Deep Learning Using Havrda-Charvat Entropy for Classification of Pulmonary Optical Endomicroscopy

1) ObjectivePulmonary optical endomicroscopy (POE) is an imaging technology in real time. It allows to examine pulmonary alveoli at a microscopic level. Acquired in clinical settings, a POE image sequence can have as much as 25% of the sequence being uninformative frames (i.e. pure-noise and motion artifacts). For future data analysis, these uninformative frames must be first removed from the sequence. Therefore, the objective of our work is to develop an automatic detection method of uninformative images in endomicroscopy images.2) Material and methodsWe propose to take the detection problem as a classification one. Considering advantages of deep learning methods, a classifier based on CNN (Convolutional Neural Network) is designed with a new loss function based on Havrda-Charvat entropy which is a parametrical generalization of the Shannon entropy. We propose to use this formula to get a better hold on all sorts of data since it provides a model more stable than the Shannon entropy.3) ResultsOur method is tested on one POE dataset including 3895 distinct images and is showing better results than using Shannon entropy and behaves better with regard to the problem of overfitting. We obtain 70% of accuracy with Shannon entropy versus 77 to 79% with Havrda-Charvat.4) ConclusionWe can conclude that Havrda-Charvat entropy is better suited for restricted and or noisy datasets due to its generalized nature. It is also more suitable for classification in endomicroscopy datasets.     

胞外酸性与肿瘤的浸润转移

组织缺氧是实体瘤的一个主要特征,它引起肿瘤细胞胞外酸性环境的形成.肿瘤细胞通过质子感知的G蛋白偶联受体（G protein-coupled receptors,GPCRs）或质子感知的离子通道感知其胞外的酸性环境,并激活多条细胞内信号通路,影响细胞功能. 肿瘤最致命的方面在于其转移能力,肿瘤转移程度与肿瘤细胞迁移能力呈正相关. 因此,对胞外酸性与肿瘤细胞迁移扩散之间关系的深入研究将有助于发现更多新的抗肿瘤转移药物.本文就肿瘤酸性微环境的形成、肿瘤细胞的质子感知制、胞外酸性环境对肿瘤浸润转移的影响及如何将肿瘤pH调节应用于癌症治疗等方面的内容予以综述.     

MicroRNA-409-3p inhibits migration and invasion of bladder cancer cells via targeting c-Met

There is increasing evidence suggesting that dysregulation of certain microRNAs (miRNAs) may contribute to tumor progression and metastasis. Previous studies have shown that miR-409-3p is dysregulated in some malignancies, but its role in bladder cancer is still unknown. Here, we find that miR-409-3p is down-regulated in human bladder cancer tissues and cell lines. Enforced expression of miR-409-3p in bladder cancer cells significantly reduced their migration and invasion without affecting cell viability. Bioinformatics analysis identified the pro-metastatic gene c-Met as a potential miR-409-3p target. Further studies indicated that miR-409-3p suppressed the expression of c-Met by binding to its 3′-untranslated region. Silencing of c-Met by small interfering RNAs phenocopied the effects of miR-409-3p overexpression, whereas restoration of c-Met in bladder cancer cells bladder cancer cells overexpressing miR-409-3p, partially reversed the suppressive effects of miR-409-3p. We further showed that MMP2 and MMP9 may be downstream effector proteins of miR-409-3p. These findings indicate that miR-409-3p could be a potential tumor suppressor in bladder cancer.     

Proangiogenic TIE2+/CD31+ macrophages are the predominant population of tumor-associated macrophages infiltrating metastatic lymph nodes

Tumor-associated macrophages (TAMs) accumulate in various cancers and promote tumor angiogenesis and metastasis, and thus may be ideal targets for the clinical diagnosis of tumor metastasis with high specificity. However, there are few specific markers to distinguish between TAMs and normal or inflammatory macrophages. Here, we show that TAMs localize in green fluorescent protein-labeled tumors of metastatic lymph nodes (MLNs) from B16F1 melanoma cells but not in necrotic tumor regions, suggesting that TAMs may promote the growth of tumor cells and the progression of tumor metastasis. Furthermore, we isolated pure populations of TAMs from MLNs and characterized their gene expression signatures compared to peritoneal macrophages (PMs), and found that TAMs significantly overexpress immunosuppressive cytokines such as IL-4, IL-10, and TGF-β as well as proangiogenic factors such as VEGF, TIE2, and CD31. Notably, immunological analysis revealed that TIE2⁺/CD31⁺ macrophages constitute the predominant population of TAMs that infiltrate MLNs, distinct from tissue or inflammatory macrophages. Importantly, these TIE2⁺/CD31⁺ macrophages also heavily infiltrated MLNs from human breast cancer biopsies but not reactive hyperplastic LNs. Thus, TIE2⁺/CD31⁺ macrophages may be a unique histopathological biomarker for detecting metastasis in clinical diagnosis, and a novel and promising target for TAM-specific cancer therapy.     

Circulating tumor cells: detection,molecular profiling and future prospects

Disseminated malignancy is responsible for the vast majority of cancer-related deaths. During this process, circulating tumor cells (CTC) are generated, spread from the primary tumor, colonize distant organs and lead to overt metastatic disease. CTC are essential for establishing metastasis; however, they are not sufficient as this process is highly inefficient and most will fail to grow in target sites. Several CTC die during migration while others remain dormant for several years and very few grow into macrometastases. CTC have been well documented in the bloodstream of cancer patients; however, the clinical relevance of this detection is still the subject of controversies and their biology is poorly understood. Indeed, available markers fail to distinguish between subgroups of CTC, and several current methods lack sensitivity, specificity or reproducibility in CTC characterization and detection. The advent of more precise technologies is renewing the interest in CTC biology. We will review herein recent findings on CTC biology, on the role of host–tumor interactions in CTC shedding and implantation, available methods of CTC detection and future perspectives for the molecular characterization of the CTC subset(s) responsible for the development of metastasis. Ultimately, understanding CTC biology and host–tumor ‘complementarities’ will help define metastasis-related biomarkers providing formidable and tailored novel therapeutic targets.     

Case of lung carcinoma revealed by vulvar metastasis associated with systemic scleroderma and literature review

Metastatic carcinoma to the vulva is rare, where the incidence is believed to be between 5% and 8%.However, malignant tumors have been described in 3–11% of systemic scleroderma (SSc) cases.We report the case of one patient, a 66-year-old postmenopausal woman, whose medical history was marked with rheumatic vascular disease (systemic scleroderma) since 1993 without muscular, renal, cardiac lesions or HTA (arterial hypertension) and without tobacco history.The woman presented with a new vulvar mass of the right labia in December 2011 that had progressively enlarged in size.CT scan of the abdominopelvic region demonstrated a lobular mass of the right labia with central necrosis, 7 cm on the wide axis, and the rectum and the vaginal wall were normal. No inguinal or iliac lymphadenopathy was noted.An outpatient excisional biopsy revealed a poorly differentiated malignant tumor suggestive of carcinoma.IHC: CK7+/CK20−, estrogen receptors−, AE 1 AE 3+, vimentine+, S100−, Desmina−, CD34−, KI 67: 20%.The thoracic scan revealed a large mass of 4 cm × 3 cm in the right lung base with right paratracheal lymphadenopathy 3 cm × 2 cm.A bronchoscopy revealed discrete stenosis of the mediastinal portion of the right bronchial tree.The bronchial biopsy also revealed poorly differentiated lung carcinoma, non-small cell, which was identical with the vulvar tumor.

Conclusion

The presence of the single lung lesion with only one lymphadenopathy paratracheal with pathological and immunohistochemical (IHC) profile similar to the vulvar lesion, and a particular IHC profile with CK7+ and CK20− was detected – that is more specific to the primitive pulmonary cancer, and the presence of only one sarcoma marker vementine+, desmine and actine−. Also the presence of KI 67: 20%, predicted the proliferative and great metastatic power of the lung tumor was observed.Additionally, lung cancer was the most frequent type and may develop in scleroderma as reported in most studies.This allows to conclude for primitive lung carcinoma revealed with vulvar metastasis after elimination of the possibility of vulvar sarcoma.The patient was treated by chemotherapy (Taxol/Platin) with partial response from the lung after 3 cycles and palliative radiotherapy in the vulva with a good response.This case described primary lung carcinoma associated with scleroderma, revealed by a vulvar metastasis, which may be related to the aggressiveness of lung cancer when the lung fibrosis follow-up is not performed well to detect early the development of lung tumors in the patient with systemic scleroderma.     

Cell Budding from Normal Appearing Epithelia: A Predictor of Colorectal Cancer Metastasis?

Background: Colorectal carcinogenesis is believed to be a multi-stage process that originates with a localized adenoma, which linearly progresses to an intra-mucosal carcinoma, to an invasive lesion, and finally to metastatic cancer. This progression model is supported by tissue culture and animal model studies, but it is difficult to reconcile with several well-established observations, principally among these are that up to 25% of early stage (Stage I/II), node-negative colorectal cancer (CRC) develop distant metastasis, and that circulating CRC cells are undetectable in peripheral blood samples of up to 50% of patients with confirmed metastasis, but more than 30% of patients with no detectable metastasis exhibit such cells. The mechanism responsible for this diverse behavior is unknown, and there are no effective means to identify patients with pending, or who are at high risk for, developing metastatic CRC.Novel findings: Our previous studies of human breast and prostate cancer have shown that cancer invasion arises from the convergence of a tissue injury, the innate immune response to that injury, and the presence of tumor stem cells within tumor capsules at the site of the injury. Focal degeneration of a capsule due to age or disease attracts lymphocyte infiltration that degrades the degenerating capsules resulting in the formation of a focal disruption in the capsule, which selectively favors proliferating or “budding” of the underlying tumor stem cells. Our recent studies suggest that lymphocyte infiltration also triggers metastasis by disrupting the intercellular junctions and surface adhesion molecules within the proliferating cell buds causing their dissociation. Then, lymphocytes and tumor cells are conjoined through membrane fusion to form tumor-lymphocyte chimeras (TLCs) that allows the tumor stem cell to avail itself of the lymphocyte''s natural ability to migrate and breach cell barriers in order to intravasate and to travel to distant organs. Our most recent studies of human CRC have detected nearly identical focal capsule disruptions, lymphocyte infiltration, budding cells, and the formation of TLCs. Our studies have further shown that age- and type-matched node-positive and -negative CRC have a significantly different morphological and immunohistochemical profile and that the majority of lymphatic ducts with disseminated cells are located within the mucosa adjacent to morphologically normal appearing epithelial structures that express a stem cell-related marker.New hypothesis: Based on these findings and the growth patterns of budding cells revealed by double immunohistochemistry, we further hypothesize that metastatic spread is an early event of carcinogenesis and that budding cells overlying focal capsule disruptions represent invasion- and metastasis-initiating cells that follow one of four pathways to progress: (1) to undergo extensive in situ proliferation leading to the formation of tumor nests that subsequently invade the submucosa, (2) to migrate with associated lymphocytes functioning as “seeds” to grow in new sites, (3) to migrate and intravasate into pre-existing vascular structures by forming TLCs, or (4) to intravasate into vascular structures that are generated by the budding cells themselves. We also propose that only node-positive cases harbor stem cells with the potential for multi-lineage differentiation and unique surface markers that permit intravasation.