Search

Thyroid eye disease (TED) remains challenging for clinicians to evaluate and manage. Novel therapies have recently emerged, and their specific roles are still being determined. Most patients with TED develop eye manifestations while being treated for hyperthyroidism and under the care of endocrinologists. Endocrinologists, therefore, have a key role in diagnosis, initial management, and selection of patients who require referral to specialist care. Given that the need for guidance to endocrinologists charged with meeting the needs of patients with TED transcends national borders, and to maximize an international exchange of knowledge and practices, the American Thyroid Association and European Thyroid Association joined forces to produce this consensus statement.

Thyroid eye disease (TED) remains challenging for clinicians to evaluate and manage. Novel therapies have recently emerged, and their specific roles are still being determined. Most patients with TED develop eye manifestations while being treated for hyperthyroidism and under the care of endocrinologists. Endocrinologists, therefore, have a key role in diagnosis, initial management, and selection of patients who require referral to specialist care. Given that the need for guidance to endocrinologists charged with meeting the needs of patients with TED transcends national borders, and to maximize an international exchange of knowledge and practices, the American Thyroid Association and European Thyroid Association joined forces to produce this Consensus Statement.

Pulmonary arterial hypertension (PAH) is a fatal disease characterized by increased mean pulmonary arterial pressure. Elevated plasma and lung concentrations of oxidized lipids, including 15-hydroxyeicosatetraenoic acid (15-HETE), have been demonstrated in patients with PAH and animal models. We previously demonstrated that feeding mice with 15-HETE is sufficient to induce pulmonary hypertension, but the mechanisms remain unknown. RNA sequencing data from the mouse lungs on 15-HETE diet revealed significant activation of pathways involved in both antigen processing and presentation and T cell-mediated cytotoxicity. Analysis of human microarray from patients with PAH also identified activation of identical pathways compared with controls. We show that in both 15-HETE-fed mice and patients with PAH, expression of the immunoproteasome subunit 5 is significantly increased, which was concomitant with an increase in the number of CD8/CD69 (cluster of differentiation 8 / cluster of differentiation 69) double-positive cells, as well as pulmonary arterial endothelial cell apoptosis in mice. Human pulmonary arterial endothelial cells cultured with 15-HETE were more prone to apoptosis when exposed to CD8 cells. Cultured intestinal epithelial cells secreted more oxidized lipids in response to 15-HETE, which is consistent with accumulation of circulating oxidized lipids in 15-HETE-fed mice. Administration of an apoA-I (apolipoprotein A-I) mimetic peptide, Tg6F (transgenic 6F), which is known to prevent accumulation of circulating oxidized lipids, not only inhibited pulmonary arterial endothelial cell apoptosis but also prevented and rescued 15-HETE-induced pulmonary hypertension in mice. In conclusion, our results suggest that (1) 15-HETE diet induces pulmonary hypertension by a mechanism that involves oxidized lipid-mediated T cell-dependent pulmonary arterial endothelial cell apoptosis and (2) Tg6F administration may be a novel therapy for treating PAH.

We present an integrated approach for structure and parameter estimation inlatent tree graphical models. Our overall approach follows a"divide-and-conquer" strategy that learns models over small groups of variablesand iteratively merges onto a global solution. The structure learning involvescombinatorial operations such as minimum spanning tree construction and localrecursive grouping; the parameter learning is based on the method of momentsand on tensor decompositions. Our method is guaranteed to correctly recover theunknown tree structure and the model parameters with low sample complexity forthe class of linear multivariate latent tree models which includes discrete andGaussian distributions, and Gaussian mixtures. Our bulk asynchronous parallelalgorithm is implemented in parallel and the parallel computation complexityincreases only logarithmically with the number of variables and linearly withdimensionality of each variable.

We present an integrated approach for structure and parameter estimation inlatent tree graphical models. Our overall approach follows a"divide-and-conquer" strategy that learns models over small groups of variablesand iteratively merges onto a global solution. The structure learning involvescombinatorial operations such as minimum spanning tree construction and localrecursive grouping; the parameter learning is based on the method of momentsand on tensor decompositions. Our method is guaranteed to correctly recover theunknown tree structure and the model parameters with low sample complexity forthe class of linear multivariate latent tree models which includes discrete andGaussian distributions, and Gaussian mixtures. Our bulk asynchronous parallelalgorithm is implemented in parallel and the parallel computation complexityincreases only logarithmically with the number of variables and linearly withdimensionality of each variable.