Latest Research
All publications from the Cancer3.AI database, newest first.
Fibromodulin positively regulated by Androgen Receptor, promotes prostate cancer progression via the PI3K/AKT signaling pathway and epithelial-medenchymal transition.
Fu J, et al
Researchers investigated the role of fibromodulin (FMOD), a proteoglycan protein, in driving prostate cancer progression and characterized how it is controlled by the androgen receptor (AR), the central molecular engine of prostate cancer growth. Using gene-silencing and overexpression techniques in prostate cancer cell lines, the team demonstrated that high FMOD expression is directly switched on by AR, and that reducing FMOD levels halts cancer cell proliferation, causes cell cycle arrest, and sharply limits the cells' ability to migrate and invade surrounding tissue. Animal xenograft experiments confirmed that tumors with depleted FMOD grew significantly more slowly, validating the in vitro findings. Mechanistically, FMOD was shown to fuel cancer progression by activating the PI3K/AKT signaling pathway and by promoting epithelial-mesenchymal transition, a process that allows cancer cells to become more aggressive and spread. These findings establish a previously uncharacterized AR-FMOD-PI3K/AKT signaling axis as a key driver of prostate cancer and position FMOD as a promising novel therapeutic target that could complement existing androgen-deprivation strategies in the clinic.
Molecular biology reports
Source →Combined radiomics, PI-RADS, and clinical model improve significant prostate cancer prediction and guide biopsy decision.
Antolin A, et al
Researchers conducted a large retrospective study using 1,497 MRI cases from 1,395 men to develop and validate machine learning models capable of predicting significant prostate cancer, defined as Gleason Grade 7 or higher, which represents disease requiring treatment rather than watchful waiting. Four models were compared: one based solely on radiomic features extracted from MRI images, one based on the standard PI-RADS scoring system used by radiologists, a combined PI-RADS plus radiomics model, and a fully integrated model that also incorporated clinical variables such as PSA levels. Although radiomics alone did not significantly outperform PI-RADS (AUC 0.838 vs. 0.833), the fully integrated model combining all three components achieved the highest diagnostic accuracy, with an AUC of 0.891, significantly surpassing each individual approach. Critically, this combined model also demonstrated the highest biopsy avoidance rate of 18.15%, meaning that nearly one in five patients could potentially be spared an unnecessary and invasive biopsy procedure. These findings indicate that supplementing the widely used PI-RADS system with radiomic data and clinical information can meaningfully reduce false positives and support more precise, patient-friendly clinical decision-making in prostate cancer diagnosis.
Insights into imaging
Source →Integrating multimodal clinical data with a large model for prostate cancer diagnosis.
Wang C, et al
Researchers developed Prost-LM, a multimodal large language model that integrates MRI-derived imaging features, prostate-specific antigen (PSA) blood test values, and free-text clinical reports into a single unified diagnostic system to tackle the long-standing challenge of accurately diagnosing prostate cancer. The model was trained and validated on a large multi-center cohort of 3,940 patients, achieving an area under the curve (AUC) of 0.954 for distinguishing prostate cancer from benign conditions, significantly outperforming MRI-only models, which reached an AUC of only 0.868. For detecting clinically significant prostate cancer defined by a Gleason score of 7 or higher, which indicates more aggressive disease requiring prompt treatment, Prost-LM achieved an AUC of 0.955. Importantly, the model also generates interpretable diagnostic explanations, allowing clinicians to understand and verify the reasoning behind each AI-assisted decision. These results demonstrate that combining multiple clinical data types through advanced AI can substantially improve automated prostate cancer diagnosis and help guide more personalized, precision-based oncology care.
NPJ digital medicine
Source →Insulin-like growth factor-binding protein 5 promotes prostate cancer metastasis and osteoblastic activity by inducing chemokines and activating NF-κB signaling.
Gong Z, et al
Researchers investigated the molecular mechanisms by which insulin-like growth factor-binding protein 5 (IGFBP5) drives prostate cancer progression and metastatic spread, with a particular focus on bone involvement. The study found that IGFBP5 promotes cancer cell invasion and metastasis by stimulating the production of chemokines — signaling proteins that direct cell migration and facilitate tumor dissemination — while simultaneously activating the NF-κB signaling pathway, a master regulator of inflammation and cell survival. These combined actions were also shown to enhance osteoblastic activity, the aberrant stimulation of bone-forming cells that characterizes skeletal metastases frequently observed in advanced prostate cancer. By identifying IGFBP5 as a key molecular driver linking tumor aggressiveness to bone colonization, the findings highlight this protein and its downstream effectors as potential targets for new therapies aimed at preventing or treating bone metastases in prostate cancer patients.
Cancer cell international
Source →Investigation of angiogenesis and epithelial-mesenchymal transition markers (CD31, VEGF, and E-Cadherin) in prostatic cancer patients: a multi-omics study.
Shahshenas S, et al
Researchers published a multi-omics investigation examining three key molecular markers — CD31, VEGF, and E-Cadherin — in prostate cancer patients to better understand the processes of angiogenesis (tumor-driven blood vessel formation) and epithelial-mesenchymal transition (EMT), the mechanism by which cancer cells gain invasive and metastatic potential. CD31 was used to quantify tumor vascularity, VEGF as an indicator of pro-angiogenic signaling, and E-Cadherin as a marker whose loss is closely associated with aggressive tumor behavior. By integrating multiple layers of biological data in a multi-omics framework, the study aimed to characterize how these markers interact and vary across prostate cancer cases. The findings contribute to a growing understanding of how tumor vascularization and cellular plasticity collaborate to drive prostate cancer progression. Such insights are clinically relevant because they could help stratify patients by disease aggressiveness and identify novel therapeutic targets within angiogenic and EMT pathways. Ultimately, this research supports the broader effort to develop more personalized treatment strategies for prostate cancer patients based on their specific tumor biology.
World journal of surgical oncology
Source →