Invariant feature based label correction for DNN when Learning with Noisy Labels.

Learning with Noisy Labels (LNL) methods have been widely studied in recent years, which aims to improve the performance of Deep Neural Networks (DNNs) when the training dataset contains incorrectly annotated labels. Popular existing LNL methods rely on semantic features extracted by the DNN to detect and mitigate label noise. However, these extracted features are often spurious and contain unstable correlations with the label across different environments (domains), which can occasionally lead to incorrect prediction and compromise the efficacy of LNL methods. To mitigate this insufficiency, we propose Invariant Feature based Label Correction (IFLC), which reduces spurious features and accurately utilizes the learned invariant features that contain stable correlation to correct label noise. To the best of our knowledge, this is the first attempt to mitigate the issue of spurious features for LNL methods. IFLC consists of two critical processes: The Label Disturbing (LD) process and the Representation Decorrelation (RD) process. The LD process aims to encourage DNN to attain stable performance across different environments, thus reducing the captured spurious features. The RD process strengthens independence between each dimension of the representation vector, thus enabling accurate utilization of the learned invariant features for label correction. We then utilize robust linear regression for the feature representation to conduct label correction. We evaluated the effectiveness of our proposed method and compared it with state-of-the-art (sota) LNL methods on four benchmark datasets, CIFAR-10, CIFAR-100, Animal-10N, and Clothing1M. The experimental results show that our proposed method achieved comparable or even better performance than the existing sota methods. The source codes are available at https://github.com/yangbo1973/IFLC.

What should be measured and reported in clinical trials for the treatment of patients with acute pancreatitis? A study protocol for establishing a core outcome set.

INTRODUCTION: Acute pancreatitis (AP) is characterised by inflammation of the exocrine pancreas, which potentially leads to local complications and organ failure resulting in significant morbidity and mortality. A long-term follow-up by an experienced team is needed. Currently, a variety of outcome measures are used in clinical trials for patients with AP. However, due to heterogeneous and selective outcome reporting across trials of interventions, it is hard to combine or compare the trial results compromising systematic evaluations of effectiveness and safety. A core outcome set is demanded to standardise reporting for the management of AP in clinical trials, so as to conduct systematic reviews and to improve the quality of the existing evidence base on the management of AP. We designed a study to establish a core outcome set (COS) on what indicators should be measured and reported in clinical trials of patients with AP (COS-AP). METHODS AND ANALYSIS: This study protocol outlines the following five phases: Phase I will be a systematic review of randomised control trials and semistructured interviews with patients to initially establish a preliminary list of potential outcomes. Phase II will be the recruitment of key stakeholders' groups comprising experts in pancreatic disease, clinical researchers, methodologists, journal editors and patients. Phase III will be two rounds of the Delphi surveys with key stakeholder groups. Phase IV will be a consensus on the outcomes that should be included in a final COS-AP. Phase V will be dissemination of COS-AP. ETHICS AND DISSEMINATION: Ethical approval for this study was obtained from the Biomedical Research Ethics Committee (BREC) of West China Hospital of Sichuan University (2020 No.691). The findings will be disseminated in peer-reviewed journals and meetings. TRIAL REGISTRATION: This study was registered with Core Outcome Measures in Effectiveness Trials (COMET) database as study 2573.

A predictive model for residual lesions after LEEP surgery in CIN III patients.

Background and aims: The residual lesions after Loop Electrosurgical Excision Procedure (LEEP) contributes to poor prognosis in patients with Cervical Intraepithelial Neoplasia Grade 3 (CIN3). The aim of this study is to establish an effective clinical predictive model for residual lesions in CIN3 patients after LEEP. Methods: A retrospective analysis was performed on 436 CIN3 patients who underwent total hysterectomy within 3 months after LEEP. Based on the post-hysterectomy pathologic, the patients were divided into the no residual group and residual group. Clinical parameters were compared between the two groups, and univariate and multivariate logistic regression analyses were conducted to identify independent risk factors for residual lesions in CIN3 patients after LEEP. Using R software, a nomogram model was established and its effectiveness was evaluated using calibration plots. Results: There were 178 cases in the residual group and 258 cases in the no residual group. The two groups had no significant difference in general characteristics (p > 0.05). It was found that Post-LEEP follow-up HPV, Post-LEEP follow-up TCT, and the Gland involvement were independent risk factors for residual lesions in CIN3 patients after LEEP (all p < 0.05). The consistency index (C-index) of the nomogram model for predicting residual lesions was 0.975 (0.962-0.988). Conclusion: The Post-LEEP follow-up HPV, Post-LEEP follow-up TCT, and Gland involvement are independent risk factors related to residual tissue after LEEP surgery in CIN3 patients. The constructed nomogram can effectively predict the presence of residual tissue after LEEP surgery in CIN3 patients and has good practical value.

Deep Learning Models for Severity Prediction of Acute Pancreatitis in the Early Phase From Abdominal Nonenhanced Computed Tomography Images.

OBJECTIVES: To develop and validate deep learning (DL) models for predicting the severity of acute pancreatitis (AP) by using abdominal nonenhanced computed tomography (CT) images. METHODS: The study included 978 AP patients admitted within 72 hours after onset and performed abdominal CT on admission. The image DL model was built by the convolutional neural networks. The combined model was developed by integrating CT images and clinical markers. The performance of the models was evaluated by using the area under the receiver operating characteristic curve. RESULTS: The clinical, Image DL, and the combined DL models were developed in 783 AP patients and validated in 195 AP patients. The combined models possessed the predictive accuracy of 90.0%, 32.4%, and 74.2% for mild, moderately severe, and severe AP. The combined DL model outperformed clinical and image DL models with 0.820 (95% confidence interval, 0.759-0.871), the sensitivity of 84.76% and the specificity of 66.67% for predicting mild AP and the area under the receiver operating characteristic curve of 0.920 (95% confidence interval, 0.873-0.954), the sensitivity of 90.32%, and the specificity of 82.93% for predicting severe AP. CONCLUSIONS: The DL technology allows nonenhanced CT images as a novel tool for predicting the severity of AP.

Predicting Persistent Acute Respiratory Failure in Acute Pancreatitis: The Accuracy of Two Lung Injury Indices.

BACKGROUND/AIMS: Early and accurate identification of patients with acute pancreatitis (AP) at high risk of persistent acute respiratory failure (PARF) is crucial. We sought to determine the accuracy of simplified Lung Injury Prediction Score (sLIPS) and simplified Early Acute Lung Injury (sEALI) for predicting PARF in ward AP patients. METHODS: Consecutive AP patients in a training cohort from West China Hospital of Sichuan University (n = 912) and a validation cohort from The First Affiliated Hospital of Nanchang University (n = 1033) were analyzed. PARF was defined as oxygen in arterial blood/fraction of inspired oxygen < 300 mmHg that lasts for > 48 h. The sLIPS was composed by shock (predisposing condition), alcohol abuse, obesity, high respiratory rate, low oxygen saturation, high oxygen requirement, hypoalbuminemia, and acidosis (risk modifiers). The sEALI was calculated from oxygen 2 to 6 L/min, oxygen > 6 L/min, and high respiratory rate. Both indices were calculated on admission. RESULTS: PARF developed in 16% (145/912) and 22% (228/1033) (22%) of the training and validation cohorts, respectively. In these patients, sLIPS and sEALI were significantly increased. sLIPS ≥ 2 predicted PARF in the training (AUROC 0.87, 95% CI 0.84-0.89) and validation (AUROC 0.81, 95% CI 0.78-0.83) cohorts. sLIPS was significantly more accurate than sEALI and current clinical scoring systems in both cohorts (all P < 0.05). CONCLUSIONS: Using routinely available clinical data, the sLIPS can accurately predict PARF in ward AP patients and outperforms the sEALI and current existing clinical scoring systems.

Optimising the measurement of intra-bladder pressure in patients with predicted severe acute pancreatitis.

BACKGROUND: Measuring intra-abdominal pressure (IAP) is important for management of patients with severe acute pancreatitis (SAP). Intra-bladder pressure (IBP) is an indirect index that reflects IAP, but measuring techniques vary. We sought to optimise IBP measuring techniques in predicted SAP patients. METHODS: Predicted SAP patients consecutively admitted between June 2018 and January 2020 were scrutinised. Eligible patients had their IBP monitored for the first 72 h at 6-h intervals, and were then sequentially allocated into three research scenarios: (1) in the supine position along with head of bed elevation(HoBE)of 0, 15 and 30° at various points including the iliac crest the midaxillary line, pubic symphysis, and right atrium level, instilled with 25 mL normal saline (NS) at room temperature (RT); (2) NS instillation volume from 0, 10, 25, 40-50 mL at the iliac crest with HoBE15 at RT; and (3) NS instillation (25 mL) at either RT or 37 °C with HoBE15. RESULTS: The dynamic IBP values measured at the pubic symphysis and iliac crest were fairly similar between HoBE0 and HoBE15 (all P > 0.05), but greatly increased at HoBE30 (all P < 0.01). IBP was significantly increased with escalating instillation volumes of NS (all P < 0.01 versus 0 mL NS), while there was no significant difference between 25 mL and 10 mL (P = 0.055). IBP was similar between NS at RT and under 37 °C (P = 0.643). CONCLUSION: In predicted SAP patients, measuring IBP at the iliac crest with HoBE15 after instilling 10 mL of NS seems to be appropriate for monitoring IAP.

AGI grade-guided chaiqin chengqi decoction treatment for predicted moderately severe and severe acute pancreatitis (CAP trial): study protocol of a randomised, double-blind, placebo-controlled, parallel-group, pragmatic clinical trial.

BACKGROUND: Acute pancreatitis (AP) is a common digestive disease with increased incidence globally but without internationally licenced pharmacological therapy. Moderately severe and severe acute pancreatitis (MSAP/SAP) contributes predominately for its morbidities and mortality and has been managed in West China Hospital for decades using the traditional Chinese medicinal formula chaiqin chengqi decoction (CQCQD). The current study tests whether the early administration of CQCQD will result in improved clinical outcomes in predicted MSAP/SAP patients. METHODS: This is a single-centre, randomised, controlled, double-blind pragmatic clinical trial. AP patients aged 18-75 admitted within 72 h of onset will be assessed at admission for enrolment. We excluded the predicted mild acute pancreatitis (Harmless Acute Pancreatitis Score > 2 at admission) and severe organ failure (Sequential Organ Failure Assessment [SOFA] score of respiratory, cardiovascular, or renal systems > 3) at admission. Eligible patients will be randomly allocated on a 1:1 basis to CQCQD or placebo control administration based on conventional therapy. The administration of CQCQD and placebo is guided by the Acute Gastrointestinal Injury grade-based algorithm. The primary outcome measure will be the duration of respiratory failure (SOFA score of respiratory system ≥ 2) within 28 days after onset. Secondary outcome measures include occurrence of new-onset any organ failure (SOFA score of respiratory, cardiovascular, or renal system ≥ 2) and new-onset persistent organ failure (organ failure lasts > 48 h), dynamic surrogate biochemical markers and clinical severity scores, gut-centred treatment modalities, local complications status, intensive care need and duration, surgical interventions, mortality, and length of hospital stay. Follow-up will be scheduled on 6, 12, and 26 weeks after enrolment to assess AP recurrence, local complications, the requirement for surgical interventions, all-cause mortality, and patient-reported outcomes. DISCUSSION: The results of this study will provide high-quality evidence to appraise the efficacy of CQCQD for the early management of AP patients. TRIAL REGISTRATION: Chictr.org.cn Registry ( ChiCTR2000034325 ). Registered on 2 July, 2020.

Impact of admission and early persistent stress hyperglycaemia on clinical outcomes in acute pancreatitis.

Background: To determine the impact of glucose levels at admission and during first week (early phase) on clinical outcomes in patients with acute pancreatitis (AP) and to investigate the relationship between stress hyperglycaemia (SHG) and hypertriglyceridaemia (HTG). Methods: Two independent and prospective databases were retrospectively analysed (n = 1792). Patients admitted with pain of less than 48 hours and confirmed AP were included. SHG was defined as admission blood glucose ≥ 10.00 mmol/L (non-diabetic) or ≥ 16.67 mmol/L (diabetic). Blood glucose records for the first week were inspected to determine whether SHG lasted ≥ 48 hours (persistent) or < 48 hours (transient). Clinical outcomes were compared between designated patient groups using multivariate and trend analyses. The correlation between SHG and HTG (serum triglyceride ≥ 5.65 mmol/L) was also analysed. Results: On admission, SHG was present in 27.8% (499/1792) patients; during the first 48 hours of admission, transient and persistent SHG was found in 31% (556/1792) and 8.0% (144/1792) patients, respectively. Admission SHG was associated with higher incidence of persistent organ failure, acute necrotic collection, major infection, and mortality as well as prolonged length of hospital stay (all P < 0.05). Duration of SHG was also associated with worsened clinical outcomes (all P < 0.05). In HTG-AP patients, more severe clinical outcomes were observed in those who concomitantly had SHG (P < 0.05). Conclusions: Admission and persistent SHG during the first week of admission worsens clinical outcomes of AP patients. These effects are more pronounced when admission HTG co-existed.

Safety and efficacy of intravenous hydromorphone patient-controlled analgesia versus intramuscular pethidine in acute pancreatitis: An open-label, randomized controlled trial.

Background: Hydromorphone patient-controlled analgesia (PCA) provides satisfactory postoperative pain therapy, but its effect has not been assessed in acute pancreatitis (AP). Aim: To assess the safety and efficacy of intravenous hydromorphone PCA for pain relief in AP. Methods: This open-label trial included AP patients admitted within 72 h of symptom onset, aged 18-70 years old, and with Visual Analog Scale (VAS) for pain intensity ≥5. They were randomized to receive intravenous hydromorphone PCA (0.05 mg/h with 0.2 mg on-demand) or intramuscular pethidine (50 mg as required) for three consecutive days. Intramuscular dezocine (5 mg on demand) was the rescue analgesia. The primary outcome was the change of VAS score recorded every 4 h for 3 days. Interim analysis was conducted by an Independent Data and Safety Monitoring Committee (IDSMC). Results: From 26 July 2019 to 15 January 2020, 77 patients were eligible for the intention-to-treat analysis in the interim analysis (39 in the hydromorphone group and 38 in the pethidine group). Baseline parameters were comparable between groups. No difference in VAS between the two groups was found. Hydromorphone PCA was associated with higher moderately severe to severe cases (82.1% vs. 55.3%, p = 0.011), acute peripancreatic fluid collections (53.9% vs. 28.9%, p = 0.027), more cumulative opioid consumption (median 46.7 vs. 5 mg, p < 0.001), higher analgesia costs (median 85.5 vs. 0.5 $, p < 0.001) and hospitalization costs (median 3,778 vs. 2,273 $, p = 0.007), and more adverse events (20.5% vs. 2.6%, p = 0.087). The per-protocol analysis did not change the results. Although a sample size of 122 patients was planned, the IDSMC halted further recruitment as disease worsening or worse clinical outcomes between the groups in the interim analysis. Conclusion: Hydromorphone PCA was not superior to pethidine in relieving pain in AP patients and might have worse clinical outcomes. Therefore, its use is not recommended. Clinical Trial Registration: Chictr.org.cn. ChiCTR1900025971.

Optimising fluid requirements after initial resuscitation: A pilot study evaluating mini-fluid challenge and passive leg raising test in patients with predicted severe acute pancreatitis.

BACKGROUND: The goals and approaches to fluid therapy vary through different stages of resuscitation. This pilot study was designed to test the safety and feasibility of a fluid therapy protocol for the second or optimisation stage of resuscitation in patients with predicted severe acute pancreatitis (SAP). METHODS: Spontaneously breathing patients with predicted SAP were admitted after initial resuscitation and studied over a 24-h period in a tertiary hospital ward. Objective clinical assessment (OCA; heart rate, mean arterial pressure, urine output, and haematocrit) was done at 0, 4, 8, 12, 18-20, and 24 h. All patients had mini-fluid challenge (MFC; 250 ml intravenous normal saline within 10 min) at 0 h and repeated at 4 and 8 h if OCA score ≥2. Patients who were fluid responsive (>10% change in stroke volume after MFC) received 5-10 ml/kg/h, otherwise 1-3 ml/kg/h until the next time point. Passive leg raising test (PLRT) was done at each time point and compared with OCA for assessing volume status and predicting fluid responsiveness. RESULTS: This fluid therapy protocol based on OCA, MFC, and PLRT and designed for the second stage of resuscitation was safe and feasible in spontaneously breathing predicted SAP patients. The PLRT was superior to OCA (at 0 and 8 h) for predicting fluid responsiveness and guiding fluid therapy. CONCLUSIONS: This pilot study found that a protocol for intravenous fluid therapy specifically for the second stage of resuscitation in patients with predicted SAP was safe, feasible, and warrants further investigation.

Anti-Inflammatory Effects and Molecular Mechanisms of Shenmai Injection in Treating Acute Pancreatitis: Network Pharmacology Analysis and Experimental Verification.

Background: Acute pancreatitis (AP) is an inflammatory disorder of the exocrine pancreas without specific treatment. Shenmai injection (SMI) was reported to eliminate the severity of experimental AP. This study aimed to explore the mechanisms underlying the synergistic protective effects of SMI on AP based on network pharmacology and experimental validation. Methods: Network pharmacology analysis and molecular docking based on identified components were performed to construct the potential therapeutic targets and pathways. The principal components of SMI were detected via ultra-high-performance liquid chromatography-coupled with quadrupole time-of-flight mass spectrometry (UHPLC-QTOF/MS). Effect of SMI and the identified components on cellular injury and IL6/STAT3 signaling was assessed on mouse pancreatic acinar cell line 266-6 cells. Finally, 4% sodium taurocholate (NaT) was used to induce AP model to assess the effects of SMI in treating AP and validate the potential molecular mechanisms. Results: By searching the TCMSP and ETCM databases, 119 candidate components of SMI were obtained. UHPLC-QTOF/MS analysis successfully determined the representative components of SMI: ginsenoside Rb1, ginsenoside Rg1, ginsenoside Re, and ophiopogonin D. Fifteen hub targets and eight related pathways were obtained to establish the main pharmacology network. Subnetwork analysis and molecular docking indicated that the effects of these four main SMI components were mostly related to the interleukin (IL) 6/STAT3 pathway. In vitro, SMI, ginsenoside Rb1, ginsenoside Rg1, ginsenoside Re, and ophiopogonin D increased the cell viability of NaT-stimulated mouse pancreatic acinar 266-6 cells and decreased IL6 and STAT3 expression. In vivo, 10 mL/kg SMI significantly alleviated the pancreatic histopathological changes and the expression of IL6 and STAT3 in the AP mice. Conclusion: This study demonstrated SMI may exert anti-inflammatory effects against AP by suppressing IL6/STAT3 activation, thus providing a basis for its potential use in clinical practice and further study in treating AP.

Predicting persistent organ failure on admission in patients with acute pancreatitis: development and validation of a mobile nomogram.

BACKGROUND: Early prediction of persistent organ failure (POF) is important for triage and timely treatment of patients with acute pancreatitis (AP). METHODS: All AP patients were consecutively admitted within 48 h of symptom onset. A nomogram was developed to predict POF on admission using data from a retrospective training cohort, validated by two prospective cohorts. The clinical utility of the nomogram was defined by concordance index (C-index), decision curve analysis (DCA), and clinical impact curve (CIC), while the performance by post-test probability. RESULTS: There were 816, 398, and 880 patients in the training, internal and external validation cohorts, respectively. Six independent predictors determined by logistic regression analysis were age, respiratory rate, albumin, lactate dehydrogenase, oxygen support, and pleural effusion and were included in the nomogram (web-based calculator: https://shina.shinyapps.io/DynNomapp/). This nomogram had reasonable predictive ability (C-indexes 0.88/0.91/0.81 for each cohort) and promising clinical utility (DCA and CIC). The nomogram had a positive likelihood ratio and post-test probability of developing POF in the training, internal and external validation cohorts of 4.26/31.7%, 7.89/39.1%, and 2.75/41%, respectively, superior or equal to other prognostic scores. CONCLUSIONS: This nomogram can predict POF of AP patients and should be considered for clinical practice and trial allocation.

Transcriptomics and Network Pharmacology Reveal the Protective Effect of Chaiqin Chengqi Decoction on Obesity-Related Alcohol-Induced Acute Pancreatitis via Oxidative Stress and PI3K/Akt Signaling Pathway.

Obesity-related acute pancreatitis (AP) is characterized by increasing prevalence worldwide and worse clinical outcomes compared to AP of other etiologies. Chaiqin chengqi decoction (CQCQD), a Chinese herbal formula, has long been used for the clinical management of AP but its therapeutic actions and the underlying mechanisms have not been fully elucidated. This study has investigated the pharmacological mechanisms of CQCQD in a novel mouse model of obesity-related alcohol-induced AP (OA-AP). The mouse OA-AP model was induced by a high-fat diet for 12 weeks and subsequently two intraperitoneal injections of ethanol, CQCQD was administered 2 h after the first injection of ethanol. The severity of OA-AP was assessed and correlated with changes in transcriptomic profiles and network pharmacology in the pancreatic and adipose tissues, and further docking analysis modeled the interactions between compounds of CQCQD and their key targets. The results showed that CQCQD significantly reduced pancreatic necrosis, alleviated systemic inflammation, and decreased the parameters associated with multi-organ dysfunction. Transcriptomics and network pharmacology analysis, as well as further experimental validation, have shown that CQCQD induced Nrf2/HO-1 antioxidant protein response and decreased Akt phosphorylation in the pancreatic and adipose tissues. In vitro, CQCQD protected freshly isolated pancreatic acinar cells from H2O2-elicited oxidative stress and necrotic cell death. The docking results of AKT1 and the active compounds related to AKT1 in CQCQD showed high binding affinity. In conclusion, CQCQD ameliorates the severity of OA-AP by activating of the antioxidant protein response and down-regulating of the PI3K/Akt signaling pathway in the pancreas and visceral adipose tissue.

Vitamin D and Pancreatitis: A Narrative Review of Current Evidence.

Emerging research indicates that vitamin D metabolic disorder plays a major role in both acute pancreatitis (AP) and chronic pancreatitis (CP). This has been demonstrated by studies showing that vitamin D deficiency is associated with pancreatitis and its anti-inflammatory and anti-fibrotic effects by binding with the vitamin D receptor (VDR). However, the role of vitamin D assessment and its management in pancreatitis remains poorly understood. In this narrative review, we discuss the recent advances in our understanding of the molecular mechanisms involved in vitamin D/VDR signaling in pancreatic cells; the evidence from observational studies and clinical trials that demonstrate the connection among vitamin D, pancreatitis and pancreatitis-related complications; and the route of administration of vitamin D supplementation in clinical practice. Although further research is still required to establish the protective role of vitamin D and its application in disease, evaluation of vitamin D levels and its supplementation should be important strategies for pancreatitis management according to currently available evidence.

Extracellular histones cause intestinal epithelium injury and disrupt its barrier function in vitro and in vivo.

Extracellular histones are cytotoxic to various cells and have been extensively proven a vital mediator of multiple organ injuries. However, the effect of extracellular histones on the intestine remains largely unknown. This study aimed to clarify the effect of extracellular histones on the intestine. IEC-6, a cell line of rat small intestinal epithelial crypt, and C57BL/6 or ICR mice were treated with histones. The IEC-6 cells treated with histones from 20 µg/mL to 200 µg/mL for 0-24 h displayed a decline of cell viability and an increase of cell death in a concentration- and time-dependent manner. Moreover, histones (100 µg/mL) induced IEC-6 apoptosis through activating caspase 3 and necroptosis through up-regulation of receptor-interacting serine/threonine protein kinase 1 and 3 (RIPK1 and RIPK3), phosphorylated mixed-lineage kinase domain-like protein (p-MLKL) along with the decrease of caspase-8. Histones treatment disturbed zonular occludens 1 (ZO-1) expression and increased permeability of IEC-6 cell monolayer. In vivo, histones 50 mg/kg injection caused mice intestinal edema, loss apex of villus, epithelial lifting down the sides of the villi, and increased neutrophil infiltration. Elevation of serum intestinal fatty acid binding protein (I-FABP), d-lactate, or Diamine oxidase (DAO) and loss of tight junction protein, ZO-1, at 3 h and 6 h after histones injection strongly indicated severe intestinal epithelium injury, which led to increased permeability of the intestine. In conclusion, extracellular histones cause intestinal epithelial damage via direct cytotoxicity. Consequently, intestinal epithelial tight junction and barrier integrity are disrupted, which may play pivotal roles in diverse diseases.

Stress Hyperglycemia Is Independently Associated with Persistent Organ Failure in Acute Pancreatitis.

BACKGROUND/AIMS: Stress hyperglycemia is common in critical illness but it has not been clearly studied in patients with acute pancreatitis (AP). This study aimed to investigate the specific blood glucose (BG) level that defines stress hyperglycemia and to determine the impact of stress hyperglycemia on clinical outcomes in AP patients. METHODS: AP patients admitted ≤ 48 h after abdominal pain onset were retrospectively analyzed. Patients were stratified by pre-existing diabetes and stress hyperglycemia was defined using stratified BG levels for non-diabetes and diabetes with clinical outcomes compared. RESULTS: There were 967 non-diabetic and 114 diabetic (10.5%) patients met the inclusion criteria and the clinical outcomes between these two groups were not significantly different. In non-diabetes, the cut-off BG level of ≥ 180 mg/dl was selected to define stress hyperglycemia with an 8.8-fold higher odds ratio for persistent organ failure (POF) (95% CI 5.4-14.3; P < 0.001). For diabetes, ≥ 300 mg/dl was selected with a 7.5-fold higher odds ratio for POF (95% CI 1.7-34.3; P = 0.009). In multivariable logistic regression, stress hyperglycemia was independently associated with POF, acute necrotic collection, major infection and mortality. The combination of BG and systemic inflammatory response syndrome (SIRS) score in predicting POF was better than SIRS or Glasgow score alone. CONCLUSIONS: This study identifies a cut-off BG level of ≥ 180 mg/dl and ≥ 300 mg/dl was optimal to define stress hyperglycemia for non-diabetic and diabetic AP patients, respectively. There was a significant relationship between stress hyperglycemia and adverse clinical outcomes.

Growth differentiation factor 15 is an early predictor for persistent organ failure and mortality in acute pancreatitis.

OBJECTIVES: Early prediction of persistent organ failure (POF) is crucial for patients with acute pancreatitis (AP). Growth differentiation factor 15 (GDF15), also known as macrophage inhibitory cytokine 1 (MIC-1), is associated with inflammatory responses. We investigated changes in plasma GDF15 and assessed its predictive value in AP. METHODS: The study included 290 consecutive patients with AP admitted within 36 h after symptoms onset. Clinical data obtained during hospitalization were collected. Plasma GDF15 levels were determined using enzyme-linked immunosorbent assays. The predictive value of GDF15 for POF was analyzed. RESULTS: There were 105 mild, 111 moderately severe, and 74 severe AP patients. Plasma GDF15 peak level were measured on admission, and significantly declined on the 3rd and 7th day. Admission GDF15 predicted POF and mortality with areas under the curve (AUC) of 0.847 (95% confidence interval [CI] 0.798-0.895) and 0.934 (95% CI 0.887-0.980), respectively. Admission GDF15, Bedside Index of Severity in Acute Pancreatitis, and hematocrit were independent factors for POF by univariate and multivariate logistic regression, and the nomogram built on these variables showed good performance (optimism-corrected c-statistic = 0.921). The combined predictive model increased the POF accuracy with an AUC 0.925 (95% CI 0.894-0.956), a net reclassification improvement of 0.3024 (95% CI: 0.1482-0.4565, P < 0.001), and an integrated discrimination index of 0.11 (95% CI 0.0497-0.1703; P < 0.001). CONCLUSIONS: Plasma GDF15 measured within 48 h of symptom onset could help predict POF and mortality in AP patients.

The unique pancreatic stellate cell gene expression signatures are associated with the progression from acute to chronic pancreatitis.

Chronic pancreatitis (CP) is characterized by irreversible fibro-inflammatory changes induced by pancreatic stellate cell (PSC). Unresolved or recurrent injury causes dysregulation of biological process following AP, which would cause CP. Here, we systematically identify genes whose expressions are unique to PSC by comparing transcriptome profiles among total pancreas, pancreatic stellate, acinar, islet and immune cells. We then identified candidate genes and correlated them with the pancreatic disease continuum by performing intersection analysis among total PSC and activated PSC genes, and genes persistently differentially expressed during acute pancreatitis (AP) recovery. Last, we examined the association between candidate genes and AP, and substantiated their potential as biomarkers in experimental AP and recurrent AP (RAP) models. A total of 68 genes were identified as highly and uniquely expressed in PSC. The PSC signatures were highly enriched with extracellular matrix remodeling genes and were significantly enriched in AP pancreas compared to healthy control tissues. Among PSC signature genes that comprised a fibrotic phenotype, 10 were persistently differentially expressed during AP recovery. SPARC was determined as a candidate marker for the pancreatic disease continuum, which was not only persistently differentially expressed even five days after AP injury, but also highly expressed in two clinical datasets of CP. Sparc was also validated as highly elevated in RAP compared to AP mice. This work highlights the unique transcriptional profiles of PSC. These PSC signatures' expression may help to identify patients with high risk of AP progression to CP.

Buyang Huanwu Decoction Enhances Revascularization via Akt/GSK3ß/NRF2 Pathway in Diabetic Hindlimb Ischemia.

BACKGROUND: Peripheral arterial disease (PAD) is a typical disease of atherosclerosis, most commonly influencing the lower extremities. In patients with PAD, revascularization remains a preferred treatment strategy. Buyang Huanwu decoction (BHD) is a popular Chinese herbal prescription which has showed effects of cardiovascular protection through conducting antioxidant, antiapoptotic, and anti-inflammatory effects. Here, we intend to study the effect of BHD on promoting revascularization via the Akt/GSK3ß/NRF2 pathway in diabetic hindlimb ischemia (HLI) model of mice. MATERIALS AND METHODS: All db/db mice (n = 60) were randomly divided into 6 groups by table of random number. (1) Sham group (N = 10): 7-0 suture thread passed through the underneath of the femoral artery and vein without occlusion. The remaining 5 groups were treated differently on the basis of the HLI (the femoral artery and vein from the inguinal ligament to the knee joint were transected and the vascular stump was ligated with 7-0 silk sutures) model: (2) HLI+NS group (N = 15): 0.2 ml NS was gavaged daily for 3 days before modeling and 14 days after occlusion; (3) HLI+BHD group (N = 15): 0.2 ml BHD (20 g/kg/day) was gavaged daily for 3 days before modeling and 14 days after occlusion; (4) HLI+BHD+sh-NC group (N = 8): local injection of adenovirus vector carrying the nonsense shRNA (Ad-GFP) in the hindlimbs of mice before treatment; (5) HLI+BHD+sh-NRF2 group (N = 8): knockdown of NRF2 in the hindlimbs of mice by local intramuscular injection of adenovirus vector carrying NRF2 shRNA (Ad-NRF2-shRNA) before treatment; and (6) HLI+BHD+LY294002 group (N = 4): intravenous injection of LY294002 (1.5 mg/kg) once a day for 14 days on the basis of the HLI+BHD group. Laser Doppler examination, vascular cast, and immunofluorescence staining were applied to detect the revascularization of lower limbs in mice. Western blot analysis was used to detect the expression of vascular endothelial growth factor (VEGF), interleukin-1beta (IL-1ß), interleukin-6 (IL-6), tumor necrosis factor- (TNF-) α, heme oxygenase-1 (HO-1), NAD(P)H dehydrogenase quinone-1 (NQO-1), catalase (CAT), glyceraldehyde-3-phosphate dehydrogenase (GAPDH), phosphorylated protein kinase B (p-AKT), and phosphorylated glycogen synthase kinase-3 beta (p-GSK3ß). HE staining was used to assess the level of muscle tissue damage and inflammation in the lower extremities. Local multipoint injection of Ad-NRF2-shRNA was used to knock down NRF2, and qPCR was applied to detect the mRNA level of NRF2. The blood glucose, triglyceride, cholesterol, MDA, and SOD levels of mice were tested using corresponding kits. The SPSS 20.0 software and GraphPad Prism 6.05 were used to do all statistics. Values of P < 0.05 were considered as statistically significant. Results and Conclusions. BHD could enhance the revascularization of lower limbs in HLI mice, while BHD has no effect on blood glucose and lipid level in db/db mice (P > 0.05). BHD could elevate the protein expression of VEGF, HO-1, NQO-1, and CAT (P < 0.05) and decrease the expression of IL-1ß, IL-6, and TNF-α (P < 0.05) in HLI mice. Meanwhile, BHD could activate NRF2 and promote the phosphorylation of AKT/GSK3ß during revascularization (P < 0.05). In contrast, knockdown of NRF2 impaired the protective effects of BHD on HLI (P < 0.05). LY294002 inhibited the upregulation of NRF2 activated by BHD through inhibiting the phosphorylation of the AKT/GSK3ß pathway (P < 0.05). The present study demonstrated that BHD could promote revascularization on db/db mice with HLI through targeting antioxidation, anti-inflammation, and angiogenesis via the AKT/GSK3ß/NRF2 pathway.

IHG-MA: Inductive heterogeneous graph multi-agent reinforcement learning for multi-intersection traffic signal control.

Multi-agent deep reinforcement learning (MDRL) has been widely applied in multi-intersection traffic signal control. The MDRL algorithms produce the decentralized cooperative traffic-signal policies via specialized multi-agent settings in certain traffic networks. However, the state-of-the-art MDRL algorithms seem to have some drawbacks. (1) It is desirable that the traffic-signal policies can be smoothly transferred to diverse traffic networks, however, the adopted specialized multi-agent settings hinder the traffic-signal policies to transfer and generalize to new traffic networks. (2) Existing MDRL algorithms which are based on deep neural networks cannot flexibly tackle a time-varying number of vehicles traversing the traffic networks. (3) Existing MDRL algorithms which are based on homogeneous graph neural networks fail to capture the heterogeneous features of objects in traffic networks. Motivated by the above observations, in this paper, we propose an algorithm, referred to as Inductive Heterogeneous Graph Multi-agent Actor-critic (IHG-MA) algorithm, for multi-intersection traffic signal control. The proposed IHG-MA algorithm has two features: (1) It conducts representation learning using a proposed inductive heterogeneous graph neural network (IHG), which is an inductive algorithm. The proposed IHG algorithm can generate embeddings for previously unseen nodes (e.g., new entry vehicles) and new graphs (e.g., new traffic networks). But unlike the algorithms based on the homogeneous graph neural network, IHG algorithm not only encodes heterogeneous features of each node, but also encodes heterogeneous structural (graph) information. (2) It also conducts policy learning using a proposed multi-agent actor-critic(MA), which is a decentralized cooperative framework. The proposed MA framework employs the final embeddings to compute the Q-value and policy, and then optimizes the whole algorithm via the Q-value and policy loss. Experimental results on different traffic datasets illustrate that IHG-MA algorithm outperforms the state-of-the-art algorithms in terms of multiple traffic metrics, which seems to be a new promising algorithm for multi-intersection traffic signal control.