Artificial Intelligence-Derived Risk Prediction: A Novel Risk Calculator Using Office and Ambulatory Blood Pressure

BACKGROUND

Quantification of total cardiovascular risk is essential for individualizing hypertension treatment. This study aimed to develop and validate a novel, machine-learning-derived model to predict cardiovascular mortality risk using office blood pressure (OBP) and ambulatory blood pressure (ABP).

METHODS

The performance of the novel risk score was compared with existing risk scores, and the possibility of predicting ABP phenotypes utilizing clinical variables was assessed. Using data from 59 124 patients enrolled in the Spanish ABP Monitoring registry, machine-learning approaches (logistic regression, gradient-boosted decision trees, and deep neural networks) and stepwise forward feature selection were used.

RESULTS

For the prediction of cardiovascular mortality, deep neural networks yielded the highest clinical performance. The novel mortality prediction models using OBP and ABP outperformed other risk scores. The area under the curve achieved by the novel approach, already when using OBP variables, was significantly higher when compared with the area under the curve of the Framingham risk score, Systemic Coronary Risk Estimation 2, and Atherosclerotic Cardiovascular Disease score. However, the prediction of cardiovascular mortality with ABP instead of OBP data significantly increased the area under the curve (0.870 versus 0.865; P=3.61×10-28), accuracy, and specificity, respectively. The prediction of ABP phenotypes (ie, white-coat, ambulatory, and masked hypertension) using clinical characteristics was limited.

CONCLUSIONS

The receiver operating characteristic curves for cardiovascular mortality using ABP and OBP with deep neural network models outperformed all other risk metrics, indicating the potential for improving current risk scores by applying state-of-the-art machine learning approaches. The prediction of cardiovascular mortality using ABP data led to a significant increase in area under the curve and performance metrics.

SingmiR: a single-cell miRNA alignment and analysis tool

Single-cell RNA sequencing (RNA-seq) has revolutionized our understanding of cell biology, developmental and pathophysiological molecular processes, paving the way toward novel diagnostic and therapeutic approaches. However, most of the gene regulatory processes on the single-cell level are still unknown, including post-transcriptional control conferred by microRNAs (miRNAs). Like the established single-cell gene expression analysis, advanced computational expertise is required to comprehensively process newly emerging single-cell miRNA-seq datasets. A web server providing a workflow tailored for single-cell miRNA-seq data with a self-explanatory interface is currently not available. Here, we present SingmiR, enabling the rapid (pre-)processing and quantification of human miRNAs from noncoding single-cell samples. It performs read trimming for different library preparation protocols, generates automated quality control reports and provides feature-normalized count files. Numerous standard and advanced analyses such as dimension reduction, clustered feature heatmaps, sample correlation heatmaps and differential expression statistics are implemented. We aim to speed up the prototyping pipeline for biologists developing single-cell miRNA-seq protocols on small to medium-sized datasets. SingmiR is freely available to all users without the need for a login at https://www.ccb.uni-saarland.de/singmir.

Experimental capture of miRNA targetomes: disease-specific 3'UTR library-based miRNA targetomics for Parkinson's disease

The identification of targetomes remains a challenge given the pleiotropic effect of miRNAs, the limited effects of miRNAs on individual targets, and the sheer number of estimated miRNA-target gene interactions (MTIs), which is around 44,571,700. Currently, targetome identification for single miRNAs relies on computational evidence and functional studies covering smaller numbers of targets. To ensure that the targetome analysis could be experimentally verified by functional assays, we employed a systematic approach and explored the targetomes of four miRNAs (miR-129-5p, miR-129-1-3p, miR-133b, and miR-873-5p) by analyzing 410 predicted target genes, both of which were previously associated with Parkinson's disease (PD). After performing 13,536 transfections, we validated 442 of the 705 putative MTIs (62,7%) through dual luciferase reporter assays. These analyses increased the number of validated MTIs by at least 2.1-fold for miR-133b and by a maximum of 24.3-fold for miR-873-5p. Our study contributes to the experimental capture of miRNA targetomes by addressing i) the ratio of experimentally verified MTIs to predicted MTIs, ii) the sizes of disease-related miRNA targetomes, and iii) the density of MTI networks. A web service to support the analyses on the MTI level is available online ( https://ccb-web.cs.uni-saarland.de/utr-seremato ), and all the data have been added to the miRATBase database ( https://ccb-web.cs.uni-saarland.de/miratbase ).

Impact of adjuvant radiotherapy and chemotherapy on thymoma

PURPOSE

Thymoma is a rare tumour. The most common treatment for thymoma is surgical resection, while the use of radiotherapy and chemotherapy remains controversial.

PATIENTS AND METHODS

We conducted a monocentric observational study of 31 patients diagnosed with thymoma from June 2004 to July 2020 at cancer centre in Strasbourg, France. We analysed the outcomes of the patients.

RESULTS

The 2- and 5- year locoregional relapse-free survival rates were 96.3% (95% confidence interval [CI]: 76.5-99.5%) and 68.0% (95% CI: 43.8-83.5%), respectively. Radiotherapy and chemotherapy significantly improved local tumour control (P=0.0008 and 0.04, respectively), while a larger initial tumour size significantly worsened local control rates (P=0.04). The 5- and 10-year overall survival rates were 87.1% (95% CI: 69.2-95%) and 81.7% (95% CI: 60.3-92.2%), respectively. The median overall survival was not reached, and no favourable factor was retrieved. For relapsed patients, the median overall survival after relapse was 115 months.

CONCLUSION

Despite the inherent limitations of retrospective studies with a limited patient sample size, we demonstrated that chemotherapy and radiotherapy in addition to surgery were effective in achieving local control and contributed to improving patient outcomes in thymoma. Notably, an aggressive treatment strategy at the time of relapse resulted in favourable outcomes for retreated patients.

Time series of chicken stool metagenomics and egg metabolomics in changing production systems: preliminary insights from a proof-of-concept

BACKGROUND

Different production systems of livestock animals influence various factors, including the gut microbiota.

METHODS

We investigated whether changing the conditions from barns to free-range chicken farming impacts the microbiome over the course of three weeks. We compared the stool microbiota of chicken from industrial barns after introducing them either in community or separately to a free-range environment.

RESULTS

Over the six time points, 12 taxa-mostly lactobacilli-changed significantly. As expected, the former barn chicken cohort carries more resistances to common antibiotics. These, however, remained positive over the observed period. At the end of the study, we collected eggs and compared metabolomic profiles of the egg white and yolk to profiles of eggs from commercial suppliers. Here, we observed significant differences between commercial and fresh collected eggs as well as differences between the former barn chicken and free-range chicken.

CONCLUSION

Our data indicate that the gut microbiota can undergo alterations over time in response to changes in production systems. These changes subsequently exert an influence on the metabolites found in the eggs. The preliminary results of our proof-of-concept study motivate larger scale observations with more individual chicken and longer observation periods.

SRF transcriptionally regulates the oligodendrocyte cytoskeleton during CNS myelination

Myelination of neuronal axons is essential for nervous system development. Myelination requires dramatic cytoskeletal dynamics in oligodendrocytes, but how actin is regulated during myelination is poorly understood. We recently identified serum response factor (SRF)-a transcription factor known to regulate expression of actin and actin regulators in other cell types-as a critical driver of myelination in the aged brain. Yet, a major gap remains in understanding the mechanistic role of SRF in oligodendrocyte lineage cells. Here, we show that SRF is required cell autonomously in oligodendrocytes for myelination during development. Combining ChIP-seq with RNA-seq identifies SRF-target genes in oligodendrocyte precursor cells and oligodendrocytes that include actin and other key cytoskeletal genes. Accordingly, SRF knockout oligodendrocytes exhibit dramatically reduced actin filament levels early in differentiation, consistent with its role in actin-dependent myelin sheath initiation. Surprisingly, oligodendrocyte-restricted loss of SRF results in upregulation of gene signatures associated with aging and neurodegenerative diseases. Together, our findings identify SRF as a transcriptional regulator that controls the expression of cytoskeletal genes required in oligodendrocytes for myelination. This study identifies an essential pathway regulating oligodendrocyte biology with high relevance to brain development, aging, and disease.

The miRNA-target interactions: An underestimated intricacy

MicroRNAs (miRNAs) play indispensable roles in posttranscriptional gene regulation. Their cellular regulatory impact is determined not solely by their sheer number, which likely amounts to >2000 individual miRNAs in human, than by the regulatory effectiveness of single miRNAs. Although, one begins to develop an understanding of the complex mechanisms underlying miRNA-target interactions (MTIs), the overall knowledge of MTI functionality is still rather patchy. In this critical review, we summarize key features of mammalian MTIs. We especially highlight latest insights on (i) the dynamic make-up of miRNA binding sites including non-canonical binding sites, (ii) the cooperativity between miRNA binding sites, (iii) the adaptivity of MTIs through sequence modifications, (iv) the bearing of intra-cellular miRNA localization changes and (v) the role of cell type and cell status specific miRNA interaction partners. The MTI biology is discussed against the background of state-of-the-art approaches with particular emphasis on experimental strategies for evaluating miRNA functionality.

ZEBRA: a hierarchically integrated gene expression atlas of the murine and human brain at single-cell resolution

The molecular causes and mechanisms of neurodegenerative diseases remain poorly understood. A growing number of single-cell studies have implicated various neural, glial, and immune cell subtypes to affect the mammalian central nervous system in many age-related disorders. Integrating this body of transcriptomic evidence into a comprehensive and reproducible framework poses several computational challenges. Here, we introduce ZEBRA, a large single-cell and single-nucleus RNA-seq database. ZEBRA integrates and normalizes gene expression and metadata from 33 studies, encompassing 4.2 million human and mouse brain cells sampled from 39 brain regions. It incorporates samples from patients with neurodegenerative diseases like Alzheimer's disease, Parkinson's disease, and Multiple sclerosis, as well as samples from relevant mouse models. We employed scVI, a deep probabilistic auto-encoder model, to integrate the samples and curated both cell and sample metadata for downstream analysis. ZEBRA allows for cell-type and disease-specific markers to be explored and compared between sample conditions and brain regions, a cell composition analysis, and gene-wise feature mappings. Our comprehensive molecular database facilitates the generation of data-driven hypotheses, enhancing our understanding of mammalian brain function during aging and disease. The data sets, along with an interactive database are freely available at https://www.ccb.uni-saarland.de/zebra.

ABC-HuMi: the Atlas of Biosynthetic Gene Clusters in the Human Microbiome

The human microbiome has emerged as a rich source of diverse and bioactive natural products, harboring immense potential for therapeutic applications. To facilitate systematic exploration and analysis of its biosynthetic landscape, we present ABC-HuMi: the Atlas of Biosynthetic Gene Clusters (BGCs) in the Human Microbiome. ABC-HuMi integrates data from major human microbiome sequence databases and provides an expansive repository of BGCs compared to the limited coverage offered by existing resources. Employing state-of-the-art BGC prediction and analysis tools, our database ensures accurate annotation and enhanced prediction capabilities. ABC-HuMi empowers researchers with advanced browsing, filtering, and search functionality, enabling efficient exploration of the resource. At present, ABC-HuMi boasts a catalog of 19 218 representative BGCs derived from the human gut, oral, skin, respiratory and urogenital systems. By capturing the intricate biosynthetic potential across diverse human body sites, our database fosters profound insights into the molecular repertoire encoded within the human microbiome and offers a comprehensive resource for the discovery and characterization of novel bioactive compounds. The database is freely accessible at https://www.ccb.uni-saarland.de/abc_humi/.

Characterizing expression changes in noncoding RNAs during aging and heterochronic parabiosis across mouse tissues

Molecular mechanisms of organismal and cell aging remain incompletely understood. We, therefore, generated a body-wide map of noncoding RNA (ncRNA) expression in aging (16 organs at ten timepoints from 1 to 27 months) and rejuvenated mice. We found molecular aging trajectories are largely tissue-specific except for eight broadly deregulated microRNAs (miRNAs). Their individual abundance mirrors their presence in circulating plasma and extracellular vesicles (EVs) whereas tissue-specific ncRNAs were less present. For miR-29c-3p, we observe the largest correlation with aging in solid organs, plasma and EVs. In mice rejuvenated by heterochronic parabiosis, miR-29c-3p was the most prominent miRNA restored to similar levels found in young liver. miR-29c-3p targets the extracellular matrix and secretion pathways, known to be implicated in aging. We provide a map of organism-wide expression of ncRNAs with aging and rejuvenation and identify a set of broadly deregulated miRNAs, which may function as systemic regulators of aging via plasma and EVs.

Feasibility Study of Using Augmented Mirrors for Alignment Task during Orthopaedic Procedures in Mixed Reality

Accurate depth estimation poses a significant challenge in egocentric Augmented Reality (AR), particularly for precision-dependent tasks in the medical field, such as needle or tool insertions during percutaneous procedures. Augmented Mirrors (AMs) provide a unique solution to this problem by offering additional non-egocentric viewpoints that enhance spatial understanding of an AR scene. Despite the perceptual advantages of using AMs, their practical utility has yet to be thoroughly tested. In this work, we present results from a pilot study involving five participants tasked with simulating epidural injection procedures in an AR environment, both with and without the aid of an AM. Our findings indicate that using AM contributes to reducing mental effort while improving alignment accuracy. These results highlight the potential of AM as a powerful tool for AR-enabled medical procedures, setting the stage for future exploration involving medical professionals.

Zfp362 potentiates murine colonic inflammation by constraining Treg cell function rather than promoting Th17 cell differentiation

Mucosal barrier integrity and pathogen clearance is a complex process influenced by both Th17 and Treg cells. Previously, we had described the DNA methylation profile of Th17 cells and identified Zinc finger protein (Zfp)362 to be uniquely demethylated. Here, we generated Zfp362-/- mice to unravel the role of Zfp362 for Th17 cell biology. Zfp362-/- mice appeared clinically normal, showed no phenotypic alterations in the T-cell compartment, and upon colonization with segmented filamentous bacteria, no effect of Zfp362 deficiency on Th17 cell differentiation was observed. By contrast, Zfp362 deletion resulted in increased frequencies of colonic Foxp3+ Treg cells and IL-10+ and RORγt+ Treg cell subsets in mesenteric lymph nodes. Adoptive transfer of naïve CD4+ T cells from Zfp362-/- mice into Rag2-/- mice resulted in a significantly lower weight loss when compared with controls receiving cells from Zfp362+/+ littermates. However, this attenuated weight loss did not correlate with alterations of Th17 cells but instead was associated with an increase of effector Treg cells in mesenteric lymph nodes. Together, these results suggest that Zfp362 plays an important role in promoting colonic inflammation; however, this function is derived from constraining the effector function of Treg cells rather than directly promoting Th17 cell differentiation.

Long-Term Urinary Toxicity Follow-Up of Combined External Beam Radiation and Cs-131 LDR Brachytherapy Boost for Prostate Cancer

PURPOSE/OBJECTIVE(S)

To evaluate and update our institutional data on chronic patient-reported urinary symptoms in patients with prostate cancer treated with combined external beam radiation therapy (EBRT) and Cesium-131 (Cs-131) LDR brachytherapy boost.

MATERIALS/METHODS

Patients treated from September 2006-July 2022 and with at least 6 months of follow-up were analyzed. Patients filled out the Expanded Prostate Cancer Index Composite (EPIC) questionnaire pre-treatment and at each scheduled follow-up. Mean EPIC urinary scores were calculated for urinary summary, urinary function, urinary bother, urinary incontinence, and urinary irritative/obstructive domains. Changes of 7-points or more in mean domain scores were deemed clinically significant. The prevalence of urinary bleeding was calculated and bleeding occurring more than "about once a week" ("more than once a week", "about once a day", and "more than once a day") was deemed clinically significant. Outcomes were analyzed at each follow-up timepoint using paired t-test.

RESULTS

A total of 341 patients with prostate cancer were included in the analysis of which 183 were intermediate-risk and 158 high-risk. 174 patients were treated with tri-modality therapy, and SpaceOAR was used in 108 patients. Long-term statistically significant decreases in mean EPIC QOL scores were seen in urinary summary, urinary function, urinary bother, urinary incontinence, and urinary irritative/obstructive domains at 6 months, with clinically and statistically significant decreases in urinary function and urinary incontinence at 6 months (86±17.4, 83.4±22.2 respectively) and up to 18 months post-treatment (84.8±17.4, 83.6±20.7 respectively) compared with pre-treatment scores (93.1±10.7, 90.8±14.8 respectively), with both domains returning to clinical baseline at 24 months. Urinary irritative/obstructive mean scores clinically and significantly decreased at 9 months (77±17.4) compared to pre-treatment (84.7±13.5), returning to clinical baseline at 12 months. The prevalence of clinically significant urinary bleeding at 6, 12, 24, 36, 48, and 60 months was 1.3%, 2.6%, 3.7%, 0.7%, 0.9%, and 2.3% respectively. At the median follow-up of 42 months (IQR 54, 18-72), mean EPIC scores were clinically improved (≥7-point increase) from pre-treatment baseline in 15-30% of patients across the analyzed urinary domains.

CONCLUSION

Long-term patient-reported clinically significant decreases in urinary function and urinary incontinence mean EPIC scores occurred up to 18 months following combined external beam radiation and Cs-131 prostate brachytherapy and returned to baseline at 24 months, while clinically significant decrease in urinary irritability/obstructive mean scores occurred at 9 months and returned to baseline at 12 months. At 42 months follow-up, mean urinary domain scores improved from pre-treatment baseline in 15-30% of patients.

Atlas of the aging mouse brain reveals white matter as vulnerable foci

Aging is the key risk factor for cognitive decline, yet the molecular changes underlying brain aging remain poorly understood. Here, we conducted spatiotemporal RNA sequencing of the mouse brain, profiling 1,076 samples from 15 regions across 7 ages and 2 rejuvenation interventions. Our analysis identified a brain-wide gene signature of aging in glial cells, which exhibited spatially defined changes in magnitude. By integrating spatial and single-nucleus transcriptomics, we found that glial aging was particularly accelerated in white matter compared with cortical regions, whereas specialized neuronal populations showed region-specific expression changes. Rejuvenation interventions, including young plasma injection and dietary restriction, exhibited distinct effects on gene expression in specific brain regions. Furthermore, we discovered differential gene expression patterns associated with three human neurodegenerative diseases, highlighting the importance of regional aging as a potential modulator of disease. Our findings identify molecular foci of brain aging, providing a foundation to target age-related cognitive decline.

Cas9-mediated knockout of Ndrg2 enhances the regenerative potential of dendritic cells for wound healing

Chronic wounds impose a significant healthcare burden to a broad patient population. Cell-based therapies, while having shown benefits for the treatment of chronic wounds, have not yet achieved widespread adoption into clinical practice. We developed a CRISPR/Cas9 approach to precisely edit murine dendritic cells to enhance their therapeutic potential for healing chronic wounds. Using single-cell RNA sequencing of tolerogenic dendritic cells, we identified N-myc downregulated gene 2 (Ndrg2), which marks a specific population of dendritic cell progenitors, as a promising target for CRISPR knockout. Ndrg2-knockout alters the transcriptomic profile of dendritic cells and preserves an immature cell state with a strong pro-angiogenic and regenerative capacity. We then incorporated our CRISPR-based cell engineering within a therapeutic hydrogel for in vivo cell delivery and developed an effective translational approach for dendritic cell-based immunotherapy that accelerated healing of full-thickness wounds in both non-diabetic and diabetic mouse models. These findings could open the door to future clinical trials using safe gene editing in dendritic cells for treating various types of chronic wounds.

miEAA 2023: updates, new functional microRNA sets and improved enrichment visualizations

MicroRNAs (miRNAs) are small non-coding RNAs that play a critical role in regulating diverse biological processes. Extracting functional insights from a list of miRNAs is challenging, as each miRNA can potentially interact with hundreds of genes. To address this challenge, we developed miEAA, a flexible and comprehensive miRNA enrichment analysis tool based on direct and indirect miRNA annotation. The latest release of miEAA includes a data warehouse of 19 miRNA repositories, covering 10 different organisms and 139 399 functional categories. We have added information on the cellular context of miRNAs, isomiRs, and high-confidence miRNAs to improve the accuracy of the results. We have also improved the representation of aggregated results, including interactive Upset plots to aid users in understanding the interaction among enriched terms or categories. Finally, we demonstrate the functionality of miEAA in the context of ageing and highlight the importance of carefully considering the miRNA input list. MiEAA is free to use and publicly available at https://www.ccb.uni-saarland.de/mieaa/.

Occurrence, resistance patterns and management of carbapenemase-producing bacteria in war-wounded refugees from Ukraine

We analysed consecutive clinical cases of infections due to carbapenemase-producing Gram-negative bacteria detected in war-wounded patients from Ukraine who were treated at one University medical centre in southwest Germany between June and December 2022. Isolates of multiresistant Gram-negative bacteria were subjected to a thorough microbiological characterisation and whole-genome sequencing. We identified five war-wounded Ukrainian patients who developed infections with NDM-1-positive Klebsiella pneumoniae. Two isolates also carried OXA-48 carbapenemases. The bacteria were resistant to novel antibiotics such as ceftazidime/avibactam and cefiderocol. Employed treatment strategies included combinations of ceftazidime/avibactam + aztreonam, colistin, or tigecycline. Whole-genome sequencing suggested transmission during primary care in Ukraine. We conclude that there is an urgent need for thorough surveillance of multiresistant pathogens in patients from war zones.

The Effect of a Planetary Health Diet on the Human Gut Microbiome: A Descriptive Analysis

In 2019, researchers from the EAT-Lancet Commission developed the 'Planetary Health (PH) diet'. Specifically, they provided recommendations pertaining to healthy diets derived from sustainable food systems. Thus far, it has not been analysed how such a diet affects the human intestinal microbiome, which is important for health and disease development. Here, we present longitudinal genome-wide metagenomic sequencing and mass spectrometry data on the gut microbiome of healthy volunteers adhering to the PH diet, as opposed to vegetarian or vegan (VV) and omnivorous (OV) diets. We obtained basic epidemiological information from 41 healthy volunteers and collected stool samples at inclusion and after 2, 4, and 12 weeks. Individuals opting to follow the PH diet received detailed instructions and recipes, whereas individuals in the control groups followed their habitual dietary pattern. Whole-genome DNA was extracted from stool specimens and subjected to shotgun metagenomic sequencing (~3 GB per patient). Conventional bacterial stool cultures were performed in parallel and bacterial species were identified with matrix-assisted laser desorption/ionization time-of-flight (MALDI-TOF) mass spectrometry. We analysed samples from 16 PH, 16 OV, and 9 VV diet patterns. The α-diversity remained relatively stable for all dietary groups. In the PH group, we observed a constant increase from 3.79% at inclusion to 4.9% after 12 weeks in relative abundance of Bifidobacterium adolescentis. Differential PH abundance analysis highlighted a non-significant increase in possible probiotics such as Paraprevotella xylaniphila and Bacteroides clarus. The highest abundance of these bacteria was observed in the VV group. Dietary modifications are associated with rapid alterations to the human gut microbiome, and the PH diet led to a slight increase in probiotic-associated bacteria at ≥4 weeks. Additional research is required to confirm these findings.

miR-34a-5p as molecular hub of pathomechanisms in Huntington's disease

BACKGROUND

Although a pivotal role of microRNA (miRNA, miR) in the pathogenesis of Huntington's disease (HD) is increasingly recognized, the molecular functions of miRNAs in the pathomechanisms of HD await further elucidation. One of the miRNAs that have been associated with HD is miR-34a-5p, which was deregulated in the mouse R6/2 model and in human HD brain tissues.

METHODS

The aim of our study was to demonstrate interactions between miR-34a-5p and HD associated genes. By computational means we predicted 12 801 potential target genes of miR-34a-5p. An in-silico pathway analysis revealed 22 potential miR-34a-5p target genes in the KEGG (Kyoto Encyclopedia of Genes and Genomes) pathway "Huntington's disease".

RESULTS

Using our high-throughput miRNA interaction reporter assay (HiTmIR) we identified NDUFA9, TAF4B, NRF1, POLR2J2, DNALI1, HIP1, TGM2 and POLR2G as direct miR-34a-5p target genes. Direct binding of miR-34a-5p to target sites in the 3'UTRs of TAF4B, NDUFA9, HIP1 and NRF1 was verified by a mutagenesis HiTmIR assay and by determining endogenous protein levels for HIP1 and NDUFA9. STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) analysis identified protein-protein interaction networks associated with HD like "Glutamine Receptor Signaling Pathway" and "Calcium Ion Transmembrane Import Into Cytosol".

CONCLUSION

Our study demonstrates multiple interactions between miR-34a-5p and HD associated target genes and thereby lays the ground for future therapeutic interventions using this miRNA.

Artificial-intelligence-based decision support tools for the differential diagnosis of colitis

BACKGROUND

Whereas Artificial Intelligence (AI) based tools have recently been introduced in the field of gastroenterology, application in inflammatory bowel disease (IBD) is in its infancies. We established AI-based algorithms to distinguish IBD from infectious and ischemic colitis using endoscopic images and clinical data.

METHODS

First, we trained and tested a Convolutional Neural Network (CNN) using 1796 real-world images from 494 patients, presenting with three diseases (IBD [n = 212], ischemic colitis [n = 157], and infectious colitis [n = 125]). Moreover, we evaluated a Gradient Boosted Decision Trees (GBDT) algorithm using five clinical parameters as well as a hybrid approach (CNN + GBDT). Patients and images were randomly split into two completely independent datasets. The proposed approaches were benchmarked against each other and three expert endoscopists on the test set.

RESULTS

For the image-based CNN, the GBDT algorithm and the hybrid approach global accuracies were .709, .792, and .766, respectively. Positive predictive values were .602, .702, and .657. Global areas under the receiver operating characteristics (ROC) and precision recall (PR) curves were .727/.585, .888/.823, and .838/.733, respectively. Global accuracy did not differ between CNN and endoscopists (.721), but the clinical parameter-based GBDT algorithm outperformed CNN and expert image classification.

CONCLUSIONS

Decision support systems exclusively based on endoscopic image analysis for the differential diagnosis of colitis, representing a complex clinical challenge, seem not yet to be ready for primetime and more diverse image datasets may be necessary to improve performance in future development. The clinical value of the proposed clinical parameters algorithm should be evaluated in prospective cohorts.

Changes in the characteristics of patients treated for brain metastases with repeat stereotactic radiotherapy (SRT): a retrospective study of 184 patients

PURPOSE

Brain metastases (BMs) are the leading cause of intracranial malignant neoplasms in adults. WHO, Karnofsky performance status (KPS), age, number of BMs, extracerebral progression (ECP), recursive partitioning analysis (RPA), diagnosis-specific graded prognostic assessment (Ds-GPA) are validated prognostic tools to help clinicians decide on treatment. No consensus exists for repeat stereotactic radiotherapy (SRT) for BMs. The aim of this study was to review the changes in patient characteristics treated with repeated SRTs.

METHODS AND MATERIALS

The data of patients treated between 2010 and 2020 with at least two courses of SRT without previous whole brain radiotherapy (WBRT) were reviewed. Age, WHO, KPS, ECP, type of systemic treatment, number of BMs were recorded. RPA, Ds-GPA and brain metastasis velocity (BMV) were calculated.

RESULTS

184 patients were treated for 915 BMs and received two to six SRTs for local or distant brain recurrence. The median number of BMs treated per SRT was 1 (range: 1-6), for a median of 4 BMs treated during all sessions (range: 2-19). WHO, Ds-GPA and RPA were stable between each session of SRT, whereas KPS was significantly better in SRT1 than in the following SRT. The number of BMs was not significantly different between each SRT, but there was a tendency for more BM at SRT1 (p = 0.06). At SRT1, patients had largest BM and undergo more surgery than during the following SRT (p < 0.001). 6.5%, 37.5% and 56% of patients were classified as high, intermediate, and low BMV, respectively, at the last SRT session. There was almost perfect concordance between the BMV-grade calculated at the last SRT session and at SRT2 (r = 0.89; p < 0.001).

CONCLUSION

Repeated SRT doesn't lead to a marked alteration in the general condition, KPS was maintained at over 70% for more than 95% of patients during all SRTs. Long survival can be expected, especially in low-grade BMV patients. WBRT shouldn't be aborted, especially for patients developing more than twelve BMs annually.

Outside the limit: questioning the distance restrictions for cooperative miRNA binding sites

Among the concepts in biology that are widely taken granted is a potentiated cooperative effect of multiple miRNAs on the same target. This strong hypothesis contrasts insufficient experimental evidence. The quantity as well as the quality of required side constraints of cooperative binding remain largely hidden. For miR-21-5p and miR-155-5p, two commonly investigated regulators across diseases, we selected 15 joint target genes. These were chosen to represent various neighboring 3'UTR binding site constellations, partially exceeding the distance rules that have been established for over a decade. We identified different cooperative scenarios with the binding of one miRNA enhancing the binding effects of the other miRNA and vice versa. Using both, reporter assays and whole proteome analyses, we observed these cooperative miRNA effects for genes that bear 3'UTR binding sites at distances greater than the previously defined limits. Astonishingly, the experiments provide even stronger evidence for cooperative miRNA effects than originally postulated. In the light of these findings the definition of targetomes specified for single miRNAs need to be refined by a concept that acknowledges the cooperative effects of miRNAs.

Acute toxicities and cumulative dose to the brain of repeated sessions of stereotactic radiotherapy (SRT) for brain metastases: a retrospective study of 184 patients

BACKGROUND

Stereotactic radiation therapy (SRT) is a focal treatment for brain metastases (BMs); thus, 20 to 40% of patients will require salvage treatment after an initial SRT session, either because of local or distant failure. SRT is not exempt from acute toxicity, and the acute toxicities of repeated SRT are not well known. The objective of this study was to analyze the acute toxicities of repeated courses of SRT and to determine whether repeated SRT could lead to cumulative brain doses equivalent to those of whole-brain radiotherapy (WBRT).

MATERIAL AND METHODS

Between 2010 and 2020, data from 184 patients treated for 915 BMs via two to six SRT sessions for local or distant BM recurrence without previous or intercurrent WBRT were retrospectively reviewed. Patients were seen via consultations during SRT, and the delivered dose, the use of corticosteroid therapy and neurological symptoms were recorded and rated according to the CTCAEv4. The dosimetric characteristics of 79% of BMs were collected, and summation plans of 76.6% of BMs were created.

RESULTS

36% of patients developed acute toxicity during at least one session. No grade three or four toxicity was registered, and grade one or two cephalalgy was the most frequently reported symptom. There was no significant difference in the occurrence of acute toxicity between consecutive SRT sessions. In the multivariate analysis, acute toxicity was associated with the use of corticosteroid therapy before irradiation (OR = 2.6; p = 0.01), BMV grade (high vs. low grade OR = 5.17; p = 0.02), and number of SRT sessions (3 SRT vs. 2 SRT: OR = 2.64; p = 0.01). The median volume equivalent to the WBRT dose (V_WBRT) was 47.9 ml. In the multivariate analysis, the V_WBRT was significantly associated with the total GTV (p < 0.001) and number of BMs (p < 0.001). Even for patients treated for more than ten cumulated BMs, the median BED to the brain was very low compared to the dose delivered during WBRT.

CONCLUSION

Repeated SRT for local or distant recurrent BM is well tolerated, without grade three or four toxicity, and does not cause more acute neurological toxicity with repeated SRT sessions. Moreover, even for patients treated for more than ten BMs, the V_WBRT is low.

Dynamic and static circulating cancer microRNA biomarkers - a validation study

For cancers and other pathologies, early diagnosis remains the most promising path to survival. Profiling of longitudinal cohorts facilitates insights into trajectories of biomarkers. We measured microRNA expression in 240 serum samples from patients with colon, lung, and breast cancer and from cancer-free controls. Each patient provided at least two serum samples, one prior to diagnosis and one following diagnosis. The median time interval between the samples was 11.6 years. Using computational models, we evaluated the circulating profiles of 21 microRNAs. The analysis yielded two sets of biomarkers, static ones that show an absolute difference between certain cancer types and controls and dynamic ones where the level over time provided higher diagnostic information content. In the first group, miR-99a-5p stands out for all three cancer types. In the second group, miR-155-5p allows to predict lung cancers and colon cancers. Classification in samples from cancer and non-cancer patients using gradient boosted trees reached an average accuracy of 79.9%. The results suggest that individual change over time or an absolute value at one time point may predict a disease with high specificity and sensitivity.

Ageing-associated small RNA cargo of extracellular vesicles

Previous work on murine models and humans demonstrated global as well as tissue-specific molecular ageing trajectories of RNAs. Extracellular vesicles (EVs) are membrane vesicles mediating the horizontal transfer of genetic information between different tissues. We sequenced small regulatory RNAs (sncRNAs) in two mouse plasma fractions at five time points across the lifespan from 2-18 months: (1) sncRNAs that are free-circulating (fc-RNA) and (2) sncRNAs bound outside or inside EVs (EV-RNA). Different sncRNA classes exhibit unique ageing patterns that vary between the fcRNA and EV-RNA fractions. While tRNAs showed the highest correlation with ageing in both fractions, rRNAs exhibited inverse correlation trajectories between the EV- and fc-fractions. For miRNAs, the EV-RNA fraction was exceptionally strongly associated with ageing, especially the miR-29 family in adipose tissues. Sequencing of sncRNAs and coding genes in fat tissue of an independent cohort of aged mice up to 27 months highlighted the pivotal role of miR-29a-3p and miR-29b-3p in ageing-related gene regulation that we validated in a third cohort by RT-qPCR.

Genetic variation in the human olfactory receptor OR5AN1 associates with the perception of musks

Humans have significant individual variations in odor perception, derived from their experience or sometimes from differences in the olfactory receptor (OR) gene repertoire. In several cases, the genetic variation of a single OR affects the perception of its cognate odor ligand. Musks are widely used for fragrance and are known to demonstrate specific anosmia. It, however, remains to be elucidated whether the OR polymorphism contributes to individual variations in musk odor perception. Previous studies reported that responses of the human musk receptor OR5AN1 to a variety of musks in vitro correlated well with perceptual sensitivity to those odors in humans and that the mouse ortholog, Olfr1440 (MOR215-1), plays a critical role in muscone perception. Here, we took advantage of genetic variation in OR5AN1 to examine how changes in receptor sensitivity are associated with human musk perception. We investigated the functional differences between OR5AN1 variants in an in vitro assay and measured both perceived intensity and detection threshold in human subjects with different OR5AN1 genotypes. Human subjects homozygous for the more sensitive L289F allele had a lower detection threshold for muscone and found macrocyclic musks to be more intense than subjects homozygous for the reference allele. These results demonstrate that the genetic variation in OR5AN1 contributes to perceptual differences for some musks. In addition, we found that the more functional variant of OR5A1, a receptor involved in β-ionone perception, is associated with the less functional variant of OR5AN1, suggesting that the perceived intensities of macrocyclic musks and β-ionone are inversely correlated.

The Impact of Focus and Context Visualization Techniques on Depth Perception in Optical See-Through Head-Mounted Displays

Estimating the depth of virtual content has proven to be a challenging task in Augmented Reality (AR) applications. Existing studies have shown that the visual system makes use of multiple depth cues to infer the distance of objects, occlusion being one of the most important ones. The ability to generate appropriate occlusions becomes particularly important for AR applications that require the visualization of augmented objects placed below a real surface. Examples of these applications are medical scenarios in which the visualization of anatomical information needs to be observed within the patient's body. In this regard, existing works have proposed several focus and context (F+C) approaches to aid users in visualizing this content using Video See-Through (VST) Head-Mounted Displays (HMDs). However, the implementation of these approaches in Optical See-Through (OST) HMDs remains an open question due to the additive characteristics of the display technology. In this article, we, for the first time, design and conduct a user study that compares depth estimation between VST and OST HMDs using existing in-situ visualization methods. Our results show that these visualizations cannot be directly transferred to OST displays without increasing error in depth perception tasks. To tackle this gap, we perform a structured decomposition of the visual properties of AR F+C methods to find best-performing combinations. We propose the use of chromatic shadows and hatching approaches transferred from computer graphics. In a second study, we perform a factorized analysis of these combinations, showing that varying the shading type and using colored shadows can lead to better depth estimation when using OST HMDs.

Hemodynamic changes during conventional and daylight photodynamic therapy of actinic keratoses - a randomized controlled trial

BACKGROUND

Photodynamic therapy (PDT) is an effective treatment for actinic keratosis (AK). However, pain and hypertension are important side effects of conventional PDT (c-PDT). Several studies have demonstrated that daylight PDT (dl-PDT) is less painful while being as effective as c-PDT.

OBJECTIVE

To observe the effect of c-PDT and dl-PDT on different hemodynamic parameters (systolic blood pressure and diastolic blood pressure, pulse rate, and peripheral oxygen saturation).

METHODS

Fifty patients with AK on the head were enrolled into this prospective, randomized, controlled study and treated with c-PDT or dl-PDT in a 1:1 ratio. Hemodynamic parameters were measured at four different time points during treatment. Pain was quantified using a visual analog scale. AK was counted before treatment and after one month.

RESULTS

C-PDT is associated with significantly more pain, a significant increase in blood pressure and a higher rate of patients with grade 3 hypertension. Whereas dl-PDT is almost painless and does not lead to any changes in hemodynamic parameters. For both treatments, a similar lesion response rate was found after one month.

CONCLUSIONS

dl-PDT has a better tolerability while being as effective as c-PDT and therefore may be the more favorable treatment option in certain patient groups.

isomiRdb: microRNA expression at isoform resolution

A significant fraction of mature miRNA transcripts carries sequence and/or length variations, termed isomiRs. IsomiRs are differentially abundant in cell types, tissues, body fluids or patients' samples. Not surprisingly, multiple studies describe a physiological and pathophysiological role. Despite their importance, systematically collected and annotated isomiR information available in databases remains limited. We thus developed isomiRdb, a comprehensive resource that compiles miRNA expression data at isomiR resolution from various sources. We processed 42 499 human miRNA-seq datasets (5.9 × 1011 sequencing reads) and consistently analyzed them using miRMaster and sRNAbench. Our database provides online access to the 90 483 most abundant isomiRs (>1 RPM in at least 1% of the samples) from 52 tissues and 188 cell types. Additionally, the full set of over 3 million detected isomiRs is available for download. Our resource can be queried at the sample, miRNA or isomiR level so users can quickly answer common questions about the presence/absence of a particular miRNA/isomiR in tissues of interest. Further, the database facilitates to identify whether a potentially interesting new isoform has been detected before and its frequency. In addition to expression tables, isomiRdb can generate multiple interactive visualisations including violin plots and heatmaps. isomiRdb is free to use and publicly available at: https://www.ccb.uni-saarland.de/isomirdb.

Forecasting football injuries by combining screening, monitoring and machine learning

Identifying players or circumstances associated with an increased risk of injury is fundamental for successful risk management in football. So far, time-constant and volatile risk factors are generally considered separately in either a screening (constant) or a monitoring (volatile) approach each resulting in a restricted set of explanatory variables. Consequently, improvements in predictive accuracy may be expected when screening and monitoring data are combined, especially when analysed with current machine learning (ML) techniques. This trial was designed as a prospective observational cohort study aiming to forecast non-contact time-loss injuries in male professional football (soccer). Injuries were registered according to the Fuller consensus. Gradient boosting with ROSE upsampling within a leave-one-out cross-validation was used for data analysis. The hierarchical data structure was considered throughout. Different splits of the original dataset were used to probe the robustness of results. Data of 88 players from 4 teams and 51 injuries could be analysed. The cross-validated performance of the gradient boosted model (ROC area under the curve 0.61) was promising and higher compared to models without integration of screening data. Importantly, holdout test set performance was similar (ROC area under the curve 0.62) indicating prospect of generalizability to new cases. However, the variation of predictive accuracy and feature importance with different splits of the original dataset reflects the relatively low number of events. It is concluded that ML-based injury forecasting based on the integration of screening and monitoring data is promising. However, external prospective verification and continued model development are required.

Radionecrosis after repeated courses of radiotherapy under stereotactic conditions for brain metastases: Analysis of clinical and dosimetric data from a retrospective cohort of 184 patients

PURPOSE

Between 10 and 40% of patients with cancer will develop one or more brain metastases (BMs). Stereotactic radiotherapy (SRT) is part of the therapeutic arsenal for the treatment of de novo or recurrent BM. Its main interest is to delay whole brain radiation therapy (WBRT), which may cause cognitive toxicity. However, SRT is not exempt from long-term toxicity, and the most widely known SRT is radionecrosis (RN). The objective of this study was to analyze the occurrence of RN per BM and per patient.

MATERIAL AND METHODS

Between 2010 and 2020, data from 184 patients treated for 915 BMs by two to six SRT sessions for local or distant brain recurrence without previous or intercurrent WBRT were retrospectively reviewed. RN was examined on trimestral follow-up MRI and potentially confirmed by surgery or nuclear medicine. For each BM and SRT session plan, summation V_12Gy, V_14Gy, V_21Gy and V_23Gy isodoses were collected. Volumes of intersections were created between the 12Gy isodose at the first SRT and the 18Gy isodose of the following SRT (V_18-12Gy).

RESULTS

At the end of follow-up, 23.0% of patients presented RN, and 6.3% of BM presented RN. Median follow-up of BM was 13.3 months (95%CI 18.3-20.8). The median interval between BM irradiation and RN was 8.7 months (95% CI 9.2-14.7). Six-, 12- and 24-month RN-free survival rates per BM were 75%, 54% and 29%, respectively. The median RN-free survival per patient was 15.3 months (95% CI 13.6-18.1). In multivariate analysis, the occurrence of RN per BM was statistically associated with local reirradiation (P<0.001) and the number of SRTs (P<0.001). In univariate analysis, the occurrence of RN per patient was statistically associated with the sum of all V_18-12Gy (P=0.02). No statistical association was found in multivariate analysis. A sum of all V_18-12Gy of less than 1.5ml was associated with a 14.6% risk of RN, compared with 35.6% when the sum of all V_18-12Gy was superior to 1.5ml. The sum of all V_18-12Gy larger than 1.5ml was associated with a 74% specificity and 53% sensitivity of RN (P<0.001).

CONCLUSION

Based on these results, a small number of BMs show RN during repeated SRT for local or distant recurrent BMs. Local reirradiation was the most predictive factor of brain RN. A V_18-12Gy larger than 7.6ml in the case of local reirradiation or larger than 1.5ml in proximity reirradiation were prognostic factors of RN. The more BM patients need radiation therapy, and the longer they survive after irradiation, the higher their individual risk of developing RN.

Disrupting mechanotransduction decreases fibrosis and contracture in split-thickness skin grafting

Burns and other traumatic injuries represent a substantial biomedical burden. The current standard of care for deep injuries is autologous split-thickness skin grafting (STSG), which frequently results in contractures, abnormal pigmentation, and loss of biomechanical function. Currently, there are no effective therapies that can prevent fibrosis and contracture after STSG. Here, we have developed a clinically relevant porcine model of STSG and comprehensively characterized porcine cell populations involved in healing with single-cell resolution. We identified an up-regulation of proinflammatory and mechanotransduction signaling pathways in standard STSGs. Blocking mechanotransduction with a small-molecule focal adhesion kinase (FAK) inhibitor promoted healing, reduced contracture, mitigated scar formation, restored collagen architecture, and ultimately improved graft biomechanical properties. Acute mechanotransduction blockade up-regulated myeloid CXCL10-mediated anti-inflammation with decreased CXCL14-mediated myeloid and fibroblast recruitment. At later time points, mechanical signaling shifted fibroblasts toward profibrotic differentiation fates, and disruption of mechanotransduction modulated mesenchymal fibroblast differentiation states to block those responses, instead driving fibroblasts toward proregenerative, adipogenic states similar to unwounded skin. We then confirmed these two diverging fibroblast transcriptional trajectories in human skin, human scar, and a three-dimensional organotypic model of human skin. Together, pharmacological blockade of mechanotransduction markedly improved large animal healing after STSG by promoting both early, anti-inflammatory and late, regenerative transcriptional programs, resulting in healed tissue similar to unwounded skin. FAK inhibition could therefore supplement the current standard of care for traumatic and burn injuries.

sRNAbench and sRNAtoolbox 2022 update: accurate miRNA and sncRNA profiling for model and non-model organisms

The NCBI Sequence Read Archive currently hosts microRNA sequencing data for over 800 different species, evidencing the existence of a broad taxonomic distribution in the field of small RNA research. Simultaneously, the number of samples per miRNA-seq study continues to increase resulting in a vast amount of data that requires accurate, fast and user-friendly analysis methods. Since the previous release of sRNAtoolbox in 2019, 55 000 sRNAbench jobs have been submitted which has motivated many improvements in its usability and the scope of the underlying annotation database. With this update, users can upload an unlimited number of samples or import them from Google Drive, Dropbox or URLs. Micro- and small RNA profiling can now be carried out using high-confidence Metazoan and plant specific databases, MirGeneDB and PmiREN respectively, together with genome assemblies and libraries from 441 Ensembl species. The new results page includes straightforward sample annotation to allow downstream differential expression analysis with sRNAde. Unassigned reads can also be explored by means of a new tool that performs mapping to microbial references, which can reveal contamination events or biologically meaningful findings as we describe in the example. sRNAtoolbox is available at: https://arn.ugr.es/srnatoolbox/.

Young CSF restores oligodendrogenesis and memory in aged mice via Fgf17

Recent understanding of how the systemic environment shapes the brain throughout life has led to numerous intervention strategies to slow brain ageing^1-3. Cerebrospinal fluid (CSF) makes up the immediate environment of brain cells, providing them with nourishing compounds^4,5. We discovered that infusing young CSF directly into aged brains improves memory function. Unbiased transcriptome analysis of the hippocampus identified oligodendrocytes to be most responsive to this rejuvenated CSF environment. We further showed that young CSF boosts oligodendrocyte progenitor cell (OPC) proliferation and differentiation in the aged hippocampus and in primary OPC cultures. Using SLAMseq to metabolically label nascent mRNA, we identified serum response factor (SRF), a transcription factor that drives actin cytoskeleton rearrangement, as a mediator of OPC proliferation following exposure to young CSF. With age, SRF expression decreases in hippocampal OPCs, and the pathway is induced by acute injection with young CSF. We screened for potential SRF activators in CSF and found that fibroblast growth factor 17 (Fgf17) infusion is sufficient to induce OPC proliferation and long-term memory consolidation in aged mice while Fgf17 blockade impairs cognition in young mice. These findings demonstrate the rejuvenating power of young CSF and identify Fgf17 as a key target to restore oligodendrocyte function in the ageing brain.

Ischemic Stroke–A Scientometric Analysis

Background

Stroke is the second leading cause of death world-wide. A comprehensive scientometric study regarding ischemic stroke research has not been performed yet. This study aims at investigating the global research output on ischemic stroke research.

Methods

All 21,115 articles regarding ischemic stroke were retrieved from the Web-of-Science-Core-Collection and analyzed regarding regional differences, the authors' sex, subtopics of stroke, as well as international research collaborations.

Results

A total of 132 different countries participated, with the USA contributing most publications with 4,614 (21.9%), followed by China with 3,872 (18.3%), and Germany with 1,120 (5.3%). Analyzing the scientific quality of different countries by H-index, the USA ranked first with an H-index of 202, followed by Germany (H-index 135) and the United Kingdom (UK;H-index 129). The most frequently used topic was “Clinical Neurology” with 9,028 publications. Among all first authors attributed to their sex, 32.3% of all first authors were female and 67.7% were male (4,335 vs. 9,097). The proportion of female last authors was comparatively lower at 22.4% (3,083 publications) compared with 77.6% male authors (10,658 publications). There was a broad network of international collaborations.

Conclusions

Research in ischemic stroke has substantially increased over time. Scientists from the USA have the highest number of publications, followed by China and Germany. Measured by the H-index, the USA held the highest publication quality, followed by Germany and the UK. The scientific landscape was male-dominated with 67.7% of all first authors being male. Worldwide international collaborations play a major role in ischemic stroke research.

Systematic Cross-biospecimen Evaluation of DNA Extraction Kits for Long- and Short-read Multi-metagenomic Sequencing Studies

High-quality DNA extraction is a crucial step in metagenomic studies. Bias by different isolation kits impairs the comparison across datasets. A trending topic is, however, the analysis of multiple metagenomes from the same patients to draw a holistic picture of microbiota associated with diseases. We thus collected bile, stool, saliva, plaque, sputum, and conjunctival swab samples and performed DNA extraction with three commercial kits. For each combination of the specimen type and DNA extraction kit, 20-gigabase (Gb) metagenomic data were generated using short-read sequencing. While profiles of the specimen types showed close proximity to each other, we observed notable differences in the alpha diversity and composition of the microbiota depending on the DNA extraction kits. No kit outperformed all selected kits on every specimen. We reached consistently good results using the Qiagen QiAamp DNA Microbiome Kit. Depending on the specimen, our data indicate that over 10 Gb of sequencing data are required to achieve sufficient resolution, but DNA-based identification is superior to identification by mass spectrometry. Finally, long-read nanopore sequencing confirmed the results (correlation coefficient > 0.98). Our results thus suggest using a strategy with only one kit for studies aiming for a direct comparison of multiple microbiotas from the same patients.

Effects of Resistant Starch on Symptoms, Fecal Markers, and Gut Microbiota in Parkinson's Disease - The RESISTA-PD Trial

The composition of the gut microbiota is linked to multiple diseases, including Parkinson's disease (PD). Abundance of bacteria producing short-chain fatty acids (SCFAs) and fecal SCFA concentrations are reduced in PD. SCFAs exert various beneficial functions in humans. In the interventional, monocentric, open-label clinical trial "Effects of Resistant Starch on Bowel Habits, Short Chain Fatty Acids and Gut Microbiota in Parkinson'sDisease" (RESISTA-PD; ID: NCT02784145), we aimed at altering fecal SCFAs by an 8-week prebiotic intervention with resistant starch (RS). We enrolled 87 subjects in three study-arms: 32 PD patients received RS (PD + RS), 30 control subjects received RS, and 25 PD patients received solely dietary instructions. We performed paired-end 100 bp length metagenomic sequencing of fecal samples using the BGISEQ platform at an average of 9.9 GB. RS was well-tolerated. In the PD + RS group, fecal butyrate concentrations increased significantly, and fecal calprotectin concentrations dropped significantly after 8 weeks of RS intervention. Clinically, we observed a reduction in non-motor symptom load in the PD + RS group. The reference-based analysis of metagenomes highlighted stable alpha-diversity and beta-diversity across the three groups, including bacteria producing SCFAs. Reference-free analysis suggested punctual, yet pronounced differences in the metagenomic signature in the PD + RS group. RESISTA-PD highlights that a prebiotic treatment with RS is safe and well-tolerated in PD. The stable alpha-diversity and beta-diversity alongside altered fecal butyrate and calprotectin concentrations call for long-term studies, also investigating whether RS is able to modify the clinical course of PD.

Emerging concepts of miRNA therapeutics: from cells to clinic

MicroRNAs (miRNAs) are very powerful genetic regulators, as evidenced by the fact that a single miRNA can direct entire cellular pathways via interacting with a broad spectrum of target genes. This property renders miRNAs as highly interesting therapeutic tools to restore cell functions that are altered as part of a disease phenotype. However, this strength of miRNAs is also a weakness because their cellular effects are so numerous that off-target effects can hardly be avoided. In this review, we point out the main challenges and the strategies to specifically address the problems that need to be surmounted in the push toward a therapeutic application of miRNAs. Particular emphasis is given to approaches that have already found their way into clinical studies.

Is Substantial Lymphovascular Space Invasion Prognostic for Clinical Outcomes in Type II Endometrial Cancer?

AIMS

Substantial lymphovascular space invasion (LVSI) compared with none or focal LVSI is predictive of lymph node involvement and worse clinical outcomes in endometrioid-type endometrial carcinoma. We aimed to quantify the incidence of substantial LVSI in type II (clear cell and serous) endometrial cancer and correlate the extent of LVSI with clinical outcomes.

MATERIALS AND METHODS

A retrospective review was conducted on type II endometrial cancer patients who underwent surgical management from July 2017 to December 2019 using the three-tier LVSI scoring system. Binary logistic regression and Cox regression were used to analyse predictors of lymph node involvement or survival outcomes, respectively. The Kaplan-Meier method and Log-rank test were used to analyse differences in locoregional disease-free survival (LR-DFS), distant metastasis disease-free survival (DM-DFS) and overall survival between patients with substantial versus none/focal LVSI.

RESULTS

In 79 patients with type II endometrial carcinoma, no LVSI, focal LVSI and substantial LVSI was present in 48.1%, 15.2% and 36.7% of patients, respectively. Lymph nodes were involved in 0.0% with no LVSI, 20.0% with focal LVSI and 60.0% with substantial LVSI (P < 0.001). The median follow-up was 22.2 months. In patients with none/focal versus substantial LVSI, the 2-year LR-DFS and DM-DFS rates were 91.5% versus 71.4% (P = 0.01) and 90.2% versus 63.8% (P = 0.005), respectively. On univariate analysis, myometrial invasion ≥50%, tumour size ≥3.6 cm, substantial versus none/focal LVSI, lymph node involvement and omission of adjuvant radiotherapy were significant predictors for worse LR-DFS and DM-DFS (P < 0.05).

DISCUSSION

Substantial LVSI has a high incidence in type II pathology at our institution and predicts for lymph node involvement and worse clinical outcomes.

Deep learning-based detection of eosinophilic esophagitis

BACKGROUND

For eosinophilic esophagitis (EoE), a substantial diagnostic delay is still a clinically relevant phenomenon. Deep learning-based algorithms have demonstrated potential in medical image analysis. Here we establish a convolutional neuronal network (CNN)-based approach that can distinguish the appearance of EoE from normal findings and candida esophagitis.

METHODS

We trained and tested a CNN using 484 real-world endoscopic images from 134 subjects consisting of three classes (normal, EoE, and candidiasis). Images were split into two completely independent datasets. The proposed approach was evaluated against three trainee endoscopists using the test set. Model-explainability was enhanced by deep Taylor decomposition.

RESULTS

Global accuracy (0.915 [95 % confidence interval (CI) 0.880-0.940]), sensitivity (0.871 [95 %CI 0.819-0.910]), and specificity (0.936 [95 %CI 0.910-0.955]) were significantly higher than for the endoscopists on the test set. Global area under the receiver operating characteristic curve was 0.966 [95 %CI 0.954-0.975]. Results were highly reproducible. Explainability analysis found that the algorithm identified the characteristic signs also used by endoscopists.

CONCLUSIONS

Complex endoscopic classification tasks including more than two classes can be solved by CNN-based algorithms. Therefore, our algorithm may assist clinicians in making the diagnosis of EoE.