Isocitrate dehydrogenase () mutations in glioma patients confer longer survival and may guide treatment decision making. We aimed to predict the status of gliomas from MR imaging by applying a residual convolutional neural network to preoperative radiographic data. Preoperative imaging was acquired for 201 patients from the Hospital of University of Pennsylvania (HUP), 157 patients from Brigham and Women's Hospital (BWH), and 138 patients from The Cancer Imaging Archive (TCIA) and divided into training, validation, and testing sets. We trained a residual convolutional neural network for each MR sequence (FLAIR, T2, T1 precontrast, and T1 postcontrast) and built a predictive model from the outputs. To increase the size of the training set and prevent overfitting, we augmented the training set images by introducing random rotations, translations, flips, shearing, and zooming. With our neural network model, we achieved IDH prediction accuracies of 82.8% (AUC = 0.90), 83.0% (AUC = 0.93), and 85.7% (AUC = 0.94) within training, validation, and testing sets, respectively. When age at diagnosis was incorporated into the model, the training, validation, and testing accuracies increased to 87.3% (AUC = 0.93), 87.6% (AUC = 0.95), and 89.1% (AUC = 0.95), respectively. We developed a deep learning technique to noninvasively predict genotype in grade II-IV glioma using conventional MR imaging using a multi-institutional data set. .

No enhancement and a smooth non-enhancing margin on MRI were predictive of longer PFS, while a smooth non-enhancing margin was a significant predictor of longer OS in LGGs. Textural analyses of MR imaging data predicted IDH1 mutation, 1p/19q codeletion, histological grade, and tumor progression with high accuracy.

MicroRNAs are tiny RNA molecules which serve as important posttranscriptional regulators of gene expression. Dysregulated expression of microRNAs has been observed in human cancers, indicating that microRNAs may function as oncogenes or as tumor suppressors. To date, the microRNAs encoded by the oncogenic miR-17-92 cluster, and its paralog the miR-106b-25 cluster, are among those which are differentially expressed in human cancers. In this study, we examined and confirmed the over-expression of these clusters in hepatocellular carcinoma and in hepatoma-derived cells. At least 50% of the tumor samples showed a greater than two-fold increase in the expression for miR-18 and for the miR-106b-25 cluster when compared with the corresponding paired non-tumor samples. Knock-down studies for the miR-106b-25 cluster, which includes miR-106b, miR-93 and miR-25, showed that the expression of the cluster is necessary for cell proliferation and for anchorageindependent growth. In tumors with high expression of this cluster, reduced expression of the BH3-only protein Bim, a miR-25 target, was observed. We further identified the transcription factor E2F1 as a target gene for miR-106b and miR-93 and it is likely that one of the roles of the miR-106b-25 cluster is to prevent excessively high E2F1 expression, which may then cause apoptosis. We conclude that there is aberrant expression of microRNAs encoded by the oncogenic miR-17-92 cluster and the miR-106b-25 cluster in hepatocellular carcinoma. The consistent overexpression of the miR-106b-25 cluster and its role in cell proliferation and anchorage-independent growth points to the oncogenic potential of this cluster. (Cancer Sci 2009; 100: 1234-1242) M icroRNAs (miRNAs) are small, endogenous RNA molecules which are transcribed as long primary transcripts, processed to the precursor miRNAs (pre-miRNA), and further processed to yield approximately 22-nucleotide duplexes of which one arm gives rise to the mature miRNA.(1) Depending on the degree of complementarity, binding of miRNA to its target mRNA can result in either cleavage of the target or in translation repression.(2-4) Despite a comprehensive registry of known miRNAs and the genes which encode them (http://microrna.sanger.ac.uk/sequences/), knowledge of exactly which mRNA and processes each miRNA regulates is incomplete. Studies have shown that miRNAs can influence diverse pathways and affect many physiological processes including metabolism, development, differentiation, and apoptosis.(5-8) Altered expression of miRNAs has been demonstrated in various human cancers including chronic lymphocytic leukemia, (9) lung cancer, (10)(11)(12) colorectal neoplasia, (13,14) breast cancer, (15) and several types of lymphoma. (16)(17)(18)(19) This suggests that miRNAs might function as oncogenes or as tumor suppressors. (20) MicroRNA genes can occur in clusters and these are transcribed as polycistronic transcripts. To date, the miR-17-92 cluster is the best characterized. The miRNAs encoded by this cluster are overexpressed in many ...