Abstract Creatine supplementation is widely used in sports and increasingly popular among exercising individuals. Although the physiological role of creatine has been extensively studied, the creatine biology in pathological conditions remains poorly understood. Here we report that exogenous creatine supplementation promotes tumor metastasis via platelet activation mechanism in various mouse models and humans. Mechanistically, creatine supplementation increases megakaryocyte creatine levels and upregulates creatine kinase B (CKB). Unbiased phosphoproteomics reveals that a CKB-downstream, non-canonical STAT5B phosphorylation instigates various platelet functional genes, leading to hyperactive, metastasis-promoting platelets. Megakaryocyte-specific knockout of the creatine transporter Slc6a8 or Stat5b , as well as pharmacological inhibition of STAT5, ablates the creatine-augmented platelet hyperactivity and prevents consequent metastasis in mice. Importantly, creatine supplementation in healthy volunteers results in hyperactive peripheral platelets that increase metastasis risks. Together, our study sheds mechanistic insights into the creatine-induced metastasis and provides an anti-metastatic therapeutic paradigm by targeting megakaryocyte creatine metabolism.

Objective Recent studies have shown that 68Ga-prostate-specific membrane antigen (68Ga-PSMA) positron emission tomography/computed tomography (PET/CT) can open a non-invasive diagnostic pathway in prostate cancer (PCa). However, the relatively small number of enrolled patients in these studies limits statistical validation and causes bias to some extent. In addition, the performance of 68Ga-PSMA PET/CT radiomics analysis in detecting PCa has not been widely evaluated. Hence, the present multicenter study endeavored to develop and validate radiomics models based on 68Ga-PSMA PET/CT for the detection of primary PCa in a relatively larger cohort. Methods This study enrolled consecutive patients with suspected PCa who underwent systematic biopsy (SB) and PSMA PET/CT-targeted biopsy (TB) after 68Ga-PSMA PET/CT in three medical centers. The whole prostate gland was adopted as the volume of interest (VOI). Eight machine learning (ML) algorithms were utilized to develop models with selected radiomics features separately, after which the best-performing radiomics model was integrated with maximum standardized uptake value (SUVmax) to create the combined model. The receiver operating characteristic (ROC) curves and area under the curve (AUC) value, accuracy, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for each model. Results Overall, 609 patients were recruited, including 195 patients with benign prostate diseases (BPD), 30 patients with clinically insignificant prostate cancer (ciPCa, Gleason score [GS] = 3 + 3), and 384 patients with clinically significant prostate cancer (csPCa, GS ≥ 3 + 4). For csPCa prediction, the radiomics model developed by the eXtreme Gradient Boosting (XGBoost) algorithm demonstrated the best performance. After integrating with SUVmax, the combined model achieved the highest AUC of 0.921 in internal validation cohort. External validation cohort 1 and 2 also showed promising results with AUC of 0.906 and 0.898, respectively. For PCa prediction, the XGBoost algorithm combined with SUVmax also peformed well in three validation cohorts with the AUC ranging from 0.860 to 0.918. Conclusions This is the largest multicenter study to date, providing a noninvasive and quantitative method based on 68Ga-PSMA PET/CT radiomics analysis modeled with ML for predicting PCa. This method has the potential to gauge the risk of PCa before biopsy.

Bladder cancer (BLC) is one of the most common malignancies of the urinary system and represents a major public health burden. Myeloid cells are key components of the tumor microenvironment and play critical roles in tumor progression and therapeutic response; however, their prognostic significance in BLC remains incompletely understood. The prognostic value of individual myeloid cell markers was evaluated using immunohistochemistry and survival analyses. A myeloid-based prognostic model was constructed using least absolute shrinkage and selection operator (LASSO) regression. The model was validated across multiple independent cohorts. Multiplex immunofluorescence staining and RNA sequencing were performed to investigate immune landscape alterations and signaling pathways associated with different risk groups. High CD68 expression within tumor regions was associated with favorable prognosis, whereas elevated expression of CD14, CD74, CD163, and S100A12 correlated with poor survival outcomes in BLC patients. A myeloid risk score (MRS) was subsequently established and demonstrated robust prognostic performance across validation cohorts. Transcriptomic analysis revealed significant activation of the PI3K–AKT signaling pathway in the MRS-high group. Furthermore, MRS-high tumors exhibited increased expression of PD-L1, FOXP3, and CD11b, along with reduced CD8⁺ T-cell infiltration, indicating a highly immunosuppressive tumor microenvironment. Potential therapeutic targets and candidate agents for MRS-high patients were also identified. We developed a robust myeloid cell–based prognostic model that effectively stratifies BLC patients by risk and reveals distinct immunosuppressive mechanisms in high-risk tumors. This model may facilitate personalized prognostic assessment and guide precision therapeutic strategies for patients with bladder cancer.

Pericytes, also known as mural cells, are cells embedded between endothelial cells and the basement membrane of capillaries, where they orchestrate the morphological and functional homeostasis of blood vessels. Within the tumor microenvironment, pericytes interact closely with various cellular components, including tumor cells, stromal cells, and immune cells. Through these dynamic interactions, pericytes are activated and subsequently transform into tumor-associated pericytes (TPCs). The origin of TPCs varies depending on the tissue and tumor type, contributing to their phenotypic and functional heterogeneity. TPCs play pivotal roles in facilitating tumor progression, metastasis, immune evasion, and therapeutic resistance by promoting angiogenesis, engaging in reciprocal interactions with tumor cells, remodeling the extracellular matrix, and fostering an immunosuppressive microenvironment. This review synthesizes the latest significant advancements in targeted therapies against TPCs. It underscores the challenges inherent in developing effective anti-TPC therapies, which include the heterogeneity and pluripotency of TPCs, the absence of specific markers for precise TPC targeting, and the limited understanding of how current anti-tumor therapies affect TPCs and vice versa. This review furnishes a comprehensive understanding of the origins, markers, and functions of TPCs, and their interplays within the tumor microenvironment, providing prospective strategies for more effective anti-tumor therapy.

Linking clinically derived risk signals to reproducible molecular states across independent cohorts remains a major challenge in translational bioinformatics. Existing approaches often rely on cohort-specific model fitting, limiting cross-dataset comparability and downstream biological interpretation. We developed a cross-cohort projection framework that maps baseline clinical variables to a clinically anchored latent risk coordinate, $\mu$, enabling application across external datasets without refitting. The fixed projector was trained in a local imaging cohort and applied unchanged to independent cohorts. Projected $\mu$ was evaluated across multiple molecular layers, including bulk transcriptomics, single-cell-guided deconvolution, spatial transcriptomics, and circulating cell-free DNA (cfDNA). In an independent external cohort, projected $\mu$ preserved separation of time to castration resistance across predefined strata (P = .002), with 30-month risk increasing from 0.13 to 0.86 across ordered $\mu$ bins. In bulk transcriptomics, higher projected $\mu$ was associated with increased proliferation-related signaling and reduced androgen receptor/lineage programs ($\rho$ = 0.40 and -0.26; both P < .001). Deconvolution analyses linked higher projected $\mu$ to reduced AR-high epithelial cell fractions ($\rho$ = -0.17, P = .001). Spatial transcriptomics demonstrated organized tissue-level structure of prespecified molecular programs. In cfDNA, higher projected $\mu$ was associated with a more negative RB1 copy-number signal in the detectable subset ($\rho$ = -0.49, P = .0278). This study presents a projection-based framework for cross-cohort translation of clinically anchored latent risk into interpretable multi-omics context. By enabling reuse of a fixed coordinate without refitting, the approach provides a practical strategy for linking clinical risk to molecular programs and blood-based readouts across datasets.

Large language models (LLMs) are increasingly explored as tools for medical education. However, evidence remains limited regarding their pedagogical quality and real-world utility in prostate cancer teaching within urology residency training. We conducted a two-phase study. Phase 1 benchmarked three LLMs (ChatGPT-4o, DeepSeek R1, and Gemini 2.0) on a structured 40-item prostate cancer education question bank using standardized prompts and blinded expert ratings. Phase 2 implemented the top-performing model in a pilot randomized teaching trial among urology residents (n = 34) using stratified block randomization based on a pre-admission standardized test score, with allocation concealment implemented through a centralized web-based system. Both groups received identical offline instruction with a time-matched lecture structure. The control group committed to avoiding LLM use for course-related questions during the teaching period. DeepSeek R1 ranked highest in expert ratings, with clearer advantages on higher-order and innovation-oriented questions. In the pilot randomized teaching trial (n = 34), the AI-assisted group achieved higher closed-book examination scores than controls (68.47 ± 12.78 vs. 57.91 ± 10.47; MD 10.56, 95

Hepatocellular carcinoma (HCC), a malignancy of the digestive system, presents limited therapeutic options at advanced stages. In this context, inhibitors of cyclic nucleotide phosphodiesterase 4 (PDE4) have emerged as promising novel agents for patients with advanced HCC. We synthesized a potent PDE4A inhibitor, designated as MG5b, derived from the natural bioactive compound alpha-mangostin, and evaluated its antitumor efficacy in HCC. To assess the impact of MG5b on the proliferation of HCC cells, we conducted MTT and colony formation assays. Flow cytometry was utilized to analyze cell cycle progression, apoptosis, and mitochondrial membrane potential in HCC cells. Intracellular reactive oxygen species (ROS) levels were quantified using a DCF-DA probe. Network pharmacology was employed to predict the associated signaling pathways. Furthermore, Western blotting and immunohistochemistry were utilized to determine the effects of MG5b on protein expression levels. The antitumor efficacy of MG5b was further evaluated in Huh7 xenograft models. MG5b demonstrated a significant inhibitory effect on the proliferation, migration, and invasion of HCC cells, induced cell cycle arrest at the G0/G1 phase, and promoted apoptosis. Mechanistically, MG5b activated the cyclic adenosine monophosphate (cAMP) induced protein kinase A (PKA) signaling pathway, resulting in increased intracellular ROS production and the subsequent suppression of the EGFR-PI3K-AKT pathway. In vivo studies indicated that MG5b markedly inhibited tumor growth while exhibiting minimal toxicity. These findings suggest that MG5b may function as a novel PDE4A inhibitor and hold potential as a promising drug candidate for the treatment of advanced HCC.

Pancreatic cancer treatment is severely hampered by poor drug delivery through its dense stromal barrier and an immunosuppressive tumor microenvironment, which together contribute to resistance to immune checkpoint blockade. To address these two major obstacles, we rationally designed and synthesized a novel conjugate, termed EVO-FAPI. This molecule covalently links a 3-fluoro-10-hydroxyl-evodiamine (EVO) derivative, a potent cytotoxic alkaloid and inducer of immunogenic cell death (ICD), to a fibroblast activation protein inhibitor (FAPI) that selectively targets fibroblast activation protein alpha (FAPα)-expressing stromal fibroblasts abundant in pancreatic tumors. By selectively inducing ICD in both cancer cells and the tumor stroma, EVO‑FAPI promotes ICD‑mediated activation of the anti‑tumor immune response. Importantly, when combined with α-PD-1, EVO-FAPI increased the tumor growth inhibition rate from 35% to 67%, markedly improving therapeutic efficacy compared with α-PD-1 monotherapy. Collectively, these findings identify EVO-FAPI as a rationally engineered dual-function agent that integrates targeted stromal modulation with immunotherapy sensitization, thereby overcoming resistance to immune checkpoint blockade in pancreatic cancer.

Abstract Background: Previous MCED studies have shown differences in detection performance across different aggressiveness cancer subtypes. However, the methylation profiles and ctDNA shedding patterns across different aggressiveness subtypes have not been analyzed. We systematically compared cell-free DNA (cfDNA) tumor fraction and methylation profiles across aggressiveness subtypes in three cancers to identify epigenetic biomarkers that differentiate tumor aggressiveness. Methods: Blood samples from a case-control study (NCT06217900) including 757 cancer cases were analyzed using a targeted methylation-based MCED test. We compared cfDNA tumor fraction and methylation profiles between aggressiveness subtypes, stage-matched small cell lung cancer (SCLC) (n=137) versus (vs) non-small cell lung cancer (NSCLC) (n=137), stage I invasive lung adenocarcinoma (IAC) (n=81) vs minimally invasive adenocarcinoma (MIA) (n=81), triple-negative breast cancer (TNBC) (n=151) vs non-TNBC (n=151), and intermediate/high-grade prostate cancer (Gleason Grade Group[GG] 2-5)(n=14) vs low-grade prostate cancer(GG1)(n=6). Tumor fraction was estimated using zero-inflated Negative Binomial model based on the distribution of methylation signals, and methylation profiles were assessed by calculating methylated/unmethylated (M/U) ratios between different aggressiveness subtypes. Marker comparisons were conducted using the Mann-Whitney U test with multiple testing corrections. Results: The sensitivity is higher in more aggressive subtypes (lung cancer 93.73% vs SCLC 99.27%; breast cancer 70.76% vs TNBC 81.46%; prostate cancer 40% vs intermediate and high-grad prostate cancer 42.86%). In lung cancer, SCLC exhibited a significantly higher cfDNA tumor fraction than NSCLC (Wilcoxon p = 2.24×10-24), with 41,136 markers (79.18% hypomethylated) showing elevated M/U ratios. In stage I lung adenocarcinoma, IAC demonstrated a higher cfDNA tumor fraction (Wilcoxon p = 2.55×10-6) and systematic methylation differences compared to MIA. In breast cancer, TNBC had a significantly higher cfDNA tumor fraction than non-TNBC (Wilcoxon p = 2.55×10-6), with 25,717 markers (80.32% hypomethylated) displaying increased M/U ratios. Among prostate cancers of the same stage, GG2-5 cases exhibited higher cfDNA tumor fraction (Wilcoxon p = 0.038) and systematic methylation alterations compared to GG1. Conclusion: Aggressive tumor subtypes consistently display elevated cfDNA tumor fraction and characteristic methylation profiles marked by increased M/U ratios. These findings suggest cfDNA methylation patterns are promising epigenetic biomarkers for distinguishing tumor aggressiveness, potentially improving cancer screening precision and reducing overdiagnosis. Citation Format: Kezhong Chen, Jian Huang, Dahong Zhang, Shu Wang, Hongxu Liu, Wenzhao Zhong, Xiangnan Li, Qiang Zhang, Zhigao Li, Jiaqi Liu, Ziqing Tian, Fei Zhou, Gongsheng Jin, Xudong Xiang, Zhigang Li, Hui Xie, Ya Wei, Guochun Zhang, Guolin Ye, Ming Cai, Junfeng Wang, Yan Zhang, Chao Cheng, Hefei Li, Desong Yang, Jianhong Lian, Sheng Huang, Tao Xu, Zengjun Wang, Xi Guo, Zhuowei Liu, Minfeng Chen, Yang Wang, Yue An, Yanzhan Yang, Min Li, Jing Liu, Baoliang Zhu, Yonghui Li, Xiaohui Wu, Fan Yang, Jun Wang. Epigenetic profiling identifies markers of aggressive cancer subtype using a targeted methylation-based multi-cancer early detection (MCED) blood test [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1100.

Background Bladder cancer (BLCA) is a prevalent malignancy characterized by high recurrence and poor prognosis, particularly muscle-invasive bladder cancer (MIBC). Histopathology, the gold standard for assessing muscle invasion, often suffers from sampling errors and operator dependency, underscoring the need for non-invasive, accurate preoperative assessment methods. This study aimed to develop and validate a hybrid artificial intelligence (AI) model based on computed tomography (CT) radiomics and deep learning (DL) to predict MIBC and overall survival (OS) preoperatively in BLCA patients. Methods A total of 1370 patients from 6 academic medical centers were retrospectively included. Preoperative contrast-enhanced CT scans were analyzed to extract handcrafted radiomic features using PyRadiomics and DL features using ResNet101, followed by machine learning (ML)-based modeling for prediction. A hybrid model combining radiomic and DL features was constructed and validated in internal and external cohorts. Model performance was evaluated using metrics such as the area under the curve (AUC) and Cox proportional hazards analysis for OS prediction. Results The DL radiomics nomogram (DLRN) model demonstrated superior diagnostic performance, achieving an AUC of 0.807 in the internal validation cohort and 0.783 in the external multi-center validation cohort for predicting muscle invasion. The DLRN generated an imaging-derived risk score (DLRN score), which was subsequently incorporated as one covariate into a multivariable Cox proportional hazards model together with clinicopathological variables to evaluate OS. Using this approach, patients were effectively stratified into high- and low-risk groups for OS, showing robust generalizability across diverse clinical settings. AI-assisted diagnostics significantly improved the sensitivity and accuracy of urologists, particularly among less experienced clinicians. Conclusion The DLRN model provides a reliable, non-invasive tool for preoperative assessment of muscle invasion and prognosis in BLCA. Addressing histopathology limitations, it offers valuable insights for personalized treatment strategies, paving the way for precision oncology in real-world clinical applications.

Multiparametric MRI (mpMRI) and ^68 Ga-PSMA PET/CT are widely used for prostate cancer (PCa) diagnosis but remain limited by false positives and modest specificity, particularly in distinguishing benign prostate diseases (BPDs) and clinically significant PCa (csPCa). Existing studies often rely on small, single-center cohorts with limited generalizability. This study aimed to develop and externally validate a multimodal radiomics model integrating PET/CT and mpMRI for automated PCa diagnosis, and to evaluate the impact of prostate VOI delineation strategies. A total of 488 patients with suspected PCa who underwent both ^68 Ga-PSMA PET/CT and mpMRI (T2 and DWI) followed by biopsy were retrospectively enrolled from two centers (366 for model development and ten-fold internal validation; 41 for external validation cohort 1; 81 for external validation cohort 2). Radiomics features were extracted from both modalities, and six classical machine learning classifiers (LR, SVM, Random Forest, Extra Trees, XGBoost, LightGBM) were trained for three tasks: (1) csPCa diagnosis, (2) overall PCa detection, and (3) comparison between expert-drawn and deep learning generated prostate VOIs. Model performance was assessed using AUC, sensitivity, specificity, accuracy, PPV, and NPV. Among 407 patients, 137 had BPD, 25 had clinically insignificant PCa, and 250 had csPCa. The multimodal PET/mpMRI radiomics model achieved the best performance with LightGBM (AUC = 0.91 internally; 0.825 externally). Automatically segmented VOIs achieved comparable diagnostic accuracy to expert annotations, with AUC differences within 3–8

4609 Background: Radical cystectomy with pelvic lymph node dissection (RC+PLND) has always been the standard treatment for MIBC. Perioperative treatment, especially cisplatin-based neoadjuvant therapy, could reduce the risk of tumor recurrence effectively and prolong the survival of patients. BFv, a novel Nectin-4 antibody-drug conjugate, combined with toripalimab, has demonstrated promising efficacy and safety profile in advanced urothelial cancer patients. This provides possibilities for the perioperative treatment of MIBC. Here we report the preliminary results of BFv plus toripalimab in patients with perioperative MIBC (cohort A). Methods: Patients with MIBC (cT2-4aN0-1M0) can be enrolled in cohort A. Eligible patients will receive BFv plus toripalimab for 4 cycles in neoadjuvant stage, and then proceed with RC+PLND. After surgery, patients will receive BFv plus toripalimab for 6 cycles and toripalimab only for the next 11 cycles in the adjuvant stage. The primary endpoint was pathological complete response (pCR). Main secondary endpoints were pathological downstaging (PDS) rate and disease-free survival (DFS), overall survival (OS) and safety. Results: As of 4 Jan 2026, 32 patients were enrolled in the cohort A. 7 patients had completed neoadjuvant therapy and 6 of them were performed with RC+PLND. One patient achieved CR in the neoadjuvant stage and refused to undergo the surgery. pCR rate was 66.7% (4/6, 95%CI 22.3-95.7), PDS rate was 83.3% (5/6, 95%CI 35.9-99.6). DFS and OS were not mature. Safety profile was consist with previous study. No new safety signals of BFv or toripalimab were observed in this study. Conclusions: This study showed remarkable preliminary efficacy of BFv plus toripalimab in perioperative patients with MIBC. It may offer a new choice for MIBC patients and deserve to be further investigation. Clinical trial information: NCT07314723 .

Background Pelvic lymph node dissection (PLND) remains controversial in the management of prostate cancer. Although it provides the most accurate pathological staging, its therapeutic value beyond staging has long been debated due to conflicting evidence and concerns regarding procedure-related morbidity.Objective To critically evaluate the contemporary role of PLND, particularly extended pelvic lymph node dissection (ePLND), in prostate cancer management in the context of modern imaging, risk stratification tools, and evolving oncologic endpoints.Evidence acquisition A narrative review of recent literature was conducted, focusing on high-level evidence including randomized trials, observational studies, and contemporary guideline recommendations addressing the indications, extent, oncologic outcomes, and complications of PLND.Evidence synthesis Recent randomized and observational studies suggest that ePLND improves nodal staging accuracy and may be associated with modest improvements in metastasis-free survival (MFS) in selected patients with intermediate- and high-risk prostate cancer, although the absolute benefit remains limited and causality is not definitively established. Advances in molecular imaging, particularly prostate-specific membrane antigen (PSMA) PET/CT, together with multiparametric MRI, validated nomograms, and emerging genomic classifiers, now allow more precise identification of patients most likely to benefit from ePLND. The integration of these tools supports a more individualized surgical strategy, including image-guided and sentinel lymph node approaches designed to maximize staging accuracy while minimizing unnecessary dissection.Conclusions In the contemporary PSMA imaging era, ePLND continues to play an important role in nodal staging and may contribute to improved oncologic outcomes in carefully selected patients.

Immunotherapy has emerged as a clinically pivotal approach in cancer treatment, but its application remains limited to a small subset of patients. Current data indicate that less than 10% of patients respond to mono-immunotherapy, and acquired resistance further restricts its efficacy. Elucidating the mechanisms of non-response and developing strategies to increase therapeutic sensitivity are critical for advancing immunotherapy. Natural products have long played a vital role in anticancer drug discovery because of their structural diversity and demonstrated biological activities. In this review, we provide a systematic classification of immunotherapies and summarize mechanisms associated with treatment resistance. We highlight promising natural products shown to enhance immunotherapeutic efficacy through the modulation of immune resistance pathways and discuss translational challenges hindering the clinical application of natural products, proposing their potential as synergistic agents to overcome current limitations in cancer immunotherapy. Harnessing natural products may thus open new avenues to overcome immune resistance and broaden the therapeutic effects of cancer immunotherapy.

Hematogenous metastasis is the leading cause of cancer mortality, with dysfunction of pericytes, key components of tumor vessels, playing a central role in facilitating metastatic spread. Although anti-pericyte therapies are gaining recognition for treating metastasis, current strategies that directly eliminate tumor pericytes (TPCs) may increase vascular leakiness, which paradoxically promotes further metastasis. Here, we identify a TPC-specific transcription factor heterodimer, TCF21-TCF3, which drives metastasis by enhancing collagen hydroxylation and extracellular matrix deposition. Based on the TCF21 residues that interact with TCF3, we rationally design a peptide to disrupt their dimerization and downregulate TCF21-TCF3-dependent collagen deposition. Notably, in murine models of colorectal cancer and osteosarcoma, the TCF21-derived peptide significantly inhibits metastasis by restoring the physiological gatekeeper function of pericytes on vessels, offering a potential therapeutic strategy to target TPCs and suppress metastasis. Our findings reveal a TPC-specific transcription factor heterodimer and provide a promising pericyte-targeting strategy for preventing hematogenous metastasis.