ARID1A deficiency promotes progression and potentiates therapeutic antitumour immunity in hepatitis B virus-related hepatocellular carcinoma.

BACKGROUND: Exploring predictive biomarkers and therapeutic strategies of ICBs has become an urgent need in clinical practice. Increasing evidence has shown that ARID1A deficiency might play a critical role in sculpting tumor environments in various tumors and might be used as pan-cancer biomarkers for immunotherapy outcomes. The current study aims to explored the immune-modulating role of ARID1A deficiency in Hepatitis B virus (HBV) related hepatocellular carcinoma (HBV-HCC) and its potential immunotherapeutic implications. METHODS: In the current study, we performed a comprehensive analysis using bioinformatics approaches and pre-clinical experiments to evaluate the ARID1A regulatory role on the biological behavior, and immune landscape of Hepatitis B virus (HBV) related hepatocellular carcinoma (HBV-HCC). A total of 425 HBV-related hepatocellular carcinoma patients from TCGA-LIHC, AMC and CHCC-HBV cohort were enrolled in bioinformatics analysis. Immunohistochemical staining of HBV-HCC specimens and ARID1A deficiency cellular models were used to validate the results of the analysis. RESULTS: Our results have shown that ARID1A deficiency promoted tumor proliferation and metastasis. More importantly, ARID1A deficiency in HBV-HCC was associated with the higher TMB, elevated immune activity, and up-regulated expression of immune checkpoint proteins, especially TIM-3 in HBV-HCC. Further, the expression of Galectin-9, which is the ligand of TIM-3, was elevated in the ARID1A knockout HBV positive cell line. CONCLUSION: To conclude, we have shown that the ARID1A deficiency was correlated with more active immune signatures and higher expression of immune checkpoints in HBV-HCC. Additionally, the present study provides insights to explore the possibility of the predictive role of ARID1A in HBV-HCC patients responsive to immunotherapy.

CRISPR screens uncover protective effect of PSTK as a regulator of chemotherapy-induced ferroptosis in hepatocellular carcinoma.

BACKGROUND: Hepatocellular carcinoma (HCC) is among the most common forms of cancer and is associated with poor patient outcomes. The emergence of therapeutic resistance has hampered the efficacy of targeted treatments employed to treat HCC patients to date. In this study, we conducted a series of CRISPR/Cas9 screens to identify genes associated with synthetic lethality capable of improving HCC patient clinical responses. METHODS: CRISPR-based loss-of-function genetic screens were used to target 18,053 protein-coding genes in HCC cells to identify chemotherapy-related synthetic lethal genes in these cells. Synergistic effects were analyzed through in vitro and in vivo analyses, while related mechanisms were explored through RNA-seq and metabolomics analyses. Potential inhibitors of identified genetic targets were selected through high-throughput virtual screening. RESULTS: The inhibition of phosphoseryl-tRNA kinase (PSTK) was found to increase HCC cell sensitivity to chemotherapeutic treatment. PSTK was associated with the suppression of chemotherapy-induced ferroptosis in HCC cells, and the depletion of PSTK resulted in the inactivation of glutathione peroxidative 4 (GPX4) and the disruption of glutathione (GSH) metabolism owing to the inhibition of selenocysteine and cysteine synthesis, thus enhancing the induction of ferroptosis upon targeted chemotherapeutic treatment. Punicalin, an agent used to treat hepatitis B virus (HBV), was identified as a possible PSTK inhibitor that exhibited synergistic efficacy when applied together with Sorafenib to treat HCC in vitro and in vivo. CONCLUSIONS: These results highlight a key role for PSTK as a mediator of resistance to targeted therapeutic treatment in HCC cells that functions by suppressing ferroptotic induction. PSTK inhibitors may thus represent ideal candidates for overcoming drug resistance in HCC.

Nuclei-Guided Network for Breast Cancer Grading in HE-Stained Pathological Images.

Breast cancer grading methods based on hematoxylin-eosin (HE) stained pathological images can be summarized into two categories. The first category is to directly extract the pathological image features for breast cancer grading. However, unlike the coarse-grained problem of breast cancer classification, breast cancer grading is a fine-grained classification problem, so general methods cannot achieve satisfactory results. The second category is to apply the three evaluation criteria of the Nottingham Grading System (NGS) separately, and then integrate the results of the three criteria to obtain the final grading result. However, NGS is only a semiquantitative evaluation method, and there may be far more image features related to breast cancer grading. In this paper, we proposed a Nuclei-Guided Network (NGNet) for breast invasive ductal carcinoma (IDC) grading in pathological images. The proposed nuclei-guided attention module plays the role of nucleus attention, so as to learn more nuclei-related feature representations for breast IDC grading. In addition, the proposed nuclei-guided fusion module in the fusion process of different branches can further enable the network to focus on learning nuclei-related features. Overall, under the guidance of nuclei-related features, the entire NGNet can learn more fine-grained features for breast IDC grading. The experimental results show that the performance of the proposed method is better than that of state-of-the-art method. In addition, we released a well-labeled dataset with 3644 pathological images for breast IDC grading. This dataset is currently the largest publicly available breast IDC grading dataset and can serve as a benchmark to facilitate a broader study of breast IDC grading.

Breast cancer histopathological image classification using a hybrid deep neural network.

Even with the rapid advances in medical sciences, histopathological diagnosis is still considered the gold standard in diagnosing cancer. However, the complexity of histopathological images and the dramatic increase in workload make this task time consuming, and the results may be subject to pathologist subjectivity. Therefore, the development of automatic and precise histopathological image analysis methods is essential for the field. In this paper, we propose a new hybrid convolutional and recurrent deep neural network for breast cancer histopathological image classification. Based on the richer multilevel feature representation of the histopathological image patches, our method integrates the advantages of convolutional and recurrent neural networks, and the short-term and long-term spatial correlations between patches are preserved. The experimental results show that our method outperforms the state-of-the-art method with an obtained average accuracy of 91.3% for the 4-class classification task. We also release a dataset with 3771 breast cancer histopathological images to the scientific community that is now publicly available at http://ear.ict.ac.cn/?page_id=1616. Our dataset is not only the largest publicly released dataset for breast cancer histopathological image classification, but it covers as many different subclasses spanning different age groups as possible, thus providing enough data diversity to alleviate the problem of relatively low classification accuracy of benign images.

Richer fusion network for breast cancer classification based on multimodal data.

BACKGROUND: Deep learning algorithms significantly improve the accuracy of pathological image classification, but the accuracy of breast cancer classification using only single-mode pathological images still cannot meet the needs of clinical practice. Inspired by the real scenario of pathologists reading pathological images for diagnosis, we integrate pathological images and structured data extracted from clinical electronic medical record (EMR) to further improve the accuracy of breast cancer classification. METHODS: In this paper, we propose a new richer fusion network for the classification of benign and malignant breast cancer based on multimodal data. To make pathological image can be integrated more sufficient with structured EMR data, we proposed a method to extract richer multilevel feature representation of the pathological image from multiple convolutional layers. Meanwhile, to minimize the information loss for each modality before data fusion, we use the denoising autoencoder as a way to increase the low-dimensional structured EMR data to high-dimensional, instead of reducing the high-dimensional image data to low-dimensional before data fusion. In addition, denoising autoencoder naturally generalizes our method to make the accurate prediction with partially missing structured EMR data. RESULTS: The experimental results show that the proposed method is superior to the most advanced method in terms of the average classification accuracy (92.9%). In addition, we have released a dataset containing structured data from 185 patients that were extracted from EMR and 3764 paired pathological images of breast cancer, which can be publicly downloaded from http://ear.ict.ac.cn/?page_id=1663 . CONCLUSIONS: We utilized a new richer fusion network to integrate highly heterogeneous data to leverage the structured EMR data to improve the accuracy of pathological image classification. Therefore, the application of automatic breast cancer classification algorithms in clinical practice becomes possible. Due to the generality of the proposed fusion method, it can be straightforwardly extended to the fusion of other structured data and unstructured data.

Identification of NINJ1 as a novel prognostic predictor for retroperitoneal liposarcoma.

BACKGROUND: Retroperitoneal liposarcoma (RPLS) is known for its propensity for local recurrence and short survival time. We aimed to identify a credible and specific prognostic biomarker for RPLS. METHODS: Cases from The Cancer Genome Atlas (TCGA) sarcoma dataset were included as the training group. Co-expression modules were constructed using weighted gene co-expression network analysis (WGCNA) to explore associations between modules and survival. Survival analysis of hub genes was performed using the Kaplan-Meier method. In addition, independent external validation was performed on a cohort of 135 Chinese RPLS patients from the REtroperitoneal SArcoma Registry (RESAR) study (NCT03838718). RESULTS: A total of 19 co-expression modules were constructed based on the expression levels of 26,497 RNAs in the TCGA cohort. Among these modules, the green module exhibited a positive correlation with overall survival (OS, p = 0.10) and disease-free survival (DFS, p = 0.06). Gene set enrichment analysis showed that the green module was associated with endocytosis and soft-tissue sarcomas. Survival analysis demonstrated that NINJ1, a hub gene within the green module, was positively associated with OS (p = 0.019) in the TCGA cohort. Moreover, in the validation cohort, patients with higher NINJ1 expression levels displayed a higher probability of survival for both OS (p = 0.023) and DFS (p = 0.012). Multivariable Cox analysis further confirmed the independent prognostic significance of NINJ1. CONCLUSIONS: We here provide a foundation for the establishment of a consensus prognostic biomarker for RPLS, which should not only facilitate medical treatment but also guide the development of novel targeted drugs.

A better overall survival (OS) for total (ipsilateral) retroperitoneal lipectomy than standard complete resection in patients with retroperitoneal liposarcoma: a comparative multi-institutional study.

Background: Complete resection (CR) serves as the standard of surgical treatment for retroperitoneal liposarcoma (RPLS). Unfortunately, even at referral centers, recurrence rates are high, and CR may not address multifocal diseases, which are a common phenomenon in RPLS. We sought to retrospectively compare the clinical outcomes of RPLS patients treated with total (ipsilateral) retroperitoneal lipectomy (TRL) and CR. Because TRL remove potentially multifocal tumors in the fat, patients may have a better prognosis than CR. Methods: Patients with primary/first-recurrent RPLS who had been treated at 5 referral centers were recruited from December 2014 to June 2018. Multivariable Cox regression analyses were conducted to determine the effects of demographic, operative, and clinicopathological variables on the following primary endpoints: local recurrence (LR), local recurrence-free survival (LRFS), and overall survival (OS). Results: A total of 134 patients were enrolled in this retrospective study, 53 of whom underwent TRL, and 81 of whom underwent CR. The 2 groups were comparable in terms of age, gender, presentation (primary vs. first-recurrent RPLS), number of tumors (unifocal vs. multifocal) at presentation, and Fédération Nationale des Centres de Lutte Contre le Cancer (FNCLCC) grade. The TRL group had higher levels of preoperative hemoglobin (Hb) (13 vs. 12.5 g/dL; P=0.008) and a lower amount of intraoperative blood loss (400 vs. 500 mL; P=0.034), but there were no significant differences in the length of hospital stay (23 vs. 22 d; P=0.47) or complications (32 vs. 30; P=0.82) between the 2 groups. In a subset of patients with multifocal tumors at initial presentation, OS was more prolonged in those treated with TRL than those treated with CR (P=0.0272). Based on the multivariable analysis, primary liposarcoma and a low FNCLCC grade were associated with decreased LR and improved OS. Conclusions: TRL is a safe procedure that positively affects the OS of patients with multifocal RPLS. This novel strategy deserves further investigation in prospective studies.

Decreased expression of ARID1A invasively downregulates the expression of ribosomal proteins in hepatocellular carcinoma.

Background: There was increasing evidence showing that ARID1A alterations correlated with higher tumor mutational burden, but there were limited studies focusing on the adaptive mechanisms for tumor cells to survive under excessive genomic alterations. Materials & methods: To further explore the adaptive mechanisms under ARID1A alterations, we performed RNA sequencing in ARID1A knockdown hepatocellular carcinoma cell lines, and demonstrated that decreased expression of ARID1A controlled global ribosomal proteins synthesis. The results were further confirmed by quantitative reverse transcription-PCR and bioinformatic analysis in The Cancer Genome Atlas Liver Hepatocellular Carcinoma database. Conclusion: The present study was the first to demonstrate that ARID1A might be involved in the translation pathway and served as an adaptive mechanism for tumor cells to survive under stress.

Implications of driver genes associated with a high tumor mutation burden identified using next-generation sequencing on immunotherapy in hepatocellular carcinoma.

Immune checkpoint blockade (ICB) therapy is a treatment strategy for hepatocellular carcinoma (HCC); however, its clinical efficacy is limited to a select subset of patients. Next-generation sequencing has identified the value of tumor mutation burden (TMB) as a predictor for ICB efficacy in multiple types of tumor, including HCC. Specific driver gene mutations may be indicative of a high TMB (TMB-H) and analysis of such mutations may provide novel insights into the underlying mechanisms of TMB-H and potential therapeutic strategies. In the present study, a hybridization-capture method was used to target 1.45 Mb of the genomic sequence (coding sequence, 1 Mb), analyzing the somatic mutation landscape of 81 HCC tumor samples. Mutations in five genes were significantly associated with TMB-H, including mutations in tumor protein 53 (TP53), Catenin®1 (CTNNB1), AT-rich interactive domain-containing protein 1A (ARID1A), myeloid/lymphoid or mixed-lineage leukemia (MLL) and nuclear receptor co-repressor 1 (NCOR1). Further analysis using The Cancer Genome Atlas Liver Hepatocellular Carcinoma database showed that TP53, CTNNB1 and MLL mutations were positively correlated with TMB-H. Meanwhile, mutations in ARID1A, TP53 and MLL were associated with poor overall survival of patients with HCC. Overall, TMB-H and associated driver gene mutations may have potential as predictive biomarkers of ICB therapy efficacy for treatment of patients with HCC.

Methylation of p16 CpG islands correlated with metastasis and aggressiveness in papillary thyroid carcinoma.

BACKGROUND: Inactivation of p16 by methylation of CpG islands is a frequent early event in human cancers, including papillary thyroid carcinoma (PTC). This study was to observe the methylation status of the p16 gene in papillary thyroid carcinoma (PTC) and its correlation with clinical parameters. METHODS: Methylation-specific PCR (MSP) was used to analyze the methylation status of the p16 gene in 74 PTCs and 21 adjacent normal thyroid tissues. RESULTS: Hypermethylation of p16 gene was observed in 27.0% (20/74) of PTC. None of the normal thyroid tissues was methylated, when compared to the PTCs (p < 0.05). There was no marked relationship between the methylation of p16 gene and the patients' age, gender, size of cancer, histological subtypes and occurrence of recurrent disease (p > 0.05). The methylation of p16 gene was positively associated with metastasis, a high AMES (age, metastasis to distant sites, extrathyroidal invasion, size) risk group (p < 0.05) and advanced pathological tumor-lymph node-metastasis stages. CONCLUSION: The methylation of the p16 gene, one event of significance in molecular biology, was common and correlated with biological metastasis and histological features in PTC, and may be involved in thyroid tumorigenesis and aggressiveness.