Search

Activation of the kynurenine pathway (KP) has been shown to predict outcome in cardiac arrest (CA) patients. We validated these findings in a Swiss cohort.We measured admission tryptophan and kynurenine levels in 270 consecutive CA patients (38 in-hospital CA) and investigated associations with in-hospital mortality and neurological outcome at hospital discharge.120 of 270 (44%) patients died in the hospital. Compared to survivors, non-survivors showed higher median initial kynurenine levels (5.28 μmol/l [IQR 2.91 to 7.40] vs 3.58 μmol/l [IQR 2.47 to 5.46]; p 0.001) and a higher median kynurenine/tryptophan ratio (0.10 μmol/l [IQR 0.07 to 0.17] vs 0.07 μmol/l [IQR 0.05 to 0.1]; p 0.001). In a model adjusted for age, gender and comorbidities, kynurenine (OR 1.16, 95% CI 1.05 to 1.27; p = 0.001) and kynurenine/tryptophan ratio (OR 1.19, 95% CI 1.08 to 1.31; p = 0.003) were significantly associated with mortality. Results were similar for neurological outcome.Our findings validate a previous study and show associations of the activation of the KP with unfavorable outcomes after CA. Future studies should evaluate whether therapeutic modulation of the KP may impact clinical outcomes after CA.

AIMS OF THE STUDY: There is increasing interest in better understanding of long COVID, a condition characterised by long-term sequelae — appearing or persisting after the typical convalescence period — of coronavirus disease 2019 (COVID-19). Herein, we describe long-term outcomes regarding residual symptoms and psychological distress in hospitalised patients 1 year after COVID-19. METHODS: This prospective cohort study included consecutive adult patients hospitalised for confirmed COVID-19 in two Swiss tertiary-care hospitals between March and June 2020. The primary endpoint was evidence of long COVID 1 year after discharge, defined as ≥1 persisting or new symptom related to COVID-19, from a predefined list of symptoms. Secondary endpoints included psychological distress and symptoms of post-traumatic stress disorder (PTSD). RESULTS: Among 90 patients included in the study, 63 (70%) had symptoms of long COVID 1 year after hospitalisation, particularly fatigue (46%), concentration difficulties (31%), shortness of breath (21%) and post-exertion malaise (20%). Three predictors, namely duration of hospitalisation (odds ratio [OR] 1.11, 95% confidence interval [CI] 1.00–1.22; p = 0.041), severity of illness (OR 1.19, 95% CI 1.04–1.37; p = 0.013), and self-perceived overall health status 30 days after hospitalisation (OR 0.97, 95% CI 0.94–1.00; p = 0.027) were associated with long COVID. Regarding secondary endpoints, 16 (18%) experienced psychological distress and 3 (3.3%) patients had symptoms of PTSD. CONCLUSION: A high proportion of COVID-19 patients report symptoms of long COVID 1 year after hospitalisation, which negatively affects their quality of life. The most important risk factors were severe initial presentation of COVID-19 with long hospital stays.

Since wine is not continuously stirred during or after fermentation, it is common to use movable mixers on a variety of tank sizes or to install small, fast-rotating propeller mixers in large tanks. Many practically relevant optimization problems relate to processes occurring directly after the start of the propeller (flow field initialization period), which require 3D instationary models if CFD simulations are used. Based on traditional instationary models, these simulations are very time-consuming even on fast computers and limit the part of the flow field initialization period that can be computed. In this work, a new method was developed to speed up the transient simulation of mechanical mixing in large wine tanks. The method uses the traditional CFD techniques of moving reference frame (MRF) and several ensuing propeller rotations using sliding mesh (SM), to calculate a steady-state flow field and to compute the flow behavior in the propeller region. Subsequently, the flow’s key figures are extracted for the propeller region at multiple time-points during one rotation and continuously mapped to a new simulation starting from a flow field with zero velocity. This new approach using a transient boundary condition (BC), was validated with tank mixing experiments and literature data. In contrast to the MRF method, it allows e.g. for an estimation of minimum mixing times similar to the SM method, while computation times are reduced by a factor of up to 20 for tank-mixer scenarios evaluated in this work. For a mixing simulation in a wine tank, the new method takes only 3 days compared to more than 60 days using the traditional SM approach.

A key obstacle in automated analytics and meta-learning is failing to recognize when different datasets contain measurements of the same variable. Because provided attribute labels are often uninformative in practice, this task may be more robustly addressed by leveraging the data values themselves, rather than relying on their arbitrarily selected variable names. Here, we present a computationally efficient method to identify high-confidence variable matches between a given set of data values and a large repository of previously encountered datasets. Our approach enjoys numerous advantages over distributional similarity based techniques because we leverage learned vector embeddings of datasets which adaptively account for natural forms of data variation encountered in practice. Based on the neural architecture of deep sets, our embeddings can be computed for both numeric and string data, and empirically outperform standard statistical techniques for dataset search.

Motivation The precise targeting of antibodies and other protein therapeutics is required for their proper function and the elimination of deleterious off-target effects. Often the molecular structure of a therapeutic target is unknown and randomized methods are used to design antibodies without a model that relates antibody sequence to desired properties. Results Here, we present Ens-Grad, a machine learning method that can design complementarity determining regions of human Immunoglobulin G antibodies with target affinities that are superior to candidates derived from phage display panning experiments. We also demonstrate that machine learning can improve target specificity by the modular composition of models from different experimental campaigns, enabling a new integrative approach to improving target specificity. Our results suggest a new path for the discovery of therapeutic molecules by demonstrating that predictive and differentiable models of antibody binding can be learned from high-throughput experimental data without the need for target structural data. Availability and implementation Sequencing data of the phage panning experiment are deposited at NIH’s Sequence Read Archive (SRA) under the accession number SRP158510. We make our code available at https://github.com/gifford-lab/antibody-2019. Supplementary information Supplementary data are available at Bioinformatics online.

The inaccuracy of neural network models on inputs that do not stem from the training data distribution is both problematic and at times unrecognized. Model uncertainty estimation can address this issue, where uncertainty estimates are often based on the variation in predictions produced by a diverse ensemble of models applied to the same input. Here we describe Maximize Overall Diversity (MOD), a straightforward approach to improve ensemble-based uncertainty estimates by encouraging larger overall diversity in ensemble predictions across all possible inputs that might be encountered in the future. When applied to various neural network ensembles, MOD significantly improves predictive performance for out-of-distribution test examples without sacrificing in-distribution performance on 38 Protein-DNA binding regression datasets, 9 UCI datasets, and the IMDB-Wiki image dataset. Across many Bayesian optimization tasks, the performance of UCB acquisition is also greatly improved by leveraging MOD uncertainty estimates., 10 pages, 3 figures

Aim Neuron-specific enolase (NSE) increases in response to brain injury and is recommended for outcome prediction in cardiac arrest patients. Our aim was to investigate whether NSE measured at different days after a cardiac arrest and its kinetics would improve the prognostic ability of two cardiac arrest specific risk scores. Methods Within this prospective observational study, we included consecutive adult patients after cardiac arrest. We calculated the Out-of-hospital cardiac arrest (OHCA) score and the Cardiac Arrest Hospital Prognosis (CAHP) score upon ICU admission and measured serum NSE upon admission and days 1, 2, 3, 5 and 7. We calculated logistic regression models to study associations of scores and NSE levels with neurological outcome defined by Cerebral Performance Category (CPC) scale and in-hospital death. Results From 336 included patients, 180 (54%) survived until hospital discharge, of which 150 (45%) had a good neurological outcome. NSE at day 3 showed the highest prognostic accuracy (discrimination) for neurological outcome (area under the curve (AUC) 0.89) and in-hospital mortality (AUC 0.88). These results were robust in reclassification statistics and across different subgroups. NSE kinetics with admission levels serving as a baseline did not further improve prognostication. NSE on day 3 significantly improved discrimination of both clinical risk scores (CAHP from AUC 0.81 to 0.91; OHCA from AUC 0.79 to 0.89). Conclusion NSE measured at day 3 significantly improves clinical risk scores for outcome prediction in cardiac arrest patients and may thus add to clinical decision making about escalation or withdrawal of therapy in this vulnerable patient population.

Text attribute transfer aims to automatically rewrite sentences such that they possess certain linguistic attributes, while simultaneously preserving their semantic content. This task remains challenging due to a lack of supervised parallel data. Existing approaches try to explicitly disentangle content and attribute information, but this is difficult and often results in poor content-preservation and ungrammaticality. In contrast, we propose a simpler approach, Iterative Matching and Translation (IMaT), which: (1) constructs a pseudo-parallel corpus by aligning a subset of semantically similar sentences from the source and the target corpora; (2) applies a standard sequence-to-sequence model to learn the attribute transfer; (3) iteratively improves the learned transfer function by refining imperfections in the alignment. In sentiment modification and formality transfer tasks, our method outperforms complex state-of-the-art systems by a large margin. As an auxiliary contribution, we produce a publicly-available test set with human-generated transfer references., EMNLP 2019 (Long Paper)

A 56-year-old man presented with a two month history of increasing anterior knee pain without previous trauma. As usual we recommended physiotherapy with stretching exercises of the quadriceps muscle. Since symptoms did not improve after 6 weeks MRI was performed. Surprisingly a hyperintense lobulated mass of the patella with small fluid-filled cavities at the inferior pole was revealed. We performed an open biopsy to exclude any malignancy and diagnosed an aneurysmal bone cyst. Further examination with CT scans showed an aggressive behaviour with cortical breakthrough. We performed an intralesional curettage with additional high-speed burring and bone cement packing. Sixteen months later the patient was free from any complaints and without signs of local recurrence. Primary bone tumors of the patella are extremely rare and occurrence of aneurysmal bone cysts in this localization is very uncommon. This case report indicates that although anterior knee pain is a very frequent and usually harmless symptom, it is essential to consider that it might also be caused by more severe disorders such as bone tumors.

Aplasia of both cruciate ligaments is a rare congenital disorder. A 28-year-old male presented with pain and the feeling of instability of his right knee after trauma. The provided MRI and previous arthroscopy reports did not indicate any abnormalities except cruciate ligament tears. He was referred to us for reconstruction of both cruciate ligaments. The patient again underwent arthroscopy which revealed a hypoplasia of the medial trochlea and an extremely narrow intercondylar notch. The tibia revealed a missing anterior cruciate ligament (ACL) footprint and a single bump with a complete coverage with articular cartilage. There was no room for an ACL graft. A posterior cruciate ligament could not be identified. The procedure was ended since a ligament reconstruction did not appear reasonable. A significant notch plasty if not a partial resection of the condyles would have been necessary to implant a ligament graft. It is most likely that this would not lead to good knee stability. If the surgeon would have retrieved the contralateral hamstrings at the beginning of the planned ligament reconstruction a significant damage would have occurred to the patient. Even in seemingly clear diagnostic findings the arthroscopic surgeon should take this rare abdnormality into consideration and be familiar with the respective radiological findings. We refer the abnormal finding of only one tibial spine to as the "dromedar-sign" as opposed to the two (medial and a lateral) tibial spines in a normal knee. This may be used as a hint for aplasia of the cruciate ligaments.

e22114 Background: Due to chronic hypoxia cancer cells are growing in a more acidic environment compared to physiological tissue. Recent preclinical in vitro and in vivo data have shown direct pH-sensitive anti-tumor activity by pore formation and apoptosis when synergistically acting combinations of Diflunisal, Paraaminosalicylic acid (PAS) and Aspirin were applied. Therefore, a pilot study was performed to evaluate its clinical relevance and its update is presented. Methods: A total of 65 patients (n=65) with advanced solid tumors and the lack of standard treatments were treated with a protocol consisting of 3-5 weeks with 2 daily i.v. infusions of Diflunisal and Aspirin for 4 days per week after they had signed an informed consent form. Response evaluation was assessed by PET/CT differentiating in metabolic and metric (RECIST) response and tumor marker where available. Results: Out of 65 patients treated 8 patients suffered from colorectal cancer, 17 from breast cancer, 6 from ovarian cancer, 3 from lung cancer, 3 from gastric cancer, 3 from glioblastoma, 3 from CUP, 1 from hemangioendothelioma, 1 from pleuramesothelioma, 1 from non-Hodgkin lymphoma, 2 from pancreatic cancer, 1 from choledochus cancer, 3 from prostate cancer, 5 from sarcomas, 2 from head and neck cancer, 1 from melanoma, 1 from cervical cancer, and 1 from thyroid cancer, respectively. Side effects related to the therapy have been fatigue grade I (30%), nausea grade I (13,3%), tinnitus (25%), hypertension (2,5%), dyspnea (3,3%) and burning sensation in tumor areas (65%). 33 patients were evaluable for metric response, 20 for metabolic and 19 for tumor marker response assessment. For tumor marker follow up 11 % had a CR, 53 % a PR, 26 % a SD of > 3 month and 11 % a PD. For metric follow up 3 % had a CR, 33 % a PR, 39 % a SD of > 3 month and 24 % a PD. For metabolic follow up 10 % had a CR, 35 % a PR, 35 % a SD of > 3 month and 20 % a PD. Conclusions: This continuing pilot study confirms a direct pH-sensitive synergistic anti-tumor activity of intravenous Diflunisal, PAS and Aspirin in patients with advanced solid tumors. Therefore, a further evaluation in a controlled clinical phase II study should be performed.

e13508 Background: Due to chronic hypoxia cancer cells are growing in a more acidic environment compared to physiological tissue. Using this differentiation a new therapy method has been developed for targeted chemotherapy. Recent preclinical in vitro and in vivo data have shown direct pH-sensitive anti-tumor activity by pore formation and apoptosis when synergistically acting combinations of diflunisal, paraaminosalicylic acid (PAS) and aspirin were applied. Therefore, a pilot study was performed to evaluate its clinical relevance. Methods: A total of 30 patients (n=30) with advanced solid tumors and the lack of standard treatments were treated with a protocol consisting of 3-5 weeks with 2 daily i.v. infusions of diflunisal and aspirin for 4 days per week after they had signed an informed consent form. Response evaluation was assessed by PET/CT differentiating in metabolic and metric (RECIST) response. Results: Out of 30 patients treated 4 patients suffered from colorectal cancer, 11 from breast cancer, 2 from ovarian cancer, 2 from lung cancer, 2 from gastric cancer, 2 from glioblastoma, 1 from CUP, 1 from hemangioendothelioma, 1 from pleuramesothelioma, 1 from non-Hodgkin lymphoma, 1 from pancreatic cancer, 1 from choledochus cancer, 1 from prostate cancer, respectively. Side effects related to the therapy have been fatigue (6,7%), nausea (13,3%), tinnitus (10%), hypertension (6,7%), dyspnea (3,3%) and burning sensation in tumor areas (3,3%). Out of 30 patients treated, until today 14 were evaluable for metric response, 12 for metabolic and 13 for tumor marker response assessment. For tumor marker follow up 2 patientis (6,7 %) had a CR, 6 (20 %) a PR, 4 (13,3 %) a SD of > 3 month and 1 (3,3 %) a PD. For metric follow up 0 patients (0 %) had a CR, 6 (20 %) a PR, 5 (16,7 %) a SD of > 3 month and 3 (10 %) a PD. For metabolic follow up 1 patients (3,3 %) had a CR, 3 (10 %) a PR, 5 (16,7 %) a SD of > 3 month and 3 (10 %) a PD. Conclusions: This pilot study confirms a direct pH-sensitive synergistic anti-tumor activity of intravenous salicylic acids in patients with advanced solid tumors. Therefore, a further evaluation in a controlled clinical phase II study will be performed.