Gliomas are the most common primary brain malignancies. Accurate and robust tumor segmentation and prediction of patients' overall survival are important for diagnosis, treatment planning and risk factor identification. Here we present a deep learning-based framework for brain tumor segmentation and survival prediction in glioma, using multimodal MRI scans. For tumor segmentation, we use ensembles of three different 3D CNN architectures for robust performance through a majority rule. This approach can effectively reduce model bias and boost performance. For survival prediction, we extract 4,524 radiomic features from segmented tumor regions, then, a decision tree and cross validation are used to select potent features. Finally, a random forest model is trained to predict the overall survival of patients. The 2018 MICCAI Multimodal Brain Tumor Segmentation Challenge (BraTS), ranks our method at 2nd and 5th place out of 60+ participating teams for survival prediction tasks and segmentation tasks respectively, achieving a promising 61.0% accuracy on the classification of short-survivors, mid-survivors and long-survivors.

ObjectivesHepatoid adenocarcinoma of the stomach (HAS) is characterized by histological resemblance to hepatocellular carcinoma and a poor prognosis. The aim of this study is to elucidate the clinicopathological and molecular characteristics of HAS.MethodsForty-two patients with HAS who received gastrectomy were enrolled in this study. Based on a panel of 483 cancer-related genes, targeted sequencing of 24 HAS and 22 clinical parameter-matched common gastric cancer (CGC) samples was performed. Prognostic factors for overall survival (OS) and disease-free survival (DFS) were analysed with the Kaplan–Meier method.ResultsThe most frequently mutated gene in both HAS and CGC was TP53, with a mutation rate of 30%. Additionally, CEBPA, RPTOR, WISP3, MARK1, and CD3EAP were identified as genes with high-frequency mutations in HAS (10–20%). Copy number gains (CNGs) at 20q11.21-13.12 occurred frequently in HAS, nearly 50% of HAS tumours harboured at least one gene with a CNG at 20q11.21-13.12. This CNG tended to be related to more adverse biobehaviour, including poorer differentiation, greater vascular and nerve invasion, and greater liver metastasis. Pathway enrichment analysis revealed that the HIF-1 signalling pathway and signalling pathways regulating stem cell pluripotency were specifically enriched in HAS. The survival analysis showed that a preoperative serum AFP level ≥ 500 ng/ml was significantly associated with poorer OS (p = 0.007) and tended to be associated with poorer DFS (p = 0.05).ConclusionCNGs at 20q11.21-13.12 happened frequently in HAS and tended to be related to more adverse biobehaviour. The preoperative serum AFP level was a sensitive prognostic biomarker for DFS and OS.Electronic supplementary materialThe online version of this article (10.1007/s10120-019-00965-5) contains supplementary material, which is available to authorized users.

No abstract

No abstract

Background Cancer poses a serious threat to the health of Chinese people, resulting in a major challenge for public health work. Today, people can obtain relevant information from not only medical workers in hospitals, but also the internet in any place in real-time. Search behaviors can reflect a population’s awareness of cancer from a completely new perspective, which could be driven by the underlying cancer epidemiology. However, such Web-retrieved data are not yet well validated or understood. Objective This study aimed to explore whether a correlation exists between the incidence and mortality of cancers and normalized internet search volumes on the big data platform, Baidu. We also assessed whether the distribution of people who searched for specific types of cancer differed by gender. Finally, we determined whether there were regional disparities among people who searched the Web for cancer-related information. Methods Standard Boolean operators were used to choose search terms for each type of cancer. Spearman’s correlation analysis was used to explore correlations among monthly search index values for each cancer type and their monthly incidence and mortality rates. We conducted cointegration analysis between search index data and incidence rates to examine whether a stable equilibrium existed between them. We also conducted cointegration analysis between search index data and mortality data. Results The monthly Baidu index was significantly correlated with cancer incidence rates for 26 of 28 cancers in China (lung cancer: r =.80, P <.001; liver cancer: r =.28, P =.016; stomach cancer: r =.50, P <.001; esophageal cancer: r =.50, P <.001; colorectal cancer: r =.81, P <.001; pancreatic cancer: r =.86, P <.001; breast cancer: r =.56, P <.001; brain and nervous system cancer: r =.63, P <.001; leukemia: r =.75, P <.001; Non-Hodgkin lymphoma: r =.88, P <.001; Hodgkin lymphoma: r =.91, P <.001; cervical cancer: r =.64, P <.001; prostate cancer: r =.67, P <.001; bladder cancer: r =.62, P <.001; gallbladder and biliary tract cancer: r =.88, P <.001; lip and oral cavity cancer: r =.88, ...

Transforming growth factor-β1 is considered a key contributor to the progression of breast cancer. MicroRNAs are important factors in the development and progression of many malignancies. In the present study, upon studies of breast cancer cell lines and tissues, we showed that microRNA -196a-3p is decreased by transforming growth factor-β1 in breast cancer cells and associated with breast cancer progression. We identified neuropilin-2 as a target gene of microRNA -196a-3p and showed that it is regulated by transforming growth factor-β1. Moreover, transforming growth factor-β1-mediated inhibition of microRNA -196a-3p and activation of neuropilin-2were required for transforming growth factor-β1-induced migration and invasion of breast cancer cells. In addition, neuropilin-2 expression was suppressed in breast tumors, particularly in triple-negative breast cancers. Collectively, our findings strongly indicate that microRNA -196a-3p is a predictive biomarker of breast cancer metastasis and patient survival and a potential therapeutic target in metastatic breast cancer.

BackgroundGATA-binding protein 3 (GATA3) has been identified as a sensitive marker for breast carcinoma but its sensitivity in primary genital extramammary Paget diseases (EMPDs) has not been well studied.MethodsHere we investigated immunohistochemical expression of GATA3 in 72 primary genital EMPDs (35 from female, 37 from male; 45 with intraepithelial disease only, 26 with both intraepithelial disease and invasive adenocarcinoma including 14 also metastasis, 1 with metastatic adenocarcinoma only for study). We also compared GATA3 to gross cystic disease fluid protein 15 (GCDFP15) for their sensitivity.ResultsPositive GATA3 staining was seen in all 71 (100%) intraepithelial diseases, 25/26 (96%; female 10/10, male 15/16) invasive adenocarcinomas and 14/15 (93%; female 3/3, male 11/12) metastatic adenocarcinomas, respectively. Positive GCDFP15 staining was seen in 46/71 (65%; female 28/34 or 82%, male 18/37 or 49%) intraepithelial diseases, 20/26 (77%; female 9/10, male 11/16) invasive adenocarcinomas, and 12/15 (80%; female 2/3, male 10/12) metastatic adenocarcinomas, respectively (GATA3 versus GCDFP15: p < 0.01 for both intraepithelial disease and invasive adenocarcinoma, p = 0.28 for metastatic adenocarcinoma). In positive-stained cases, GATA3 stained more tumor cells than GCDFP15 (79% versus 25% for intraepithelial disease, 71% vs 34% for invasive adenocarcinoma, 73% vs 50% for metastatic adenocarcinoma, p < 0.01 for all 3 components).ConclusionsOur findings indicate that GATA3 is a very sensitive marker for primary genital EMPDs and is more sensitive than GCDFP15.

Accurate evaluation of tumor response to preoperative chemotherapy is crucial for assigning appropriate patients with colorectal liver metastases (CRLM) to surgery or conservative therapy. However, there is no well‐recognized method for predicting pathological response before surgery. Our study constructed and validated a deep learning algorithm using prechemotherapy and postchemotherapy magnetic resonance imaging (MRI) to predict pathological response in CRLM. CRLM patients from center one who had ≤5 lesions and were scheduled to receive preoperative chemotherapy followed by liver resection between January 2013 and November 2016, were included prospectively and chronologically divided into a training cohort (80% of patients) and a testing cohort (20% of patients). Patients from center two were included January 2017 and December 2018 as an external validation cohort. MRI‐based models were constructed to discriminate according to pathology tumor regression grade (TRG) between the response (TRG1/2) and nonresponse (TRG3/4/5) groups at the lesion level. From center one, 155 patients (328 lesions) were included; chronologically, 101 (264 lesions) in the training cohort and 54 (64 lesions) in the testing cohort. The model achieved better accuracy (0.875 vs 0.578) and AUC (0.849 vs 0.615) than RECIST for discriminating response; it also distinguished the survival outcomes after hepatectomy better than the RECIST criteria. Evaluations of the external validation cohort (25 patients, 61 lesions) also showed good ability with an AUC of 0.833. In conclusion, the MRI‐based deep learning model provided accurate prediction of pathological tumor response to preoperative chemotherapy in patients with CRLM and may inform individualized treatment.