Extraordinary levels of per- and polyfluoroalkyl substances (PFAS) in vertebrate animals at a New Mexico desert oasis: Multiple pathways for wildlife and human exposure

Per- and polyfluoroalkyl substances (PFAS) in the environment pose persistent and complex threats to human and wildlife health. Around the world, PFAS point sources such as military bases expose thousands of populations of wildlife and game species, with potentially far-reaching implications for population and ecosystem health. But few studies shed light on the extent to which PFAS permeate food webs, particularly ecologically and taxonomically diverse communities of primary and secondary consumers. Here we conducted >2000 assays to measure tissue-concentrations of 17 PFAS in 23 species of mammals and migratory birds at Holloman Air Force Base (AFB), New Mexico, USA, where wastewater catchment lakes form biodiverse oases. PFAS concentrations were among the highest reported in animal tissues, and high levels have persisted for at least three decades. Twenty of 23 species sampled at Holloman AFB were heavily contaminated, representing middle trophic levels and wetland to desert microhabitats, implicating pathways for PFAS uptake: ingestion of surface water, sediments, and soil; foraging on aquatic invertebrates and plants; and preying upon birds or mammals. The hazardous long carbon-chain form, perfluorooctanosulfonic acid (PFOS), was most abundant, with liver concentrations averaging >10,000 ng/g wet weight (ww) in birds and mammals, respectively, and reaching as high 97,000 ng/g ww in a 1994 specimen. Perfluorohexanesulfonic acid (PFHxS) averaged thousands of ng/g ww in the livers of aquatic birds and littoral-zone house mice, but one order of magnitude lower in the livers of upland desert rodent species. Piscivores and upland desert songbirds were relatively uncontaminated. At control sites, PFAS levels were strikingly lower on average and different in composition. In sum, legacy PFAS at this desert oasis have permeated local aquatic and terrestrial food webs across decades, severely contaminating populations of resident and migrant animals, and exposing people via game meat consumption and outdoor recreation.

A diet-dependent host metabolite shapes the gut microbiota to protect from autoimmunity

Diet can protect from autoimmune disease; however, whether diet acts via the host and/or microbiome remains unclear. Here, we use a ketogenic diet (KD) as a model to dissect these complex interactions. A KD rescued the experimental autoimmune encephalomyelitis (EAE) mouse model of multiple sclerosis in a microbiota-dependent fashion. Dietary supplementation with a single KD-dependent host metabolite (β-hydroxybutyrate, βHB) rescued EAE whereas transgenic mice unable to produce βHB in the intestine developed more severe disease. Transplantation of the βHB-shaped gut microbiota was protective. Lactobacillus sequence variants were associated with decreased T helper 17 (Th17) cell activation in vitro . Finally, we isolated a L. murinus strain that protected from EAE, which was phenocopied by the Lactobacillus metabolite indole lactic acid. Thus, diet alters the immunomodulatory potential of the gut microbiota by shifting host metabolism, emphasizing the utility of taking a more integrative approach to study diet-host-microbiome interactions.

Assessing potential human health and ecological implications of PFAS from leave-in dental products

Contaminated water and food are the main sources of documented per- and polyfluoroalkyl substances (PFAS) exposure in humans. However, other sources may contribute to the overall PFAS intake. While several studies documented the presence of PFAS in consumer products, PFAS evaluation in dental products has been limited to floss and tape to date. This study estimated PFAS exposures from a convenience sample of leave-in dental products (night guards and whitening trays), which remain in contact with the mouth for longer durations than previously evaluated dental products. This analysis evaluated whether consumer usage of these dental products meaningfully contributes to oral exposure of PFAS. Leaching of PFAS upon disposal of products was also considered. Out of 24 PFAS measured, perfluorobutanoic acid (PFBA; 3.24-4.17 ng/product or 0.67-0.83 ng/g) and perfluorooctanesulfonic acid (PFOS; 7.25-16.45 ng/product or 1.2-2.3 ng/g) were detected in night guards, and no PFAS were detected in whitening trays. Non-targeted analysis showed additional possible PFAS, which could not be characterized. The findings showed that PFOS and/or PFBA present in night guards were unlikely to pose a health concern. From an ecological perspective, the dental products examined were shown to constitute a negligible contribution to environmental PFAS. In conclusion, the examined dental products do not represent a significant source of exposure to PFAS for humans or the environment. The study demonstrates how risk assessment can be integrated by the industry into product stewardship programs to evaluate the potential health and environmental impacts of chemicals in consumer products.

Self-Collection Blood Test for PFASs: Comparing Volumetric Microsamplers with a Traditional Serum Approach

A remote sampling approach was developed at Eurofins for quantifying per- and polyfluoroalkyl substances (PFASs) in whole blood samples collected using volumetric absorptive microsamplers (VAMSs), which allow for self-collection of blood using a finger prick. This study compares PFAS exposure measured by self-collection of blood using VAMSs to the standard venous serum approach. Blood samples were collected from participants (n = 53) in a community with prior PFAS drinking water contamination using a venous blood draw as well as participant self-collection using VAMSs. Whole blood from the venous tubes was also loaded onto VAMSs to compare differences in capillary vs venous whole blood PFAS levels. Samples were quantified for PFASs using liquid chromatography tandem mass spectrometry and online solid-phase extraction. PFAS levels in serum were highly correlated with measurements in capillary VAMSs (r ≥ 0.91 and p < 0.05). Serum PFAS levels were generally twofold higher than whole blood, reflecting expected differences in their composition. Of interest, FOSA was detected in whole blood (both venous and capillary VAMSs) but not in serum. Overall, these findings indicate that VAMSs are useful self-collection tools for assessing elevated human exposure to PFASs.

A field-validated equilibrium passive sampler for the monitoring of per- and polyfluoroalkyl substances (PFAS) in sediment pore water and surface water

A simple equilibrium passive sampler, consisting of water in an inert container capped with a rate-limiting barrier, for the monitoring of per- and polyfluoroalkyl substances (PFAS) in sediment pore water and surface water was developed and tested through a series of laboratory and field experiments. The objectives of the laboratory experiments were to determine (1) the membrane type that could serve as the sampler's rate-limiting barrier, (2) the mass transfer coefficient of environmentally relevant PFAS through the selected membrane, and (3) the performance reference compounds (PRCs) that could be used to infer the kinetics of PFAS diffusing into the sampler. Of the membranes tested, the polycarbonate (PC) membrane was deemed the most suitable rate-limiting barrier, given that it did not appreciably adsorb the studied PFAS (which have ≤8 carbons), and that the migration of these compounds through this membrane could be described by Fick's law of diffusion. When employed as the PRC, the isotopically labelled PFAS M₂PFOA and M₄PFOS were able to predict the mass transfer coefficients of the studied PFAS analytes. In contrast, the mass transfer coefficients were underpredicted by Br^- and M₃PFPeA. For validation, the PC-based passive samplers consisting of these four PRCs, as well as two other PRCs (i.e., M₈PFOA and C₈H₁₇SO₃^-), were deployed in the sediment and water at a PFAS-impacted field site. The concentration-time profiles of the PRCs indicated that the samplers deployed in the sediment required at least 6 to 7 weeks to reach 90% equilibrium. If the deployment times are shorter (e.g., 2 to 4 weeks), PFAS concentrations at equilibrium could be estimated based on the concentrations of the PRCs remaining in the sampler at retrieval. All PFAS concentrations determined via this approach were within a factor of two compared to those measured in the mechanically extracted sediment pore water and surface water samples obtained adjacent to the sampler deployment locations. Neither biofouling of the rate-limiting barrier nor any physical change to it was observed on the sampler after retrieval. The passive sampler developed in this study could be a promising tool for the monitoring of PFAS in pore water and surface water.

Robust Losses for Learning Value Functions

Most value function learning algorithms in reinforcement learning are based on the mean squared (projected) Bellman error. However, squared errors are known to be sensitive to outliers, both skewing the solution of the objective and resulting in high-magnitude and high-variance gradients. To control these high-magnitude updates, typical strategies in RL involve clipping gradients, clipping rewards, rescaling rewards, or clipping errors. While these strategies appear to be related to robust losses-like the Huber loss-they are built on semi-gradient update rules which do not minimize a known loss. In this work, we build on recent insights reformulating squared Bellman errors as a saddlepoint optimization problem and propose a saddlepoint reformulation for a Huber Bellman error and Absolute Bellman error. We start from a formalization of robust losses, then derive sound gradient-based approaches to minimize these losses in both the online off-policy prediction and control settings. We characterize the solutions of the robust losses, providing insight into the problem settings where the robust losses define notably better solutions than the mean squared Bellman error. Finally, we show that the resulting gradient-based algorithms are more stable, for both prediction and control, with less sensitivity to meta-parameters.

Computational Methods Summarizing Mutational Patterns in Cancer: Promise and Limitations for Clinical Applications

Since the rise of next-generation sequencing technologies, the catalogue of mutations in cancer has been continuously expanding. To address the complexity of the cancer-genomic landscape and extract meaningful insights, numerous computational approaches have been developed over the last two decades. In this review, we survey the current leading computational methods to derive intricate mutational patterns in the context of clinical relevance. We begin with mutation signatures, explaining first how mutation signatures were developed and then examining the utility of studies using mutation signatures to correlate environmental effects on the cancer genome. Next, we examine current clinical research that employs mutation signatures and discuss the potential use cases and challenges of mutation signatures in clinical decision-making. We then examine computational studies developing tools to investigate complex patterns of mutations beyond the context of mutational signatures. We survey methods to identify cancer-driver genes, from single-driver studies to pathway and network analyses. In addition, we review methods inferring complex combinations of mutations for clinical tasks and using mutations integrated with multi-omics data to better predict cancer phenotypes. We examine the use of these tools for either discovery or prediction, including prediction of tumor origin, treatment outcomes, prognosis, and cancer typing. We further discuss the main limitations preventing widespread clinical integration of computational tools for the diagnosis and treatment of cancer. We end by proposing solutions to address these challenges using recent advances in machine learning.

Hospital triage and skin disease: hospital outcomes are differentially associated with cutaneous morphology

BACKGROUND

Determining the exact dermatologic diagnosis is difficult in the inpatient setting.

OBJECTIVE

Determine whether morphologic classification rather than specific diagnosis is associated with hospital outcomes.

METHODS

Retrospective single-center study. Information from 1798 inpatient dermatology consults at The Ohio State University Wexner Medical Center from 2012 to 2014 was queried. Dermatologic diseases were categorized into 16 groups based on appearance. Logistic regression was performed comparing mortality rate vs morphology. Linear regression was performed comparing the length of stay (LOS) vs morphology.

RESULTS

Morphology was associated with a mortality rate (P = 0.038). The morphologic subgroups acneiform/follicular/occlusion (P = 0.011), blistering disorders (P = 0.009), retiform purpura (P = 0.011), and vasculitis/vascular (P = 0.007) were associated with increased mortality. Morphology was associated with LOS (P = 0.004), and the morbilliform subgroup was associated with increased LOS (P < 0.001).

CONCLUSION

This study demonstrated the importance of morphologic diagnosis and its association with mortality rate and LOS. This information may help triage cutaneous disorders in the inpatient setting and determine the relative risk of dermatologic conditions when assessing the need for hospital transfers and more aggressive therapies.

Mutated processes predict immune checkpoint inhibitor therapy benefit in metastatic melanoma

Immune Checkpoint Inhibitor (ICI) therapy has revolutionized treatment for advanced melanoma; however, only a subset of patients benefit from this treatment. Despite considerable efforts, the Tumor Mutation Burden (TMB) is the only FDA-approved biomarker in melanoma. However, the mechanisms underlying TMB association with prolonged ICI survival are not entirely understood and may depend on numerous confounding factors. To identify more interpretable ICI response biomarkers based on tumor mutations, we train classifiers using mutations within distinct biological processes. We evaluate a variety of feature selection and classification methods and identify key mutated biological processes that provide improved predictive capability compared to the TMB. The top mutated processes we identify are leukocyte and T-cell proliferation regulation, which demonstrate stable predictive performance across different data cohorts of melanoma patients treated with ICI. This study provides biologically interpretable genomic predictors of ICI response with substantially improved predictive performance over the TMB.

POS1127 USE OF ARTIFICIAL INTELLIGENCE AND CLOUD-BASED INFRASTRUCTURE TO IMPROVE THE SPEED AND ACCURACY OF ELIGIBILITY READS IN OSTEOARTHRITIS TRIALS

Background

The Kellgren-Lawrence grading (KLG) system is used in clinical trials of osteoarthritis (OA) to define the structural severity of the disease as part of patient eligibility assessment. However, the use of KLG system has proven to be challenging due to considerable inter-reader variability[1], [2], which may result in recruitment of sub-optimal patient cohort or delays in patient recruitment.

Objectives

The objective of this study was to determine the impact of an AI-assisted, cloud-based data management system on the rate of adjudication and the speed of OA patient recruitment.

Methods

A total of 3855 bilateral fixed-flexion posteroanterior radiographs of the tibiofemoral joints from global multi-centre trials were included in this study. Two experienced readers performed an initial KLG assessment of both knees; the adjudication was performed by a third experienced reader. A cloud-based imaging data management system was deployed, the readers could access the data simultaneously and adjudication was automatically triggered.

We quantified the adjudication rate and the distribution of disagreements in KLG scores provided by the initial readers. Furthermore, the delay in delivery time of the KLG reports to the recruiting site was recorded.

Results

48% (1836) of the initial reads required adjudication. Approximately 70% of the disagreements affected the conventional KLG 2-3 inclusion range of OA clinical trials. Use of the cloud-based data management allowed 41% of the reports to be delivered within 24 hours, if no adjudication was required vs an average of 5 days as estimated based on the readers’ prior experience.

Table 1 provides details on the distribution of disagreements resulting in adjudication reads. Figure 1 shows the delivery time for KLG with and without adjudication.

Table 1.

Distribution of disagreements of initial reads resulting in an adjudication read being triggered.

Disagreement triggering AdjudicationNumber of CasesPercentageKLG 0 – 171327.8%KLG 1 – 282031.9%KLG 2 – 348218.7%KLG 3 – 444617.4%Other1104.3%Total2569100%Figure 1.

Time to deliver eligibility report

Nehrer et al. showed that assisting the readers with AI generated KLG scoring reports, the agreement rate between readers for KLG assessment increased by 21% [2]. Adding to this, 30% of the adjudications (stemming from KLG 0 –1 disagreements between readers) could be avoided when using AI generated reports.

Conclusion

We assess the rate of adjudication and speed of reporting of KLG of data from multi-centre OA clinical trials. Future work is planned to assess the effect of AI-assisted OA grading systems within our cloud-based data management system on reader agreement and recruitment speed in global clinical trials.

References

[1]D. J. Hunter et al., “OARSI Clinical Trials Recommendations: Knee imaging in clinical trials inosteoarthritis,” Osteoarthritis and Cartilage, vol. 23, no. 5. W.B. Saunders Ltd, pp. 698–715, 2015. doi: 10.1016/j.joca.2015.03.012.

[2]S. Nehrer et al., “Automated Knee Osteoarthritis Assessment Increases Physicians’ Agreement Rate and Accuracy: Data from the Osteoarthritis Initiative,” Cartilage, vol. 13, no. 1_suppl, pp. 957S-965S, Dec. 2021, doi: 10.1177/1947603519888793.

Disclosure of Interests

Olga Kubassova Grant/research support from: Takeda, Lilly, Abbvie, Pfizer, Mikael Boesen Speakers bureau: Lilly, Novartis, Abbvie, Pfizer, Cristiano Pereira da Costa: None declared, Julia O’Lynn: None declared, Andrew Patterson: None declared, Dimitri Kessler: None declared

Defective Humoral Immunity Disrupts Bile Acid Homeostasis which Promotes Inflammatory Disease of the Small Bowel

Mucosal antibodies maintain gut homeostasis by promoting spatial segregation between host tissues and luminal microbes. Whether and how mucosal antibody responses influence gut health through modulation of microbiota composition is unclear. Here, we use a CD19−/− mouse model of antibody-deficiency to demonstrate that a relationship exists between dysbiosis, defects in bile acid homeostasis, and gluten-sensitive enteropathy of the small intestine. The gluten-sensitive small intestine enteropathy that develops in CD19−/− mice is associated with alterations to luminal bile acid composition in the SI, marked by significant reductions in the abundance of conjugated bile acids. Manipulation of bile acid availability, adoptive transfer of functional B cells, and ablation of bacterial bile salt hydrolase activity all influence the severity of small intestine enteropathy in CD19−/− mice. Collectively, results from our experiments support a model whereby mucosal humoral immune responses limit inflammatory disease of the small bowel by regulating bacterial BA metabolism.

J.L.K. was supported by a NIAID K22 Award (AI123481), a Jeffrey Modell Network Specific Diseases Research Award, a University of South Carolina ASPIRE-I Award, a pilot project award through the University of South Carolina Center for Alternative Medicine COBRE program (P20GM103641; awarded to Drs. Mitzi and Prakash Nagarkatti), an R21 (AI142409), and 1R01AI155887.

Predicting Clinical Events Based on Raw Text: From Bag-of-Words to Attention-Based Transformers

Identifying which patients are at higher risks of dying or being re-admitted often happens to be resource- and life- saving, thus is a very important and challenging task for healthcare text analytics. While many successful approaches exist to predict such clinical events based on categorical and numerical variables, a large amount of health records exists in the format of raw text such as clinical notes or discharge summaries. However, the text-analytics models applied to free-form natural language found in those notes are lagging behind the break-throughs happening in the other domains and remain to be primarily based on older bag-of-words technologies. As a result, they rarely reach the accuracy level acceptable for the clinicians. In spite of their success in other domains, the superiority of deep neural approaches over classical bags of words for this task has not yet been convincingly demonstrated. Also, while some successful experiments have been reported, the most recent break-throughs due to the pre-trained language models have not yet made their ways into the medical domain. Using a publicly available healthcare dataset, we have explored several classification models to predict patients' re-admission or a fatality based on their discharge summaries and established that 1) The performance of the neural models used in our experiments convincingly exceeds those based on bag-of-words by several percentage points as measured by the standard metrics. 2) This allows us to achieve the accuracy typically acceptable by the clinicians as of practical use (area under the ROC curve above 0.70) for the majority of our prediction targets. 3) While the pre-trained attention-based transformer performed only on par with the model that averages word embeddings when applied to full length discharge summaries, the transformer still handles shorter text segments substantially better, at times with the margin of 0.04 in the area under the ROC curve. Thus, our findings extend the success of pre-trained language models reported in other domains to the task of clinical event prediction, and likely to other text-classification tasks in the healthcare analytics domain. 4) We suggest several models to overcome the transformers' major drawback (their input size limitation), and confirm that this is crucial to achieve their top performance. Our modifications are domain agnostic, and thus can be applied in other applications where the text inputs exceed 200 words. 5) We have successfully demonstrated how non-text attributes (such as patient age, demographics, type of admission etc.) can be combined with text to gain additional improvements for several prediction targets. We include extensive ablation studies showing the impact of the training size, and highlighting the tradeoffs between the performance and the resources needed.

Defective humoral immunity disrupts bile acid homeostasis which promotes inflammatory disease of the small bowel

Mucosal antibodies maintain gut homeostasis by promoting spatial segregation between host tissues and luminal microbes. Whether and how mucosal antibody responses influence gut health through modulation of microbiota composition is unclear. Here, we use a CD19^-/- mouse model of antibody-deficiency to demonstrate that a relationship exists between dysbiosis, defects in bile acid homeostasis, and gluten-sensitive enteropathy of the small intestine. The gluten-sensitive small intestine enteropathy that develops in CD19^-/- mice is associated with alterations to luminal bile acid composition in the SI, marked by significant reductions in the abundance of conjugated bile acids. Manipulation of bile acid availability, adoptive transfer of functional B cells, and ablation of bacterial bile salt hydrolase activity all influence the severity of small intestine enteropathy in CD19^-/- mice. Collectively, results from our experiments support a model whereby mucosal humoral immune responses limit inflammatory disease of the small bowel by regulating bacterial BA metabolism.

Hyperpolarized Carbon-13 MRI for Early Response Assessment of Neoadjuvant Chemotherapy in Breast Cancer Patients

Hyperpolarized 13C-MRI is an emerging tool for probing tissue metabolism by measuring 13C-label exchange between intravenously injected hyperpolarized [1-13C]pyruvate and endogenous tissue lactate. Here, we demonstrate that hyperpolarized 13C-MRI can be used to detect early response to neoadjuvant therapy in breast cancer. Seven patients underwent multiparametric 1H-MRI and hyperpolarized 13C-MRI before and 7-11 days after commencing treatment. An increase in the lactate-to-pyruvate ratio of approximately 20% identified three patients who, following 5-6 cycles of treatment, showed pathological complete response. This ratio correlated with gene expression of the pyruvate transporter MCT1 and lactate dehydrogenase A (LDHA), the enzyme catalyzing label exchange between pyruvate and lactate. Analysis of approximately 2,000 breast tumors showed that overexpression of LDHA and the hypoxia marker CAIX was associated with reduced relapse-free and overall survival. Hyperpolarized 13C-MRI represents a promising method for monitoring very early treatment response in breast cancer and has demonstrated prognostic potential. SIGNIFICANCE: Hyperpolarized carbon-13 MRI allows response assessment in patients with breast cancer after 7-11 days of neoadjuvant chemotherapy and outperformed state-of-the-art and research quantitative proton MRI techniques.

1359 Everyone Deserves to Smile: An Audit on The Success Rate of Dental Implants in Complex Maxillofacial Patients

Aim

To determine the success rates of dental implants in complex patients treated in the Oral and Maxillofacial Surgery Department at Rotherham General Hospital (RGH).

Method

All patients who received dental implants in the period January 2010 – March 2017 at RGH were included in the audit.

Standards

90% local standard

Results

68 dental implants were placed in 22 patients: 15 patients head and neck reconstruction, 5 atrophic edentulous patients, 2 trauma.

Conclusions

Survival rates for dental implants in the literature is variable ranging from 91 -96% in native bone and mucosa and 46 - 98% in irradiated tissue.

PIR-B Regulates CD4+ IL17a+ T-Cell Survival and Restricts T-Cell-Dependent Intestinal Inflammatory Responses

BACKGROUND & AIMS

CD4⁺ T cells are regulated by activating and inhibitory cues, and dysregulation of these proper regulatory inputs predisposes these cells to aberrant inflammation and exacerbation of disease. We investigated the role of the inhibitory receptor paired immunoglobulin-like receptor B (PIR-B) in the regulation of the CD4⁺ T-cell inflammatory response and exacerbation of the colitic phenotype.

METHODS

We used Il10^-/- spontaneous and CD4⁺CD45RB^hi T-cell transfer models of colitis with PIR-B-deficient (Pirb^-/-) mice. Flow cytometry, Western blot, and RNA sequencing analysis was performed on wild-type and Pirb^-/- CD4⁺ T cells. In silico analyses were performed on RNA sequencing data set of ileal biopsy samples from pediatric CD and non-inflammatory bowel disease patients and sorted human memory CD4⁺ T cells.

RESULTS

We identified PIR-B expression on memory CD4⁺ interleukin (IL)17a⁺ cells. We show that PIR-B regulates CD4⁺ T-helper 17 cell (Th17)-dependent chronic intestinal inflammatory responses and the development of colitis. Mechanistically, we show that the PIR-B- Src-homology region 2 domain-containing phosphatase-1/2 axis tempers mammalian target of rapamycin complex 1 signaling and mammalian target of rapamycin complex 1-dependent caspase-3/7 apoptosis, resulting in CD4⁺ IL17a⁺ cell survival. In silico analyses showed enrichment of transcriptional signatures for Th17 cells (RORC, RORA, and IL17A) and tissue resident memory (HOBIT, IL7R, and BLIMP1) networks in PIR-B⁺ murine CD4⁺ T cells and human CD4⁺ T cells that express the human homologue leukocyte immunoglobulin-like receptor subfamily B member 3 (LILRB3). High levels of LILRB3 expression were associated strongly with mucosal injury and a proinflammatory Th17 signature, and this signature was restricted to a treatment-naïve, severe pediatric CD population.

CONCLUSIONS

Our findings show an intrinsic role for PIR-B/LILRB3 in the regulation of CD4⁺ IL17a⁺ T-cell pathogenic memory responses.

The Pretreatment Gut Microbiome Is Associated With Lack of Response to Methotrexate in New-Onset Rheumatoid Arthritis

OBJECTIVE

Although oral methotrexate (MTX) remains the anchor drug for rheumatoid arthritis (RA), up to 50% of patients do not achieve a clinically adequate outcome. In addition, there is a lack of prognostic tools for treatment response prior to drug initiation. This study was undertaken to investigate whether interindividual differences in the human gut microbiome can aid in the prediction of MTX efficacy in new-onset RA.

METHODS

We performed 16S ribosomal RNA gene and shotgun metagenomic sequencing on the baseline gut microbiomes of drug-naive patients with new-onset RA (n = 26). Results were validated in an additional independent cohort (n = 21). To gain insight into potential microbial mechanisms, we conducted ex vivo experiments coupled with metabolomics analysis to evaluate the association between microbiome-driven MTX depletion and clinical response.

RESULTS

Our analysis revealed significant associations of the abundance of gut bacterial taxa and their genes with future clinical response (q < 0.05), including orthologs related to purine and MTX metabolism. Machine learning techniques were applied to the metagenomic data, resulting in a microbiome-based model that predicted lack of response to MTX in an independent group of patients. Finally, MTX levels remaining after ex vivo incubation with distal gut samples from pretreatment RA patients significantly correlated with the magnitude of future clinical response, suggesting a possible direct effect of the gut microbiome on MTX metabolism and treatment outcomes.

CONCLUSION

Taken together, these findings are the first step toward predicting lack of response to oral MTX in patients with new-onset RA and support the value of the gut microbiome as a possible prognostic tool and as a potential target in RA therapeutics.

High-throughput organ-on-chip platform with integrated programmable fluid flow and real-time sensing for complex tissue models in drug development workflows

Drug development suffers from a lack of predictive and human-relevant in vitro models. Organ-on-chip (OOC) technology provides advanced culture capabilities to generate physiologically appropriate, human-based tissue in vitro, therefore providing a route to a predictive in vitro model. However, OOC technologies are often created at the expense of throughput, industry-standard form factors, and compatibility with state-of-the-art data collection tools. Here we present an OOC platform with advanced culture capabilities supporting a variety of human tissue models including liver, vascular, gastrointestinal, and kidney. The platform has 96 devices per industry standard plate and compatibility with contemporary high-throughput data collection tools. Specifically, we demonstrate programmable flow control over two physiologically relevant flow regimes: perfusion flow that enhances hepatic tissue function and high-shear stress flow that aligns endothelial monolayers. In addition, we integrate electrical sensors, demonstrating quantification of barrier function of primary gut colon tissue in real-time. We utilize optical access to the tissues to directly quantify renal active transport and oxygen consumption via integrated oxygen sensors. Finally, we leverage the compatibility and throughput of the platform to screen all 96 devices using high content screening (HCS) and evaluate gene expression using RNA sequencing (RNA-seq). By combining these capabilities in one platform, physiologically-relevant tissues can be generated and measured, accelerating optimization of an in vitro model, and ultimately increasing predictive accuracy of in vitro drug screening.

General Value Function Networks

State construction is important for learning in partially observable environments. A general purpose strategy for state construction is to learn the state update using a Recurrent Neural Network (RNN), which updates the internal state using the current internal state and the most recent observation. This internal state provides a summary of the observed sequence, to facilitate accurate predictions and decision-making. At the same time, specifying and training RNNs is notoriously tricky, particularly as the common strategy to approximate gradients back in time, called truncated Back-prop Through Time (BPTT), can be sensitive to the truncation window. Further, domain-expertise—which can usually help constrain the function class and so improve trainability—can be difficult to incorporate into complex recurrent units used within RNNs. In this work, we explore how to use multi-step predictions to constrain the RNN and incorporate prior knowledge. In particular, we revisit the idea of using predictions to construct state and ask: does constraining (parts of) the state to consist of predictions about the future improve RNN trainability? We formulate a novel RNN architecture, called a General Value Function Network (GVFN), where each internal state component corresponds to a prediction about the future represented as a value function. We first provide an objective for optimizing GVFNs, and derive several algorithms to optimize this objective. We then show that GVFNs are more robust to the truncation level, in many cases only requiring one-step gradient updates.

Prevalence of common mental disorders among Syrian refugee children and adolescents in Sultanbeyli district, Istanbul: results of a population-based survey

AIMS

Research demonstrates elevated levels of common mental disorders among Syrian refugees, but the majority of studies have, to date, focused on adult populations. This study aims to estimate the prevalence of depression, anxiety and post-traumatic stress disorder (PTSD) among Syrian children and adolescents living in Sultanbeyli district of Istanbul, Turkey.

METHODS

A population-based survey among Syrian children and adolescents aged 8-17 years living in Sultanbeyli district was conducted in 2019, as part of an all-age survey of disability. 80 clusters of 50 participants (all-ages) were selected from the local municipality's refugee registration database using probability proportionate to size sampling. Children aged 8-17 years were assessed for symptoms of common mental disorders using the Child Revised Impact of Event Scale (CRIES-8) and abbreviated versions of the Center for Epidemiologic Studies Depression Scale for Children (CES-DC) and the Screen for Child Anxiety Related Disorders (SCARED).

RESULTS

Of the 852 participants, 23.7% (95% CI 19.9-27.2) screened positive for symptomatic depression, PTSD and anxiety. The prevalence estimates for depression, PTSD and anxiety were 12.5% (95% CI 9.8-15.6), 11.5% (95% CI 9.1-14.4) and 9.2% (95% CI 6.8-12.1), respectively. Depression and PTSD were significantly more common in older adolescents, whilst anxiety and PTSD were significantly more common in girls. Depression was more common in children from poorer households and those who had received no education. Children coming from larger households were less likely to show symptoms of PTSD.

CONCLUSIONS

Syrian refugee children and adolescents are vulnerable to common mental disorders, and culturally appropriate prevention and intervention support are needed for this population.

MDM2-Dependent Rewiring of Metabolomic and Lipidomic Profiles in Dedifferentiated Liposarcoma Models

Dedifferentiated liposarcoma (DDLPS) is an aggressive mesenchymal cancer marked by amplification of MDM2, an inhibitor of the tumor suppressor TP53. DDLPS patients with higher MDM2 amplification have lower chemotherapy sensitivity and worse outcome than patients with lower MDM2 amplification. We hypothesized that MDM2 amplification levels may be associated with changes in DDLPS metabolism. Six patient-derived DDLPS cell line models were subject to comprehensive metabolomic (Metabolon) and lipidomic (SCIEX 5600 TripleTOF-MS) profiling to assess associations with MDM2 amplification and their responses to metabolic perturbations. Comparing metabolomic profiles between MDM2 higher and lower amplification cells yielded a total of 17 differentially abundant metabolites across both panels (FDR < 0.05, log2 fold change < 0.75), including ceramides, glycosylated ceramides, and sphingomyelins. Disruption of lipid metabolism through statin administration resulted in a chemo-sensitive phenotype in MDM2 lower cell lines only, suggesting that lipid metabolism may be a large contributor to the more aggressive nature of MDM2 higher DDLPS tumors. This study is the first to provide comprehensive metabolomic and lipidomic characterization of DDLPS cell lines and provides evidence for MDM2-dependent differential molecular mechanisms that are critical factors in chemoresistance and could thus affect patient outcome.

Chemotherapy-associated nonbacterial thrombotic endocarditis: A radiological mimicker of cardiac amyloidosis requiring histopathologic examination for definitive diagnosis

Nonbacterial thrombotic endocarditis is a form of a thrombotic angiopathy involving the endothelial lined endocardial surfaces of the heart which includes valves and the chamber walls. Underlying etiologies for nonbacterial thrombotic endocarditis include autoimmune diseases, hypercoagulable states, in the setting of certain malignant neoplasms, and physical injury. The pathogenesis for these processes is that of primary endothelial injury resulting in a thrombotic angiopathy. We present a patient with heart failure being evaluated before hematopoietic stem cell transplantation who had previously been provided with chemotherapy and whose cardiac magnetic resonance imaging reveals findings suggestive of amyloidosis. A subsequent endomyocardial biopsy instead showed nonbacterial thrombotic endocarditis characterized by the endocardium with fibromyxoid thickening and overlying fresh fibrin. This case highlights histopathologic findings of chemotherapy-associated nonbacterial thrombotic endocarditis involving the right ventricle wall of the endocardium, therefore expanding the radiological differential in patients with cardiac magnetic resonance imaging findings suggestive of amyloidosis.

Imaging breast cancer using hyperpolarized carbon-13 MRI

Our purpose is to investigate the feasibility of imaging tumor metabolism in breast cancer patients using 13C magnetic resonance spectroscopic imaging (MRSI) of hyperpolarized 13C label exchange between injected [1-13C]pyruvate and the endogenous tumor lactate pool. Treatment-naïve breast cancer patients were recruited: four triple-negative grade 3 cancers; two invasive ductal carcinomas that were estrogen and progesterone receptor-positive (ER/PR+) and HER2/neu-negative (HER2-), one grade 2 and one grade 3; and one grade 2 ER/PR+ HER2- invasive lobular carcinoma (ILC). Dynamic 13C MRSI was performed following injection of hyperpolarized [1-13C]pyruvate. Expression of lactate dehydrogenase A (LDHA), which catalyzes 13C label exchange between pyruvate and lactate, hypoxia-inducible factor-1 (HIF1α), and the monocarboxylate transporters MCT1 and MCT4 were quantified using immunohistochemistry and RNA sequencing. We have demonstrated the feasibility and safety of hyperpolarized 13C MRI in early breast cancer. Both intertumoral and intratumoral heterogeneity of the hyperpolarized pyruvate and lactate signals were observed. The lactate-to-pyruvate signal ratio (LAC/PYR) ranged from 0.021 to 0.473 across the tumor subtypes (mean ± SD: 0.145 ± 0.164), and a lactate signal was observed in all of the grade 3 tumors. The LAC/PYR was significantly correlated with tumor volume (R = 0.903, P = 0.005) and MCT 1 (R = 0.85, P = 0.032) and HIF1α expression (R = 0.83, P = 0.043). Imaging of hyperpolarized [1-13C]pyruvate metabolism in breast cancer is feasible and demonstrated significant intertumoral and intratumoral metabolic heterogeneity, where lactate labeling correlated with MCT1 expression and hypoxia.

Optoacoustic Imaging Detects Hormone-Related Physiological Changes of Breast Parenchyma

PURPOSE

 Optoacoustic imaging with ultrasound (OPUS) can assess in-vivo perfusion/oxygenation through surrogate measures of oxy, deoxy and total hemoglobin content in tissues. The primary aim of our study was to evaluate the ability of OPUS to detect physiological changes in the breast during the menstrual cycle and to determine qualitative/quantitative metrics of normal parenchymal tissue in pre-/post-menopausal women. The secondary aim was to assess the technique's repeatability.

MATERIALS AND METHODS

 We performed a prospective ethically approved study in volunteers using OPUS (700, 800 and 850 nm wavelengths) in the proliferative/follicular and secretory phase of the menstrual cycle. Regions of interest (ROIs) were drawn on the most superficial region of fibroglandular tissue and same-day intra-observer repeatability was assessed. We used t-tests to interrogate differences in the OPUS measurements due to hormonal changes and interclass correlation coefficients/Bland-Altman plots to evaluate the repeatability of mean ROI signal intensities.

RESULTS

 22 pre-menopausal and 8 post-menopausal volunteers were recruited. 21 participants underwent repeatability examinations. OPUS intensity values were significantly higher (p < 0.0001) at all excitation wavelengths in the secretory compared to the proliferative/follicular phase. Post-menopausal volunteers showed similar optoacoustic values to the proliferative/follicular phase of pre-menopausal volunteers. The repeatability of the technique was comparable to other handheld ultrasound modalities.

CONCLUSION

 OPUS detects changes in perfusion/vascularity related to the menstrual cycle and menopausal status of breast parenchyma.

Diffusion-weighted imaging of the breast-a consensus and mission statement from the EUSOBI International Breast Diffusion-Weighted Imaging working group

The European Society of Breast Radiology (EUSOBI) established an International Breast DWI working group. The working group consists of clinical breast MRI experts, MRI physicists, and representatives from large vendors of MRI equipment, invited based upon proven expertise in breast MRI and/or in particular breast DWI, representing 25 sites from 16 countries. The aims of the working group are (a) to promote the use of breast DWI into clinical practice by issuing consensus statements and initiate collaborative research where appropriate; (b) to define necessary standards and provide practical guidance for clinical application of breast DWI; (c) to develop a standardized and translatable multisite multivendor quality assurance protocol, especially for multisite research studies; (d) to find consensus on optimal methods for image processing/analysis, visualization, and interpretation; and (e) to work collaboratively with system vendors to improve breast DWI sequences. First consensus recommendations, presented in this paper, include acquisition parameters for standard breast DWI sequences including specifications of b values, fat saturation, spatial resolution, and repetition and echo times. To describe lesions in an objective way, levels of diffusion restriction/hindrance in the breast have been defined based on the published literature on breast DWI. The use of a small ROI placed on the darkest part of the lesion on the ADC map, avoiding necrotic, noisy or non-enhancing lesion voxels is currently recommended. The working group emphasizes the need for standardization and quality assurance before ADC thresholds are applied. The working group encourages further research in advanced diffusion techniques and tailored DWI strategies for specific indications.

Key Points

• The working group considers breast DWI an essential part of a multiparametric breast MRI protocol and encourages its use.

• Basic requirements for routine clinical application of breast DWI are provided, including recommendations on b values, fat saturation, spatial resolution, and other sequence parameters.

• Diffusion levels in breast lesions are defined based on meta-analysis data and methods to obtain a reliable ADC value are detailed.

P591Insights from Exercise Cardiac Magnetic Resonance Imaging (ExMR) With Cardiopulmonary Testing in HIV Patients

Background

Cardiac magnetic resonance imaging in conjunction with exercise (exMR) and cardiopulmonary testing (CPET) to evaluate cardiopulmonary function has the potential of uncovering poorly understood mechanisms of dyspnea in patients that cannot be otherwise discerned with a single imaging modality.

Purpose

We prospectively evaluated the value of this technique in dyspneic patients with HIV without obvious cardiopulmonary etiologies to comprehensively assess mechanisms of dyspnea.

Methods

Thirty-six HIV patients with dyspnea without obvious cardiopulmonary causes [normal chest x-ray, normal resting LVEF and resting pulmonary artery (PA) pressures <40 mm Hg on echo] and exercise limitation [Modified Medical Research Council (MMRC) dyspnea scale >2] underwent testing using a novel combined exMR/CPET platform (Figure 1). Resting and exercise magnetic resonance imaging (MRI), including flow velocities across the pulmonary artery together with CPET was evaluated at pre and post peak stress. We analyzed the correlation between CPET and MRI pre and peak exercise variables.

Results

The mean age was 51 years; 60% were male; and mean absolute CD4 count was 718. Ventilator efficiency at peak exercise (Peak VO2) did not correlate with exMR parameters. Peak VE/VCO2 was negatively associated with peak LV and RV stroke volume (p values 0.002 & 0.005 respectively). There was a positive relationship between peak petCO2and LV cardiac output (p=0.02), peak exercise RV stroke volume (p=0.003), peak exercise LV stroke volume (p=0.02). Absolute CD4 count was positively correlated with post exercise pulmonary artery velocity (p=0.045). We found no correlation between absolute CD8 count and CMR and CPET exercise parameters.

Figure 1. MRI-CPET setting

Conclusion

HIV patients appear to have an impairment of ventilatory efficiency, exemplified by the association of VE/VCO2and PetCO2 with exMR parameters in the context of normal RV/LV contractile reserve and no evidence of Pulmonary arterial hypertension. A combined CPET/exMRI platform may provide new insight into cardiopulmonary function and the factors influencing exercise tolerance in symptomatic HIV patients.

Acknowledgement/Funding

7R01HL125060-03

Specialist and Generalist Fungal Parasites Induce Distinct Biochemical Changes in the Mandible Muscles of Their Host

Some parasites have evolved the ability to adaptively manipulate host behavior. One notable example is the fungus Ophiocordyceps unilateralis sensu lato, which has evolved the ability to alter the behavior of ants in ways that enable fungal transmission and lifecycle completion. Because host mandibles are affected by the fungi, we focused on understanding changes in the metabolites of muscles during behavioral modification. We used High-Performance Liquid Chromatography-Mass/Mass (HPLC-MS/MS) to detect the metabolite difference between controls and O. unilateralis-infected ants. There was a significant difference between the global metabolome of O. unilateralis-infected ants and healthy ants, while there was no significant difference between the Beauveria bassiana treatment ants group compared to the healthy ants. A total of 31 and 16 of metabolites were putatively identified from comparisons of healthy ants with O. unilateralis-infected ants and comparisons of B. bassiana with O. unilateralis-infected samples, respectively. This result indicates that the concentrations of sugars, purines, ergothioneine, and hypoxanthine were significantly increased in O. unilateralis-infected ants in comparison to healthy ants and B. bassiana-infected ants. This study provides a comprehensive metabolic approach for understanding the interactions, at the level of host muscles, between healthy ants and fungal parasites.

The open diffusion data derivatives, brain data upcycling via integrated publishing of derivatives and reproducible open cloud services

We describe the Open Diffusion Data Derivatives (O3D) repository: an integrated collection of preserved brain data derivatives and processing pipelines, published together using a single digital-object-identifier. The data derivatives were generated using modern diffusion-weighted magnetic resonance imaging data (dMRI) with diverse properties of resolution and signal-to-noise ratio. In addition to the data, we publish all processing pipelines (also referred to as open cloud services). The pipelines utilize modern methods for neuroimaging data processing (diffusion-signal modelling, fiber tracking, tractography evaluation, white matter segmentation, and structural connectome construction). The O3D open services can allow cognitive and clinical neuroscientists to run the connectome mapping algorithms on new, user-uploaded, data. Open source code implementing all O3D services is also provided to allow computational and computer scientists to reuse and extend the processing methods. Publishing both data-derivatives and integrated processing pipeline promotes practices for scientific reproducibility and data upcycling by providing open access to the research assets for utilization by multiple scientific communities.

Primary cardiac epithelioid angiosarcoma with frond-like features: a rare and ominous radiological mimicker of benign cardiac tumors

Most primary cardiac tumors are benign neoplasms, which generally can be differentiated from malignant neoplasms via certain radiological features. We present briefly a case of a 26-year-old man undergoing resection of a right atrial mass that based on preceding radiologic findings represent a myxoma. After pathologic examination, the lesion was determined to be an epithelioid angiosarcoma with unique frond-like architecture and multiple pedicular attachments to the atrial wall.