Landscape and Regulation of m6A and m6Am Methylome across Human and Mouse Tissues.

N6-methyladenosine (m6A), the most abundant internal mRNA modification, and N6,2'-O-dimethyladenosine (m6Am), found at the first-transcribed nucleotide, are two reversible epitranscriptomic marks. However, the profiles and distribution patterns of m6A and m6Am across human and mouse tissues are poorly characterized. Here, we report the m6A and m6Am methylome through profiling of 43 human and 16 mouse tissues and demonstrate strongest tissue specificity for the brain tissues. A small subset of tissue-specific m6A peaks can also readily classify tissue types. The overall m6A and m6Am level is partially correlated with the expression level of their writers and erasers. Additionally, the m6A-containing regions are enriched for SNPs. Furthermore, cross-species analysis revealed that species rather than tissue type is the primary determinant of methylation. Collectively, our study provides an in-depth resource for dissecting the landscape and regulation of the m6A and m6Am epitranscriptomic marks across mammalian tissues.

N-linked glycosylation of PD-L1/PD-1: an emerging target for cancer diagnosis and treatment.

During tumorigenesis and progression, the immune checkpoint programmed death-1 (PD-1) and its ligand programmed death ligand-1 (PD-L1) play critical roles in suppressing T cell-mediated anticancer immune responses, leading to T-cell exhaustion and subsequent tumor evasion. Therefore, anti-PD-L1/PD-1 therapy has been an attractive strategy for treating cancer over the past decade. However, the overall efficacy of this approach remains suboptimal, revealing an urgent need for novel insights. Interestingly, increasing evidence indicates that both PD-L1 on tumor cells and PD-1 on tumor-specific T cells undergo extensive N-linked glycosylation, which is essential for the stability and interaction of these proteins, and this modification promotes tumor evasion. In various preclinical models, targeting the N-linked glycosylation of PD-L1/PD-1 was shown to significantly increase the efficacy of PD-L1/PD-1 blockade therapy. Furthermore, deglycosylation of PD-L1 strengthens the signal intensity in PD-L1 immunohistochemistry (IHC) assays, improving the diagnostic and therapeutic relevance of this protein. In this review, we provide an overview of the regulatory mechanisms underlying the N-linked glycosylation of PD-L1/PD-1 as well as the crucial role of N-linked glycosylation in PD-L1/PD-1-mediated immune evasion. In addition, we highlight the promising implications of targeting the N-linked glycosylation of PD-L1/PD-1 in the clinical diagnosis and treatment of cancer. Our review identifies knowledge gaps and sheds new light on the cancer research field.

Application of a single-cell-RNA-based biological-inspired graph neural network in diagnosis of primary liver tumors.

Single-cell technology depicts integrated tumor profiles including both tumor cells and tumor microenvironments, which theoretically enables more robust diagnosis than traditional diagnostic standards based on only pathology. However, the inherent challenges of single-cell RNA sequencing (scRNA-seq) data, such as high dimensionality, low signal-to-noise ratio (SNR), sparse and non-Euclidean nature, pose significant obstacles for traditional diagnostic approaches. The diagnostic value of single-cell technology has been largely unexplored despite the potential advantages. Here, we present a graph neural network-based framework tailored for molecular diagnosis of primary liver tumors using scRNA-seq data. Our approach capitalizes on the biological plausibility inherent in the intercellular communication networks within tumor samples. By integrating pathway activation features within cell clusters and modeling unidirectional inter-cellular communication, we achieve robust discrimination between malignant tumors (including hepatocellular carcinoma, HCC, and intrahepatic cholangiocarcinoma, iCCA) and benign tumors (focal nodular hyperplasia, FNH) by scRNA data of all tissue cells and immunocytes only. The efficacy to distinguish iCCA from HCC was further validated on public datasets. Through extending the application of high-throughput scRNA-seq data into diagnosis approaches focusing on integrated tumor microenvironment profiles rather than a few tumor markers, this framework also sheds light on minimal-invasive diagnostic methods based on migrating/circulating immunocytes.

Ten-eleven translocation-2 inactivation restrains IL-10-producing regulatory B cells to enable antitumor immunity in hepatocellular carcinoma.

BACKGROUND AND AIMS: IL-10-producing regulatory B cells (IL-10 + B cells), a dominant regulatory B cell (Breg) subset, foster tumor progression. However, the mechanisms underlying their generation in HCC are poorly understood. Ten-eleven translocation-2 (TET2), a predominant epigenetic regulatory enzyme in B cells, regulates gene expression by catalyzing demethylation of 5-methylcytosine into 5-hydroxymethyl cytosine (5hmC). In this study, we investigated the role of TET2 in IL-10 + B cell generation in HCC and its prospects for clinical application. APPROACH AND RESULTS: TET2 activation in B cells triggered by oxidative stress from the HCC microenvironment promoted IL-10 expression, whereas adoptive transfer of Tet2 -deficient B cells suppressed HCC progression. The aryl hydrocarbon receptor is required for TET2 to hydroxylate Il10 . In addition, high levels of IL-10, TET2, and 5hmc in B cells indicate poor prognosis in patients with HCC. Moreover, we determined TET2 activity using 5hmc in B cells to evaluate the efficacy of anti-programmed death 1 (anti-PD-1) therapy. Notably, TET2 inhibition in B cells facilitates antitumor immunity to improve anti-PD-1 therapy for HCC. CONCLUSIONS: Our findings propose a TET2-dependent epigenetic intervention targeting IL-10 + B cell generation during HCC progression and identify the inhibition of TET2 activity as a promising combination therapy with immune checkpoint inhibitors for HCC.

A Transformer-Based microvascular invasion classifier enhances prognostic stratification in HCC following radiofrequency ablation.

BACKGROUND & AIMS: We aimed to develop a Transformer-based deep learning (DL) network for prognostic stratification in hepatocellular carcinoma (HCC) patients undergoing RFA. METHODS: A Swin Transformer DL network was trained to establish associations between magnetic resonance imaging (MRI) datasets and the ground truth of microvascular invasion (MVI) based on 696 surgical resection (SR) patients with solitary HCC ≤3 cm, and was validated in an external cohort (n = 180). The multiphase MRI-based DL risk outputs using an optimal threshold of .5 was employed as a MVI classifier for prognosis stratification in the RFA cohort (n = 180). RESULTS: Over 90% of all enrolled patients exhibited hepatitis B virus infection. Liver cirrhosis was significantly more prevalent in the RFA cohort compared to the SR cohort (72.2% vs. 44.1%, p < .001). The MVI risk outputs exhibited good performance (area under the curve values = .938 and .883) for predicting MVI in the training and validation cohort, respectively. The RFA patients at high risk of MVI classified by the MVI classifier demonstrated significantly lower recurrence-free survival (RFS) and overall survival rates at 1, 3 and 5 years compared to those classified as low risk (p < .001). Multivariate cox regression modelling of a-fetoprotein > 20 ng/mL [hazard ratio (HR) = 1.53; 95% confidence interval (95% CI): 1.02-2.33, p = .047], high risk of MVI (HR = 3.76; 95% CI: 2.40-5.88, p < .001) and unfavourable tumour location (HR = 2.15; 95% CI: 1.40-3.29, p = .001) yielded a c-index of .731 (bootstrapped 95% CI: .667-.778) for evaluating RFS after RFA. Among the three risk factors, MVI was the most powerful predictor for intrahepatic distance recurrence. CONCLUSIONS: The proposed MVI classifier can serve as a valuable imaging biomarker for prognostic stratification in early-stage HCC patients undergoing RFA.

Regulation of cohesin-mediated chromosome folding by PDS5 in mammals.

Cohesin regulates sister chromatid cohesion but also contributes to chromosome folding by promoting the formation of chromatin loops, a process mediated by loop extrusion. Although PDS5 regulates cohesin dynamics on chromatin, the exact function of PDS5 in cohesin-mediated chromatin looping remains unclear. Two paralogs of PDS5 exist in vertebrates, PDS5A and PDS5B. Here we show that PDS5A and PDS5B co-localize with RAD21 and CTCF at loop anchors. Rapid PDS5A or PDS5B degradation in liver cancer cells using an inducible degron system reduces chromatin loops and increases loop size. RAD21 enrichment at loop anchors is decreased upon depletion of PDS5A or PDS5B. PDS5B loss also reduces CTCF signals at loop anchors and has a stronger effect on loop enlargement compared with PDS5A. Co-depletion of PDS5A and PDS5B reduces RAD21 levels at loop anchors although the amount of cohesin on chromatin is increased. Our study provides insight into how PDS5 proteins regulate cohesin-mediated chromatin looping.

Robot-assisted resection of renal tumor in children and comparison with laparoscopic surgery.

BACKGROUND: Robot-assisted surgery (RAS) is being performed with increasing frequency in pediatric oncology. We report our experience with RAS for renal tumors in children and compare the outcomes between RAS and laparoscopic surgery (LAS). METHODS: A total of 23 patients with renal tumor who underwent minimally invasive surgery (MIS) between January 2020 and December 2023 were included in the study. The inclusion criteria enrolled in this study was unilateral tumors with maximum tumor diameter less than 10 cm. Patients who had enlarged lymph node, venous thrombosis, preoperative tumor rupture, bilateral renal tumor, or extrarenal extension on imaging were deemed contraindications and excluded. Patient demographics, operative details, postoperative outcomes and follow-up were recorded. RESULTS: Among these patients, 17 underwent RAS and 6 underwent LAS. In the RAS group, the median age was 64 months (range, 9-156) with a median weight of 19.48 kg (range, 8.4-46.5); the maximum tumor diameter at operation was 55.65 mm (range, 22-88); the operation time was 188.8 min (range, 120-210), the intraoperative blood loss was 20 ml (range, 5-50), and the length of postoperative hospital stay was 4 days (range 1-9). There was no significant difference in patients' age, weight, location, tumor size, histological pattern and operation time between the two groups (P > 0.05). The RAS group had a significantly less intraoperative blood loss (P = 0.026) and less length of postoperative stay (P = 0.01) than the LAS group. CONCLUSION: Our initial experience suggested that RAS in pediatric renal tumor was feasible and safe, and it reduced surgical trauma and accelerate postoperative recovery for the patients. Due to the limitations of sample size and study quality, the clinical importance of these findings still needs to be further verified.

Perioperative and oncologic outcomes of laparoscopic versus open liver resection for combined hepatocellular-cholangiocarcinoma: a propensity score matching analysis.

BACKGROUND: Laparoscopic liver resection (LLR) has now been established as a safe and minimally invasive technique that is deemed feasible for treating hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC). However, the role of LLR in treating combined hepatocellular-cholangiocarcinoma (cHCC-CC) patients has been rarely reported. This study aimed to assess the efficacy of LLR when compared with open liver resection (OLR) procedure for patients with cHCC-CC. METHODS: A total of 229 cHCC-CC patients who underwent hepatic resection (34 LLR and 195 OLR patients) from January 2014 to December 2018 in Zhongshan Hospital, Fudan University were enrolled and underwent a 1:2 propensity score matching (PSM) analysis between the LLR and OLR groups to compare perioperative and oncologic outcomes. Overall survival (OS) and recurrence-free survival (RFS) parameters were assessed by the log-rank test and the sensitivity analysis. RESULTS: A total of 34 LLR and 68 OLR patients were included after PSM analysis. The LLR group displayed a shorter postoperative hospital stay (6.61 vs. 8.26 days; p value < 0.001) when compared with the OLR group. No significant differences were observed in the postoperative complications' incidence or a negative surgical margin rate between the two groups (p value = 0.409 and p value = 1.000, respectively). The aspartate aminotransferase (AST), alanine aminotransferase (ALT), and inflammatory indicators in the LLR group were significantly lower than those in the OLR group on the first and third postoperative days. Additionally, OS and RFS were comparable in both the LLR and OLR groups (p value = 0.700 and p value = 0.780, respectively), and similar results were obtained by conducting a sensitivity analysis. CONCLUSION: LLR can impart less liver function damage, better inflammatory response attenuation contributing to a faster recovery, and parallel oncologic outcomes when compared with OLR. Therefore, LLR can be recommended as a safe and effective therapeutic modality for treating selected cHCC-CC patients, especially for those with small tumors in favorable location.

Combining γ-GT, PIVKA-II, and AFP to predict long-term prognosis in patients with hepatocellular carcinoma after hepatectomy.

Hepatocellular carcinoma (HCC) is one of the top five contributors to the cancer burden in China, with a poor prognosis and heavy disability-adjusted life year burden. The criteria used for HCC prognosis are complicated and therefore restricted in routine clinical practice. Multiple factors influence HCC malignancy and progression. In this study, we retrospectively evaluated 173 patients with HCC who underwent curative resection for 9 years to evaluate the correlation of a combination of γ-glutamyl transferase (γ-GT), protein induced by vitamin K absence or antagonist-II (PIVKA-II), and α-fetoprotein (AFP) with the long-term survival of patients with HCC. Multivariate analysis revealed that the γ-GT level was an independent prognostic factor for recurrence. The prediction rate of early recurrence with γ-GT, PIVKA-II, and AFP levels individually was 63.5%, 79.4%, and 39.7%, respectively, whereas the prediction rate of early recurrence was 95.2% with the combination of γ-GT, PIVKA-II, and AFP levels as a composite indicator. Our long-term retrospective study revealed that γ-GT, PIVKA-II, and AFP can aid in predicting long-term prognosis of HCC recurrence. The combination of γ-GT, PIVKA-II, and AFP can further aid in identifying patients with early recurrence. Together, γ-GT, PIVKA-II, and AFP may a be used to develop a new prediction method to improve the prognosis of patients with HCC, and our results indicate the requirement of more active HCC treatment strategies.

Current applications and future perspective of CRISPR/Cas9 gene editing in cancer.

Clustered regularly interspaced short palindromic repeats (CRISPR) system provides adaptive immunity against plasmids and phages in prokaryotes. This system inspires the development of a powerful genome engineering tool, the CRISPR/CRISPR-associated nuclease 9 (CRISPR/Cas9) genome editing system. Due to its high efficiency and precision, the CRISPR/Cas9 technique has been employed to explore the functions of cancer-related genes, establish tumor-bearing animal models and probe drug targets, vastly increasing our understanding of cancer genomics. Here, we review current status of CRISPR/Cas9 gene editing technology in oncological research. We first explain the basic principles of CRISPR/Cas9 gene editing and introduce several new CRISPR-based gene editing modes. We next detail the rapid progress of CRISPR screening in revealing tumorigenesis, metastasis, and drug resistance mechanisms. In addition, we introduce CRISPR/Cas9 system delivery vectors and finally demonstrate the potential of CRISPR/Cas9 engineering to enhance the effect of adoptive T cell therapy (ACT) and reduce adverse reactions.

TET2 promotes anti-tumor immunity by governing G-MDSCs and CD8+ T-cell numbers.

The host immune response is a fundamental mechanism for attenuating cancer progression. Here we report a role for the DNA demethylase and tumor suppressor TET2 in host anti-tumor immunity. Deletion of Tet2 in mice elevates IL-6 levels upon tumor challenge. Elevated IL-6 stimulates immunosuppressive granulocytic myeloid-derived suppressor cells (G-MDSCs), which in turn reduce CD8+ T cells upon tumor challenge. Consequently, systematic knockout of Tet2 in mice leads to accelerated syngeneic tumor growth, which is constrained by anti-PD-1 blockade. Removal of G-MDSCs by the anti-mouse Ly6g antibodies restores CD8+ T-cell numbers in Tet2-/- mice and reboots their anti-tumor activity. Importantly, anti-IL-6 antibody treatment blocks the expansion of G-MDSCs and inhibits syngeneic tumor growth. Collectively, these findings reveal a TET2-mediated IL-6/G-MDSCs/CD8+ T-cell immune response cascade that safeguards host adaptive anti-tumor immunity, offering a cell non-autonomous mechanism of TET2 for tumor suppression.

CircRNA UBAP2 serves as a sponge of miR-1294 to increase tumorigenesis in hepatocellular carcinoma through regulating c-Myc expression.

Circular RNAs (circRNAs) are a class of regulatory RNAs with complex roles in healthy and diseased tissues. However, the oncogenic role of circRNAs in hepatocellular carcinoma (HCC) remains poorly understood, including the mechanisms by which the circular ubiquitin-binding associated protein 2 (circUBAP2) contributes to tumorigenesis. We analyzed the expression of circUBAP2 in 20 paired samples of HCC and healthy tissue as well as in seven HCC cell lines via quantitative real-time polymerase chain reaction. Functional experiments, such as CCK8 viability assays, colony formation assays, wound healing, transwell assays and flow cytometry, were conducted to assess the effects of circUBAP2 in vitro. To further elucidate the mechanisms by which circUBAP2 acts, we conducted dual-luciferase assays, western blots, RNA pull-down assays and rescue experiments. CircUBAP2 was highly upregulated in most HCC tissues and was associated with poor prognosis. HCC patients with high circUBAP2 expression had greater vascular invasion and worse differentiation. Functionally, circUBAP2 overexpression enhanced HCC cell proliferation, migration and invasion and inhibited apoptosis. Furthermore, we found that circUBAP2 upregulated c-Myc expression by sponging miR-1294, thus contributing to hepatocarcinogenesis. Inhibiting circUBAP2 expression in HCC attenuated the oncogenic effects of c-Myc. These findings suggest that circUBAP2 promotes HCC growth and metastasis. CircUBAP2 may have value as an independent prognostic biomarker or as a new target for the treatment of HCC.

Publisher Correction: N6-Methyladenosine methyltransferase ZCCHC4 mediates ribosomal RNA methylation.

In the version of this article originally published, the references were incorrectly re-ordered during production. The hyphen in "N6-methyladenosine" in the title was also superscript. The errors have been corrected in the HTML and PDF versions of the paper.

N6-Methyladenosine methyltransferase ZCCHC4 mediates ribosomal RNA methylation.

N6-Methyladenosine (m6A) RNA modification is present in messenger RNAs (mRNA), ribosomal RNAs (rRNA), and spliceosomal RNAs (snRNA) in humans. Although mRNA m6A modifications have been extensively studied and shown to play critical roles in many cellular processes, the identity of m6A methyltransferases for rRNAs and the function of rRNA m6A modifications are unknown. Here we report a new m6A methyltransferase, ZCCHC4, which primarily methylates human 28S rRNA and also interacts with a subset of mRNAs. ZCCHC4 knockout eliminates m6A4220 modification in 28S rRNA, reduces global translation, and inhibits cell proliferation. We also find that ZCCHC4 protein is overexpressed in hepatocellular carcinoma tumors, and ZCCHC4 knockout significantly reduces tumor size in a xenograft mouse model. Our results highlight the functional significance of an rRNA m6A modification in translation and in tumor biology.

A novel WT1 gene mutation in a chinese girl with denys-drash syndrome.

OBJECTIVE: Denys-Drash syndrome (DDS) is defined by the triad of Wilms tumor, nephrotic syndrome, and/or ambiguous genitalia. Genetic testing may help identify new gene mutation sites and play an important role in clinical decision-making. METHODS: We present a patient with an XY karyotype and female appearance, nephropathy, and Wilms tumor in the right kidney. Genomic DNA was extracted from peripheral blood cells according to standard protocols. "Next-generation" sequencing (NGS) was performed to identify novel variants. The variant was analyzed with Mutation Taster, and its function was explored by a cell growth inhibition assay. RESULTS: We found the first case of Denys-Drash syndrome with the uncommon missense mutation (c.1420C>T, p.His474 Tyr) in the WT1 gene. In silico analysis, the variant was predicted "disease-causing" by Mutation Taster. The mutated variant showed a weaker effect in inhibiting tumor cells than wild-type WT1. CONCLUSIONS: The uncommon missense mutation (c.1420C>T, p.His474 Tyr) in the WT1 gene may be a crucial marker in DDS.

Do the existing staging systems for primary liver cancer apply to combined hepatocellular carcinoma-intrahepatic cholangiocarcinoma?

BACKGROUND: The incidence of combined hepatocellular carcinoma-intrahepatic cholangiocarcinoma (cHCC-ICC) is relatively low, and the knowledge about the prognosis of cHCC-ICC remains obscure. In the study, we aimed to screen existing primary liver cancer staging systems and shed light on the prognosis and risk factors for cHCC-ICC. METHODS: We retrospectively reviewed 206 cHCC-ICC patients who received curative surgical resection from April 1999 to March 2017. The correlation of survival measures with the histological types or with tumor staging systems was determined and predictive values of tumor staging systems with cHCC-ICC prognosis were compared. RESULTS: The histological type was not associated with overall survival (OS) (P = 0.338) or disease-free survival (DFS) (P = 0.843) of patients after curative surgical resection. BCLC, TNM for HCC, and TNM for ICC stages correlated with both OS and DFS in cHCC-ICC (all P < 0.05). The predictive values of TNM for HCC and TNM for ICC stages were similar in terms of predicting postoperative OS (P = 0.798) and DFS (P = 0.191) in cHCC-ICC. TNM for HCC was superior to BCLC for predicting postoperative OS (P = 0.022) in cHCC-ICC. CONCLUSION: The TNM for HCC staging system should be prioritized for clinical applications in predicting cHCC-ICC prognosis.

Cancer cell-derived exosomal circUHRF1 induces natural killer cell exhaustion and may cause resistance to anti-PD1 therapy in hepatocellular carcinoma.

OBJECTIVE: Natural killer (NK) cells play a critical role in the innate antitumor immune response. Recently, NK cell dysfunction has been verified in various malignant tumors, including hepatocellular carcinoma (HCC). However, the molecular biological mechanisms of NK cell dysfunction in human HCC are still obscure. METHODS: The expression of circular ubiquitin-like with PHD and ring finger domain 1 RNA (circUHRF1) in HCC tissues, exosomes, and cell lines was detected by qRT-PCR. Exosomes were isolated from the culture medium of HCC cells and plasma of HCC patients using an ultracentrifugation method and the ExoQuick Exosome Precipitation Solution kit and then characterized by transmission electronic microscopy, NanoSight and western blotting. The role of circUHRF1 in NK cell dysfunction was assessed by ELISA. In vivo circRNA precipitation, RNA immunoprecipitation, and luciferase reporter assays were performed to explore the molecular mechanisms of circUHRF1 in NK cells. In a retrospective study, the clinical characteristics and prognostic significance of circUHRF1 were determined in HCC tissues. RESULTS: Here, we report that the expression of circUHRF1 is higher in human HCC tissues than in matched adjacent nontumor tissues. Increased levels of circUHRF1 indicate poor clinical prognosis and NK cell dysfunction in patients with HCC. In HCC patient plasma, circUHRF1 is predominantly secreted by HCC cells in an exosomal manner, and circUHRF1 inhibits NK cell-derived IFN-γ and TNF-α secretion. A high level of plasma exosomal circUHRF1 is associated with a decreased NK cell proportion and decreased NK cell tumor infiltration. Moreover, circUHRF1 inhibits NK cell function by upregulating the expression of TIM-3 via degradation of miR-449c-5p. Finally, we show that circUHRF1 may drive resistance to anti-PD1 immunotherapy in HCC patients. CONCLUSIONS: Exosomal circUHRF1 is predominantly secreted by HCC cells and contributes to immunosuppression by inducing NK cell dysfunction in HCC. CircUHRF1 may drive resistance to anti-PD1 immunotherapy, providing a potential therapeutic strategy for patients with HCC.

Prevention and control strategies in the diagnosis and treatment of solid tumors in children during the COVID-19 pandemic.

Since the emergence of patients with COVID-19 in December 2019, the virus has rapidly spread worldwide. Children with malignant tumors are more prone to be infected and develop severe infections. Prevention and control of the pandemic and ensuring the progress of the cancer diagnosis and treatment is a significant problem in the current scenario. This article proposes a scientific management system for patients with solid tumors to guarantee emergency surgery, rationally arrange limited-term surgery, appropriately defer elective surgery, and guarantee regular chemotherapy, while protecting children from SARS-CoV-2 infection and ensuring the continuity of comprehensive diagnosis and treatment.

Genome-wide mapping of 5-hydroxymethylcytosines in circulating cell-free DNA as a non-invasive approach for early detection of hepatocellular carcinoma.

OBJECTIVE: The lack of highly sensitive and specific diagnostic biomarkers is a major contributor to the poor outcomes of patients with hepatocellular carcinoma (HCC). We sought to develop a non-invasive diagnostic approach using circulating cell-free DNA (cfDNA) for the early detection of HCC. DESIGN: Applying the 5hmC-Seal technique, we obtained genome-wide 5-hydroxymethylcytosines (5hmC) in cfDNA samples from 2554 Chinese subjects: 1204 patients with HCC, 392 patients with chronic hepatitis B virus infection (CHB) or liver cirrhosis (LC) and 958 healthy individuals and patients with benign liver lesions. A diagnostic model for early HCC was developed through case-control analyses using the elastic net regularisation for feature selection. RESULTS: The 5hmC-Seal data from patients with HCC showed a genome-wide distribution enriched with liver-derived enhancer marks. We developed a 32-gene diagnostic model that accurately distinguished early HCC (stage 0/A) based on the Barcelona Clinic Liver Cancer staging system from non-HCC (validation set: area under curve (AUC)=88.4%; (95% CI 85.8% to 91.1%)), showing superior performance over α-fetoprotein (AFP). Besides detecting patients with early stage or small tumours (eg, ≤2.0 cm) from non-HCC, the 5hmC model showed high capacity for distinguishing early HCC from high risk subjects with CHB or LC history (validation set: AUC=84.6%; (95% CI 80.6% to 88.7%)), also significantly outperforming AFP. Furthermore, the 5hmC diagnostic model appeared to be independent from potential confounders (eg, smoking/alcohol intake history). CONCLUSION: We have developed and validated a non-invasive approach with clinical application potential for the early detection of HCC that are still surgically resectable in high risk individuals.

Predicting overall survival of patients with hepatocellular carcinoma using a three-category method based on DNA methylation and machine learning.

Hepatocellular carcinoma (HCC) is closely associated with abnormal DNA methylation. In this study, we analyzed 450K methylation chip data from 377 HCC samples and 50 adjacent normal samples in the TCGA database. We screened 47,099 differentially methylated sites using Cox regression as well as SVM-RFE and FW-SVM algorithms, and constructed a model using three risk categories to predict the overall survival based on 134 methylation sites. The model showed a 10-fold cross-validation score of 0.95 and satisfactory predictive power, and correctly classified 26 of 33 samples in testing set obtained by stratified sampling from high, intermediate and low risk groups.