https://scholar.dgist.ac.kr/handle/20.500.11750/1157 2026-06-09T19:59:44Z https://scholar.dgist.ac.kr/handle/20.500.11750/60409 Title: Deciphering the Heterogeneity of Cancer-Associated Fibroblasts in Prostate Cancer: From Stromal Biology to Clinical Translation Author(s): Truong, Ho Trong Tan; Kwon, Whi-An; Woo, Hyeong Jung; Kim, Minseok S.; Tran, Nhu Quang; Joung, Jae Young Abstract: Prostate cancer (PCa) progression and treatment resistance are driven by tumor-intrinsic mechanisms and adaptive remodeling of the tumor microenvironment, in which cancer-associated fibroblasts (CAFs) play a crucial role. Although CAF biology is increasingly recognized, a major translational gap remains: CAFs are highly heterogeneous, and comprise distinct functional states with divergent effects on disease progression, immune regulation, and therapeutic resistance. To bridge this gap, we synthesize evidence from single-cell and spatial transcriptomic studies, tissue-based pathology, liquid biopsy assays, and molecular imaging to construct an evidence-tiered, decision-oriented translational framework that connects stromal mechanisms, translational measurement strategies, and therapeutic interventions in PCa. Single-cell and spatial transcriptomic analyses have consistently identified multiple CAF programs, including matrix-remodeling, inflammatory, immunoregulatory, antigen-presenting, and therapy-imprinted states, each with distinct functional outputs and clinical correlates. Tissue-based readouts, including reactive stromal grade (RSG) and fibroblast activation protein (FAP) immunohistochemistry, provide practical proxies for stromal activation and correlate with disease-specific mortality and imaging phenotypes. Circulating CAFs (cCAFs) represent an emerging liquid biopsy modality for longitudinal stromal monitoring, although technical standardization is required before clinical implementation. FAP-targeted PET imaging and emerging dual prostate-specific membrane antigen (PSMA)/FAP-targeted theranostic strategies provide noninvasive tools for patient selection and response assessment, particularly in PSMA-discordant or tracer-heterogeneous disease. Androgen receptor (AR)-targeted therapy can reprogram stromal states toward resistance-promoting circuits, highlighting the dynamic and plastic nature of the CAF compartment. A state-based CAF framework organizes stromal biology into testable translational hypotheses rather than immediate clinical standards. RSG and FAP-based tissue or imaging readouts are practical markers of stromal activation, whereas spatial CAF-immune signatures and cCAF assays remain investigational and require assay harmonization and prospective validation. Future trials should pre-specify stromal biomarkers as enrichment or pharmacodynamic variables when matched to the intervention and should avoid treating CAFs as a uniform therapeutic target. 2026-04-30T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/59943 Title: CD45+hybrid circulating cells may reflect tumor-immune interactions and serve as transcriptomic indicators of metastatic potential in prostate cancer Author(s): Kim, Baek Gil; Jang, Yeonsue; Kim, Min Gyu; Song, Dongwook; Jung, Jungchan; Jung, Jihee; Yoo, Ayoung; Lee, Jongsoo; Cho, Nam Hoon; Woo, Hyeong Jung; Kim, Woon-Hae; Shin, Hyun Young; Kim, Minseok S.; Han, Hyun Ho; Joung, Jae Young Abstract: Rationale: Circulating hybrid cells expressing both epithelial and immune markers have emerged as indicators of dynamic tumor-immune interactions. This study aimed to characterize circulating hybrid cells co-expressing KRT18 (pan-cytokeratin) and PTPRC (CD45), termed KP_Pos, in metastatic prostate cancer (mPCa), and to assess their molecular features, tumor microenvironmental (TME) origins, and clinical relevance. Methods: Imaging mass cytometry (IMC) was used to examine spatial relationships between CK+tumor and CD45+ immune cells in metastatic prostate tissues. Single-cell RNA sequencing (scRNA-seq) datasets from mPCa were analyzed to identify KP_Pos cells and characterize their transcriptional heterogeneity across epithelial and immune lineages. Differentially expressed genes (DEGs) between KP_Pos and other cells were used to generate predictive gene signatures. Random forest (RF) and extreme gradient boosting (XGB) models were applied to evaluate metastatic classification performance, and high-performing signatures were validated in bulk RNA-seq datasets and correlated with clinical parameters. Results: IMC revealed frequent spatial proximity between tumor and immune compartments, supporting a TME-derived hybrid phenotype. KP_Pos cells were detected across multiple immune and epithelial clusters, showing heterogeneity and enrichment of immune response and epithelial-mesenchymal transition (EMT)-related genes. Machine learning-based classifiers using KP_Pos-derived DEGs achieved high predictive accuracy (AUC >= 0.7) for metastasis, and selected combinations further improved performance in internal validation sets. Signature scores significantly correlated with PSA and Gleason grade, and CD45+ hybrid circulating cells were more abundant in patients with advanced disease burden. Conclusions: CD45+ KRT18+ hybrid circulating cells (KP_Pos) represent biologically distinct populations shaped by tumor-immune interactions within the TME. Their transcriptomic features and derived gene signatures may serve as biomarkers of metastatic potential and indicators of disease progression in prostate cancer. However, their causal role in metastasis and impact on survival remain to be determined. 2025-12-31T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/58957 Title: Robust Automated Separation of Circulating Tumor Cells and Cancer-Associated Fibroblasts for Enhanced Liquid Biopsy in Breast Cancer Author(s): Woo, Hyeong Jung; Rademacher, Paul N.; Shin, Hyun Young; Lee, Jungmin; Tasnuf, Aseer Intisar; Warkiani, Majid E.; Joung, Jae Young; Rzhevskiy, Alexey; Coith, Cornelia; Harten, Frederike; Hilpert, Felix; Schmalfeldt, Barbara; Riethdorf, Sabine; Pantel, Klaus; Joosse, Simon A.; Kim, Minseok S. Abstract: Circulating tumor cells (CTCs) are a key biomarker in cancer diagnostics, offering critical insights into metastasis and treatment responses. Although several automated CTC isolation systems have been developed, a thorough comparison of their performance with diverse cell types remains lacking. In addition to CTCs, simultaneous tumor microenvironment (TME) analysis can be valuable for formulating cancer treatment strategies. This includes investigating circulating cancer-associated fibroblasts (cCAFs), which offer a minimally invasive, real-time status of the TME, enabling frequent monitoring of cancer metastasis and treatment response. However, the automated and simultaneous isolation of CTCs and cCAFs has been unexplored. This research systematically evaluated the performance of FDA-registered automated CTC isolation systems with cancer cells of heterogeneous phenotypes, a breast cancer CTC cell line, as well as clinical samples from 27 breast cancer patients. The continuous centrifugal microfluidic system (CTCeptor) demonstrated superior recovery rates and enriched CTCs with broader size and surface marker heterogeneity compared to other positive selection-based technologies, isolating significantly more CTCs from the blood of cancer patients and achieving high detection rates. Notably, since the system relies on an unbiased isolation method, it also isolated cCAFs from patient blood, which were detected at frequencies 10 times higher than CTCs in early-stage breast cancer patients. For the first time, this study identified key CAF markers, highlighting the potential of cCAFs as a biomarker for early diagnosis and prognosis. The ability of this automated system to efficiently isolate both CTCs and cCAFs represents a significant advancement in liquid biopsy and precision oncology. 2025-07-31T15:00:00Z https://scholar.dgist.ac.kr/handle/20.500.11750/58229 Title: Quantitative analysis of the dexamethasone side effect on human-derived young and aged skeletal muscle by myotube and nuclei segmentation using deep learning Author(s): Park, Seonghwan; Kim, Min Young.; Jeong, Jaewon; Yang, Sohae; Kim, Minseok S.; Moon, Inkyu Abstract: Motivation Skeletal muscle cells (skMCs) combine together to create long, multi-nucleated structures called myotubes. By studying the size, length, and number of nuclei in these myotubes, we can gain a deeper understanding of skeletal muscle development. However, human experimenters may often derive unreliable results owing to the unusual shape of the myotube, which causes significant measurement variability.Results We propose a new method for quantitative analysis of the dexamethasone side effect on human-derived young and aged skeletal muscle by simultaneous myotube and nuclei segmentation using deep learning combined with post-processing techniques. The deep learning model outputs myotube semantic segmentation, nuclei semantic segmentation, and nuclei center, and post-processing applies a watershed algorithm to accurately distinguish overlapped nuclei and identify myotube branches through skeletonization. To evaluate the performance of the model, the myotube diameter and the number of nuclei were calculated from the generated segmented images and compared with the results calculated by human experimenters. In particular, the proposed model produced outstanding outcomes when comparing human-derived primary young and aged skMCs treated with dexamethasone. The proposed standardized and consistent automated image segmentation system for myotubes is expected to help streamline the drug-development process for skeletal muscle diseases.Availability and implementation The code and the data are available at https://github.com/tdn02007/QA-skMCs-Seg 2024-12-31T15:00:00Z