Partial advances in the diagnosis and treatment of gastrointestinal cancer

Gastrointestinal cancers are difficult to be cured with a high recurrence rate accounting for more than 50% of global cancer-related morbidity and mortality.¹ Many patients with gastrointestinal cancer are diagnosed at a late stage. Over the past few decades, basic and clinical research with new technologies have made significant progress in the diagnosis and treatment of gastrointestinal cancers, which significantly improved the quality of life and prolonged the survival of patients. Here, we briefly highlight several advances in diagnosing and treating gastrointestinal tumors, mainly gastric cancer and colorectal cancer from multiple perspectives.

Early diagnosis is crucial for the treatment of gastrointestinal cancers. With the continuous advances in molecular biology, genomics, and epigenetics, individualized diagnosis of gastric cancer and colorectal cancer holds promise for basic research and clinical applications. Liquid biopsy, including detection of circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), tumor-related extracellular vesicles (exosomes and microvesicles), tumor-educated platelets, proteins as well as metabolites in a range of bodily fluids offers a cost-efficient and noninvasive approach to screening tumor and monitor relapse and response to treatment.² CTCs are tumor cells in peripheral blood, falling off from the solid tumor focus (primary focus or metastatic focus) due to spontaneous or diagnostic and treatment procedures.³ Most of the CTCs undergo apoptosis or phagocytosis and are engulfed by immune cells after entering the peripheral blood; only a few can escape and invade a distant organ to form metastatic foci, increasing the risk of death in patients with gastrointestinal cancers.³ Recent studies have shown that CTCs are heterogenic and can proliferate in vivo and in vitro. These, together with the findings from the single-cell molecular analysis, have provided unique insights into the biology of cancer metastasis and therapeutic response.

ctDNA may reflect tumor-specific abnormalities, and analysis of ctDNA can be applied in the diagnosis, therapeutic response, and prognosis of cancer patients. The mutations in specific genes have been detected in the plasma of patients with several types of gastrointestinal cancers, suggesting that ctDNA may be a possible biomarker of gastrointestinal cancers. The minimal residual disease (MRD) is a microscopic focus of treatment-insensitive tumor cells or the early recurrence of tumors.⁴ Hence, precise evaluation of MRD is especially significant during or after the treatment of gastrointestinal tumors.⁵ The detection of ctDNA and CTCs can help identify and explain the nature of MRD and may provide a new strategy for the prevention and treatment of tumor metastasis.^{3, 5} Nevertheless, ctDNA fragments usually have a short half-life, and currently, there are no convincing cut-off values to discriminate the boundary between high and low ctDNA concentrations. Therefore, the clinical value in the detection of ctDNA has not been gained wide recognition yet.⁶ Another attractive lipid biopsy item is exosomes derived from cancer cells which can promote tumor cell invasion and metastasis by mediating the interaction between tumor cells and the intercellular matrix; exosomes are also closely related to epithelial-mesenchymal transformation (EMT), tumor angiogenesis, and chemotherapeutic drug resistance.⁷ The Food and Drug Administration (FDA) has approved several single-gene and multigene tests as complementary diagnoses for cancer-specific molecular targeted therapy, marking the widespread clinical use of liquid biopsies, especially in patients with advanced cancer.^{2, 7}

AI, especially machine learning and deep learning, has opened up new avenues for scientific and clinical research. With increasingly multidimensional data generated in daily treatment, AI can aid clinicians to form a personalized perspective on patients' treatment paths and ultimately guide clinical decision-making. These decisions rely on integrating different and complex data streams, including clinical manifestations, patient history, pathological results, genomics, medical imaging, and matching these data with the conclusions of the growing scientific literature. Early data and computational limitations often require the simplification of unstructured patient data, such as medical images and biopsies, to a set of discrete observations of the degree of disease that humans can understand. A prominent example of this simplification is the cancer staging system, most notably the TNM classification of the American Joint Commission on Cancer (AJCC).⁸ AI has been used to improve the accuracy of liquid biopsy analysis of the data from CTC and ctDNA tests, which may facilitate its integration into clinical workflows in the future. For example, machine learning has been applied to identify CTCs in cancer patients,⁹ analyze ctDNA for cancer detection and location, comprehensive multigroup analysis, liquid biopsy tests and other clinical genetics, metabolomics, immunology, microbiology, and steady-state data to guide treatment decisions. Currently, the FDA has approved a few AI applications for oncology indications and many applications are underway. There is a great interest in simplifying the gap between the development of AI and clinical transformation. Consequently, the FDA is designing specific guidelines for AI and machine learning for the approved clinical applications.¹⁰ The prospect of AI is bright, but there are still many challenges in successfully integrating AI into clinical oncology, it may be only suitable for some particular cancer problems and its specific value may stem from specific cases in clinics.¹¹ Future studies should analyze the prospects, successes, and failures of AI as well as translate them clinically on a case-by-case basis. The further development of AI applications in cancer diagnosis and treatment may benefit the clinical management of gastrointestinal cancers.

In 2018, the European Society for Medical Oncology (ESMO) and oncologists from Asian countries revised the guidelines 2016 version, based on scientific evidence, for better management of gastric cancer. The main change in the 2019 ESMO metastatic gastric cancer guidelines updated immunotherapy as a treatment for gastric cancer.¹² According to the tumor immunoediting theory, immune checkpoint inhibitors (ICIs) have become effective strategies for the treatment of gastrointestinal tumors. The most commonly used ICIs include monoclonal antibodies programmed cell death protein-1 (PD-1)/programmed death protein ligand-1 (PD-L1) and cytotoxic T-lymphocyte-associated protein-4 (CTLA-4). In addition, other immunotherapies are available, including adoptive T cell therapies (ACT), such as adoptive transfer of chimeric antigen receptor (CAR)- and T cell receptor (TCR)-T cells, as well as bulk tumor-infiltrating lymphocyte (TIL) therapy. Unfortunately, the response rates of immunotherapy are not satisfactory, and some patients may lose their therapeutic effects after a period of treatment. Moreover, gastrointestinal cancers usually have a low response rate to immunotherapies because cancer cells may evade the immune response, contributing to immune tolerance.¹³ Further exploration of the mechanisms by which cancer cells are resistant to immunotherapies may provide a basis for the development of new promising immunotherapies for gastrointestinal cancers.¹⁴ A previous study has revealed that the gut microbiota, a complex community of trillions of microorganisms, plays a crucial role in immunotherapeutic responses in patients with gastrointestinal cancer.¹⁵ Besides physiological functions, the gut microbiota can regulate oncogenesis and resistance to antitumor therapies. During antitumor therapies, many medicines are commonly used for reducing adverse events and these medicines include antidiabetics, aspirin, proton pump inhibitors, psychotropic drugs, and nonsteroidal anti-inflammatory drugs, and analgesics such as opioids. Interestingly, treatment with these drugs can significantly modulate the composition of gut microbiota. Hence, understanding the regulatory effect of gut microbiota on antitumor immunotherapies may help improve their therapeutic efficacy in gastrointestinal cancers. Nanoparticles represent a promising strategy to enhance efficacy and reduce the toxicity of immunotherapy. In the context of treatment of colorectal cancer, nanoparticles can be used to transmit various substances with the potential to affect immune cells and their immunoreactivity within the tumor immune microenvironment, targeting the immune system, and activating multiple immune cells including, but not limited to, T cells.¹⁶

Besides tumor cells, there are several types of immune cells, endothelial cells, fibroblasts, extracellular matrix as well as others in the TME of gastrointestinal tumors. These components in TME are the key factors affecting tumor initiation, progression, and immune responses. The balance between cellular and humoral components as well as various inflammatory responses in the TME is critical for tumor growth. Cellular immunotherapy and pharmacological immunotherapy can modulate the TME to enhance tumor immunogenicity and antitumor immune responses to limit tumor growth. However, the efficacy of immunotherapies, such as ICIs, is impaired by the dynamic changes in the TME, which may induce acquired resistance in gastrointestinal cancers. Thus, further understanding of the dynamic changes in the TME may help develop new therapeutic strategies for improvement of immunotherapies and prevention of therapeutic resistance in gastrointestinal cancers. To determine the interaction between TME and tumor cells, further exploration of the mechanism underlying immune escape and immune tolerance, and effectively reshaping the TME are the leading solutions for cancer immunotherapies. Bioprinting of 3D cell culture models can help monitor complex cell behaviors under physical simulation conditions, modulate multiple types of cells in an easy-to-handle system, and study cell interaction in real-time.^{17, 18} With improved access to clinical tissues and better technologies to visualize cells in situ and dissect heterocellular signaling, the use of bioprinting to understand the TME holds great promise to address translational research questions. Furthermore, the 3D cell culture models are valuable for the development of anticancer drugs. Personalized cancer models can also be used to screen personalized antitumor drugs and promote basic cancer research, drug development, and precise medicines. Despite the attractiveness of 3D bioprinting, there are several technical challenges in the printing platforms, cells, and materials used to build 3D models because of the inherent challenge of comprehensively replicating the cellular behaviors and structural complexity of natural tissues.

Emerging biomarkers, particularly ctDNA, CTCs, and exosomes, are increasingly driving diagnosis advances, allowing for the early identification and dynamic monitoring of gastrointestinal cancers. Apart from the other useful treatment modalities (such as endoscopic therapy, surgery, radiotherapy, chemotherapy, and targeted therapy), immunotherapy, with its unique and valuable benefits, has made significant progress in the management of gastrointestinal cancers, which is our primary focus area. Concomitant nano-based therapeutics represent a promising strategy to improve the efficacy of immunotherapy. Fortunately, the collaboration in nonmedical disciplines, such as AI, 3D bioprinting, and materials science, can enable in-depth characterization of the TME, improving the diagnosis and treatment of gastrointestinal cancers and fostering multidisciplinary treatment strategies in the future. However, limitations are yet undeniable, which implies that new advances are often accompanied by formidable challenges to overcome.

Kai Li and Zhenning Wang conceptualized and designed the study. Mingguang Ju drafted the first version of the manuscript. All authors read and edited the manuscript.

The authors declare no conflict of interest. Professor Kai Li and Zhenning Wang are members of Chronic Diseases and Translational Medicine editorial board and are not involved in the peer review process of this article.

None.