Vessel co-option has been demonstrated to mediate colorectal cancer liver metastasis (CRCLM) resistance to antiangiogenic therapy. The current mechanisms underlying vessel co-option have mainly focused on “hijacker” tumor cells, whereas the function of the “hijackee” sinusoidal blood vessels has not been explored. Here, we found that the occurrence of vessel co-option in bevacizumab-resistant CRCLM xenografts was associated with increased expression of fibroblast activation protein α (FAPα) in the co-opted hepatic stellate cells (HSCs), which was dramatically attenuated in HSC-specific conditional Fap-knockout mice bearing CRCLM allografts. Mechanistically, bevacizumab treatment induced hypoxia to upregulate the expression of fibroblast growth factor–binding protein 1 (FGFBP1) in tumor cells. Gain- or loss-of-function experiments revealed that the bevacizumab-resistant tumor cell–derived FGFBP1 induced FAPα expression by enhancing the paracrine FGF2/FGFR1/ERK1/-2/EGR1 signaling pathway in HSCs. FAPα promoted CXCL5 secretion in HSCs, which activated CXCR2 to promote the epithelial-mesenchymal transition of tumor cells and the recruitment of myeloid-derived suppressor cells. These findings were further validated in tumor tissues derived from patients with CRCLM. Targeting FAPα⁺ HSCs effectively disrupted the co-opted sinusoidal blood vessels and overcame bevacizumab resistance. Our study highlights the role of FAPα⁺ HSCs in vessel co-option and provides an effective strategy to overcome the vessel co-option–mediated bevacizumab resistance.

In-situ oral delivery of therapeutic antibodies, like monoclonal antibody, for chronic inflammation treatment is the most convenient approach compared with other administration routes. Moreover, the abundant links between the gut microbiota and colonic inflammation indicate that the synergistic or antagonistic effect of gut microbiota to colonic inflammation. However, the antibody activity would be significantly affected while transferring through the gastrointestinal tract due to hostile conditions. Moreover, these antibodies have short serum half-lives, thus, require to be frequently administered with high doses to be effective, leading to low patient tolerance. Here, we develop a strategy utilizing thin shell hydrogel microcapsule fabricated by microfluidic technique as the oral delivering carrier. By encapsulating antibodies in these microcapsules, antibodies survive in the hostile gastrointestinal environment and rapidly release into the small intestine through oral administration route, achieving the same therapeutic effect as the intravenous injection evaluated by a colonic inflammation disease model. Moreover, the abundance of some intestinal microorganisms as the indication of the improvement of inflammation has remarkably altered after in-situ antibody-laden microcapsules delivery, implying the restoration of micro-ecology of the intestine. These findings prove our microcapsules are exploited as an efficient oral delivery agent for antibodies with programmable function in clinical application.

Neutrophils are integral to the frontline defense against pathogenic bacterial and fungal invasions. Beyond their traditional roles, these cells are increasingly recognized for their dualistic contributions to the pathology of autoimmune and inflammatory diseases, as well as their complex involvement in cancer progression. Neutrophils interact with different disease states, highlighting their potential as therapeutic targets. Within tumor microenvironment (TME), tumor-associated neutrophils (TANs) exhibit a functional dichotomy, capable of either fostering or impeding tumor growth and metastasis. This binary functional potential of TANs, under certain conditions, suggests a reversible state that could transition from tumor-promoting to tumor-eradicating phenotypes. Despite the critical implications of such functional plasticity, systematic studies of TAN behavioral shifts in the context of cancer immunotherapy remain scarce. Herein, we review recent advancements in the understanding of TANs within the TME, highlighting their binary regulatory effects on solid tumors. Leveraging the latest insights from experimental and clinical research, this review elucidates the complex roles of TANs in tumor development and explores their molecular interactions as potential therapeutic targets. The elucidation of these mechanisms holds promise for novel cancer treatment strategies, aiming to improve patient outcomes by manipulating the tumor-promoting or -suppressing functions of TANs.

The precise matching of medication regimens to individual patients, known as personalized medication, is critical for the effective management of chronic diseases. Traditional machine learning‐based models for personalized medication regimens typically rely solely on either clinical macro‐phenotypes or molecular‐level drug characteristics. It remains challenging to capture the patient‐medication relationship from a comprehensive perspective that integrates individual patient characteristics with macro‐ and micro‐level properties of the medication. Determining patient‐medication relationships constitutes a three‐stage sequential decision process from a clinical decision‐making perspective. Therefore, inspired by Chain‐of‐Thought prompting, which simulates the decision‐making process of human experts, a Multimodal Data‐Driven Chain‐of‐Decisions (MDD‐CoD) framework is proposed, where three‐stage deep learning tasks are sequentially organized to reflect upstream–downstream logical dependencies, thereby forming a coherent clinical decision‐making process. The model incorporates multimodal clinical phenotype data, multi‐attribute medication data, and insights from clinical experts. Performance evaluation of the model involved comprehensive experiments utilizing five datasets covering four chronic diseases sourced from three hospitals. The dataset comprises information from chronic kidney disease (CKD), membranous nephropathy (MN), rheumatoid arthritis (RA), colorectal cancer (CRC), and knee osteoarthritis (KOA), totaling 3173 unimodal, 502 multimodal, and 2187 medication records from 3675 patients. Experimental results demonstrate that the framework achieves enhanced predictive performance in personalized medication decision‐making based on individual patient disease characteristics, surpassing the strongest baseline across all tasks. This framework serves as a foundational model for clinical mixed data, with improved generalization and interpretability in cross‐disease personalized decision‐making tasks. It offers a scalable solution for the implementation of personalized medication regimens for chronic diseases.