Prognostic factors for liver metastasis in patients with small intestinal stromal tumor: A retrospective analysis of surveillance, epidemiology, and end results.

BACKGROUND: Liver metastasis (LIM) is the most common distant site of metastasis in small intestinal stromal tumors (SISTs). The aim of this study was to determine the risk and prognostic factors associated with LIM in patients with SISTs. METHODS: Patients diagnosed with gastrointestinal stromal tumors between 2010 and 2019 were selected from the Surveillance, Epidemiology, and End Results (SEER) database. Univariate and multivariate logistic regression models, as well as a Cox regression model were used to explore the risk factors associated with the development and prognosis of LIM. Additionally, the overall survival (OS) of patients with LIM was analyzed using the Kaplan-Meier method. Furthermore, a predictive nomogram was constructed, and the model's performance was evaluated using receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA). RESULTS: A total of 1582 eligible patients with SISTs were included, among whom 146 (9.2%) were diagnosed with LIM. Poor tumor grade, absence of surgery, later T-stage, and no chemotherapy were associated with an increased risk of developing LIM. The nomogram prediction model achieved an AUC of 0.810, 95% Confidence Interval (CI) 0.773-0.846, indicating good performance, and the calibration curve showed excellent accuracy in predicting LIM. The OS rate of patients with LIM was significantly lower than that of patients without LIM (p < 0.001). CONCLUSIONS: Patients with SISTs who are at high risk of developing LIM deserve more attention during follow-up, as LIM can significantly affect patient prognosis. The nomogram demonstrated good calibration and discrimination for predicting LIM.

Automated machine learning for early prediction of acute kidney injury in acute pancreatitis.

BACKGROUND: Acute kidney injury (AKI) represents a frequent and grave complication associated with acute pancreatitis (AP), substantially elevating both mortality rates and the financial burden of hospitalization. The aim of our study is to construct a predictive model utilizing automated machine learning (AutoML) algorithms for the early prediction of AKI in patients with AP. METHODS: We retrospectively analyzed patients who were diagnosed with AP in our hospital from January 2017 to December 2021. These patients were randomly allocated into a training set and a validation set at a ratio of 7:3. To develop predictive models for each set, we employed the least absolute shrinkage and selection operator (LASSO) algorithm along with AutoML. A nomogram was developed based on multivariate logistic regression analysis outcomes. The model's efficacy was assessed using receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis (DCA). Additionally, the performance of the model constructed via AutoML was evaluated using decision curve analysis (DCA), feature importance, SHapley Additive exPlanations (SHAP) plots, and locally interpretable model-agnostic explanations (LIME). RESULTS: This study incorporated a total of 437 patients who met the inclusion criteria. Out of these, 313 were assigned to the training cohort and 124 to the validation cohort. In the training and validation cohorts, AKI occurred in 68 (21.7%) and 29(23.4%) patients, respectively. Comparative analysis revealed that the AutoML models exhibited enhanced performance over traditional logistic regression (LR). Furthermore, the deep learning (DL) model demonstrated superior predictive accuracy, evidenced by an area under the ROC curve of 0.963 in the training set and 0.830 in the validation set, surpassing other comparative models. The key variables identified as significant in the DL model within the training dataset included creatinine (Cr), urea (Urea), international normalized ratio (INR), etiology, smoking, alanine aminotransferase (ALT), hypertension, prothrombin time (PT), lactate dehydrogenase (LDH), and diabetes. CONCLUSION: The AutoML model, utilizing DL algorithm, offers considerable clinical significance in the early detection of AKI among patients with AP.

High-fat diet promotes colitis-associated tumorigenesis by altering gut microbial butyrate metabolism.

Background: Dietary fat intake is associated with an increased risk of colitis associated cancer (CAC). A high-fat diet (HFD) leads to systemic low-grade inflammation. The colon is believed to be the first organ suffering from inflammation caused by the infiltration of pro-inflammatory macrophages, and promotes CAC progression. We explored the role of HFD in driving CAC by altering gut microbial butyrate metabolism. Methods: Changes in the gut microbiota caused by HFD were investigated via HFD treatment or fecal microbiota transplantation (FMT). The underlying mechanisms were further explored by analyzing the role of gut microbiota, microbial butyrate metabolism, and NLRP3 inflammasome in colon tissues in a CAC mouse model. Results: HFD accelerated CAC progression in mice, and it could be reversed by broad-spectrum antibiotics (ABX). 16S-rRNA sequencing revealed that HFD inhibited the abundance of butyrate-producing bacteria in the gut. The level of short-chain fatty acids (SCFAs), especially butyrate, in the gut of mice treated with HFD was significantly reduced. In addition, treatment with exogenous butyrate reversed the M1 polarization of proinflammatory macrophages, aggravation of intestinal inflammation, and accelerated tumor growth induced by HFD; the NLRP3/Caspase-1 pathway activated by HFD in the colon was also significantly inhibited. In vitro, macrophages were treated with lipopolysaccharide combined with butyrate to detect the M1 polarization level and NLRP3/Caspase-1 pathway expression, and the results were consistent with those of the in vivo experiments. Conclusion: HFD drives colitis-associated tumorigenesis by inducing gut microbial dysbiosis and inhibiting butyrate metabolism to skew macrophage polarization. Exogenous butyrate is a feasible new treatment strategy for CAC, and has good prospect for clinical application.

Aberrant expression of AKR1B1 indicates poor prognosis and promotes gastric cancer progression by regulating the AKT-mTOR pathway.

Gastric cancer (GC) is a common malignant tumor in the digestive tract and a major cause of global cancer death. Due to the limited access to early screening, many patients are diagnosed with advanced GC. Therefore, postoperative radiotherapy and chemotherapy possess limited efficacy in treating GC. AKR1B1 has been associated with tumorigenesis and metastasis across various tumors, becoming a potential therapeutic target for GC. However, its role and mechanism in GC remain unclear. In this study, AKR1B1 was elevated in GC tissue, depicting a poor prognosis. AKR1B1 is closely related to age, vascular and neural invasion, lymph node metastasis, and the TNM stage of GC. The developed survival prediction model suggested that AKR1B1 expression level is crucial in the prognosis of GC patients. Moreover, the expression level of AKR1B1 in GC tissues is closely associated with the AKT-mTOR pathway. In vitro and in vivo assays functional assays helped determine the oncogenic role of AKR1B1. Additionally, the knockdown of AKR1B1 expression level in GC cell lines could effectively suppress the AKT-mTOR pathway and inhibit the proliferation and migration of tumor cells. In conclusion, this study provides a theoretical basis to establish the potential association and regulatory mechanism of AKR1B1 while offering a new strategy for GC-targeted therapy.

Survival analysis of extragastrointestinal stromal tumors based on the SEER database: a population-based study.

BACKGROUND: Extragastrointestinal stromal tumors (EGISTs) are rare mesenchymal neoplasms that originate outside the gastrointestinal tract. However, the population-level survival analysis of EGIST remains poorly grasped. Therefore, we aimed to analyze the survival of EGIST patients using the Surveillance, Epidemiology, and End Results (SEER) database. METHODS: All patients diagnosed with GIST and EGIST between 2000 and 2019 were identified through utilization of the SEER database. Missing data were handled using multiple imputation methodology. Kaplan-Meier analyses and Cox proportional hazard models were employed to assess the influence of demographic and clinical characteristics on both overall survival (OS) and cancer-specific survival (CSS). RESULTS: A total of 13,330 patients were enrolled in the study, comprising 12,627 diagnosed with GIST and 703 with EGIST. EGIST patients demonstrated significantly poorer OS [hazard ratio (HR) 1.732, 95% confidence interval (CI) 1.522-1.970, P < 0.001] and CSS (HR 2.167, 95% CI 1.821-2.577, P < 0.001) compared to GIST patients. The mean 1-year, 3-year, 5-year, and 10-year OS rates for EGIST patients were 78.3%, 61.9%, 50.5%, and 32.5%, respectively, with corresponding mean CSS rates of 84.3%, 70.8%, 61.3%, and 46.5%. Multivariate Cox regression analysis identified age, race, sex, grade, size, and surgical type as independent risk factors for OS in EGIST patients, while age, sex, year of diagnosis, grade, surgical type, and radiation therapy were identified as independent risk factors for CSS. Patients with EGIST who underwent surgical treatment exhibited significantly higher 5-year OS rates (49.0% vs. 39.9%, P = 0.035) and CSS rates (63.9% vs. 53.0%, P = 0.028) compared to those who did not undergo surgery. CONCLUSIONS: EGIST patients have a poorer prognosis compared to GIST patients; however, surgical treatment has been shown to improve the prognosis.

Application Value of the Automated Machine Learning Model Based on Modified Computed Tomography Severity Index Combined With Serological Indicators in the Early Prediction of Severe Acute Pancreatitis.

BACKGROUND AND AIMS: Machine learning (ML) algorithms are widely applied in building models of medicine due to their powerful studying and generalizing ability. To assess the value of the Modified Computed Tomography Severity Index (MCTSI) combined with serological indicators for early prediction of severe acute pancreatitis (SAP) by automated ML (AutoML). PATIENTS AND METHODS: The clinical data, of the patients with acute pancreatitis (AP) hospitalized in Hospital 1 and hospital 2 from January 2017 to December 2021, were retrospectively analyzed. Serological indicators within 24 hours of admission were collected. MCTSI score was completed by noncontrast computed tomography within 24 hours of admission. Data from the hospital 1 were adopted for training, and data from the hospital 2 were adopted for external validation. The diagnosis of AP and SAP was based on the 2012 revised Atlanta classification of AP. Models were built using traditional logistic regression and AutoML analysis with 4 types of algorithms. The performance of models was evaluated by the receiver operating characteristic curve, the calibration curve, and the decision curve analysis based on logistic regression and decision curve analysis, feature importance, SHapley Additive exPlanation Plot, and Local Interpretable Model Agnostic Explanation based on AutoML. RESULTS: A total of 499 patients were used to develop the models in the training data set. An independent data set of 201 patients was used to test the models. The model developed by the Deep Neural Net (DL) outperformed other models with an area under the receiver operating characteristic curve (areas under the curve) of 0.907 in the test set. Furthermore, among these AutoML models, the DL and gradient boosting machine models achieved the highest sensitivity values, both exceeding 0.800. CONCLUSION: The AutoML model based on the MCTSI score combined with serological indicators has good predictive value for SAP in the early stage.

Assessment of a Real-world Learning Curve for the Endoscopic Resection of Gastric Gastrointestinal Stromal Tumors.

BACKGROUND: Endoscopic resection (ER) is a proven treatment for gastric gastrointestinal stromal tumors (gGISTs). We aimed to assess the learning curve (LC) associated with ER for gGISTs and identify determinants. METHODS: We conducted an analysis of 289 patients who underwent the ER of gGISTs by an experienced endoscopist. To characterize the LC, we employed cumulative sum analysis of the duration of surgery. The participants were divided into an early phase (cases 1-50) and a later phase (case 51-289), which were compared. Furthermore, we identified risk factors for the conversion from endoscopic to laparoscopic resection (LR). RESULTS: The durations of surgery and hospitalization were shorter, and there were fewer complications and fasting days in the later phase. The conversion rates to LR were 6.0% and 2.5% in the early and later phases, respectively. The tumor diameter (≥3.0 cm) and invasion beyond the muscularis propria were significant risk factors for conversion to LR (odds ratio 17.92, 95% confidence interval 2.66-120.87; and 58.03, 6.40-525.84; respectively). CONCLUSIONS: The LC for ER of gGISTs lasts for approximately 50 cases. In addition, tumors ≥3.0 cm in diameter and those that invade beyond the muscularis propria are more likely to require conversion to LR.

Risk factors for postoperative complications in endoscopic resection of gastric gastrointestinal stromal tumors: a multi-center analysis.

BACKGROUND: Endoscopic resection (ER) is widely used in treating gastric gastrointestinal stromal tumors (gGISTs); however, complications occur frequently after resection. We aimed to determine factors associated with postoperative complications for ER of gGISTs. METHODS: This was a retrospective, multi-center, observational study. Consecutive patients who underwent ER of gGISTs at five institutes from January 2013 to December 2022 were analyzed. The risk factors for delayed bleeding and postoperative infection were assessed. RESULTS: A total of 513 cases were finally analyzed. Of 513 patients, 27 (5.3%) had delayed bleeding and 69 (13.4%) had a postoperative infection. Multivariate analysis indicated that risk factors for delayed bleeding were long operative time (OR = 50.655; 95% CI, 13.777-186.252; P < 0.001) and severe intraoperative bleeding (OR = 4.731, 95% CI, 1.139-19.658; P = 0.032), and risk factors for postoperative infection were long operative time (OR = 13.749, 95% CI, 6.884-27.461; P < 0.001) and perforation (OR = 4.339, 95% CI, 2.178-8.644; P < 0.001). CONCLUSIONS: Our study indicated the risk factors for postoperative complications in ER of gGISTs. Long operation time is a common risk factor for delayed bleeding and postoperative infection. Patients with these risk factors should be given careful observation postoperatively.

Association of serum uric acid with hepatic steatosis detected by controlled attenuation parameter in the United States population.

BACKGROUND: The relationship between serum uric acid (SUA) and nonalcoholic fatty liver disease (NAFLD) has been previously reported. Controlled attenuation parameter (CAP) has better diagnostic performance than ultrasonography for assessing hepatic steatosis. The association of SUA with hepatic steatosis detected by CAP is worth further study. METHODS: The US population aged 20 years or older from the National Health and Nutrition Examination Survey (NHANES) was assessed. Hepatic steatosis was evaluated by the controlled attenuation parameter (CAP). NAFLD status was defined as CAP values of 268 dB/m without hepatitis B or C virus infection or considerable alcohol consumption. Multiple imputations were performed to fill in the missing covariate values. Linear regression, logistic regression, and smooth curve fitting were used to examine the association. RESULTS: In total, 3919 individuals participated in this study. There was a positive association between SUA (µmol/L) and CAP (ß = 0.14, 95% CI: 0.12-0.17, P < 0.01). After stratification by sex, a significant relationship between SUA and CAP existed in both males (ß = 0.12, 95% CI: 0.09-0.16, P < 0.01) and females (ß = 0.17, 95% CI: 0.14-0.20, P < 0.01) after multiple imputation. The inflection points of the threshold effect of SUA on CAP were 487.7 µmol/L in males and 386.6 µmol/L in females. There was a positive association between SUA (mg/dL) and NAFLD (OR = 1.30, 95% CI: 1.23-1.37, P < 0.01). After stratification by race, positive relationships were also observed. Meanwhile, a positive relationship existed between hyperuricemia and NAFLD (OR = 1.94, 95% CI: 1.64-2.30, P < 0.01). The positive relationship was more significant in females than in males (P for interaction < 0.01). CONCLUSIONS: There was a positive association between SUA and CAP, as well as between SUA and NAFLD. Subgroup studies stratified by sex and ethnicity demonstrated that the effects were consistent.

Development and validation of a preoperative difficulty scoring system for endoscopic resection of gastric gastrointestinal stromal tumor: a multi-center study.

BACKGROUND: Endoscopic resection (ER) is a promising technique for resecting gastric gastrointestinal stromal tumors (gGISTs); however, ER is technically challenging. This study aimed to develop and validate a difficulty scoring system (DSS) to determine the difficulty for ER of a gGIST. METHODS: This retrospective study enrolled 555 patients with gGISTs in multi-centers from December 2010 to December 2022. Data on patients, lesions, and outcomes of ER were collected and analyzed. A difficult case was defined as an operative time ≥ 90 min, or the occurrence of severe intraoperative bleeding, or conversion to laparoscopic resection. The DSS was developed in the training cohort (TC) and validated in the internal validation cohort (IVC) and external validation cohort (EVC). RESULTS: The difficulty occurred in 97 cases (17.5%). The DSS comprised the following: tumor size ≥ 3.0 cm (3 points) or 2.0-3.0 cm (1 point); location in the upper third of the stomach (2 points); invasion depth beyond the muscularis propria (2 points); lack of experience (1 point). The area under the curve (AUC) of DSS in IVC and EVC was 0.838 and 0.864, respectively, and the negative predictive value (NPV) was 0.923 and 0.972, respectively. The proportions of difficult operation in easy (score 0-3), intermediate (score 4-5), and difficult (score 6-8) categories were 6.5%, 29.4%, and 88.2% in the TC, 7.7%, 45.8%, and 85.7% in the IVC, and 7.0%, 29.4%, and 85.7% in the EVC, respectively. CONCLUSIONS: We developed and validated a preoperative DSS for ER of gGISTs based on tumor size, location, invasion depth, and endoscopists' experience. This DSS can be used to grade the technical difficulty before surgery.

Learning curve for endoscopic resection of gastric gastrointestinal stromal tumors: a single-center experience.

BACKGROUND: endoscopic resection (ER) is widely used in the treatment of gastric gastrointestinal stromal tumors (gGISTs). However, no studies have previously described the learning curve (LC) for ER of gGISTs. This study aimed to evaluate the LC based on multifarious operative outcomes. METHODS: one hundred consecutive patients who underwent ER of gGISTs by a single endoscopist from January 2017 to December 2022 were included. Patients were analyzed in groups of ten to minimize demographic differences, and operative time (OT), conversion rate, intraoperative and postoperative complication were assessed to evaluate the LC. Meanwhile, for the OT, the LC was further analyzed using the cumulative sum (CUSUM) method and patients were organized chronologically in three phases. RESULT: there was a statistically significant decrease in OT, conversion to laparoscopic surgery, and postoperative complication after 30 cases (median 80.0 min vs 56.0 min, p < 0.001; 10.0 % vs 0 %, p = 0.025; 33.3 % vs 10.0 %, p = 0.004), rate of intraoperative complications after 20 cases (15.0 % vs 1.3 %, p = 0.025). CUSUM chart demonstrated that OT increased dramatically before around 30 cases (phase 1) and decreased after 60 cases (phase 3), with a plateau phase in the middle 30 cases (phase 2). Among the three phases, the R0 resection and conversion rate were not significantly different. However, OT, intraoperative and postoperative complications were gradually decreased (p < 0.05). CONCLUSIONS: the LC of ER of gGISTs is approximately 60 cases. However, about 30 cases were sufficient to acquire skills to reduce complications and conversion rate during the ER procedure.

Automated machine learning to predict the difficulty for endoscopic resection of gastric gastrointestinal stromal tumor.

Background: Accurate preoperative assessment of surgical difficulty is crucial to the success of the surgery and patient safety. This study aimed to evaluate the difficulty for endoscopic resection (ER) of gastric gastrointestinal stromal tumors (gGISTs) using multiple machine learning (ML) algorithms. Methods: From December 2010 to December 2022, 555 patients with gGISTs in multi-centers were retrospectively studied and assigned to a training, validation, and test cohort. A difficult case was defined as meeting one of the following criteria: an operative time ≥ 90 min, severe intraoperative bleeding, or conversion to laparoscopic resection. Five types of algorithms were employed in building models, including traditional logistic regression (LR) and automated machine learning (AutoML) analysis (gradient boost machine (GBM), deep neural net (DL), generalized linear model (GLM), and default random forest (DRF)). We assessed the performance of the models using the areas under the receiver operating characteristic curves (AUC), the calibration curve, and the decision curve analysis (DCA) based on LR, as well as feature importance, SHapley Additive exPlanation (SHAP) Plots and Local Interpretable Model Agnostic Explanation (LIME) based on AutoML. Results: The GBM model outperformed other models with an AUC of 0.894 in the validation and 0.791 in the test cohorts. Furthermore, the GBM model achieved the highest accuracy among these AutoML models, with 0.935 and 0.911 in the validation and test cohorts, respectively. In addition, it was found that tumor size and endoscopists' experience were the most prominent features that significantly impacted the AutoML model's performance in predicting the difficulty for ER of gGISTs. Conclusion: The AutoML model based on the GBM algorithm can accurately predict the difficulty for ER of gGISTs before surgery.

Risk factors for conversion from endoscopic resection to laparoscopic resection for gastric gastrointestinal stromal tumors.

OBJECTIVE: Endoscopic resection (ER) of gastric gastrointestinal stromal tumors (gGISTs) is a commonly used treatment; however, it is associated with a risk of conversion to laparoscopic resection (LR). This study was performed to identify factors influencing conversion from ER to LR and the effects of conversion on outcomes. METHODS: The clinicopathological features of patients treated for gGISTs from March 2010 to May 2021 were retrospectively collected. Endpoints included the determination of risk factors associated with LR conversion, with comparisons of surgical outcomes with and without conversion. Propensity score matching was performed to compare the two groups. RESULTS: In total, 371 gGISTs were analyzed. Sixteen patients required conversion from ER to LR. Propensity score matching demonstrated that invasion depth (muscularis propria with exophytic growth) and gGIST size (≥3 cm) were independent risk factors for conversion to LR. The procedure duration (median, 160.5 vs. 60.0 minutes), postoperative hospitalization duration (median, 8 vs. 6 days), and postoperative fasting duration (median, 5 vs. 3 days) were significantly longer in patients who underwent conversion to LR. CONCLUSIONS: Accurate preoperative measurements of tumor size and invasion depth may help determine more appropriate surgical approaches for patients with gGISTs.

An evaluation of the use of double-curved endoscopes for gastric gastrointestinal stromal tumors.

BACKGROUND: Endoscopic full-thickness resection (EFTR) is a standard treatment method for gastric gastrointestinal stromal tumors (gGISTs). Evidence of the safety and efficacy of a double-curved endoscope (DCE) in EFTR of gGISTs is limited. We aimed to compare the operative outcomes of DCE versus single-curved endoscopes (SCE) in EFTR of gGISTs. MATERIAL AND METHODS: This retrospective observational study was conducted at four Chinese tertiary institutes. From January 2019 to November 2021, 104 patients who underwent EFTR by SCE (n = 57) or DCE (n = 47) were enrolled. One-to-one propensity score matching (PSM) was performed between the two groups to compare the demographics and operative outcomes. RESULTS: All gGISTs were resected successfully with no recurrence during follow-up. The median (range) tumor size was 1.2 (0.5, 3.5) cm in DCE and 2.0 (0.6, 4.8) cm in SCE (p < .001), and the procedure time was shorter in the DCE group than in the SCE group (50.0 min vs. 62.0 min, p < .05). After PSM, 41 pairs were selected, and no difference was noted in demographics. The procedure time was also shorter in the DCE group than in the SCE group (50.0 min vs. 55.0 min, p < .05). Subgroup analysis showed that the DCE group had a shorter procedure time in the gastric fundus than the SCE group (47.0 min vs. 55.0 min, p < .05). In multiple linear regression analysis, significant factors related to prolonged procedure time were the type of endoscope of SCE and larger tumor size (p < .05). CONCLUSIONS: EFTR of gGISTs using DCE is safe and effective. Compared with SCE, DCE had an advantage in terms of operative time, especially in the gastric fundus.

Learning curve for endoscopic resection of gastric gastrointestinal stromal tumors: a single-center experience

Background: endoscopic resection (ER) is widely used in the treatment of gastric gastrointestinal stromal tumors (gGISTs). However, no studies have previously described the learning curve (LC) for ER of gGISTs. This study aimed to evaluate the LC based on multifarious operative outcomes. Methods: one hundred consecutive patients who underwent ER of gGISTs by a single endoscopist from January 2017 to December 2022 were included. Patients were analyzed in groups of ten to minimize demographic differences, and operative time (OT), conversion rate, intraoperative and postoperative complication were assessed to evaluate the LC. Meanwhile, for the OT, the LC was further analyzed using the cumulative sum (CUSUM) method and patients were organized chronologically in three phases. Result: there was a statistically significant decrease in OT, conversion to laparoscopic surgery, and postoperative complication after 30 cases (median 80.0 min vs 56.0 min, p < 0.001; 10.0 % vs 0 %, p = 0.025; 33.3 % vs 10.0 %, p = 0.004), rate of intraoperative complications after 20 cases (15.0 % vs 1.3 %, p = 0.025). CUSUM chart demonstrated that OT increased dramatically before around 30 cases (phase 1) and decreased after 60 cases (phase 3), with a plateau phase in the middle 30 cases (phase 2). Among the three phases, the R0 resection and conversion rate were not significantly different. However, OT, intraoperative and postoperative complications were gradually decreased (p < 0.05). Conclusions: the LC of ER of gGISTs is approximately 60 cases. However, about 30 cases were sufficient to acquire skills to reduce complications and conversion rate during the ER procedure (AU)