Fibroblasts and macrophages play key roles in the development of hepatocellular carcinoma (HCC). However, cross-talk between these two kinds of cells has not been well studied. Endosialin (CD248/TEM1) is a transmembrane glycoprotein that is expressed in certain cancer cells, tumor stromal cells, and pericytes. In this study, we found that endosialin is mainly expressed in cancer-associated fibroblasts (CAF) in HCC and its expression inversely correlates with patient prognosis. Endosialin interacted with CD68 to recruit macrophages and regulated expression of GAS6 in CAFs to mediate M2 polarization of macrophages. The fully human antibody IgG78 bound glycosylated endosialin and induced its internalization in CAFs, thus weakening the cross-talk between CAFs and macrophages. In subcutaneous and orthotopic xenograft models of HCC in nude mice, treatment with IgG78 significantly inhibited tumor growth. These results indicate that endosialin-positive CAFs promote HCC progression and highlight IgG78 as a promising therapeutic candidate for HCC treatment. Significance: These findings highlight CAF-expressed endosialin as a primary regulator of macrophage recruitment and polarization and demonstrate endosialin inhibition as a potential treatment strategy for HCC.

Prostate cancer rarely responds to immune-checkpoint blockade (ICB) therapies. Cancer-associated fibroblasts (CAF) are critical components of the immunologically “cold” tumor microenvironment and are considered a promising target to enhance the immunotherapy response. In this study, we aimed to reveal the mechanisms regulating CAF plasticity to identify potential strategies to switch CAFs from protumorigenic to antitumor phenotypes and to enhance ICB efficacy in prostate cancer. Integration of four prostate cancer single-cell RNA sequencing datasets defined protumorigenic and antitumor CAFs, and RNA-seq, flow cytometry, and a prostate cancer organoid model demonstrated the functions of two CAF subtypes. Extracellular matrix–associated CAFs (ECM-CAF) promoted collagen deposition and cancer cell progression, and lymphocyte-associated CAFs (Lym-CAF) exhibited an antitumor phenotype and induced the infiltration and activation of CD8+ T cells. YAP1 activity regulated the ECM-CAF phenotype, and YAP1 silencing promoted switching to Lym-CAFs. NF-κB p65 was the core transcription factor in the Lym-CAF subset, and YAP1 inhibited nuclear translocation of p65. Selective depletion of YAP1 in ECM-CAFs in vivo promoted CD8+ T-cell infiltration and activation and enhanced the therapeutic effects of anti-PD-1 treatment on prostate cancer. Overall, this study revealed a mechanism regulating CAF identity in prostate cancer and highlighted a therapeutic strategy for altering the CAF subtype to suppress tumor growth and increase sensitivity to ICB. Significance: YAP1 regulates cancer-associated fibroblast phenotypes and can be targeted to switch cancer-associated fibroblasts from a protumorigenic subtype that promotes extracellular matrix deposition to a tumor-suppressive subtype that stimulates antitumor immunity and immunotherapy efficacy.

Purpose: Multiparametric MRI is the current standard for detecting clinically significant prostate cancer (csPCa). However, men with negative or equivocal MRI often undergo unnecessary biopsies due to concerns about false-negative results. The recently proposed ⁶⁸Ga-PSMA PET/CT-based PRIMARY score exhibited good diagnostic performance for csPCa. This study aimed to externally validate the performance of the PRIMARY score and evaluate its added diagnostic value to MRI triage in detecting csPCa. Patients and Methods: This retrospective cohort study included 431 men who underwent both ⁶⁸Ga-PSMA PET/CT and MRI before biopsy. Performance was assessed using the area under the receiver operating characteristic curve and the decision curve analysis. The PRIMARY score + MRI was considered positive for either PRIMARY score 3–5 or Prostate Imaging Reporting and Data System (PI-RADS) 4/5. Results: The prevalence of csPCa was 51.7% (223/431). The area under the receiver operating characteristic curve of the 5-level PRIMARY score for csPCa was significantly higher than that of MRI (0.873 vs 0.786, P < 0.001). For the entire group, sensitivity, specificity, positive predictive value, and negative predictive value of the PRIMARY score were 90.6%, 61.1%, 71.4%, and 85.8%, respectively, which outperformed 87.9%, 49.0%, 64.9%, and 79.1% of PI-RADS on MRI. The PRIAMRY score + MRI improved sensitivity (96.0% vs 87.9%, P < 0.001) and negative predictive value (91.5% vs 79.1%, P < 0.001) without compromising specificity and positive predictive value compared with MRI alone. This combined approach avoided 24.6% (106/431) of unnecessary biopsies, while missing 4.0% (9/223) of csPCa cases. The addition of the PRIMARY score in men with PI-RADS 1–3 showed a net benefit, but not in men with PI-RADS 4/5. Conclusions: The PRIMARY score was superior to MRI in detecting csPCa, and its added diagnostic value was in men with negative or equivocal MRI results. The PRIMARY score + MRI improved negative predictive value and sensitivity for csPCa compared with MRI alone. Further prospective trials will validate whether men with clinical suspicion of csPCa but negative PRIMARY score + MRI can safely avoid unnecessary biopsies.

Neuroendocrine prostate cancer (NEPC) evolves as an aggressive phenotype during prolonged androgen deprivation therapy, lacking effective clinical management. In this study, we elucidated a reciprocal metabolic–epigenetic mechanism involving a positive feedback loop between glycolysis and the transcription factor LHX2 that promotes prostate cancer progression. Mechanistically, enzalutamide-mediated inhibition of the androgen receptor led to upregulation of key glycolytic enzymes. Elevated glycolysis resulted in lactate accumulation and subsequent histone lactylation, which in turn enhanced LHX2 expression. Reciprocally, LHX2 transcriptionally activated the lactate dehydrogenase LDHA, which further amplified lactate production. Furthermore, LHX2 augmented the expression of the lineage-determining enzyme DNMT1, potentiating neuroendocrine gene expression and tumor growth. Significantly, the antiviral agent paritaprevir could directly bind to LHX2, effectively suppressing neuroendocrine marker expression and tumor progression. These findings uncover a potential role for LHX2 in orchestrating NEPC development, offering putative therapeutic targets for treating this aggressive cancer phenotype and overcoming drug resistance. Significance: LHX2 bridges metabolic reprogramming and epigenetic rewiring to drive tumor progression via DNMT1, providing a mechanism of neuroendocrine transdifferentiation and the rationale for developing a potential therapeutic strategy for neuroendocrine prostate cancer.

Personalized medicine for gastric cancer continues to face numerous challenges, primarily due to the complexity of clinical decision making and the difficulty of integrating multimodal data. Artificial intelligence (AI), with its powerful capabilities in feature learning and pattern recognition, is emerging as a key technology to overcome these barriers. It provides critical support in areas such as early screening, histological subtyping, prediction of treatment response, and prognostic risk stratification. This review examines the application of AI in diagnosing and treating gastric cancer, with particular attention to the current mainstream AI methodologies, including feature engineering and deep learning and the rapidly evolving pretrained foundation models and multimodal large models. With the integration of medical images, digital pathology, multiomics data, and structured clinical information, AI systems are increasingly effective at capturing tumor heterogeneity and supporting complex clinical decisions in real time. On the one hand, task-specific models have demonstrated excellent performance in subtyping, staging, and prognosis assessment. On the other hand, the rise of foundation models and general-purpose large models is redefining the limits of AI in cross-task transfer, complex reasoning, and human–machine interaction. These technologies hold promise in addressing key obstacles such as data scarcity, modality heterogeneity, and fragmented clinical workflows, offering a feasible path toward a unified and efficient AI-driven diagnostic and therapeutic system for gastric cancer. As technological maturity progresses alongside the development of robust safety and ethical frameworks, AI is expected to evolve from a static auxiliary interpretation tool into an intelligent decision-making platform capable of semantic understanding, dynamic feedback, and multidisciplinary collaboration—therefore playing a pivotal role across the full spectrum of precision medicine in gastric cancer.