy donors, suggesting that these immune responses are specifically activated in CRC patients. Collectively, these results establish HLA-I phosphopeptides as targets of the tumor-resident immunity in CRC, and highlight their potential as candidates for future immunotherapeutic strategies.

There is a pressing need for novel immunotherapeutic targets in colorectal cancer (CRC). Cytotoxic T cell infiltration is well established as a key prognostic indicator in CRC, and it is known that these tumor infiltrating lymphocytes (TILs) target and kill tumor cells. However, the specific antigens that drive these CD8+ T cell responses have not been well characterized. Recently, phosphopeptides have emerged as strong candidates for tumor-specific antigens, as dysregulated signaling in cancer leads to increased and aberrant protein phosphorylation. Here, we identify 120 HLA-I phosphopeptides from primary CRC tumors, CRC liver metastases and CRC cell lines using mass spectrometry and assess the tumor-resident immunity against these posttranslationally modified tumor antigens. Several CRC tumor-specific phosphopeptides were presented by multiple patients' tumors in our cohort (21% to 40%), and many have previously been identified on other malignancies (58% of HLA-A*02 CRC phosphopeptides). These shared antigens derived from mitogenic signaling pathways, including p53, Wnt and MAPK, and are therefore markers of malignancy. The identification of public tumor antigens will allow for the development of broadly applicable targeted therapeutics. Through analysis of TIL cytokine responses to these phosphopeptides, we have established that they are already playing a key role in tumor-resident immunity. Multifunctional CD8+ TILs from primary and metastatic tumors recognized the HLA-I phosphopeptides presented by their originating tumor. Furthermore, TILs taken from other CRC patients' tumors targeted two of these phosphopeptides. In another cohort of CRC patients, the same HLA-I phosphopeptides induced higher peripheral T cell responses than they did in healthy donors, suggesting that these immune responses are specifically activated in CRC patients. Collectively, these results establish HLA-I phosphopeptides as targets of the tumor-resident immunity in CRC, and highlight their potential as candidates for future immunotherapeutic strategies. [ABSTRACT FROM AUTHOR]

Background: The progress in Colorectal cancer (CRC) screening and management has resulted in an unprecedented caseload for histopathological diagnosis. While artificial intelligence (AI) presents a potential solution, the predominant emphasis on slide-level aggregation performance without thorough verification of cancer in each location, impedes both explainability and transparency. Effectively addressing these challenges is crucial to ensuring the reliability and efficacy of AI in histology applications., Method: In this study, we created an innovative AI algorithm using transfer learning from a polyp segmentation model in endoscopy. The algorithm precisely localized CRC targets within 0.25 mm² grids from whole slide imaging (WSI). We assessed the CRC detection capabilities at this fine granularity and examined the influence of AI on the diagnostic behavior of pathologists. The evaluation utilized an extensive dataset comprising 858 consecutive patient cases with 1418 WSIs obtained from an external center., Results: Our results underscore a notable sensitivity of 90.25% and specificity of 96.60% at the grid level, accompanied by a commendable area under the curve (AUC) of 0.962. This translates to an impressive 99.39% sensitivity at the slide level, coupled with a negative likelihood ratio of <0.01, signifying the dependability of the AI system to preclude diagnostic considerations. The positive likelihood ratio of 26.54, surpassing 10 at the grid level, underscores the imperative for meticulous scrutiny of any AI-generated highlights. Consequently, all four participating pathologists demonstrated statistically significant diagnostic improvements with AI assistance., Conclusion: Our transfer learning approach has successfully yielded an algorithm that can be validated for CRC histological localizations in whole slide imaging. The outcome advocates for the integration of the AI system into histopathological diagnosis, serving either as a diagnostic exclusion application or a computer-aided detection (CADe) tool. This integration has the potential to alleviate the workload of pathologists and ultimately benefit patients., Competing Interests: Authors XX and JL declare as equity holders of Wision Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2024 Yuan, Zhou, Xiao, Zhang, Chen, Liu, Liu, Bao and Tao.)

The chimeric antigen receptor (CAR) is an artificial molecule engineered to induce cytolytic T cell reactions in tumors. Generally, this molecule combines an extracellular single-chain variable fragment (scFv) able to recognize tumor-associated epitopes together with the intracellular signaling domains that are required for T cell activation. When expressed by T cells, the CAR enables the recognition and subsequent destruction of cancer cells expressing the complementary antigen on their surface. Although the clinical application for CAR T cells is currently limited to some hematological malignancies, researchers are trying to develop CAR T cell-based therapies for the treatment of solid tumors. However, while in the case of CD19, or other targets restricted to the hematopoietic compartment, the toxicity is limited and manageable, the scarcity of specific antigens expressed by solid tumors and not by healthy cells from vital organs makes the clinical development of CAR T cells in this context particularly challenging. Here we summarize relevant research and clinical trials conducted to redirect CAR T cells to surface antigens in solid tumors and cancer stem cells with a focus on colorectal cancer and glioblastoma. Finally, we will discuss current knowledge of altered glycosylation of CSCs and cancer cells and how these novel epitopes may help to target CAR T cell-based immunotherapy in the future.

Fusobacterium nucleatum is a common oral bacterium that is enriched in colorectal adenomas and adenocarcinomas (CRC). In humans, high fusobacterial CRC abundance is associated with chemoresistance and poor prognosis. In animal models, fusobacteria accelerate CRC progression. Targeting F. nucleatum may reduce fusobacteria cancer progression and therefore determining the origin of CRC F. nucleatum and the route by which it reaches colon tumors is of biologic and therapeutic importance. Arbitrarily primed PCR performed previously on matched same-patients CRC and saliva F. nucleatum isolates, suggested that CRC F. nucleatum may originate from the oral cavity. However, the origin of CRC fusobacteria as well as the route of their arrival to the tumor have not been well-established. Herein, we performed and analyzed whole genome sequencing of paired, same-patient oral, and CRC F. nucleatum isolates and confirmed that CRC-fusobacteria originate from the oral microbial reservoir. Oral fusobacteria may translocate to CRC by descending via the digestive tract or using the hematogenous route during frequent transient bacteremia caused by chewing, daily hygiene activities, or dental procedures. Using the orthotropic CT26 mouse model we previously showed that IV injected F. nucleatum colonize CRC. Here, we compared CRC colonization by gavage vs. intravenous inoculated F. nucleatum in the MC38 and CT26 mouse orthotropic CRC models. Under the tested conditions, hematogenous fusobacteria were more successful in CRC colonization than gavaged ones. Our results therefore provide evidence that the hematogenous route may be the preferred way by which oral fusobacteria reach colon tumors.

The chimeric antigen receptor (CAR) is an artificial molecule engineered to induce cytolytic T cell reactions in tumors. Generally, this molecule combines an extracellular single-chain variable fragment (scFv) able to recognize tumor-associated epitopes together with the intracellular signaling domains that are required for T cell activation. When expressed by T cells, the CAR enables the recognition and subsequent destruction of cancer cells expressing the complementary antigen on their surface. Although the clinical application for CAR T cells is currently limited to some hematological malignancies, researchers are trying to develop CAR T cell-based therapies for the treatment of solid tumors. However, while in the case of CD19, or other targets restricted to the hematopoietic compartment, the toxicity is limited and manageable, the scarcity of specific antigens expressed by solid tumors and not by healthy cells from vital organs makes the clinical development of CAR T cells in this context particularly challenging. Here we summarize relevant research and clinical trials conducted to redirect CAR T cells to surface antigens in solid tumors and cancer stem cells with a focus on colorectal cancer and glioblastoma. Finally, we will discuss current knowledge of altered glycosylation of CSCs and cancer cells and how these novel epitopes may help to target CAR T cell-based immunotherapy in the future. [ABSTRACT FROM AUTHOR]

Fusobacterium nucleatum is a common oral bacterium that is enriched in colorectal adenomas and adenocarcinomas (CRC). In humans, high fusobacterial CRC abundance is associated with chemoresistance and poor prognosis. In animal models, fusobacteria accelerate CRC progression. Targeting F. nucleatum may reduce fusobacteria cancer progression and therefore determining the origin of CRC F. nucleatum and the route by which it reaches colon tumors is of biologic and therapeutic importance. Arbitrarily primed PCR performed previously on matched same-patients CRC and saliva F. nucleatum isolates, suggested that CRC F. nucleatum may originate from the oral cavity. However, the origin of CRC fusobacteria as well as the route of their arrival to the tumor have not been well-established. Herein, we performed and analyzed whole genome sequencing of paired, same-patient oral, and CRC F. nucleatum isolates and confirmed that CRC-fusobacteria originate from the oral microbial reservoir. Oral fusobacteria may translocate to CRC by descending via the digestive tract or using the hematogenous route during frequent transient bacteremia caused by chewing, daily hygiene activities, or dental procedures. Using the orthotropic CT26 mouse model we previously showed that IV injected F. nucleatum colonize CRC. Here, we compared CRC colonization by gavage vs. intravenous inoculated F. nucleatum in the MC38 and CT26 mouse orthotropic CRC models. Under the tested conditions, hematogenous fusobacteria were more successful in CRC colonization than gavaged ones. Our results therefore provide evidence that the hematogenous route may be the preferred way by which oral fusobacteria reach colon tumors. [ABSTRACT FROM AUTHOR]

Approximately 25% of colorectal cancer (CRC) patients experience systemic metastases, with the most frequent target organs being the liver and lung. Metabolic reprogramming has been recognized as one of the hallmarks of cancer. Here, metabolic and functional differences between two CRC cells with different metastatic organotropisms (metastatic KM12SM CRC cells to the liver and KM12L4a to the lung when injected in the spleen and in the tail vein of mice) were analysed in comparison to their parental non-metastatic isogenic KM12C cells, for a subsequent investigation of identified metabolic targets in CRC patients. Meta-analysis from proteomic and transcriptomic data deposited in databases, qPCR, WB, in vitro cell-based assays, and in vivo experiments were used to survey for metabolic alterations contributing to their different organotropism and for the subsequent analysis of identified metabolic markers in CRC patients. Although no changes in cell proliferation were observed between metastatic cells, KM12SM cells were highly dependent on oxidative phosphorylation at mitochondria, whereas KM12L4a cells were characterized by being more energetically efficient with lower basal respiration levels and a better redox management. Lipid metabolism-related targets were found altered in both cell lines, including LDLR , CD36 , FABP4 , SCD , AGPAT1 , and FASN , which were also associated with the prognosis of CRC patients. Moreover, CD36 association with lung metastatic tropism of CRC cells was validated in vivo. Altogether, our results suggest that LDLR , CD36 , FABP4 , SCD , FASN , LPL , and APOA1 metabolic targets are associated with CRC metastatic tropism to the liver or lung. These features exemplify specific metabolic adaptations for invasive cancer cells which stem at the primary tumour., Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The handling editor JI-G declared a shared parent affiliation with the author MF-A at the time of review., (Copyright © 2022 Montero-Calle, Gómez de Cedrón, Quijada-Freire, Solís-Fernández, López-Alonso, Espinosa-Salinas, Peláez-García, Fernández-Aceñero, Ramírez de Molina and Barderas.)

Advances in immunotherapy have made an unprecedented leap in treating colorectal cancer (CRC). However, more effective therapeutic regimes need a deeper understanding of molecular architectures for precise patient stratification and therapeutic improvement. We profiled patients receiving neoadjuvant chemotherapy alone or in combination with immunotherapy (PD-1 checkpoint inhibitor) using Digital Spatial Profiler (DSP), a high-plex spatial proteogenomic technology. Compartmentalization-based high-plex profiling of both proteins and mRNAs revealed pronounced immune infiltration at tumor regions associated with immunotherapy treatment. The protein and the corresponding mRNA levels within the same selected regions of those patient samples correlate, indicating an overall concordance between the transcriptional and translational levels. An elevated expression of PD-L1 at both protein and the mRNA levels was discovered in the tumor compartment of immunotherapy-treated patients compared with chemo-treated patients, indicating potential prognostic biomarkers are explorable in a spatial manner at the local tumor microenvironment (TME). An elevated expression of PD-L1 was verified by immunohistochemistry. Other compartment-specific biomarkers were also differentially expressed between the tumor and stromal regions, indicating a dynamic interplay that can potentiate novel biomarker discovery from the TME perspectives. Simultaneously, a high-plex spatial profiling of protein and mRNA in the tumor microenvironment of colorectal cancer was performed., Competing Interests: Authors NW, RW, XL, ZS, LX, and ZD were employed by the company Fynn Biotechnologies Ltd. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest., (Copyright © 2021 Wang, Wang, Li, Song, Xia, Wang, Zhang, Wu and Ding.)