SUMMARY Beyond sample curation and basic pathologic characterization, the digitized H&E-stained images of TCGA samples remain underutilized. To highlight this resource, we present mappings of tumor-infiltrating lymphocytes (TILs) based on H&E images from 13 TCGA tumor types. These TIL maps are derived through computational staining using a convolutional neural network trained to classify patches of images. Affinity propagation revealed local spatial structure in TIL patterns and correlation with overall survival. TIL map structural patterns were grouped using standard histopathological parameters. These patterns are enriched in particular T cell subpopulations derived from molecular measures. TIL densities and spatial structure were differentially enriched among tumor types, immune subtypes, and tumor molecular subtypes, implying that spatial infiltrate state could reflect particular tumor cell aberration states. Obtaining spatial lymphocytic patterns linked to the rich genomic characterization of TCGA samples demonstrates one use for the TCGA image archives with insights into the tumor-immune microenvironment.

Convolutional Neural Networks (CNN) are state-of-the-art models for many image classification tasks. However, to recognize cancer subtypes automatically, training a CNN on gigapixel resolution Whole Slide Tissue Images (WSI) is currently computationally impossible. The differentiation of cancer subtypes is based on cellular-level visual features observed on image patch scale. Therefore, we argue that in this situation, training a patch-level classifier on image patches will perform better than or similar to an image-level classifier. The challenge becomes how to intelligently combine patch-level classification results and model the fact that not all patches will be discriminative. We propose to train a decision fusion model to aggregate patch-level predictions given by patch-level CNNs, which to the best of our knowledge has not been shown before. Furthermore, we formulate a novel Expectation-Maximization (EM) based method that automatically locates discriminative patches robustly by utilizing the spatial relationships of patches. We apply our method to the classification of glioma and non-small-cell lung carcinoma cases into subtypes. The classification accuracy of our method is similar to the inter-observer agreement between pathologists. Although it is impossible to train CNNs on WSIs, we experimentally demonstrate using a comparable non-cancer dataset of smaller images that a patch-based CNN can outperform an image-based CNN.

Despite recent advances in antibiotic therapy and intensive care, sepsis is still considered to be the most common cause of death in intensive care units. Excessive production of reactive oxygen species plays an important role in the pathogenesis of sepsis. Recently, it has been suggested that molecular hydrogen (H2) exerts a therapeutic antioxidant activity by selectively reducing hydroxyl radicals (*OH, the most cytotoxic reactive oxygen species) and effectively protects against organ damage induced by I/R. Therefore, we hypothesized that H2 treatment had a beneficial effect on sepsis. In the present study, we found that H2 inhalation starting at 1 and 6 h after cecal ligation and puncture (CLP) or sham operation significantly improved the survival rate of septic mice with moderate or severe CLP in a concentration- and time-dependent manner. Furthermore, moderate or severe CLP mice showed significant multiple organ damage characterized by the increases of lung myeloperoxidase activity, wet-to-dry weight ratio, protein concentration in bronchoalveolar lavage, serum biochemical parameters, and organ histopathologic scores at 24 h after CLP operation, which was significantly attenuated by 2% H2 treatment. In addition, we found that the beneficial effects of H2 treatment on sepsis and sepsis-associated organ damage were associated with the decreased levels of oxidative product, increased activities of antioxidant enzymes, and reduced levels of high-mobility group box 1 in serum and tissue. Thus, H2 inhalation may be an effective therapeutic strategy for patients with sepsis.

Acute lung injury (ALI) is still a leading cause of morbidity and mortality in critically ill patients. Recently, our and other studies have found that hydrogen gas (H₂) treatment can ameliorate the lung injury induced by sepsis, ventilator, hyperoxia, and ischemia-reperfusion. However, the molecular mechanisms by which H₂ ameliorates lung injury remain unclear. In the current study, we investigated whether H₂ or hydrogen-rich saline (HS) could exert protective effects in a mouse model of ALI induced by intratracheal administration of lipopolysaccharide (LPS) via inhibiting the nuclear factor κB (NF-κB) signaling pathway-mediated inflammation and apoptosis. Two percent of H₂ was inhaled for 1 h beginning at 1 and 6 h after LPS administration, respectively. We found that LPS-challenged mice exhibited significant lung injury characterized by the deterioration of histopathology and histologic scores, wet-to-dry weight ratio, and oxygenation index (PaO₂/FIO₂), as well as total protein in the bronchoalveolar lavage fluid (BALF), which was attenuated by H₂ treatment. Hydrogen gas treatment inhibited LPS-induced pulmonary early and late NF-κB activation. Moreover, H₂ treatment dramatically prevented the LPS-induced pulmonary cell apoptosis in LPS-challenged mice, as reflected by the decrease in TUNEL (deoxynucleotidyl transferase dUTP nick end labeling) staining-positive cells and caspase 3 activity. Furthermore, H₂ treatment markedly attenuated LPS-induced lung neutrophil recruitment and inflammation, as evidenced by downregulation of lung myeloperoxidase activity, total cells, and polymorphonuclear neutrophils in BALF, as well as proinflammatory cytokines (tumor necrosis factor α, interleukin 1β, interleukin 6, and high-mobility group box 1) and chemokines (keratinocyte-derived chemokine, macrophage inflammatory protein [MIP] 1α, MIP-2, and monocyte chemoattractant protein 1) in BALF. In addition, i.p. injection of 10 mL/kg hydrogen-rich saline also significantly attenuated the LPS-induced ALI. Collectively, these results demonstrate that molecular hydrogen treatment ameliorates LPS-induced ALI through reducing lung inflammation and apoptosis, which may be associated with the decreased NF-κB activity. Hydrogen gas may be useful as a novel therapy to treat ALI. munosorbent assay; H₂-hydrogen gas; HMGB1-high-mobility group box 1; HS-hydrogen-rich saline; i.t.-intratracheal; KC-keratinocyte-derived chemokine; LPS-lipopolysaccharide; MCP-1-monocyte chemoattractant protein 1; MIP-1α-macrophage inflammatory protein 1α; MIP-2-macrophage inflammatory protein 2; MPO-myeloperoxidase; PBS-phosphate-buffered saline; PMNs-polymorphonuclear neutrophils; TUNEL-deoxynucleotidyl transferase dUTP nick end labeling; W/D-wet-to-dry.

No abstract

Sepsis/multiple organ dysfunction syndrome is the leading cause of death in critically ill patients. Recently, it has been suggested that hydrogen gas (H2) exerts a therapeutic antioxidant activity by selectively reducing hydroxyl radical (•OH, the most cytotoxic reactive oxygen species). We have found that H2 inhalation significantly improved the survival rate and organ damage of septic mice with moderate or severe cecal ligation and puncture. In the present study, we investigated the effects of 2% H2 treatment on survival rate and organ damage in zymosan (ZY)-induced generalized inflammation model. Here, we found that 2% H2 inhalation for 60 min starting at 1 and 6 h after ZY injection, respectively, significantly improved the 14-day survival rate of ZY-challenged mice from 10% to 70%. Furthermore, ZY-challenged mice showed significant multiple organ damage characterized by the increase in serum biochemical parameters (aspartate aminotransferase, alanine aminotransferase, blood urea nitrogen, and creatinine), as well as lung, liver, and kidney histopathological scores at 24 h after ZY injection, which was significantly attenuated by 2% H2 treatment. In addition, we found that the beneficial effects of H2 treatment on ZY-induced organ damage were associated with the decreased levels of oxidative product, increased activities of antioxidant enzyme, and reduced levels of early and late proinflammatory cytokines in serum and tissues. In conclusion, this study provides evidence that H2 treatment protects against multiple organ damages in ZY-induced generalized inflammation model, suggesting the potential use of H2 as a therapeutic agent in the therapy of conditions associated with inflammation-related multiple organ dysfunction syndrome.

No abstract

Histopathology images are crucial to the study of complex diseases such as cancer. The histologic characteristics of nuclei play a key role in disease diagnosis, prognosis and analysis. In this work, we propose a sparse Convolutional Autoencoder (CAE) for fully unsupervised, simultaneous nucleus detection and feature extraction in histopathology tissue images. Our CAE detects and encodes nuclei in image patches in tissue images into sparse feature maps that encode both the location and appearance of nuclei. Our CAE is the first unsupervised detection network for computer vision applications. The pretrained nucleus detection and feature extraction modules in our CAE can be finetuned for supervised learning in an end-to-end fashion. We evaluate our method on four datasets and reduce the errors of state-of-the-art methods up to 42%. We are able to achieve comparable performance with only 5% of the fullysupervised annotation cost.