A high-fidelity, dual site-specific integration system in CHO cells by a Bxb1 recombinase

Site-specific integration (SSI) via recombinase mediated cassette exchange (RMCE) has shown advantages over random integration methods for expression of biotherapeutics. As an extension of our previous work developing SSI host cells, we developed a dual-site SSI system having two independent integration sites at different genomic loci, each containing a unique landing pad (LP). This system was leveraged to generate and compare two RMCE hosts, one (dFRT) compatible with the Flp recombinase, the other (dBxb1) compatible with the Bxb1 recombinase. Our comparison demonstrated that the dBxb1 host was able to generate stable transfectant pools in a shorter time frame, and cells within the dBxb1 transfectant pools were more phenotypically and genotypically stable. We further improved process performance of the dBxb1 host, resulting in desired fed batch performance attributes. Clones derived from this improved host (referred as 41L-11) maintained stable expression profiles over extended generations. While the data represents a significant improvement in the efficiency of our cell line development process, the dual LP architecture also affords a high degree of flexibility for development of complex protein modalities.

Using technology to reduce critical deterioration (the DETECT study): a cost analysis of care costs at a tertiary children's hospital in the United Kingdom

BACKGROUND

Electronic early warning systems have been used in adults for many years to prevent critical deterioration events (CDEs). However, implementation of similar technologies for monitoring children across the entire hospital poses additional challenges. While the concept of such technologies is promising, their cost-effectiveness is not established for use in children. In this study we investigate the potential for direct cost savings arising from the implementation of the DETECT surveillance system.

METHODS

Data were collected at a tertiary children's hospital in the United Kingdom. We rely on the comparison between patients in the baseline period (March 2018 to February 2019) and patients in the post-intervention period (March 2020 to July 2021). These provided a matched cohort of 19,562 hospital admissions for each group. From these admissions, 324 and 286 CDEs were observed in the baseline and post-intervention period, respectively. Hospital reported costs and Health Related Group (HRG) National Costs were used to estimate overall expenditure associated with CDEs for both groups of patients.

RESULTS

Comparing post-intervention with baseline data we found a reduction in the total number of critical care days, driven by an overall reduction in the number of CDEs, however without statistical significance. Using hospital reported costs adjusted for the Covid-19 impact, we estimate a non-significant reduction of total expenditure from £16.0 million to £14.3 million (corresponding to £1.7 million of savings - 11%). Additionally, using HRG average costs, we estimated a non-significant reduction of total expenditure from £8.2 million to £ 7.2 million (corresponding to £1.1 million of savings - 13%).

DISCUSSION AND CONCLUSION

Unplanned critical care admissions for children not only impose a substantial burden on patients and families but are also costly for hospitals. Interventions aimed at reducing emergency critical care admissions can be crucial to contribute to the reduction of these episodes' costs. Even though cost reductions were identified in our sample, our results do not support the hypothesis that reducing CDEs using technology leads to a significant reduction on hospital costs.

TRIAL REGISTRATION

Current Controlled Trials ISRCTN61279068, date of registration 07/06/2019, retrospectively registered.

Effect of subclinical endometritis and flunixin meglumine administration on pregnancy in embryo recipient beef cows

Fertility of recipient beef cows with subclinical endometritis (SCE) that did or did not receive flunixin meglumine (FM) treatment were compared following transfer of d 7 embryo. The study population comprised of 600 Angus cross cows that expressed estrus following Select-Synch + CIDR (Controlled Internal Drug Release) estrus synchronization protocol. At the time of embryo transfer, approximately 3 wk after sampling for subclinical endometritis, cows were randomly allocated either to receive FM treatment (500 mg of Banamine®; n = 300) or not (Control; n = 300). The effect of subclinical endometritis (at ≥ 1% PMN on endometrial cytology by cytobrush method) and FM treatment on pregnancy/embryo transfer (P/ET, %) were evaluated by mixed model. Of the 600 cows, 323 (53.8%) became pregnant; 55.0% (165/300) cows that received FM treatment vs. 52.7% (158/300) control cows (P > 0.1), and 55.9% (266/476) normal vs. 46.0% (57/124) subclinical endometritis cows (P < 0.05). There was a trend for treatment by subclinical endometritis for P/ET (P = 0.09). Pregnancy was recorded in 55.3% (134/242) of normal and 53.4% (31/58) of subclinical endometritis cows that received FM treatment, and in 56.4% (132/234) of normal and 39.4% (26/66) of subclinical endometritis cows that did not receive FM treatment (P = 0.09). In conclusion, subclinical endometritis in recipient beef cows resulted in lower P/ET. Though not significant in cows with subclinical endometritis, FM treatment resulted in 14.0% points more pregnancy compared with control.

International Interlaboratory Comparison of Thermogravimetric Analysis of Graphene-Related Two-Dimensional Materials

Research on graphene-related two-dimensional (2D) materials (GR2Ms) in recent years is strongly moving from academia to industrial sectors with many new developed products and devices on the market. Characterization and quality control of the GR2Ms and their properties are critical for growing industrial translation, which requires the development of appropriate and reliable analytical methods. These challenges are recognized by International Organization for Standardization (ISO 229) and International Electrotechnical Commission (IEC 113) committees to facilitate the development of these methods and standards which are currently in progress. Toward these efforts, the aim of this study was to perform an international interlaboratory comparison (ILC), conducted under Versailles Project on Advanced Materials and Standards (VAMAS) Technical Working Area (TWA) 41 "Graphene and Related 2D Materials" to evaluate the performance (reproducibility and confidence) of the thermogravimetric analysis (TGA) method as a potential new method for chemical characterization of GR2Ms. Three different types of representative and industrially manufactured GR2Ms samples, namely, pristine few-layer graphene (FLG), graphene oxide (GO), and reduced graphene oxide (rGO), were used and supplied to ILC participants to complete the study. The TGA method performance was evaluated by a series of measurements of selected parameters of the chemical and physical properties of these GR2Ms including the number of mass loss steps, thermal stability, temperature of maximum mass change rate (T_p) for each decomposition step, and the mass contents (%) of moisture, oxygen groups, carbon, and impurities (organic and non-combustible residue). TGA measurements determining these parameters were performed using the provided optimized TGA protocol on the same GR2Ms by 12 participants across academia, industry stakeholders, and national metrology institutes. This paper presents these results with corresponding statistical analysis showing low standard deviation and statistical conformity across all participants that confirm that the TGA method can be satisfactorily used for characterization of these parameters and the chemical characterization and quality control of GR2Ms. The common measurement uncertainty for each parameter, key contribution factors were identified with explanations and recommendations for their elimination and improvements toward their implementation for the development of the ISO/IEC standard for chemical characterization of GR2Ms.

Fourier transform infrared spectroscopy as a non-destructive method for analysing herbarium specimens

Dried plant specimens stored in herbaria are an untapped treasure chest of information on environmental conditions, plant evolution and change over many hundreds of years. Owing to their delicate nature and irreplaceability, there is limited access for analysis to these sensitive samples, particularly where chemical data are obtained using destructive techniques. Fourier transform infrared (FTIR) spectroscopy is a chemical analysis technique which can be applied non-destructively to understand chemical bonding information and, therefore, functional groups within the sample. This provides the potential for understanding geographical, spatial and species-specific variation in plant biochemistry. Here, we demonstrate the use of mid-FTIR microspectroscopy for the chemical analysis of Drosera rotundifolia herbarium specimens, which were collected 100 years apart from different locations. Principal component and hierarchical clustering analysis enabled differentiation between three main regions on the plant (lamina, tentacle stalk and tentacle head), and between the different specimens. Lipids and protein spectral regions were particularly sensitive differentiators of plant tissues. Differences between the different sets of specimens were smaller. This study demonstrates that relevant information can be extracted from herbarium specimens using FTIR, with little impact on the specimens. FTIR, therefore, has the potential to be a powerful tool to unlock historic information within herbaria.

Letter to the editor on “Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003–2017, parts 1–3”

We would like to thank authors Reece and Hulse (2022) for their three-part article titled “Geotemporospatial and causal inferential epidemiological overview and survey of USA cannabis, cannabidiol and cannabinoid genotoxicity expressed in cancer incidence 2003-2017”, in which the authors infer that cannabis use has a causal role in the development of various cancer types. While the authors use reputable datasets and a well-established epidemiological methodology, the authors’ conclusion of a causal association is limited due to biases inherent in ecological epidemiological studies. Though the researchers attempt to overcome these biases through validation and statistical manipulations, their approaches are insufficient to create conditions suitable for causal inferencing upon examination. There are also concerns in the practical and conceptual application of the studies’ dataset that further question the validity of the authors’ inferences. Further research exploring the potential benefits and harm of cannabinoids in the context of cancer must be performed before a distinct relationship can be defined.

Global hypo-methylation in a proportion of glioblastoma enriched for an astrocytic signature is associated with increased invasion and altered immune landscape

We describe a subset of glioblastoma, the most prevalent malignant adult brain tumour, harbouring a bias towards hypomethylation at defined differentially methylated regions. This epigenetic signature correlates with an enrichment for an astrocytic gene signature, which together with the identification of enriched predicted binding sites of transcription factors known to cause demethylation and to be involved in astrocytic/glial lineage specification, point to a shared ontogeny between these glioblastomas and astroglial progenitors. At functional level, increased invasiveness, at least in part mediated by SRPX2, and macrophage infiltration characterise this subset of glioblastoma.

Implication of cell culture methods and biases on UV inactivation of viruses

Inactivation of human respiratory viruses in air and on surfaces is important to control their spread. Exposure to germicidal ultraviolet (UV-C) light damages viral nucleic acid rendering them non-infectious. Most of the recent viral inactivation studies have not considered potential artifacts caused by interactions between UV-C light and culture media used to suspend and deposit virus on surfaces. We show that the reactive oxygen and nitrogen species (ROS and RNS) form when commonly used virus culture media is exposed to 265 nm irradiation from light emitting diodes (LEDs) at UV-C doses (4 or 40 mJ/cm²) commonly considered to achieve multiple log-inactivation of virus. Surface viral inactivation values were enhanced from 0.49 to 2.92 log₁₀ of viruses in DMEM, EMEM or EMEM-F as compared to absence of culture media (only suspended in Tris-buffer). The mechanisms responsible for the enhanced surface inactivate is hypothesized to involve photo-activation of vitamins and dyes present in the culture media, deposited with the virus on surfaces to be disinfected, which produce ROS and RNS. Given the rapidly growing research and commercial markets for UV-C disinfecting devices, there is a need to establish surface disinfecting protocols that avoid viral inactivation enhancement artifacts associated with selection and use of common cell culture media in the presence of UV-C light. This study addresses this weak link in the literature and highlights that inadequate selection of virus suspension media may cause a bias (i.e., over-estimation) for the UV-C dosages required for virus inactivation on surfaces.

OA31 Successes and challenges in harmonising 4 national Juvenile Idiopathic Arthritis cohorts: an example from CLUSTER consortium

Introduction/Background

The CLUSTER consortium aims to identify biomarkers and strata that improve personalised treatments for JIA/JIA-uveitis. By bringing together knowledge and data, CLUSTER can conduct novel analyses in this rare, heterogeneous disease. Data harmonisation across existing JIA cohorts facilitates new, larger datasets that would otherwise take years to collect; however, challenges exist as datasets are often collected autonomously. Here we present progress towards a large-scale, unique JIA data resource, bringing together treatment data from 4 real-world JIA treatment studies.

Description/Method

Four studies (CAPS, CHARMS, BCRD and BSPAR-ETN; the latter two being part of the UK JIA Biologics register) contributed data into CLUSTER.

We created two clinical datasets of JIA patients starting first-line methotrexate (MTX) or tumour necrosis factor inhibitors (TNFi). Variables were selected based on a previously developed core dataset, accounting for different levels of granularity across studies. The same inclusion and exclusion criteria were agreed for both datasets, designed to allow for combined analysis of these.

OpenPseudonymiser software encrypted NHS numbers - these were matched cross-study to identify duplicates and checked against known duplicate lists. Errors in NHS numbers and existing duplicate matches were identified and corrected. Each NHS number was assigned a CLUSTER ID, meaning 1 child has the same ID across all relevant studies such that children contributing similar data across multiple studies could be identified.

Discussion/Results

A total of 7013 records (from 5435 individuals) were identified; of which 2882 (41%, corresponding to 1304 individuals) represented the same child across >1 study. 197 individuals had duplicate records within 1 study, 961 in 2 studies, 142 in 3, and 4 children had duplicate records in all 4 studies.

After removing 350 MTX and 605 TNFi duplicate entries, the final datasets contain 2899 and 2401 unique MTX and TNFi patients respectively; 1018 are in both datasets having received both treatments. Missingness across core outcome variables ranged from 10% (active joint count MTX timepoint 2) to 60% (physician VAS TNFi timepoint 2) and was not improved through combining datasets with duplicate entries. Specificity in some variables was lost to allow integration by combining data using least common denominators (e.g. ethnicity captured as Caucasian/Non-Caucasian, despite more specific categories available in some studies).

Key learning points/Conclusion

Combining data across studies has achieved dataset sizes rarely seen in JIA, which is invaluable to progressing research into personalised treatments and disease outcomes.  However, losing specificity in some variables and missingness (a known challenge in observational data) and their impact on future analyses requires further consideration. Ongoing work includes identifying patients with both clinical and biological data that can be combined for more in-depth analyses. Both datasets are available for researchers to use via the CLUSTER Consortium Data Management Committee.

Identification of a Potent, Selective, and Brain-Penetrant Rho Kinase Inhibitor and its Activity in a Mouse Model of Huntington's Disease

The Rho kinase (ROCK) pathway is implicated in the pathogenesis of several conditions, including neurological diseases. In Huntington's disease (HD), ROCK is implicated in mutant huntingtin (HTT) aggregation and neurotoxicity, and members of the ROCK pathway are increased in HD mouse models and patients. To validate this mode of action as a potential treatment for HD, we sought a potent, selective, central nervous system (CNS)-penetrant ROCK inhibitor. Identifying a compound that could be dosed orally in mice with selectivity against other AGC kinases, including protein kinase G (PKG), whose inhibition could potentially activate the ROCK pathway, was paramount for the program. We describe the optimization of published ligands to identify a novel series of ROCK inhibitors based on a piperazine core. Morphing of the early series developed in-house by scaffold hopping enabled the identification of a compound exhibiting high potency and desired selectivity and demonstrating a robust pharmacodynamic (PD) effect by the inhibition of ROCK-mediated substrate (MYPT1) phosphorylation after oral dosing.

Gyrokinetic electrostatic turbulence close to marginality in the Wendelstein 7-X stellarator

The transition from strong (fluidlike) to nearly marginal (Floquet-type) regimes of ion-temperature-gradient (ITG) driven turbulence is studied in the stellarator Wendelstein 7-X by means of numerical simulations. Close to marginality, extended (along magnetic field lines) linearly unstable modes are dominant, even in the presence of kinetic electrons, and provide a drive that results in finite turbulent transport. A total suppression of turbulence above the linear stability threshold of the ITG modes, commonly present in tokamaks and known as the "Dimits shift," is not observed. We show that this is mostly due to the peculiar radial structure of marginal turbulence, which is more localized than in the fluid case and therefore less likely to be stabilized by shearing flows.

Insensitivity of machine log files to MLC leaf backlash and effect of MLC backlash on clinical dynamic MLC motion: An experimental investigation

Purpose

Multi‐leaf‐collimator (MLC) leaf position accuracy is important for accurate dynamic radiotherapy treatment plan delivery. Machine log files have become widely utilized for quality assurance (QA) of such dynamic treatments. The primary aim is to test the sensitivity of machine log files in comparison to electronic portal imaging device (EPID)‐based measurements to MLC position errors caused by leaf backlash. The secondary aim is to investigate the effect of MLC leaf backlash on MLC leaf motion during clinical dynamic plan delivery.

Methods

The sensitivity of machine log files and two EPID‐based measurements were assessed via a controlled experiment, whereby the length of the “T” section of a series of 12 MLC leaf T‐nuts in a Varian Millennium MLC for a Trilogy C‐series type linac was reduced by sandpapering the top of the “T” to introduce backlash. The built‐in machine MLC leaf backlash test as well as measurements for two EPID‐based dynamic MLC positional tests along with log files were recorded pre‐ and post‐T‐nut modification. All methods were investigated for sensitivity to the T‐nut change by assessing the effect on measured MLC leaf positions. A reduced version of the experiment was repeated on a TrueBeam type linac with Millennium MLC.

Results

No significant differences before and after T‐nut modification were detected in any of the log file data. Both EPID methods demonstrated sensitivity to the introduced change at approximately the expected magnitude with a strong dependence observed with gantry angle. EPID‐based data showed MLC positional error in agreement with the micrometer measured T‐nut length change to 0.07 ± 0.05 mm (1 SD) using the departmental routine QA test. Backlash results were consistent between linac types.

Conclusion

Machine log files appear insensitive to MLC position errors caused by MLC leaf backlash introduced via the T‐nut. The effect of backlash on clinical MLC motions is heavily gantry angle dependent.

OP0085 CELL LINEAGE-SPECIFIC TRANSCRIPT DECONVOLUTION OF SYNOVIAL BIOPSIES FROM THE R4RA TRIAL IDENTIFIES CELL POPULATIONS ASSOCIATED WITH RESPONSE TO RITUXIMAB AND TOCILIZUMAB

Background

The R4RA trial, the first biopsy-based randomised trial in TNF-i inadequate responder patients with Rheumatoid Arthritis, showed that molecular stratification of RA synovial tissue was associated with clinical response, demonstrating that, in patients with low/absent B-cell lineage signature in synovial-tissue, tocilizumab is superior to rituximab1.

Objectives

Here, we aimed to perform cell-transcript deconvolution of pre-and post-treatment synovial biopsies from the R4RA trial.

Methods

A total of 164 patients underwent pre-treatment synovial biopsy (US-guided or arthroscopic) prior to randomization 1:1 to rituximab (83) or tocilizumab (81). 65 patients had a repeat biopsy at 16 weeks when clinical response was assessed using Clinical Disease Activity Index (CDAI) 50% improvement. RNA extracted from a minimum of 6 synovial samples/patient underwent RNA-sequencing and the abundance of tissue-infiltrating immune and stromal cell populations was estimated using the Microenvironment Cell Populations-counter (MCP-counter) method (Figure 1a).

Results

At baseline, while synovial semiquantitative immunohistochemistry scores did not differ between CDAI50% responders and non-responders, both for rituximab and tocilizumab, MCP-counter analysis showed significantly higher CD8 T-cells in responders to rituximab and higher macrophage-monocytes and myeloid dendritic cells (mDC) in responders to tocilizumab (Figure 1b). Moreover, when patients were classified according to MCP-counter scores, B-cell poor patients (MCP-counter B cell score <median value) showed significantly higher response rates to tocilizumab, while no difference was found in B-cell rich patients (Figure 1c). In contrast, macrophage and myeloid dendritic cell (mDC) rich individuals showed higher responses to tocilizumab (Figure 1d). Combined scores for lymphoid and myeloid cells demonstrated that patients poor in B-cells but rich in macrophages/mDC had a significantly higher response to tocilizumab (77% responders to tocilizumab vs 14% responders to rituximab, p=0.017, OR 16.48, 95%CI 1.29-1000.5) (Figure 1e). By analysing disease activity over time from baseline to week 16, we found a statistically significant interaction effect between treatments and time in B-cell poor (p=0.003), T-cell poor (p=0.022), mDC rich (p=0.029) and B-cell poor/Macrophages-mDC rich patients (p=0.006) (Figure 1f-g-h). Finally, by applying MCP-counter on matched pre-and post-treatment biopsies, rituximab-treated patients showed a significant reduction of B-cells, T-cells and monocyte/macrophages, while tocilizumab-treated patients showed a significant reduction of monocyte/macrophages, T-cells, but also neutrophils, myeloid dendritic cells and, interestingly, an increase in fibroblast signature (Figure 1i).

Conclusion

In silico deconvolution of the synovial tissue identify pre-treatment lymphoid cell lineages associated with response to rituximab and myeloid cells for tocilizumab. The longitudinal analysis of matched pre- and post-treatment synovial biopsies indicated that both medications have an effect on synovial immune cells, but tocilizumab can also affect stromal cells.

References

[1]Humby et al. Rituximab versus tocilizumab in anti-TNF inadequate responder patients with rheumatoid arthritis (R4RA): 16-week outcomes of a stratified, biopsy-driven, multicentre, open-label, phase 4 randomised controlled trial Lancet. 2021 Jan 23;397(10271):305-317. doi: 10.1016/S0140-6736(20)32341-2.

Acknowledgements

We would like to thank all patients and the R4RA recruiting centres and principal investigators http://www.r4ra-nihr.whri.qmul.ac.uk/recruiting_centres.php We would also like to acknowledge the UK National Institute of Health Research for funding the R4RA trial (grant reference: 11/100/76) and Versus Arthritis for providing infrastructure support through the Experimental Arthritis Treatment Centre (grant number: 20022).

Disclosure of Interests

None declared.

OP0109 EXPRESSION QUANTITATIVE TRAIT LOCI IN EARLY TREATMENT-NAÏVE RHEUMATOID ARTHRITIS

Background

To date more than 100 genetic loci have been associated with rheumatoid arthritis (RA), particularly in the human leukocyte antigen (HLA) region. Our understanding of the functional consequences of genetic variation in RA causality, however, is limited and it has been shown that a substantial portion of complex disease risk alleles modify gene expression in a cell-specific manner [1]. The Pathobiology of Early Arthritis Cohort (PEAC) is a longitudinal study looking at treatment-naiv̈e RA patients with genotyped data as well as both synovial and blood RNA-sequenced biopsies prior to treatment with disease modifying anti-rheumatic drugs (DMARDs).

Objectives

To explore expression quantitative trait loci (eQTL) in synovium and blood within PEAC and characterise the effects of genetic variation on gene expression measured by RNA-sequencing. A further goal was to investigate the role of these variants in RA disease severity and response variables.

Methods

Genotypes were generated by Illumina Human CoreExome-24 version 1-0 array in 118 RA patients. Single nucleotide polymorphisms (SNPs) in the HLA region were imputed using HLA-TAPAS. A candidate gene study was performed on variants within the HLA region using Plink v2.0. Synovial (n=85) and blood (n=51) RNA-sequenced samples then underwent cis-eQTL analysis (loci within ±5x105Mb of the variant) based on linear regression models with the matrixeQTL R package using PEER [2] and PCA eigenvectors as covariates. Differences in eQTL between tissues were determined using a linear interaction term.

Results

The candidate gene study determined several amino acids around HLA-DRB1 acting as markers for seropositivity, which replicated findings by Raychaudhuri et. al. [3]. Using eQTL analysis, around 33,000 synovial SNPs were found with genome-wide significance (p ≤ 5x10-8) and around 29,000 in blood. This corresponded to 279 unique significant genes in synovium and 417 in blood (Figure 1). There were 100 genes common to both synovium and blood, including PSORS1C3, HLA-DRB9 and ERAP2, which have known associations with autoimmune diseases and inflammatory arthritis. Notably, 92 genes showed significantly different patterns of QTL expression between synovial tissue and blood (p ≤ 5x10-8). eQTL data also confirmed the triad of genetic variants significantly driving tissue gene expression of HLA-DPB2, while both HLA-DPB2 SNPs and HLA-DPB2 RNA-sequencing synovial expression correlated highly with erythrocyte sedimentation rate (ESR).

Figure 1.

Manhattan plots for cis-expression quantitative trait loci (eQTL) analysis performed on 85 synovial samples (top) and 51 blood samples (middle). Tissue interaction eQTL (bottom) show significant differences between tissues (p ≤ 5x10-8).

Conclusion

The high significance of genes in the HLA region in both tissues is in-keeping with the strong association between HLA and susceptibility to RA, as well as other autoimmune diseases. Most notably variants linked to HLA-DPB2 synovial expression were found to be a marker for disease severity through ESR association. Additionally, the significant differences between eQTL in blood and synovium highlight the need to explore functional consequences of genetic associations in the diseased tissue directly.

References

[1]Thalayasingam et. al. (2018). CD4+ and B lymphocyte expression quantitative traits at rheumatoid arthritis risk loci in patients with untreated early arthritis: implications for causal gene identification. Arthritis & Rheumatology, 70(3), 361-370.

[2]Stegle et al. (2012). Using probabilistic estimation of expression residuals (PEER) to obtain increased power and interpretability of gene expression analyses. Nature protocols, 7(3), 500-507.

[3]Raychaudhuri et al. (2012). Five amino acids in three HLA proteins explain most of the association between MHC and seropositive rheumatoid arthritis. Nature genetics, 44(3), 291-296.

Disclosure of Interests

None declared

OP0012 INTERFERON-α MEDIATED THERAPEUTIC RESISTANCE IN EARLY RA IMPLICATES EPIGENETIC REPROGRAMMING

Background

An interferon gene signature (IGS) is present in approximately 50% of early, treatment naive rheumatoid arthritis (eRA) patients. We previously demonstrated it negatively impacts on initial disease outcomes.

Objectives

To 1) reproduce previous findings demonstrating the harmful effects of the IGS on early RA clinical outcomes, 2) identify which IFN class is responsible for the IGS and 3) seek evidence that IFN-α exposure contributes to harmful epigenetic footprint at disease onset.

Methods

In a large multicentre inception cohort (n=190) of eRA patients (RA-MAP TACERA) whole blood transcriptome, IGS (MxA, IFI44L, OAS1, ISG15, IFI6) and circulating interferons (IFN)-α, -β, -γ and -λ was examined at baseline and 6 months in conjunction with disease activity and clinical characteristics. A separate eRA cohort of paired methylome and transcriptome from CD4 T and CD19 B cells (n=41 for each) was used to explore any epigenetic influence of the IGS.

Results

The baseline IGS reproducibly and significantly negatively impacts on 6-month clinical outcomes. In the high IGS cohort there was increased DAS-28 (p=0.025) and reduced probability of achieving a good EULAR response (p=0.034) at 6-months. In addition, the IGS in eRA is shown for the first time to predominantly reflect raised circulating IFN-α protein, not other classes of IFN and examination of whole blood upstream nucleic acid sensors expression suggest a RNA trigger. Both the IGS and IFN-α significantly fell in parallel at 6 months (p<0.0001), whereas other classes of IFN remained statistically static. There was a significant association with IFN-α and RF titre but not ACPA. Comparison of CD4 T and CD19 B cells between IGS high and low eRA patients demonstrated differentially methylated CPG sites and altered transcript expression of disease relevant genes e.g. PARP9, STAT1, EPTSI1 which was similarly, and persistently, altered 6 months in the separate TACERA cohort. Differentially methylated CPGs implicated altered transcription factor binding in B cells (GATA3, ETSI, NFATC2, EZH2) and T cells (p300, HIF1α) which cumulatively suggested IFN-α induced epigenetic changes promoting increased, and sustained, lymphocyte activation, proliferation and loss of anergy in the IGS high cohort.

Conclusion

We validate that the IGS is a robust prognostic biomarker in eRA predicting poor therapeutic response. Its persistent harmful effects may be driven via epigenetic modifications. These data have relevance for other IFN-α states, such as COVID-19, but also provide a rationale for the initial therapeutic targeting of IFN-α signalling, such as with JAKi, at disease onset in stratified eRA subsets.

References

nil

Acknowledgements

JDI is a National Institute for Health Research (NIHR) Senior Investigator. The authors acknowledge the support of TACERA Principal Investigators from all contributing NHS sites and the members of the TACERA Study Steering and Data Monitoring Committee. Additional acknowledgements include patient volunteers and administrative support from Ben Hargreaves. Newcastle researchers received infrastructural support via the Versus Arthritis Research into Inflammatory Arthritis Centre (Ref 22072), funding from The Medical Research Council; Academy of Medical Sciences; British Society of Rheumatology; The Wellcome Trust; JGW Patterson Foundation; Immune-Mediated Inflammatory Disease Biobank in the UK (IMID-Bio-UK), ANR and RTCure. This work was supported by the NIHR Newcastle Biomedical Research Centre at Newcastle Hospitals Foundation Trust and Newcastle University; views expressed are the authors’ and not necessarily those of the National Health Service, the National Institute of Health Research or the Department of Health.

Disclosure of Interests

Faye Cooles Speakers bureau: Astrazeneca: December 2021, Jessica Tarn: None declared, Dennis Lendrem: None declared, Najib Naamane: None declared, Alice Lin: None declared, Ben Millar: None declared, Nicola Maney: None declared, Nishanthi Thalayasingam: None declared, Vincent Bondet: None declared, Darragh Duffy: None declared, Michael Barnes: None declared, Graham Smith: None declared, Sandra Ng: None declared, David Watson: None declared, Rafael Henkin: None declared, Andrew Cope: None declared, Louise Reynard: None declared, Arthur Pratt: None declared, RA-MAP Consortium: None declared, John Isaacs Speakers bureau: speaker/consulting fees from AbbVie, Gilead, Roche and UCB., Grant/research support from: JDI discloses research grants from Pfizer, Janssen and GSK.

OA19 Successes and challenges in harmonising four national juvenile idiopathic arthritis cohorts: an example from CLUSTER consortium

Background/Aims

The CLUSTER consortium aims to identify biomarkers and strata that improve personalised treatments for JIA/JIA-uveitis. By bringing together knowledge and data, CLUSTER can conduct novel analyses in this rare, heterogeneous disease. Data harmonisation across existing JIA cohorts facilitates new, larger datasets that would otherwise take years to collect; however, challenges exist as datasets are often collected autonomously. Here we present progress towards a large-scale, unique JIA data resource, bringing together treatment data from four real-world JIA treatment studies.

Methods

Four studies (CAPS, CHARMS, BCRD and BSPAR-ETN; the latter two being part of the UK JIA Biologics register) contributed data into CLUSTER.

We created two clinical datasets of JIA patients starting first-line methotrexate (MTX) or tumour necrosis factor inhibitors (TNFi). Variables were selected based on a previously developed core dataset, accounting for different levels of granularity across studies. The same inclusion and exclusion criteria were agreed for both datasets, designed to allow for combined analysis of these.

OpenPseudonymiser software encrypted NHS numbers - these were matched cross-study to identify duplicates and checked against known duplicate lists. Errors in NHS numbers and existing duplicate matches were identified and corrected. Each NHS number was assigned a CLUSTER ID, meaning one child has the same ID across all relevant studies such that children contributing similar data across multiple studies could be identified.

Results

A total of 7013 records (from 5435 individuals) were identified, of which 2882 (41%, corresponding to 1304 individuals) represented the same child across &gt;1 study. 197 individuals had duplicate records within one study, 961 in two studies, 142 in three, and four children had duplicate records in all four studies.

After removing 350 MTX and 605 TNFi duplicate entries, the final datasets contain 2899 and 2401 unique MTX and TNFi patients respectively; 1018 are in both datasets having received both treatments. Missingness across core outcome variables ranged from 10% (active joint count MTX timepoint 2) to 60% (physician VAS TNFi timepoint 2) and was not improved through combining datasets with duplicate entries. Specificity in some variables was lost to allow integration by combining data using least common denominators (e.g. ethnicity captured as Caucasian/Non-Caucasian, despite more specific categories available in some studies).

Conclusion

Combining data across studies has achieved dataset sizes rarely seen in JIA, which is invaluable to progressing research into personalised treatments and disease outcomes. However, losing specificity in some variables and missingness (a known challenge in observational data) and their impact on future analyses requires further consideration. Ongoing work includes identifying patients with both clinical and biological data that can be combined for more in-depth analyses. Both datasets are available for researchers to use via the CLUSTER Consortium Data Management Committee.

Disclosure

S. Lawson-Tovey: None. L.R. Wedderburn: Consultancies; L.W. reports consulting fees from Pfizer unrelated to this work. Grants/research support; CLUSTER consortium receives support from AbbVie, UCB, Pfizer, Sobi and GSK. N. Geifman: None. M. Barnes: None. K.L. Hyrich: Grants/research support; KLH reports grant income from BMS, UCB, and Pfizer. Other; KLH reports non-personal speaker's fees from Abbvie.

Comorbidities in the UK Primary Sjögren’s Syndrome Registry

Introduction

Primary Sjögren’s Syndrome (PSS) is a chronic disease characterised by symptoms of oral and ocular dryness, pain, fatigue, anxiety and depression. PSS patients can be subclassified by the pattern of severity of these five key symptoms using the Newcastle Sjögren’s Stratification Tool (NSST). Although PSS is often associated with one or more comorbidities, the relationship between comorbidities, polypharmacy, and PSS symptom burden is unclear. Using data from the UK Primary Sjögren’s Syndrome Registry (UKPSSR) we describe the landscape of polypharmacy and comorbidities in PSS.

Methods

The UKPSSR is research biobank of clinically well-defined PSS patients where clinical, demographic, comorbidities and concomitant medications data are recorded. Patients were subclassified into the four NSST subgroups: Low Symptom Burden (LSB), High Symptom Burden (HSB), Dryness Dominated Fatigue (DDF) and Pain Dominated Fatigue (PDF). Group analyses of comorbid conditions and polypharmacy scores were performed. Comorbidity and Polypharmacy Scores (CPS) were modelled as a function of age, sex, symptom duration, body mass index (BMI), current immunosuppressant and hydroxychloroquine prescriptions and NSST subgroup.

Results

There were marked differences in the number and the nature of comorbidities associated with the NSST subgroups. LSB and DDF patients were characterized by fewer comorbidities and medications. In contrast, HSB and PDF patients were associated with more comorbidities and were more likely to be prescribed multiple medications. Group analysis shows that HSB patients are more closely associated with peripheral vascular disease and infection whereas the PDF patients were associated with cardiovascular disease and gastrointestinal comorbidities. Comorbidity and polypharmacy scores increase with age and BMI regardless of symptom subgroup and symptom duration. In addition, the longer the reported symptom duration the higher the associated comorbidities and polypharmacy scores.

Conclusion

Comorbid conditions are more prevalent in some subgroups of the PSS cohort but increase with age and BMI across the entire cohort. It is unclear from these data whether specific comorbid conditions are a consequence of PSS or represent shared aetiology or pathogenetic susceptibility. Regardless, these findings may have implications for disease management and clinical trial design.

COVID-19 Health Beliefs Regarding Mask-Wearing and Vaccinations on Twitter: A Deep Learning Approach (Preprint)

Background

Amid the global COVID-19 pandemic, a worldwide infodemic also emerged with large amounts of COVID-19–related information and misinformation spreading through social media channels. Various organizations, including the World Health Organization (WHO) and the Centers for Disease Control and Prevention (CDC), and other prominent individuals issued high-profile advice on preventing the further spread of COVID-19.

Objective

The purpose of this study is to leverage machine learning and Twitter data from the pandemic period to explore health beliefs regarding mask wearing and vaccines and the influence of high-profile cues to action.

Methods

A total of 646,885,238 COVID-19–related English tweets were filtered, creating a mask-wearing data set and a vaccine data set. Researchers manually categorized a training sample of 3500 tweets for each data set according to their relevance to Health Belief Model (HBM) constructs and used coded tweets to train machine learning models for classifying each tweet in the data sets.

Results

In total, 5 models were trained for both the mask-related and vaccine-related data sets using the XLNet transformer model, with each model achieving at least 81% classification accuracy. Health beliefs regarding perceived benefits and barriers were most pronounced for both mask wearing and immunization; however, the strength of those beliefs appeared to vary in response to high-profile cues to action.

Conclusions

During both the COVID-19 pandemic and the infodemic, health beliefs related to perceived benefits and barriers observed through Twitter using a big data machine learning approach varied over time and in response to high-profile cues to action from prominent organizations and individuals.

Investigation of the Utility of Features in a Clinical De-identification Model: A Demonstration Using EHR Pathology Reports for Advanced NSCLC Patients

Background

Electronic health record (EHR) systems contain a large volume of texts, including visit notes, discharge summaries, and various reports. To protect the confidentiality of patients, these records often need to be fully de-identified before circulating for secondary use. Machine learning (ML) based named entity recognition (NER) model has emerged as a popular technique of automatic de-identification.

Objective

The performance of a machine learning model highly depends on the selection of appropriate features. The objective of this study was to investigate the usability of multiple features in building a conditional random field (CRF) based clinical de-identification NER model.

Methods

Using open-source natural language processing (NLP) toolkits, we annotated protected health information (PHI) in 1,500 pathology reports and built supervised NER models using multiple features and their combinations. We further investigated the dependency of a model's performance on the size of training data.

Results

Among the 10 feature extractors explored in this study, n-gram, prefix–suffix, word embedding, and word shape performed the best. A model using combination of these four feature sets yielded precision, recall, and F1-score for each PHI as follows: NAME (0.80; 0.79; 0.80), LOCATION (0.85; 0.83; 0.84), DATE (0.86; 0.79; 0.82), HOSPITAL (0.96; 0.93; 0.95), ID (0.99; 0.82; 0.90), and INITIALS (0.97; 0.49; 0.65). We also found that the model's performance becomes saturated when the training data size is beyond 200.

Conclusion

Manual de-identification of large-scale data is an impractical procedure since it is time-consuming and subject to human errors. Analysis of the NER model's performance in this study sheds light on a semi-automatic clinical de-identification pipeline for enterprise-wide data warehousing.

Predicting suicidal thoughts and behavior among adolescents using the risk and protective factor framework: A large-scale machine learning approach

INTRODUCTION

Addressing the problem of suicidal thoughts and behavior (STB) in adolescents requires understanding the associated risk factors. While previous research has identified individual risk and protective factors associated with many adolescent social morbidities, modern machine learning approaches can help identify risk and protective factors that interact (group) to provide predictive power for STB. This study aims to develop a prediction algorithm for STB among adolescents using the risk and protective factor framework and social determinants of health.

METHODS

The sample population consisted of more than 179,000 high school students living in Utah and participating in the Communities That Care (CTC) Youth Survey from 2011-2017. The dataset includes responses to 300+ questions from the CTC and 8000+ demographic factors from the American Census Survey for a total of 1.2 billion values. Machine learning techniques were employed to extract the survey questions that were best able to predict answers indicative of STB, using recent work in interpretable machine learning.

RESULTS

Analysis showed strong predictive power, with the ability to predict individuals with STB with 91% accuracy. After extracting the top ten questions that most affected model predictions, questions fell into four main categories: familial life, drug consumption, demographics, and peer acceptance at school.

CONCLUSIONS

Modern machine learning approaches provide new methods for understanding the interaction between root causes and outcomes, such as STB. The model developed in this study showed significant improvement in predictive accuracy compared to previous research. Results indicate that certain risk and protective factors, such as adolescents being threatened or harassed through digital media or bullied at school, and exposure or involvement in serious arguments and yelling at home are the leading predictors of STB and can help narrow and reaffirm priority prevention programming and areas of focused policymaking.

Theoretical evaluation of a novel method for producing fluorine-18 for Positron-emission-tomography (PET) applications utilizing the 3He(d,p)4He reaction

A novel method of producing fluorine-18 for Positron emission tomography (PET) scan applications is evaluated theoretically. The method is based upon the ³He(d,p)⁴He nuclear reaction from which the protons produced are used to produce fluorine-18 via the ¹⁸O(p,n)¹⁸F reaction. The potential advantage of such a system over cyclotron-based production is of lower input beam energy, which may lower the cost of the system and potentially allow for onsite production. Two theoretical designs were investigated. The first utilizes a helium-3 beam incident on a deuterated plastic target such as Mylar which is backed with an oxygen-18 heavy-water (H₂O¹⁸) target. The second design utilizes a super-heavy-water (D₂O¹⁸) target effectively combining both targets into one. Theoretical yield calculations were performed for both designs and the practicalities, primarily those of thermal and target degradation effects were assessed. To produce sufficient fluorine-18 yield a 1 MeV helium-3 beam at 100 mA was simulated. For this beam it was calculated that 310 MBq of fluorine-18 activity, as required to scan a 74 kg patient would be generated in a 69 min or an 18 min production run for the Mylar and super-heavy-water systems respectively. The simulated beam is at 100 kW power and without significant cooling would vaporize the target materials with the melting point of Mylar and boiling point of water calculated to be breached within 0.41 μs and 0.45 μs from beam-on. To achieve sufficient fluorine-18 yield the helium-3 beam power had to be increased to impractical levels making the systems technically infeasible.

P35: BODY COMPOSITION: ASSOCIATION WITH OPERATIVE AND ONCOLOGIC OUTCOMES IN THE MODERN MANAGEMENT OF RETROPERITONEAL SARCOMA – RESULTS OF A FEASIBILITY STUDY

Introduction

Sarcopenia and obesity in cancer may confer negative outcomes, but their prevalence and impact on modern regimens for retroperitoneal sarcoma (RPS) have not been systematically studied. Sarcopaenia and obesity measurements by computed tomography (CT) measurements at L3 may be confounded by the disease process itself. The aim of this study is to determine feasibility in providing the prevalence of sarcopenia and visceral obesity in the management of RPS to assess their potential impact on operative and oncologic outcomes.

Method

Consecutive sample patients undergoing treatment for RPS from our database were retrospectively studied. Total, subcutaneous and visceral fat areas (VFA), myosteatosis, skeletal muscle index (SMI) lean body mass and fat mass were determined at diagnosis by CT. Sarcopenia will be defined by CT at L3 as SMI &lt;52.4 cm2/m2 for males and SMI &lt;38.5 cm2/m2 for females, and visceral obesity as VFA &gt;163.8cm2 for men and &gt;80.1cm2 for women.

Result

40 consecutive patients, 21 [52.5%] female, 56.7±15.1 years, were studied. The most common histologic types were leiomyosarcoma (8 [20.0%]), dedifferentiated liposarcoma (7 [17.5%]), well differentiated liposarcoma (7 [17.5%]), myxoid liposarcoma (4 [10.0%]). Mean±SD body composition measures were: lean body mass, 50.4±12.0 kg; total fat mass, 27.2±8.6 kg; visceral fat area, 148.5±120.2 cm2; subcutaneous fat area, 222.2±20.4 cm2; myosteatosis, 8.9±7.3 cm2.

Conclusion

Assessment of body composition among patients with RPS is feasible. Analysis to identify the prevalence and significance of sarcopenia and visceral obesity and its relationship to operative and oncologic outcomes is ongoing.

Take-home message

Assessment of body composition in RPS is feasible and its effect on outcomes is under investigation

Artificial Intelligence-Assisted Amphiregulin and Epiregulin IHC Predicts Panitumumab Benefit in RAS Wild-Type Metastatic Colorectal Cancer

PURPOSE

High tumor mRNA levels of the EGFR ligands amphiregulin (AREG) and epiregulin (EREG) are associated with anti-EGFR agent response in metastatic colorectal cancer (mCRC). However, ligand RNA assays have not been adopted into routine practice due to issues with analytic precision and practicality. We investigated whether AREG/EREG IHC could predict benefit from the anti-EGFR agent panitumumab.

EXPERIMENTAL DESIGN

Artificial intelligence algorithms were developed to assess AREG/EREG IHC in 274 patients from the PICCOLO trial of irinotecan with or without panitumumab (Ir vs. IrPan) in RAS wild-type mCRC. The primary endpoint was progression-free survival (PFS). Secondary endpoints were RECIST response rate (RR) and overall survival (OS). Models were repeated adjusting separately for BRAF mutation status and primary tumor location (PTL).

RESULTS

High ligand expression was associated with significant PFS benefit from IrPan compared with Ir [8.0 vs. 3.2 months; HR, 0.54; 95% confidence interval (CI), 0.37-0.79; P = 0.001]; whereas low ligand expression was not (3.4 vs. 4.4 months; HR, 1.05; 95% CI, 0.74-1.49; P = 0.78). The ligand-treatment interaction was significant (P _interaction = 0.02) and remained significant after adjustment for BRAF-mutation status and PTL. Likewise, RECIST RR was significantly improved in patients with high ligand expression (IrPan vs. Ir: 48% vs. 6%; P < 0.0001) but not those with low ligand expression (25% vs. 14%; P = 0.10; P _interaction = 0.01). The effect on OS was similar but not statistically significant.

CONCLUSIONS

AREG/EREG IHC identified patients who benefitted from the addition of panitumumab to irinotecan chemotherapy. IHC is a practicable assay that may be of use in routine practice.

Electronic Tuning of Monolayer Graphene with Polymeric "Zwitterists"

Work function engineering of two-dimensional (2D) materials by application of polymer coatings represents a research thrust that promises to enhance the performance of electronic devices. While polymer zwitterions have been demonstrated to significantly modify the work function of both metal electrodes and 2D materials due to their dipole-rich structure, the impact of zwitterion chemical structure on work function modulation is not well understood. To address this knowledge gap, we synthesized a series of sulfobetaine-based zwitterionic random copolymers with variable substituents and used them in lithographic patterning for the preparation of negative-tone resists (i.e., "zwitterists") on monolayer graphene. Ultraviolet photoelectron spectroscopy indicated a significant work function reduction, as high as 1.5 eV, induced by all polymer zwitterions when applied as ultrathin films (<10 nm) on monolayer graphene. Of the polymers studied, the piperidinyl-substituted version, produced the largest dipole normal to the graphene sheet, thereby inducing the maximum work function reduction. Density functional theory calculations probed the influence of zwitterion composition on dipole orientation, while lithographic patterning allowed for evaluation of surface potential contrast via Kelvin probe force microscopy. Overall, this polymer "zwitterist" design holds promise for fine-tuning 2D materials electronics with spatial control based on the chemistry of the polymer coating and the dimensions of the lithographic patterning.

A commercial treatment planning system with a hybrid dose calculation algorithm for synchrotron radiotherapy trials

Synchrotron Radiotherapy (SyncRT) is a preclinical radiation treatment which delivers synchrotron x-rays to cancer targets. SyncRT allows for novel treatments such as Microbeam Radiotherapy, which has been shown to have exceptional healthy tissue sparing capabilities while maintaining good tumour control. Veterinary trials in SyncRT are anticipated to take place in the near future at the Australian Synchrotron's Imaging and Medical Beamline (IMBL). However, before veterinary trials can commence, a computerised treatment planning system (TPS) is required, which can quickly and accurately calculate the synchrotron x-ray dose through patient CT images. Furthermore, SyncRT TPS's must be familiar and intuitive to radiotherapy planners in order to alleviate necessary training and reduce user error. We have paired an accurate and fast Monte Carlo (MC) based SyncRT dose calculation algorithm with Eclipse^TM, the most widely implemented commercial TPS in the clinic. Using Eclipse^TM, we have performed preliminary SyncRT trials on dog cadavers at the IMBL, and verified calculated doses against dosimetric measurement to within 5% for heterogeneous tissue-equivalent phantoms. We have also performed a validation of the TPS against a full MC simulation for constructed heterogeneous phantoms in Eclipse^TM, and showed good agreement for a range of water-like tissues to within 5%-8%. Our custom Eclipse^TM TPS for SyncRT is ready to perform live veterinary trials at the IMBL.

Medical cannabis in the UK: From principle to practice

BACKGROUND

In the UK, medical cannabis was approved in November 2018, leading many patients to believe that the medicine would now be available on the NHS. Yet, to date, there have been only 12 NHS prescriptions and less than 60 prescriptions in total. In marked contrast, a recent patient survey by the Centre for Medical Cannabis (Couch, 2020) found 1.4 m people are using illicit cannabis for medical problems.

AIMS

Such a mismatch between demand and supply is rare in medicine. This article outlines some of the current controversies about medical cannabis that underpin this disparity, beginning by contrasting current medical evidence from research studies with patient-reported outcomes.

OUTCOMES

Although definite scientific evidence is scarce for most conditions, there is significant patient demand for access to medical cannabis. This disparity poses a challenge for prescribers, and there are many concerns of physicians when deciding if, and how, to prescribe medical cannabis which still need to be addressed. Potential solutions are outlined as to how the medical profession and regulators could respond to the strong demand from patients and families for access to medical cannabis to treat chronic illnesses when there is often a limited scientific evidence base on whether and how to use it in many of these conditions.

CONCLUSIONS

There is a need to maximise both clinical research and patient benefit, in a safe, cautious and ethical manner, so that those patients for whom cannabis is shown to be effective can access it. We hope our discussion and outlines for future progress offer a contribution to this process.

A Monte Carlo model of synchrotron radiotherapy shows good agreement with experimental dosimetry measurements: Data from the imaging and medical beamline at the Australian Synchrotron

Experimental measurement of Synchrotron Radiotherapy (SyncRT) doses is challenging, especially for Microbeam Radiotherapy (MRT), which is characterised by very high dynamic ranges with spatial resolutions on the micrometer scale. Monte Carlo (MC) simulation is considered a gold standard for accurate dose calculation in radiotherapy, and is therefore routinely relied upon to produce verification data. We present a MC model for Australian Synchrotron's Imaging and Medical Beamline (IMBL), which is capable of generating accurate dosimetry data to inform and/or verify SyncRT experiments. Our MC model showed excellent agreement with dosimetric measurement for Synchrotron Broadbeam Radiotherapy (SBBR). Our MC model is also the first to achieve validation for MRT, using two methods of dosimetry, to within clinical tolerances of 5% for a 20×20 mm² field size, except for surface measurements at 5 mm depth, which remained to within good agreement of 7.5%. Our experimental methodology has allowed us to control measurement uncertainties for MRT doses to within 5-6%, which has also not been previously achieved, and provides a confidence which until now has been lacking in MRT validation studies. The MC model is suitable for SyncRT dose calculation of clinically relevant field sizes at the IMBL, and can be extended to include medical beamlines at other Synchrotron facilities as well. The presented MC model will be used as a validation tool for treatment planning dose calculation algorithms, and is an important step towards veterinary SyncRT trials at the Australian Synchrotron.

Predictive quality assurance of a linear accelerator based on the machine performance check application using statistical process control and ARIMA forecast modeling

Purpose

A predictive linac quality assurance system based on the output of the Machine Performance Check (MPC) application was developed using statistical process control and autoregressive integrated moving average forecast modeling. The aim of this study is to demonstrate the feasibility of predictive quality assurance based on MPC tests that allow proactive preventative maintenance procedures to be carried out to better ensure optimal linac performance and minimize downtime.

Method and Materials

Daily MPC data were acquired for a total of 490 measurements. The initial 85% of data were used in prediction model learning with the autoregressive integrated moving average technique and in calculating upper and lower control limits for statistical process control analysis. The remaining 15% of data were used in testing the accuracy of the predictions of the proposed system. Two types of prediction were studied, namely, one‐step‐ahead values for predicting the next day's quality assurance results and six‐step‐ahead values for predicting up to a week ahead. Results that fall within the upper and lower control limits indicate a normal stage of machine performance, while the tolerance, determined from AAPM TG‐142, is the clinically required performance. The gap between the control limits and the clinical tolerances (as the warning stage) provides a window of opportunity for rectifying linac performance issues before they become clinically significant. The accuracy of the predictive model was tested using the root‐mean‐square error, absolute error, and average accuracy rate for all MPC test parameters.

Results

The accuracy of the predictive model is considered high (average root‐mean‐square error and absolute error for all parameters of less than 0.05). The average accuracy rate for indicating the normal/warning stages was higher than 85.00%.

Conclusion

Predictive quality assurance with the MPC will allow preventative maintenance, which could lead to improved linac performance and a reduction in unscheduled linac downtime.

Key considerations and challenges in the application of social-network research for environmental decision making

Attempts to better understand the social context in which conservation and environmental decisions are made has led to increased interest in human social networks. To improve the use of social-network analysis in conservation, we reviewed recent studies in the literature in which such methods were applied. In our review, we looked for problems in research design and analysis that limit the utility of network analysis. Nineteen of 55 articles published from January 2016 to June 2019 exhibited at least 1 of the following problems: application of analytical methods inadequate or sensitive to incomplete network data; application of statistical approaches that ignore dependency in the network; or lack of connection between the theoretical base, research question, and choice of analytical techniques. By drawing attention to these specific areas of concern and highlighting research frontiers and challenges, including causality, network dynamics, and new approaches, we responded to calls for increasing the rigorous application of social science in conservation.

0665 Bilateral Hypoglossal Nerve Stimulation for Treatment of Obstructive Sleep Apnea

Introduction

Hypoglossal Nerve Stimulation (HGNS) decreases obstructive sleep apnea (OSA) severity by contracting the tongue and decreasing upper airway collapsibility. This study assessed the safety and effectiveness of a new implantable device that delivers bilateral HGNS: the Genio™ system.

Methods

The BLAST OSA study (BiLAteral Hypoglossal Nerve Stimulation for Treatment of Obstructive Sleep Apnea), was a prospective, open-label, non-randomized, single arm treatment study conducted at eight centres in three countries (Australia, France, UK). Primary outcomes were the incidence of device-related Serious Adverse Events (SAEs) and change in the Apnea-Hypopnea Index (AHI). The secondary outcome was change in the 4% Oxygen Desaturation Index (ODI). Additional outcomes included measures of sleepiness, quality of life, snoring, and device use. Participants were eligible if: 21-75 years old; BMI ≤ 32 kg/m2; obstructive AHI 20-60 events/hr and combined central and mixed AHI &lt; 10 events/hr; no positional OSA; no Complete Concentric Collapse of the soft palate during Drug Induced Sleep Endoscopy; and failed to tolerate or accept Positive Airway Pressure treatments.

Results

27 participants were implanted (63% male, aged 55.9±12.0 years, BMI 27.4±3.0 kg/m2). 22 completed the protocol. At 6 months, AHI decreased from 23.7±12.2 to 12.9±10.1 events/hr [p&lt;0.001]; and ODI decreased from 19.1±11.2 to 9.8±6.9 events/hr [p&lt;0.001]. Daytime sleepiness (ESS, p=0.011) and sleep-related quality of life (FOSQ-10, p=0.016) both significantly improved. 91% of participants reported using their device &gt;5 days per week, and 77% used it &gt;5 hours per night. The number of bed partners reporting disruptive snoring decreased from 96% to 35%. No device-related SAE occurred.

Conclusion

In a targeted population of individuals with moderate-to-severe OSA, the Genio system reduced OSA severity and sleepiness, improved quality of life, and was associated with high adherence and an acceptable safety profile.

Support

This study trial was funded by Nyxoah S.A. This trial was registered with ClinicalTrials.gov, number NCT03048604.

Report on computational assessment of Tumor Infiltrating Lymphocytes from the International Immuno-Oncology Biomarker Working Group

Assessment of tumor-infiltrating lymphocytes (TILs) is increasingly recognized as an integral part of the prognostic workflow in triple-negative (TNBC) and HER2-positive breast cancer, as well as many other solid tumors. This recognition has come about thanks to standardized visual reporting guidelines, which helped to reduce inter-reader variability. Now, there are ripe opportunities to employ computational methods that extract spatio-morphologic predictive features, enabling computer-aided diagnostics. We detail the benefits of computational TILs assessment, the readiness of TILs scoring for computational assessment, and outline considerations for overcoming key barriers to clinical translation in this arena. Specifically, we discuss: 1. ensuring computational workflows closely capture visual guidelines and standards; 2. challenges and thoughts standards for assessment of algorithms including training, preanalytical, analytical, and clinical validation; 3. perspectives on how to realize the potential of machine learning models and to overcome the perceptual and practical limits of visual scoring.

Polymer Zwitterions for Stabilization of CsPbBr3 Perovskite Nanoparticles and Nanocomposite Films

Functional polymers with sulfobetaine or phosphorylcholine zwitterions as pendent groups are demonstrated as both ligands and host matrices for CsPbBr₃ perovskite nanoparticles (PNPs). These polymers produce nanocomposite films with excellent NP dispersion, optical transparency, and impressive resistance to NP degradation upon exposure to water. Multidentate interactions of the zwitterion-containing copolymers with the PNPs induce dispersed or weakly aggregated nanocomposite morphologies, depending on the extent of zwitterionic functionality in the polymer. Incorporating additional functionality into the polymers, such as benzophenone pendent groups, yields lithographically patternable films, while time-resolved photoluminescence measurements provide insight into the electronic impact of PNPs in zwitterionic polymer matrices.

M3C: Monte Carlo reference-based consensus clustering

Genome-wide data is used to stratify patients into classes for precision medicine using clustering algorithms. A common problem in this area is selection of the number of clusters (K). The Monti consensus clustering algorithm is a widely used method which uses stability selection to estimate K. However, the method has bias towards higher values of K and yields high numbers of false positives. As a solution, we developed Monte Carlo reference-based consensus clustering (M3C), which is based on this algorithm. M3C simulates null distributions of stability scores for a range of K values thus enabling a comparison with real data to remove bias and statistically test for the presence of structure. M3C corrects the inherent bias of consensus clustering as demonstrated on simulated and real expression data from The Cancer Genome Atlas (TCGA). For testing M3C, we developed clusterlab, a new method for simulating multivariate Gaussian clusters.

Visual Immunoprecipitate Assay (VIP) for Listeria monocytogenes and Related Listeria Species Detection in Selected Foods: Collaborative Study

Six foods representing a variety of food products were analyzed by the Assurance Listeria polyclonal enzyme immunoassay (EIA) and by either the Bacteriological Analytical Manual or the U.S. Department of Agriculture culture method for detecting Listeria monocytogenes and related Listeria species. Samples of each food type, at each inoculation level, were analyzed simultaneously by both methods. A total of 19 laboratories representing federal government agencies and private industry in the United States and Canada participated. Food types were inoculated with Listeria species including L. monocytogenes, with the exception of 3 lots of green beans, which were naturally contaminated. During this study, 1764 samples and controls were analyzed and confirmed, of which 492 were positive and 947 were negative by both methods. There were 159 samples that were positive by culture method but negative by the EIA and 188 that were negative by culture method but positive by EIA. Twenty-two samples were negative by EIA and by culture method but confirmed positive when Assurance selective enrichment broths were subcultured to selective agar. The Assurance polyclonal EIA for detecting L. monocytogenes and related Listeria species in foods has been adopted first action by AOAC INTERNATIONAL.

Assurance Polyclonal Enzyme Immunoassay for Detection of Listeria monocytogenes and Related Listeria Species in Selected Foods: Collaborative Study

Six foods representing a variety of food products were analyzed by the Assurance Listeria polyclonal enzyme immunoassay (EIA) and by either the Bacteriological Analytical Manual or the U.S. Department of Agriculture culture method for detecting Listeria monocytogenes and related Listeria species. Samples of each food type, at each inoculation level, were analyzed simultaneously by both methods. A total of 19 laboratories representing federal government agencies and private industry in the United States and Canada participated. Food types were inoculated with Listeria species including L. monocytogenes, with the exception of 3 lots of green beans, which were naturally contaminated. During this study, 1764 samples and controls were analyzed and confirmed, of which 492 were positive and 947 were negative by both methods. There were 159 samples that were positive by culture method but negative by the EIA and 188 that were negative by culture method but positive by EIA. Twenty-two samples were negative by EIA and by culture method but confirmed positive when Assurance selective enrichment broths were subcultured to selective agar. The Assurance polyclonal EIA for detecting L. monocytogenes and related Listeria species in foods has been adopted first action by AOAC INTERNATIONAL.

Acceptability and feasibility of recruiting women to collect a self-administered vaginal swab at a pharmacy clinic for sexually transmissible infection screening

Willingness to self-collect vaginal swabs at a pharmacy clinic is of interest as a venue to increase sexually transmissible infections (STIs) screening for chlamydia, gonorrhoea and trichomonas. Women self-collected vaginal swabs at the pharmacy, completed questionnaires and received STI results within 2 h. Women with STIs were offered free treatment. A total of 313 of 777 (40.3%) women consented and prevalence for any STI was 3.9%. Questionnaires demonstrated acceptability for self-collection at the pharmacy, with 63% (95% CI 57.3-68) and 32.3% (95% CI 27.4-37.8) indicating they 'strongly agreed' or 'agreed' that they felt comfortable with pharmacy collection, respectively. Self-collected vaginal swabs for STI testing for women who were at a pharmacy were feasible and acceptable to women.

P959 Handheld echocardiography in a real world scenario: concordances compared to standard echo reports

Background

Handheld echocardiography devices(HH) arise as a common tool in clinical examination and screening for various cardiovascular disorders. Despite of it, studies with this method are small, with unselected patients and limited scope of diagnostic comparison.

Purpose

Assess the usefulness of the new miniaturized HH and compare its concordances with the standard high definition echocardiography study(STD).

Methods

Between April and May of 2016 adult patients who were scheduled to regular STD, were also submitted to a HH exam. Experienced sonographers performed and an experienced echocardiographer reviewed the STD exam, and an experienced echocardiographer performed and reviewed the HH study - reviewers were blinded to the other study results. STD exams included 2-dimensional images, Color Doppler and hemodynamics analysis. With the HH hemodynamic assessment was not performed as the machine does not include such technology. Agreement between the reports was analyzed.

Results

110 patients were included. Mean age was 62.4 ± 16.7 years. The κ values(Table) show good correlation between HH and STD on the analysis of left ventricular global and segment functions, right ventricle size and function, mitral and aortic stenosis. On the evaluation of left ventricle hypertrophy, mitral and aortic regurgitations the correlation was modest, while poor correlation was found for pulmonary and tricuspid regurgitations.

Conclusion

In a daily practice scenario with experienced hands, HH demonstrates good results for the assessment of ventricles size and function, while the evaluation of right heart valves was the least reliable performance. Dissemination of HH should occur with considerations and caution.

K Values for Echocardiography Analysis Echocardiography analysis K value Global estimated LV dysfunction 0.85 Wall motion abnormalities 0.78 LV hypertrophy 0.6 RV size 0.83 RV function 0.82 Mitral regurgitation 0.42 Aortic regurgitation 0.56 Mitral stenosis 0.96 Aortic stenosis 0.82 Tricuspid regurgitation 0.26 Pulmonary regurgitation 0.25 LV: left ventricle; RV: right ventricle.