Translational cancer research最新文献

Background: Acute myeloid leukemia (AML) is a highly heterogeneous hematologic malignancy, with its pathogenesis closely associated with cellular states at various stages of differentiation. Existing clinical prognostic models often fail to account for this heterogeneity and lack integration of key molecular pathways. This study aimed to characterize AML differentiation-associated heterogeneity at the single-cell level, investigate the role of UNC13D in immune and dedifferentiation states, and develop a prognostic model integrating these features.

Methods: This study combined single-cell RNA sequencing data (GSE178910) with bulk RNA-sequencing (RNA-seq) datasets [The Cancer Genome Atlas-Acute Myeloid Leukemia (TCGA-LAML) and Oregon Health & Science University (OHSU)]. Seurat and Harmony were used for batch correction and unsupervised clustering, followed by cell state annotation using AddModuleScore-based scoring of lineage-specific gene sets. UNC13D expression was assessed to infer its association with differentiation stage and pathway activity. Prognostic genes within the MYC proto-oncogene signaling pathway were identified using univariate Cox and least absolute shrinkage and selection operator (LASSO) regression. An eight-gene risk model was then constructed and validated across two cohorts.

Results: We identified eleven AML cellular subpopulations, grouped into five functional differentiation states. UNC13D was predominantly expressed in common myeloid progenitor-like (CMP-like) cells and correlated with multiple oncogenic and immune-related pathways. The resulting eight-gene prognostic model (PRDX4, KPNB1, DEK, ABCE1, ODC1, GLO1, MCM5, CCNA2) demonstrated good predictive performance in both the training and validation cohorts, with stable 1- and 3-year area under the curve (AUC) values. Differential pathway enrichment revealed marked biological divergence between high- and low-risk groups, including immune signaling and cell cycle regulation.

Conclusions: Our study delineates the differentiation landscape of AML and identifies UNC13D as a potential biomarker of cellular plasticity and immune modulation. The constructed model provides a reliable prognostic tool and offers novel insights for AML stratification and precision therapy development.

Background: Existing studies provide limited knowledge of the metastatic pattern, survival rate, and prognosis of primary gastrointestinal melanoma (PGM). This study aimed to investigate the metastatic patterns, prognostic factors, and conduct deep learning model of PGM.

Methods: The Surveillance, Epidemiology, and End Results (SEER) database was analysed to determine survival time, survival rates, and metastatic patterns in PGM. Cox regression analysis identified prognostic factors associated with overall survival (OS) and cancer-specific survival (CSS). Patients were divided into discovery (80%) and validation cohorts (20%) to develop and validate deep learning-based models for predicting OS and CSS of PGMs. The area under the receiver operating characteristic curve (AUC) was used to evaluate model performance.

Results: The median OS was 18 months [95% confidence interval (CI): 15-21] and 22 months (95% CI: 19-26) at CSS. OS rates were 60% (95% CI: 56-64%), 32% (95% CI: 28-36%), and 22% (95% CI: 18-26%) at 1, 3, and 5 years. The most common metastasis sites were the liver (19%), lungs (16%), bones (5%), and brain (4%). Older age, involvement of other sites, regional or distant stage disease, and two distant metastases were associated with worse OS or CSS, whereas systemic therapy was a protective factor. The deep learning models demonstrated performance in predicting OS (AUC: 0.7757-0.8366 at 1 year and 0.8046-0.8177 at 3 years) and CSS (0.7870-0.8169 AUC at 1 year and 0.7314-0.7720 at 3 years).

Conclusions: The prognosis of PGM varies significantly among subtypes, and the models developed in this study provide accurate predictions of OS and CSS, offering potentials for clinical utility.

Background: Lung adenocarcinoma (LUAD) is the most common histological subtype of lung cancer, with a poor prognosis due to radiotherapy and chemotherapy resistance. Novel systemic treatments have limitations, highlighting the need for identifying new oncogenic genes and therapeutic targets. Beta-secretase 2 (BACE2) is involved in the progression of multiple cancers, but its role and mechanism in LUAD remain unreported. This study aimed to explore the expression pattern, biological function, and underlying mechanism of BACE2 in LUAD.

Methods: BACE2 expression was assessed in LUAD tissues via bioinformatics analysis and immunohistochemistry. Cell viability, proliferation, apoptosis, migration, and cell cycle were detected using Cell Counting Kit-8 (CCK-8), 5-ethynyl-2'-deoxyuridine (EdU), flow cytometry, Transwell, and scratch assays. Gene set enrichment analysis (GSEA) and Western blot were used to explore the downstream pathway regulated by BACE2. A xenograft model was established to verify BACE2's in vivo role.

Results: BACE2 expression was elevated in LUAD tissues and cell lines, and high BACE2 expression correlated with poor patient survival. Silencing BACE2 led to increased apoptosis, reduced cell viability, growth, and migration, and G2 phase arrest. GSEA identified the mammalian target of rapamycin complex 1 (mTORC1) signalling pathway as a downstream target of BACE2, which was confirmed by Western blot (reduced p-mTOR/mTOR and p-RPS6KB1/RPS6KB1 levels after BACE2 silencing). Inhibiting mTORC1 with rapamycin abrogated the oncogenic effects of BACE2 overexpression. In vivo, BACE2 knockdown significantly suppressed xenograft tumor growth.

Conclusions: BACE2 contributes to LUAD progression by activating the mTORC1 signalling pathway, providing a novel therapeutic target for LUAD treatment.

Background: Hepatocellular carcinoma (HCC) remains a leading cause of cancer-related deaths globally. The Like-Smith (LSM) family members are involved in RNA metabolism and tumor progression, but their role in HCC remains unclear. This study aims to construct a novel signature based on LSM family members-related genes and explore its clinical value in HCC.

Methods: Molecular patterns related to LSM family members were identified through clustering analysis. Differential expression analysis was used to identify genes with potential prognostic significance. Multivariate Cox regression analysis was performed to construct a signature with The Cancer Genome Atlas (TCGA) cohort. The International Cancer Genome Consortium (ICGC) cohort served as external validation. Kaplan-Meier curves and receiver operating characteristic (ROC) curves were used to evaluate the predictive ability. Enrichment analysis, immune infiltration assessment, and single-cell RNA sequencing (scRNA-seq) data analysis were conducted to explore the underlying mechanisms.

Results: Two genes-paired-like homeodomain 2 (PITX2) and chromogranin A (CHGA)-were ultimately identified as a novel signature for HCC. Based on the risk score derived from the signature, samples were divided into high- and low-risk groups. Results indicated that the high-risk group had significantly poorer overall survival in both TCGA and ICGC cohorts. The ROC curves demonstrated that the signature exhibits stable predictive accuracy. Enrichment analysis showed that the high-risk group was associated with tumor-related pathways. Differences in immune infiltration were observed between high- and low-risk groups. scRNA-seq analysis indicated that PITX2 and CHGA were highly expressed in hepatocytes.

Conclusions: The novel two-gene signature comprising PITX2 and CHGA effectively predicts survival outcomes in HCC patients and is closely associated with tumor metabolism and immune regulation. This signature may serve as a valuable tool for prognostic evaluation and guiding personalized treatment strategies for HCC patients.

Background: Bone tumors have diverse clinical and imaging features, rendering preoperative differentiation of benign, intermediate/malignant types challenging. Unimodal methods (medical records or X-rays) are prone to misdiagnosis/missed diagnosis due to incomplete information. While postoperative histopathology is the gold standard, there is an urgent clinical demand for a precise preoperative diagnostic tool. This study aims to develop and validate a multimodal model integrating deep learning with Dempster-Shafer (DS) evidence theory for the differential diagnosis of benign, intermediate/malignant bone tumors. Using postoperative histopathology as the reference standard, the model achieves diagnosis by integrating preoperative clinical text and radiographs.

Methods: This single-center retrospective study included 319 pathologically confirmed bone tumor patients admitted between 2020 and 2025 following selection criteria. Utilizing the patients' X-ray images and medical record text data, we constructed a fusion model based on deep learning and DS evidence theory to classify tumors into benign and intermediate/malignant categories. The performance of the model was evaluated using the receiver operating characteristic (ROC) curve along with its 95% confidence interval (CI).

Results: The dataset comprised text data and radiographs from a total of 319 patients and it was stratified by time into a training set, an internal validation set, and an external validation set. On the internal validation set, the fusion model achieved an area under the curve (AUC) of 0.821 (95% CI: 0.713-0.916), with an accuracy of 81.6%, precision of 81.3%, recall of 76.5% and an F1 score of 78.8%, outperforming both the unimodal text model with an AUC of 0.814 and accuracy of 77.6% and the image model with an AUC of 0.782 and accuracy of 72.4%. On the external validation set, the fusion model maintained robust performance: AUC reached 0.808 (95% CI: 0.667-0.928), accuracy 77.3%, and F1 score 70.6%. Compared to the proposed fusion approach, most baseline models underperformed across all metrics, with their accuracy ranging from 59.1% to 77.3% and F1 score ranging from 47.1% to 70.6%. Furthermore, the model's diagnostic performance rivals that of senior radiologists and significantly outperforms junior radiologists. McNemar's test results confirmed no significant difference in diagnostic performance between the model and senior radiologists, while a statistically significant performance gap existed between junior and senior radiologists.

Conclusions: We have developed and validated a fusion model that integrated deep learning and DS evidence theory. In the task of distinguishing between benign and intermediate/malignant bone tumors, this fusion model demonstrated encouraging performance compared to models that utilize unimodal data and other baseline fusion models.

Background: Although the global incidence of gastric cancer (GC) has declined over the past 5 years, it remains the fourth leading cause of cancer-related mortality worldwide. Given the molecular heterogeneity of GC, survival outcomes can vary significantly among patients receiving the same treatment at the same stage. Therefore, this study aimed to develop and validate a robust prognostic model for GC that complements the current staging system, to ultimately facilitate better clinical decision-making.

Methods: Utilizing gene expression data from four independent cohorts comprising 1,305 GC patients, we developed and validated the immune-related transcriptomic predictive model for gastric cancer prognosis (ITPG), which incorporates transcriptomic biomarkers and explores gene-gene interactions. Specifically, the ITPG model integrates two genes with main effects (KCNQ1, FLRT2) and two pairs of genes with gene-gene interactions (ATP4B×CD84, NPY×ITGBL1), in addition to clinical variables including age and pathological stage. Prognostic biomarkers were identified in The Cancer Genome Atlas (TCGA) cohort. The model's risk stratification ability, predictive performance, and clinical utility were subsequently evaluated in three external cohorts: GSE66229, GSE15459, and GSE84437.

Results: The ITPG demonstrated strong risk stratification potential in identifying high-risk patients. Compared to those in the lowest 25^th percentile of ITPG scores, patients in the top 90^th percentile had significantly shorter overall survival [hazard ratio (HR) =9.79, 95% confidence interval (CI): 7.25-13.21, P=2.78×10^-50]. Furthermore, ITPG exhibited robust predictive performance across four cohorts, with pooled area under the curve (AUC) values for 1-year of 0.769 (95% CI: 0.735-0.803), 3-year of 0.762 (95% CI: 0.723-0.802), and 5-year of 0.765 (95% CI: 0.704-0.826) survival, and a C-index of 0.704 (95% CI: 0.678-0.729). Additionally, the model displayed substantial clinical utility in identifying GC patients at high risk of mortality [net benefit (NB) at 1-year =1.8%, NB_3-year =15.8%, NB_5-year =23.7%; net reduction (NR) at 1-year =58.6%, NR_3-year =20.4%, NR_5-year =11.7%]. Subgroup analyses confirmed the model's robustness across different population stratifications.

Conclusions: The ITPG model is an efficient and clinically relevant tool for prognostic prediction in GC.

Background: Idiopathic pulmonary fibrosis (IPF) is a progressive and lethal lung disease associated with significant morbidity and frequent complications. Basal cell carcinoma (BCC) is a common skin malignancy often diagnosed at the intermediate to advanced stages. Emerging evidence suggests that a epidemiological link exists between these conditions. This study aimed to investigate the shared genomic landscape and causal relationship between IPF and BCC and to clarify the related underlying molecular mechanisms and therapeutic implications.

Methods: Gene expression datasets (GSE10667, GSE24206, and GSE53845) were obtained from the Gene Expression Omnibus database. After normalization and integration, differential expression analysis identified 1,333 differentially expressed genes (DEGs) between patients with IPF and controls. Subsequently, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway functional enrichment analyses were performed. Mendelian randomization (MR) analysis was conducted with summary statistics from genome-wide association studies to infer the effect of IPF on BCC risk. Furthermore, a gene-drug interaction network and a competing endogenous RNA (ceRNA) network (consisting of long noncoding RNAs, microRNAs, and messenger RNAs) were constructed via Cytoscape to identify potential therapeutic targets.

Results: Enrichment analysis indicated a significant overrepresentation of the BCC signaling pathway among the DEGs, with 12 core genes shared between IPF and BCC pathogenesis being identified. These genes involved in critical molecular pathways and are correlated with certain immune cell interactions, suggesting a mechanistic link between IPF and BCC. The MR analysis provided evidence of a genetic basis for the causal relationship: compared to the general population, individuals with a genetic predisposition to IPF have a significantly higher risk of developing BCC. The networks highlighted key regulatory nodes and potential drug targets within the shared pathophysiology of the two diseases.

Conclusions: This study integrating genomic and causal inference study demonstrated that patients with IPF are at an increased risk of developing BCC. Further MR analysis indicated that this association is underpinned by shared genetic pathways, immune-related interactions, and a causal relationship. The core genes and regulatory networks identified in this study help clarify the molecular nature of the link between these diseases and offers novel avenues for devising therapeutic strategies targeting IPF and comorbid BCC.

Background: Patient outcomes in rectal cancer with synchronous liver metastases remain heterogeneous, underscoring the need for reliable prognostic factors to guide individualized treatment strategies. This study sought to assess the influence of lymph node-related indices on the survival of rectal cancer patients with synchronous liver metastases who underwent preoperative chemoradiotherapy and to construct a novel nomogram for this patient population.

Methods: Data on rectal cancer patients with synchronous liver metastases who underwent preoperative chemoradiotherapy between 2010 and 2019 were retrieved from the Surveillance, Epidemiology, and End Results (SEER) database. X-tile software was employed to determine the optimal cutoff point of lymph node-related indices. The Kaplan-Meier method and log-rank test were used to assess survival differences. Cox univariate and multivariate analyses were conducted to identify the prognostic factors for overall survival (OS). We constructed a novel nomogram to predict individual survival probability. The concordance index (C-index) and calibration curve were used to determine the predictive accuracy and discriminative ability of the nomogram.

Results: A total of 547 rectal cancer patients with synchronous liver metastases were enrolled. The OS of patients with total number of lymph nodes (TLN) <7 was significantly worse than that of patients with TLN ≥7 (3-year OS rate: 46.6% vs. 62.7%; 5-year OS rate: 22.9% vs. 43.8%, P=0.001). Patients with number of positive lymph nodes (PLN) <7 had better OS than that of patients with PLN ≥7 significantly (3-year OS rate: 63.5% vs. 36.0%; 5-year OS rate: 44.1% vs. 12.6%, P<0.001). The 3-year OS rate was 49.7% for patients with number of negative lymph nodes (NLN) <11 compared with 68.6% for patients with NLN ≥11; and the 5-year OS rate was 23.9% and 53.6% for patients with NLN <11 and patients with NLN ≥11, respectively (P<0.001). Multivariate analysis showed that TLN, NLN and PLN were independent prognostic factors for OS [TLN: hazard ratio (HR) =0.571, 95% confidence interval (CI): 0.408-0.799, P=0.001; NLN: HR =0.593, 95% CI: 0.456-0.770, P<0.001; PLN: HR =1.736, 95% CI: 1.201-2.509, P=0.003]. A novel nomogram was established based on the independent prognostic factors indicated in the multivariate Cox analysis. The C-index for prognostic nomogram was 0.67 (95% CI: 0.63-0.71). The calibration plot for the probability of 3-year survival demonstrated an optimal agreement between the predicted and actual survival.

Conclusions: Lymph node (LN)-related indices are significant prognostic factors for OS in rectal cancer patients with synchronous liver metastases, offering insights into survival prediction and enabling personalized treatment strategies.