Dataset Information

Pancreatic cancer RNA sequencing

ABSTRACT: For optimal pancreatic cancer treatment, early and accurate diagnosis is vital. Blood-derived biomarkers and genetic predispositions can contribute to early diagnosis, but they often have limited accuracy or applicability. Here, we seek to exploit the synergy between both approaches by combining the standard clinical blood biomarker CA19-9 with novel genetic variants. Concretely, we aim to use deep sequencing and deep learning to improve pancreatic cancer diagnosis, to differentiate between resectable pancreatic ductal adenocarcinoma (rPDAC) from chronic pancreatitis (CP), and to estimate survival. We obtained samples of nucleated cells found in peripheral blood from over 300 patients suffering from resectable pancreatic ductal adenocarcinoma, non-resectable pancreatic cancer (nrPC), chronic pancreatitis or none of these. We sequenced RNA with high coverage and reduced millions of raw to hundreds of high-quality, significant genetic variants. Together with CA19-9 levels, these served as input to deep learning to separate cancer from non-cancer, resectable PDAC from chronic pancreatitis, and resectable from non-resectable cancer. All deep learning models achieved area under the curve (AUC) scores of over 90%. In particular, differentiating resectable PDAC from pancreatitis can be solved with an AUC of 98%. Moreover, we identified genetic variants to estimate survival in rPDAC patients. Overall, we show that the blood transcriptome harbours genetic variants, which substantially improve non-invasive clinical diagnosis and patient stratification in pancreatic cancer.

PROVIDER: EGAS00001004706 | EGA |

REPOSITORIES: EGA

ACCESS DATA

Similar Datasets

Project description:Pancreatic cancer is the fourth leading cause of cancer death. Lack of early detection technology for pancreatic cancer invariably leads to a typical clinical presentation of incurable disease at initial diagnosis. Oral fluid (saliva) meets the demand for non-invasive, accessible, and highly efficient diagnostic medium. The level of salivary analytes, such as mRNA and microflora, vary upon disease onset; thus possess valuable signatures for early detection and screening. In this study, we evaluated the performance and translational utilities of the salivary transcriptomic and microbial biomarkers for non-invasive detection of early pancreatic cancer. Two biomarker discovery technologies were used to profile transcriptome in saliva supernatant and microflora in saliva pellet. The Affymetrix Human Genome U133 Plus 2.0 Array was used to discover altered gene expression in saliva supernatant. The Human Oral Microbe Identification Microarray (HOMIM) was used to investigate microflora shift in saliva pellet. Biomarkers selected from both studies were subjected to an independent clinical validation using a cohort of 30 early pancreatic cancer, 30 chronic pancreatitis and 30 healthy matched-control saliva samples. Two panels of salivary biomarkers, including eleven mRNA biomarkers and two microbial biomarkers were discovered and validated for pancreatic cancer detection. The logistic regression model with the combination of three mRNA biomarkers (ACRV1, DMXL2 and DPM1) yielded a ROC-plot AUC value of 0.974 (95% CI, 0.896 to 0.997; P < 0.0001) with 93.3% sensitivity and 90% specificity in distinguishing pancreatic cancer patients from healthy subjects. The logistic regression model with the combination of two bacterial biomarkers (Neisseria elongata and Streptococcus mitis) yielded a ROC-plot AUC value of 0.895 (95% CI, 0.784 to 0.961; P < 0.0001) with 96.4% sensitivity and 82.1% specificity in distinguishing pancreatic cancer patients from healthy subjects. Importantly, the logistic regression model with the combination of four biomarkers (mRNA biomarkers, ACRV1, DMXL2 and DPM1; bacterial biomarker, S. mitis) could differentiate pancreatic cancer patients from all non-cancer subjects (chronic pancreatitis and healthy control), yielding a ROC-plot AUC value of 0.949 (95% CI, 0.877 to 0.985; P < 0.0001) with 92.9% sensitivity and 85.5% specificity. This study comprehensively compared the salivary transcriptome and microflora between pancreatic cancer and control subjects. We have discovered and validated eleven mRNA biomarkers and two microbial biomarkers for early detection of pancreatic cancer in saliva. The logistic regression model with four salivary biomarkers can detect pancreatic cancer specifically without the complication of chronic pancreatitis. This is the first report demonstrating the value of multiplex salivary biomarkers for the non-invasive detection of a high impact systemic cancer. Keywords: Salivary biomarker, pancreatic cancer, early detection, salivary transcriptome, salivary microflora This study, which was approved by the UCLA Institutional Review Board, started sample collection in February 2006 and ended in April 2008. It had a discovery and verification phase, followed by an independent validation phase. All patients with clinically diagnosed early or locally advanced pancreatic cancer, chronic pancreatitis and matched healthy control were recruited from the UCLA Medical Center. The controls were matched for gender, age, ethnicity and smoking history. We chose early and locally advanced pancreatic cancers because they were considered early stage diseases. All patients were recently diagnosed with primary disease, and had not received any prior treatment in the form of chemotherapy, radiotherapy, surgery, or alternative remedies. No subjects had a history of prior malignancy, immunodeficiency, autoimmune disorders, hepatitis, or HIV infection. Written informed consent was obtained from all patients who agreed to serve as saliva donors. Unstimulated saliva samples were collected and processed as previously described. The saliva bank of pancreatic disease at the UCLA Dental Research Institute has collected 283 saliva samples since 2006. Of these, 114 samples, from 42 pancreatic cancer patients, 30 chronic pancreatitis patients and 42 healthy control subjects, met the following eligibility criteria: the disease subjectsâ?? samples were obtained immediately after primary diagnosis from patients without evidence of metastasis; the saliva sample was free of blood (Table 1 in accompanying manuscript). Of the 114 samples, 12 pancreatic cancer samples and 12 healthy control samples were chosen for the discovery and verification phase. The transcriptomic approach profiled the saliva supernatant samples from 12 pancreatic cancer patients and 12 healthy control subjects using the Affymetrix Human Genome U133 Plus 2.0 Array platform.

Project description:BACKGROUND & AIMS: Pancreatic ductal adenocarcinoma (PDAC) is usually incurable. Contrary to genetic mechanisms involved in PDAC pathogenesis, epigenetic alterations are ill defined. Here we determine the contribution of epigenetically silenced genes to the development of PDAC. METHODS: We investigated methylated DNAs from PDACs, chronic pancreatitis and normal pancreatic tissues using Methyl-CpG immunoprecipitation followed by microarray hybridization. Promoter methylation of selected genes was confirmed with the Epityper assay. Expression levels were evaluated by quantitative RT-PCR. WNK2 was further investigated in tissue microarrays, methylation analysis of early pancreatic intraepithelial neoplasia (PanINs), mouse models for PDAC and pancreatitis, re-expression studies after demethylation, and cell growth assays using WNK2 overexpression. RESULTS: A total of 3.8% of 27.800 interrogated CpG islands were hypermethylated in PDAC versus normal and chronic pancreatitis tissues. Hypermethylation was confirmed in 12 out of 13 selected islands and was associated with gene silencing in 4 of them. The most prominently hypermethylated gene, WNK2, was further investigated. Demethylation assays confirmed the link between methylation and expression. WNK2 hypermethylation was higher in pancreatic tumor cells than in surrounding inflamed tissues and was observed in PanIN lesions as well as in a PDAC mouse model. WNK2 mRNA and protein expression were lower in PDAC and chronic pancreatitis compared to normal tissues both in patients and mouse models. Overexpression of WNK2 led to a reduced cell growth and WNK2 expression in tissues correlated negatively with the expression of pERK1/2, a downstream target of WNK2 responsible for cell proliferation. CONCLUSIONS: WNK2 is downregulated by promoter hypermethylation early in PDAC pathogenesis and may support tumor cell growth via the ERK-MAPK pathway. 3 types of pancreatic tissue samples: 5 normals, 2 chronic pancreatitis, 7 tumors (PDAC)

Project description:Pancreatic ductal adenocarcinoma is one of the most lethal cancers, with a 5 year-survival rate below 5%. Lack of curative treatment and failure of targeted therapies urge the need to identify novel efficient therapeutic strategy. Achievement of this goal will be obtained through the identification of diagnosis and prognosis biomarkers, identification of novel therapeutic targets and the knowledge of resistance mechanisms induced by these targeted therapies. PI3K/Akt/mTOR signalling, one of the most altered in cancers, is overactivated in pancreatic cancer and correlated with poor prognosis. In the Vertebrates, the family of class I phosphoinoitide-3-kinase (PI3K) includes four isoforms: p110α, p110β, p110δ and p110γ. Although they all perform the same biochemical reaction (phosphorylation of PIP2 in PIP3, a membrane lipid messenger), each isoform were demonstrated to have specific physiological roles. Global PI3K inhibitors are currently being tested in phase I/II clinical trials in advanced solid cancers, but show at maximal doses tolerated a limited therapeutic benefit. Isoform-selective PI3K inhibitors are currently the most promising agents because, at low doses, they are more efficient to inhibit one PI3K isoform, and thus, less toxic than pan-PI3K inhibitors. The objectives of this thesis are to determine isoform-specific PI3K roles and the therapeutic interest to target one or more isoforms in pancreatitis and PDAC, by the identification of isoform-specific pathways and the study of adaptive responses induced by targeting of one or all isoforms of PI3K. In a first part, my work has highlighted, validated and completed results obtained in the team, to demonstrate the significance of PI3K/Akt signalling in two physiological processes: chronic pancreatitis and initiation of pancreatic carcinogenesis. Precisely, the overactivation of PI3K/Akt pathway measured on human and murine chronic pancreatitis samples is correlated with a specific p110α activation gene expression signature. Moreover, genetic and pharmacologic inactivation of p110α during pancreatic chronic inflammation or cancerogenesis (by oncogenic Kras) prevents the formation of acino-ductal metaplasia, structures at the origin of pancreatic carcinogenesis initiation. Development of in vitro acino-ductal transdifferentiation protocol allowed me to demonstrate that only p110α is necessary at this initial step of pancreatic carcinogenesis by the regulation of Rho small GTPases, further regulating actin remodelling. In the second part, by a phosphoproteomic-based approach, I quantified PI3K downstream phosphorylation-regulated targets in a pancreatic cancer cell line treated or not by a pan- or selective PI3K inhibitor at different times. I demonstrated for the first time existence of targets, signalling pathways and adaptive responses regulated by each PI3K isoform. To conclude, all these results demonstrate the rational of combinatorial use of isoform-specific PI3K inhibitors in patients with pancreatic cancer for better clinical response

Dataset Information

Pancreatic cancer RNA sequencing

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets