Search

Background: Gastrointestinal mucosal damage due to human immunodeficiency virus (HIV) infection leads to microbial translocation and immune activation, contributing to the development of non-infectious comorbidities (NICM) in people living with HIV (PLWH). Additionally, persistent proviral HIV-1 in the gut-associated lymphatic tissue (GALT) can trigger immunological changes in the epithelial environment, impacting the mucosal barrier. However, the role of zonulin, a modulator of epithelial tight junctions in GALT during HIV infection, remains poorly understood. Methods: We measured zonulin in serum and intestinal tissue sections from five treatment-naive (HIV+NAIVE) and 10 cART-treated (HIV+cART) HIV+ individuals, along with 11 controls (CTRL). We compared zonulin levels with clinical characteristics, inflammatory markers (IFN-α, CXCR3, and PD-1), and the viral reservoir in peripheral blood (PB) and terminal ileum (TI). Results: Upon HIV infection, TI was found to harbor more HIV DNA than PB. Circulating zonulin levels were highest in HIV+NAIVE compared to HIV+cART or CTRL. Surprisingly, in the gut tissue sections, zonulin levels were higher in CTRL than in HIV+ individuals. Elevated circulating zonulin levels were found to be correlated with CD4+T-cell depletion in PB and TI, and with intestinal IFN-α. Conclusions: The findings of this study indicate a shift in zonulin levels from the gut to the bloodstream in response to HIV infection. Furthermore, elevated systemic zonulin levels are associated with the depletion of intestinal CD4+ T cells and increased gut inflammation, suggesting a potential link between systemic zonulin and intestinal damage. Gaining insight into the regulation of gut tight junctions during HIV infection could offer valuable understanding for preventing NICM in PLWH. [ABSTRACT FROM AUTHOR]

Simple Summary: The prediction of tumor response after neoadjuvant FLOT therapy is highly necessary. The use of deep learning on gastroesophageal biopsies enabled us to extract predictive information. This prediction model could be easily applied in clinical decision making. Patients could avoid unnecessary treatment or receive an intensified FLOT therapy. Background: The aim of this study was to establish a deep learning prediction model for neoadjuvant FLOT chemotherapy response. The neural network utilized clinical data and visual information from whole-slide images (WSIs) of therapy-naïve gastroesophageal cancer biopsies. Methods: This study included 78 patients from the University Hospital of Cologne and 59 patients from the University Hospital of Heidelberg used as external validation. Results: After surgical resection, 33 patients from Cologne (42.3%) were ypN0 and 45 patients (57.7%) were ypN+, while 23 patients from Heidelberg (39.0%) were ypN0 and 36 patients (61.0%) were ypN+ (p = 0.695). The neural network had an accuracy of 92.1% to predict lymph node metastasis and the area under the curve (AUC) was 0.726. A total of 43 patients from Cologne (55.1%) had less than 50% residual vital tumor (RVT) compared to 34 patients from Heidelberg (57.6%, p = 0.955). The model was able to predict tumor regression with an error of ±14.1% and an AUC of 0.648. Conclusions: This study demonstrates that visual features extracted by deep learning from therapy-naïve biopsies of gastroesophageal adenocarcinomas correlate with positive lymph nodes and tumor regression. The results will be confirmed in prospective studies to achieve early allocation of patients to the most promising treatment. [ABSTRACT FROM AUTHOR]

While current guidelines recommend performing endoscopy within 24 h in case of acute upper gastrointestinal bleeding (AUGIB), the precise timing remains an issue of debate. Lactate is an established parameter for risk stratification in a variety of medical emergencies. This study evaluated the predictive ability of elevated lactate levels in identifying patients with UGIB, who may benefit from emergent endoscopy. We retrospectively analyzed all patients with elevated lactate levels, who presented to our emergency department between 01 January 2015 and 31 December 2019 due to suspected AUGIB. Of 134 included cases, 81.3% had an Charlson comorbidity index of ≥3 and 50.4% presented with shock. Fifteen (11.2%) patients died and mortality rates rose with increasing lactate levels. Emergent endoscopy within 6 h (EE) and non-EE were performed in 64 (47.8%) and 70 (52.2%) patients, respectively. Patients who underwent EE had lower systolic blood pressure (107.6 mmHg vs. 123.2 mmHg; p = 0.001) and received blood transfusions more frequently (79.7% vs 64.3%; p = 0.048), but interestingly need for endoscopic intervention (26.6% vs 20.0%; p = 0.37), rebleeding (17.2% vs. 15.7%; p = 0.82) and mortality (9.4% vs. 11.4%; p = 0.7) did not differ significantly. In conclusion, our findings support the recommendations of current guidelines to perform non-EE after sufficient resuscitation and management of comorbid illnesses. [ABSTRACT FROM AUTHOR]

Backgound Autosomal dominant polycystic kidney disease (ADPKD) is the most common inherited kidney disease, and leads to a steady loss of kidney function in adulthood. The variable course of the disease makes it necessary to identify the patients with rapid disease progression who will benefit the most from targeted therapies and interventions. Currently, magnetic resonance imaging–based volumetry of the kidney is the most commonly used tool for this purpose. Biomarkers that can be easily and quantitatively determined, which allow a prediction of the loss of kidney function, have not yet been established in clinical practice. The glycoprotein Dickkopf 3 (DKK3) which is secreted in the renal tubular epithelium upon stress and contributes to tubulointerstitial fibrosis via the Wnt signaling pathway, was recently described as a biomarker for estimating risk of kidney function loss, but has not been investigated for ADPKD. This study aimed to obtain a first insight into whether DKK3 may indeed improve outcome prediction in ADPKD in the future. Methods In 184 ADPKD patients from the AD(H)PKD registry and 47 healthy controls, the urinary DKK3 (uDKK3) levels were determined using ELISA. Multiple linear regression was used to examine the potential of these values in outcome prediction. Results ADPKD patients showed significantly higher uDKK3 values compared with the controls (mean 1970 ± 5287 vs 112 ± 134.7 pg/mg creatinine). Furthermore, there was a steady increase in uDKK3 with an increase in the Mayo class (A/B 1262 ± 2315 vs class D/E 3104 ± 7627 pg/mg creatinine), the best-established biomarker of progression in ADPKD. uDKK3 also correlated with estimated glomerular filtration rate (eGFR). Patients with PKD1 mutations show higher uDKK3 levels compared with PKD2 patients (PKD1 : 2304 ± 5119; PKD2 : 506.6 ± 526.8 pg/mg creatinine). Univariate linear regression showed uDKK3 as a significant predictor of future eGFR slope estimation. In multiple linear regression this effect was not significant in models also containing height-adjusted total kidney volume and/or eGFR. However, adding both copeptin levels and the interaction term between copeptin and uDKK3 to the model resulted in a significant predictive value of all these three variables and the highest R2 of all models examined (∼0.5). Conclusion uDKK3 shows a clear correlation with the Mayo classification in patients with ADPKD. uDKK3 levels correlated with kidney function, which could indicate that uDKK3 also predicts a disproportionate loss of renal function in this collective. Interestingly, we found an interaction between copeptin and uDKK3 in our prediction models and the best model containing both variables and their interaction term resulted in a fairly good explanation of variance in eGFR slope compared with previous models. Considering the limited number of patients in these analyses, future studies will be required to confirm the results. Nonetheless, uDKK3 appears to be an attractive candidate to improve outcome prediction of ADPKD in the future. [ABSTRACT FROM AUTHOR]

The goal of the present study was to develop a convolutional neural network for the detection of bleedings in capsule endoscopy videos using realistic clinical data from one single-centre. Capsule endoscopy videos from all 133 patients (79 male, 54 female; meanage = 53.73 years, SDage = 26.13) who underwent capsule endoscopy at our institution between January 2014 and August 2018 were screened for pathology. All videos were screened for pathology by two independent capsule experts and confirmed findings were checked again by a third capsule expert. From these videos, 125 pathological findings (individual episodes of bleeding spanning a total of 5696 images) and 103 non-pathological findings (sections of normal mucosal tissue without pathologies spanning a total of 7420 images) were used to develop and validate a neural network (Inception V3) using transfer learning. The overall accuracy of the model for the detection of bleedings was 90.6% [95%CI: 89.4%–91.7%], with a sensitivity of 89.4% [95%CI: 87.6%–91.2%] and a specificity of 91.7% [95%CI: 90.1%–93.2%]. Our results show that neural networks can detect bleedings in capsule endoscopy videos under realistic, clinical conditions with an accuracy of 90.6%, potentially reducing reading time per capsule and helping to improve diagnostic accuracy. [ABSTRACT FROM AUTHOR]