Dual-feature Fusion Attention Network for Small Object Segmentation.

Accurate segmentation of medical images is an important step during radiotherapy planning and clinical diagnosis. However, manually marking organ or lesion boundaries is tedious, time-consuming, and prone to error due to subjective variability of radiologist. Automatic segmentation remains a challenging task owing to the variation (in shape and size) across subjects. Moreover, existing convolutional neural networks based methods perform poorly in small medical objects segmentation due to class imbalance and boundary ambiguity. In this paper, we propose a dual feature fusion attention network (DFF-Net) to improve the segmentation accuracy of small objects. It mainly includes two core modules: the dual-branch feature fusion module (DFFM) and the reverse attention context module (RACM). We first extract multi-resolution features by multi-scale feature extractor, then construct DFFM to aggregate the global and local contextual information to achieve information complementarity among features, which provides sufficient guidance for accurate small objects segmentation. Moreover, to alleviate the degradation of segmentation accuracy caused by blurred medical image boundaries, we propose RACM to enhance the edge texture of features. Experimental results on datasets NPC, ACDC, and Polyp demonstrate that our proposed method has fewer parameters, faster inference, and lower model complexity, and achieves better accuracy than more state-of-the-art methods.

Interleukin-21 enhances rituximab activity in a cynomolgus monkey model of B cell depletion and in mouse B cell lymphoma models.

Rituximab, a monoclonal antibody targeting CD20 on B cells, is currently used to treat many subtypes of B cell lymphomas. However, treatment is not curative and response rates are variable. Recombinant interleukin-21 (rIL-21) is a cytokine that enhances immune effector function and affects both primary and transformed B cell differentiation. We hypothesized that the combination of rIL-21 plus rituximab would be a more efficacious treatment for B cell malignancies than rituximab alone. We cultured human and cynomolgus monkey NK cells with rIL-21 and found that their activity was increased and proteins associated with antibody dependent cytotoxicity were up-regulated. Studies in cynomolgus monkeys modeled the effects of rIL-21 on rituximab activity against CD20 B cells. In these studies, rIL-21 activated innate immune effectors, increased ADCC and mobilized B cells into peripheral blood. When rIL-21 was combined with rituximab, deeper and more durable B cell depletion was observed. In another series of experiments, IL-21 was shown to have direct antiproliferative activity against a subset of human lymphoma cell lines, and combination of murine IL-21 with rituximab yielded significant survival benefits over either agent alone in xenogeneic mouse tumor models of disseminated lymphoma. Therefore, our results do suggest that the therapeutic efficacy of rituximab may be improved when used in combination with rIL-21.

A glycoprotein hormone expressed in corticotrophs exhibits unique binding properties on thyroid-stimulating hormone receptor.

Corticotroph-derived glycoprotein hormone (CGH), also referred to as thyrostimulin, is a noncovalent heterodimer of glycoprotein hormone alpha 2 (GPHA2) and glycoprotein hormone beta 5 (GPHB5). Here, we demonstrate that both subunits of CGH are expressed in the corticotroph cells of the human anterior pituitary, as well as in skin, retina, and testis. CGH activates the TSH receptor (TSHR); (125)I-CGH binding to cells expressing TSHR is saturable, specific, and of high affinity. In competition studies, unlabeled CGH is a potent competitor for (125)I-TSH binding, whereas unlabeled TSH does not compete for (125)I-CGH binding. Binding and competition analyses are consistent with the presence of two binding sites on the TSHR transfected baby hamster kidney cells, one that can interact with either TSH or CGH, and another that binds CGH alone. Transgenic overexpression of GPHB5 in mice produces elevations in serum T(4) levels, reductions in body weight, and proptosis. However, neither transgenic overexpression of GPHA2 nor deletion of GPHB5 produces an overt phenotype in mice. In vivo administration of CGH to mice produces a dose-dependent hyperthyroid phenotype including elevation of T(4) and hypertrophy of cells within the inner adrenal cortex. However, the distinctive expression patterns and binding characteristics of CGH suggest that it has endogenous biological roles that are discrete from those of TSH.

Platelet-derived growth factor C induces liver fibrosis, steatosis, and hepatocellular carcinoma.

Members of the platelet-derived growth factor (PDGF) ligand family are known to play important roles in wound healing and fibrotic disease. We show that both transient and stable expression of PDGF-C results in the development of liver fibrosis consisting of the deposition of collagen in a pericellular and perivenular pattern that resembles human alcoholic and nonalcoholic fatty liver disease. Fibrosis in PDGF-C transgenic mice, as demonstrated by staining and hydroxyproline content, is preceded by activation and proliferation of hepatic stellate cells, as shown by collagen, alpha-smooth muscle actin and glial fibrillary acidic protein staining and between 8 and 12 months of age is followed by the development of liver adenomas and hepatocellular carcinomas. The hepatic expression of a number of known profibrotic genes, including type beta1 TGF, PDGF receptors alpha and beta, and tissue inhibitors of matrix metalloproteinases-1 and -2, increased by 4 weeks of age. Increased PDGF receptor alpha and beta protein levels were associated with activation of extracellular regulated kinase-1 and -2 and protein kinase B. At 9 months of age, PDGF-C transgenic mice had enlarged livers associated with increased fibrosis, steatosis, cell dysplasia, and hepatocellular carcinomas. These studies indicate that hepatic expression of PDGF-C induces a number of profibrotic pathways, suggesting that this growth factor may act as an initiator of fibrosis. Moreover, PDGF-C transgenic mice represent a unique model for the study of hepatic fibrosis progressing to tumorigenesis.

Fibroblast growth factor-18 is a trophic factor for mature chondrocytes and their progenitors.

OBJECTIVE: The aim of this study was to examine the effects of recombinant human Fgf18 on chondrocyte proliferation and matrix production in vivo and in vitro. In addition, the expressions of Fgf18 and Fgf receptors (Fgfr) in adult human articular cartilage were examined. METHODS: Adenovirus-mediated transfer of Fgf18 into murine pinnae and addition of FGF18 to primary cultures of adult articular chondrocytes were used to assess the effects of FGF18 on chondrocytes. In situ hybridization was used to examine the expression of Fgf18 and Fgfr s in adult human articular cartilage. RESULTS: Expression of Fgf18 by adenovirus-mediated gene transfer in murine pinnae resulted in a significant increase in chondrocyte number. Chondrocytes were identified by staining with toluidine blue and a monoclonal antibody directed against type II collagen. Fgf18, Fgfr 2-(IIIc), Fgfr 3-(IIIc), and Fgfr 4 mRNAs were detected within these cells by in situ hybridization. The nuclei of the chondrocytes stained with antibodies to PCNA and FGF receptor (FGFR) 2. Addition of FGF18 to the culture media of primary articular chondrocytes increased the proliferation of these cells and increased their production of extracellular matrix. To assess the receptor selectivity of FGF18, BaF3 cells stably expressing the genes for the major splice variants of Fgfr1-3 were used. Proliferation of cells expressing Fgfr 3-(IIIc) or Fgfr 2-(IIIc) was increased by incubation with FGF18. Using FGFR-Fc fusion proteins and BaF3 cells expressing Fgfr 3-(IIIc), only FGFR 3-(IIIc)-Fc, FGFR 2-(IIIc)-Fc or FGFR 4-Fc reduced FGF18-mediated cell proliferation. Expression of Fgf18, Fgfr 3-(IIIc) and Fgfr 2-(IIIc) mRNAs was localized to chondrocytes of human articular cartilage by in situ hybridization. CONCLUSION: These data demonstrate that Fgf18 can act as a trophic factor for elastic chondrocytes and their progenitors in vivo and articular chondrocytes cultured in vitro. Expression of Fgf18 and the genes for two of its receptors in chondrocytes suggests that Fgf18 may play an autocrine role in the biology of normal articular cartilage.