The association between mutations of key driver genes and colorectal cancer (CRC) metastasis has been investigated by many studies. However, the results of these studies have been contradictory. Here, we perform a comprehensive analysis to screen key driver genes from the TCGA database and validate the roles of these mutations in CRC metastasis. Using bioinformatics analysis, we identified six key driver genes, namely APC, KRAS, BRAF, PIK3CA, SMAD4 and p53. Through a systematic search, 120 articles published by November 30, 2017, were included, which all showed roles for these gene mutations in CRC metastasis. A meta-analysis showed that KRAS mutations (combined OR 1.18, 95% CI 1.05-1.33) and p53 mutations (combined OR 1.49, 95% CI 1.23-1.80) were associated with CRC metastasis, including lymphatic and distant metastases. Moreover, CRC patients with a KRAS mutation (combined OR 1.29, 95% CI 1.13-1.47), p53 mutation (combined OR 1.35, 95% CI 1.06-1.72) or SMAD4 mutation (combined OR 2.04, 95% CI 1.41-2.95) were at a higher risk of distant metastasis. Subgroup analysis stratified by ethnic populations indicated that the BRAF mutation was related to CRC metastasis (combined OR 1.42, 95% CI 1.18-1.71) and distant metastasis (combined OR 1.51, 95% CI 1.20-1.91) in an Asian population. No significant association was found between mutations of APC or PIK3CA and CRC metastasis. In conclusion, mutations of KRAS, p53, SMAD4 and BRAF play significant roles in CRC metastasis and may be both potential biomarkers of CRC metastasis as well as therapeutic targets.

Background/Aims: Elevated serum cholesterol levels were linked to a higher risk of colorectal adenoma and colorectal cancer (CRC), while the effect of cholesterol on CRC metastasis has not been widely studied. Methods: CRC patients were enrolled to evaluate the association between low-density lipoprotein cholesterol (LDL) and CRC metastases, and LDL receptor (LDLR) level of the CRC tissue was assessed by immunohistochemistry. The effects of LDL on cell proliferation, migration and stemness were assessed in CRC cells in vitro, and the effects of high fat diet (HFD) on tumor growth and intestinal tumorigenicity were investigated in vivo. ROS assays, gene expression array analysis and western blot were used to explore the mechanisms of LDL in CRC progression. Results: The level of LDL was positively correlated with liver metastases, and a higher level of LDL receptor (LDLR) expression was associated with advanced N and M stages of CRC. In vitro, LDL promoted the migration and sphere formation of CRC cells and induced upregulated expression of “stemness” genes including Sox2, Oct4, Nanog and Bmi 1. High-fat diet (HFD) significantly enhanced tumor growth in vivo, and was associated with a shorter intestinal length in azoxymethane/dextran sodium sulfate (AOM/DSS)-treated mice. Furthermore, LDL significantly elevated reactive oxygen species (ROS) levels and Whole Human Genome Microarray found 87 differentially expressed genes between LDL-treated CRC cells and controls, which were largely clustered in the MAP kinase (MAPK) signaling pathway. Conclusions: LDL enhances intestinal inflammation and CRC progression via activation of ROS and signaling pathways including the MAPK pathway. Inflammation is strongly associated with cancer initiation, and the role of LDL in intestinal tumorigenicity should be further explored.

Recent advances on 3D object detection heavily rely on how the 3D data are represented, i.e., voxel-based or point-based representation. Many existing high performance 3D detectors are point-based because this structure can better retain precise point positions. Nevertheless, point-level features lead to high computation overheads due to unordered storage. In contrast, the voxel-based structure is better suited for feature extraction but often yields lower accuracy because the input data are divided into grids. In this paper, we take a slightly different viewpoint --- we find that precise positioning of raw points is not essential for high performance 3D object detection and that the coarse voxel granularity can also offer sufficient detection accuracy. Bearing this view in mind, we devise a simple but effective voxel-based framework, named Voxel R-CNN. By taking full advantage of voxel features in a two-stage approach, our method achieves comparable detection accuracy with state-of-the-art point-based models, but at a fraction of the computation cost. Voxel R-CNN consists of a 3D backbone network, a 2D bird-eye-view (BEV) Region Proposal Network, and a detect head. A voxel RoI pooling is devised to extract RoI features directly from voxel features for further refinement. Extensive experiments are conducted on the widely used KITTI Dataset and the more recent Waymo Open Dataset. Our results show that compared to existing voxel-based methods, Voxel R-CNN delivers a higher detection accuracy while maintaining a real-time frame processing rate, i.e., at a speed of 25 FPS on an NVIDIA RTX 2080 Ti GPU. The code is available at https://github.com/djiajunustc/Voxel-R-CNN.

These findings provide further indication that postdiagnosis aspirin therapy improved CRC overall survival, especially for patients with positive PTGS2 (COX-2) expression and mutated PIK3CA tumours.

Supplemental Digital Content is available in the text

In the recent years, fluorine 18 fluorodeoxyglucose ( 18 F-FDG) positron emission tomography (PET)/computed tomography (CT) has emerged as a new modality for staging non-small-cell lung cancer (NSCLC) patients. The aim of this meta-analysis was to assess the diagnostic value of 18 F-FDG PET/CT in detecting metastatic lesions in NSCLC patients. Meta-analysis methods were used to pool sensitivity, specificity, positive and negative likehood ratios, diagnostic odd ratios and to construct a summary receiver-operating characteristic curve. Data from included studies were pooled to compare the diagnostic accuracy between PET/CT and PET or CT alone in nodal staging. Totally, 56 studies involving 8,699 patients met the inclusion criteria. Lung cancer is the leading cause of tumor-related deaths worldwide, and non-small-cell lung cancer (NSCLC) accounts for about 80% of all lung cancers. 1 The current criteria for staging NSCLC is based on the TNM system, 2 which determines treatment options and predicts survival. 3 So, staging NSCLC is important, and accurate clinical methods are in great need. Various diagnostic methods have been used for staging NSCLC. [4][5][6][7] Computed tomography (CT) has been widely used to evaluate the nodal status of lung cancer based on the size or shape of the lymph nodes. 8 However, the sensitivity and the specificity were relatively low, because the lymph node size may not correlate with the presence of metastatic disease. 9 Based on the fact that malignant cells show higher rates of glycolysis than most surrounding normal structures, 10 fluorine 18 fluorodeoxyglucose (FDG) positron emission tomography (PET) has been introduced and developed as an effective modality for tumor staging in a variety of cancers. 11 However, PET has relatively poor spatial resolution, thus limits its anatomical localization of lesions. 12 Mediastinoscopy, endobronchial and endoscopic ultrasound, and other surgical procedures are frequently used for thoracic nodal evaluation in lung cancer, but these techniques are invasive and cumbersome. 7,13 In the past recent years, 18 F-FDG PET/CT has emerged as a new modality for staging NSCLC patients. 4,14 The aim of this study was to evaluate the diagnostic accuracy of 18 F-FDG PET/CT in detecting metastases in NSCLC patients. Material and Methods Selection criteriaSelection criteria were given as follows: (1) 18 F-FDG PET/CT was used for the evaluation of lymph node (LN) metastases or extrapulmonary metastases in NSCLC; (2) histologic assessment should be applied as reference standard for LN metastases; for studies evaluated extrathoracic metastases, other reference standards such as imaging modality, radiological and clinical follow-up were also applied; (3) absolute numbers of true-positive (TP), false-positive (FP), true-

Recent advances on 3D object detection heavily rely on how the 3D data are represented, i.e., voxel-based or point-based representation. Many existing high performance 3D detectors are point-based because this structure can better retain precise point positions. Nevertheless, point-level features lead to high computation overheads due to unordered storage. In contrast, the voxel-based structure is better suited for feature extraction but often yields lower accuracy because the input data are divided into grids. In this paper, we take a slightly different viewpoint -we find that precise positioning of raw points is not essential for high performance 3D object detection and that the coarse voxel granularity can also offer sufficient detection accuracy. Bearing this view in mind, we devise a simple but effective voxel-based framework, named Voxel R-CNN. By taking full advantage of voxel features in a two stage approach, our method achieves comparable detection accuracy with state-of-the-art point-based models, but at a fraction of the computation cost. Voxel R-CNN consists of a 3D backbone network, a 2D bird-eye-view (BEV) Region Proposal Network and a detect head. A voxel RoI pooling is devised to extract RoI features directly from voxel features for further refinement. Extensive experiments are conducted on the widely used KITTI Dataset and the more recent Waymo Open Dataset. Our results show that compared to existing voxel-based methods, Voxel R-CNN delivers a higher detection accuracy while maintaining a real-time frame processing rate, i.e., at a speed of 25 FPS on an NVIDIA RTX 2080 Ti GPU. The code will be make available soon.

Plenty of studies have assessed the association between intestinal metaplasia (IM) and gastric cancer risk, while the results were inconsistent. We aimed to assess the risk of gastric cancer among patients with IM. Systematic literature searches were conducted in PubMed, Embase and Cochrane databases. Baseline characteristics and outcomes from the included studies were extracted independently by two investigators. Either a fixed-effects or a random-effects model was used to composite the pooled OR for gastric cancer risk. Finally, a total of 21 studies, which comprised 402,636 participants and 4,535 gastric cancer patients, were finally included in the current meta-analysis. Compared with those participants without IM, IM patients were at a higher risk of gastric cancer (pooled OR = 3.58, 95% CI 2.71-4.73). We observed that incomplete IM (pooled OR = 9.48, 95% CI 4.33-20.78) but not complete IM (pooled OR = 1.55, 95% CI 0.91-2.65) was significantly associated with a higher gastric cancer risk. Besides, it appeared that gastric cancer risk was higher among patients with IM in the corpus (pooled OR = 7.39, 95% CI 4.94-11.06) than those with IM in the antrum only (pooled OR = 4.06, 95% CI 2.79-5.91). And the pooled ORs for gastric noncardia cancer and gastric cardia cancer were 4.98 (95% CI 3.12-7.95) and 1.93 (95% CI 1.15-3.24), respectively. In conclusion, patients with IM were at a higher risk of gastric cancer, especially for incomplete IM and IM in the corpus. The current evidence supports the use of IM subtypes in the surveillance of gastric cancer.