Tumor genetic heterogeneity may underlie poor clinical outcomes because diverse subclones could be comprised of metastatic and drug resistant cells. Targeted deep sequencing has been used widely as a diagnostic tool to identify actionable mutations in cancer patients. In this study, we evaluated the clinical utility of estimating tumor heterogeneity using targeted panel sequencing data. We investigated the prognostic impact of a tumor heterogeneity (TH) index on clinical outcomes, using mutational profiles from targeted deep sequencing data acquired from 1,352 patients across 8 cancer types. The TH index tended to be increased in high pathological stage disease in several cancer types, indicating clonal expansion of cancer cells as tumor progression proceeds. In colorectal cancer patients, TH index values also correlated significantly with clinical prognosis. Integration of the TH index with genomic and clinical features could improve the power of risk prediction for clinical outcomes. In conclusion, deep sequencing to determine the TH index could serve as a promising prognostic indicator in cancer patients.

Cancer cells use multiple mechanisms to evade the effects of glutamine metabolism inhibitors. The pathways that govern responses to alterations in glutamine availability within the tumor may represent therapeutic targets for combinatorial strategies with these inhibitors. Here, we showed that targeting glutamine utilization stimulated Yes-associated protein (YAP) signaling in cancer cells by reducing cyclic adenosine monophosphate/protein kinase A (PKA)–dependent phosphorylation of large tumor suppressor (LATS). Elevated YAP activation induced extracellular matrix (ECM) deposition by increasing the secretion of connective tissue growth factor that promoted the production of fibronectin and collagen by surrounding fibroblasts. Consequently, inhibiting YAP synergized with inhibition of glutamine utilization to effectively suppress tumor growth in vivo, along with a concurrent decrease in ECM deposition. Blocking ECM remodeling also augmented the tumor suppressive effects of the glutamine utilization inhibitor. Collectively, these data reveal mechanisms by which targeting glutamine utilization increases ECM accumulation and identify potential strategies to reduce ECM levels and increase the efficacy of glutamine metabolism inhibitors. Significance: Blocking glutamine utilization activates YAP to promote ECM deposition by fibroblasts, highlighting the potential of YAP inhibitors and antifibrotic strategies as promising approaches for effective combination metabolic therapies in cancer.

Background: Colorectal cancer (CRC) is one of the leading causes of cancer-related deaths worldwide. Although early diagnosis and treatment is the most successful strategy for improving patient survival, feasible and sensitive blood biomarkers for CRC screening remain elusive. Methods: Sixty-five CRC patients and thirty-three healthy individuals were enrolled. Peripheral blood (PB) and tumor tissues from CRC patients, and PB from healthy individuals were subjected to immunophenotyping and phospho-flow analysis of cytokine-induced phosphorylated STAT (CIPS). Logistic regression was used as a classifier that separates CRC patients from healthy individuals. Results: The proportion of regulatory T cells was increased in PB from CRC patients compared to PB from healthy individuals (p < 0.05). Interestingly, peripheral T cells share several cytokine-induced phosphorylated STAT (CIPS) signatures with T cells from CRC tumor-sites. Additionally, a classifier was made using two signatures distinct between T cells from CRC patients and T cells from healthy individuals. The AUCs (area under curves) of the classifier were 0.88 in initial cohort and 0.94 in the additional validation cohort. Overall AUC was 0.94 with sensitivity of 91% and specificity of 88%. Conclusion: This study highlights that immune cell signatures in peripheral blood could offer a new type of biomarker for CRC screening.

Among the various types of breast cancer, the luminal B subtype is the most common in young women, and ESR1-CCDC170 (E:C) fusion is the most frequent oncogenic fusion driver of the luminal B subtype. Nevertheless, treatments targeting E:C fusion has not been well established yet. Hence, the aim of this study is to investigate potential therapies targeting E:C fusion based on systematic bioinformatical analysis of the Cancer Genome Atlas (TCGA) data. One thousand related genes were extracted using transcriptome analysis, and major signaling pathways associated with breast cancer were identified with over-representation analysis. Then, we conducted drug-target network analysis based on the OncoKB and CIViC databases, and finally selected potentially applicable drug candidates. Six major cancer-related signaling pathways (p53, ATR/ATM, FOXM1, hedgehog, cell cycle, and Aurora B) were significantly altered in E:C fusion-positive cases of breast cancer. Further investigation revealed that nine genes (AURKB, HDAC2, PLK1, CENPA, CHEK1, CHEK2, RB1, CCNA2, and MDM2) in coordination with E:C fusion were found to be common denominators in three or more of these pathways, thereby making them promising gene biomarkers for target therapy. Among the 21 putative actionable drugs inferred by drug-target network analysis, palbociclib, alpelisib, ribociclib, dexamethasone, checkpoint kinase inhibitor AXD 7762, irinotecan, milademetan tosylate, R05045337, cisplatin, prexasertib, and olaparib were considered promising drug candidates targeting genes involved in at least two E:C fusion-related pathways.

Prostate cancer continues to pose a global health challenge as one of the most prevalent malignancies. Mutations of the Forkhead box A1 (FOXA1) gene have been linked to unique oncogenic features in prostate cancer. In this study, we aimed to unravel the intricate molecular characteristics of FOXA1 mutant prostate cancer through comprehensive in silico analysis of transcriptomic data from The Cancer Genome Atlas (TCGA). A comparison between FOXA1 mutant and control groups unearthed 1525 differentially expressed genes (DEGs), which map to eight intrinsic and six extrinsic signaling pathways. Interestingly, the majority of intrinsic pathways, but not extrinsic pathways, were validated using RNA-seq data of 22Rv1 cells from the GEO123619 dataset, suggesting complex biology in the tumor microenvironment. As a result of our in silico research, we identified novel therapeutic targets and potential drug candidates for FOXA1 mutant prostate cancer. KDM1A, MAOA, PDGFB, and HSP90AB1 emerged as druggable candidate targets, as we found that they have approved drugs throughout the drug database CADDIE. Notably, as most of the approved drugs targeting MAOA and KDM1A were monoamine inhibitors used for mental illness or diabetes, we suggest they have a potential to cure FOXA1 mutant primary prostate cancer without lethal side effects.

Acute myeloid leukemia (AML) patients are at risk of bleeding due to disease-related lack of platelets and systemic coagulopathy. Platelets play a role in hemostasis. Leukemic blasts have been shown to alter platelet activation in vitro. Here we investigated biomarkers associated with thrombocytopenia in normal karyotype AML (NK-AML). From The Cancer Genome Atlas database, case-control study was performed between normal karyotype (NK) platelet-decreased AML (PD-AML, platelet count < 100 × 10⁹/L, n = 24) and NK platelet-not-decreased AML (PND-AML, with platelet count ≥ 100 × 10⁹/L, n = 13). Differentially expressed gene analysis, pathway analysis and modelling for predicting platelet decrease in AML were performed. DEG analysis and pathway analysis revealed 157 genes and eight pathways specific for PD-AML, respectively. Most of the eight pathways were significantly involved in G-protein-coupled receptor-related pathway, cytokine-related pathway, and bone remodeling pathway. Among the key genes involved in at least one pathway, three genes including CSF1R, TNFSF15 and CLEC10A were selected as promising biomarkers for predicting PD-AML (0.847 of AUC in support vector machine model). This is the first study that identified biomarkers using RNA expression data analysis and could help understand the pathophysiology in AML with low platelet count.