OBJECTIVE: Endoscopic full-thickness resection (EFTR) for submucosal tumors (SMTs) has been technically challenging. This retrospective study aimed to evaluate the feasibility, safety, and efficacy of EFTR for upper gastrointestinal (GI) SMTs, including extraluminal lesions.
METHODS: We retrospectively investigated 232 patients with SMTs who underwent EFTR from January 2014 to August 2023. Clinicopathologic characteristics, procedure-related parameters, adverse events (AEs), and follow-up outcomes were assessed in all patients.
RESULTS: The en-bloc resection and en-bloc with R0 resection rates were 98.7% and 96.1%, respectively. The average endoscopic tumor size measured 17.2 ± 8.7 mm, ranging from 6 to 50 mm. The resection time and suture time were 49.0 ± 19.4 min and 22.5 ± 11.6 min, respectively. In all, 39 lesions (16.8%) exhibited predominantly extraluminal growth. Gastrointestinal stromal tumors (GISTs) were the predominant pathology, accounting for 78.4% of the cases. Twenty-one patients (9.1%) encountered complications, including pneumothorax (1/232, 0.43%), hydrothorax (1/232, 0.43%), localized peritonitis (3/232, 1.29%), and fever (16/232, 6.9%). Although the incidence of postoperative fever was notably higher in the predominantly extraluminal group (7/39, 17.9%) compared to the predominantly intraluminal group (9/193, 4.7%, P = 0.008), there were no significant differences in outcomes of the EFTR procedure. No instances of recurrence were observed during the mean follow-up period of 3.7 ± 2.3 years.
CONCLUSIONS: EFTR was found to be feasible, safe, and effective for resecting upper GI SMTs, including lesions with predominantly extraluminal growth. Further validation in a prospective study is warranted.

BACKGROUND: Nonadherence to imatinib is common in patients with gastrointestinal stromal tumor (GIST), which is associated with poor prognosis and financial burden. The primary aim of this study was to investigate the adherence rate in patients with GIST and subsequently develop a model based on machine learning (ML) and deep learning (DL) techniques to identify the associated factors and predict the risk of imatinib nonadherence.
METHODS: All eligible patients completed four sections of questionnaires. After the data set was preprocessed, statistically significance variables were identified and further processed to modeling. Six ML and four DL algorithms were applied for modeling, including eXtreme gradient boosting, light gradient boosting machine (LGBM), categorical boosting, random forest, support vector machine, artificial neural network, multilayer perceptron, NaiveBayes, TabNet, and Wide&Deep. The optimal ML model was used to identify potential factors for predicting adherence.
RESULTS: A total of 397 GIST patients were recruited. Nonadherence was observed in 185 patients (53.4%). LGBM exhibited superior performance, achieving a mean f1_score of 0.65 and standard deviation of 0.12. The predominant indicators for nonadherent prediction of imatinib were cognitive functioning, whether to perform therapeutic drug monitoring (if_TDM), global health status score, social support, and gender.
CONCLUSIONS: This study represents the first real-world investigation using ML techniques to predict risk factors associated with imatinib nonadherence in patients with GIST. By highlighting the potential factors and identifying high-risk patients, the multidisciplinary medical team can devise targeted strategies to effectively address the daily challenges of treatment adherence.

BACKGROUND: Imatinib (IMA) has received approval as the primary treatment for gastrointestinal stromal tumors (GIST). Nonetheless, approximately half of the patients with advanced GIST show disease advancement following IMA treatment. Presently, the efficacy of secondary and tertiary medications in addressing various GIST secondary mutations is somewhat restricted. Consequently, there is a significant medical demand for the creation of kinase inhibitors that extensively block secondary drug-resistant mutations in advanced GIST. Ripretinib (RPT) is a new, switch-control tyrosine kinase inhibitors that can suppress different mutations of KIT and PDGFRA via a dual mechanism of action.
OBJECTIVE: To investigate the literature on RPT to assess an effective, safe, and successful treatment strategy against advanced GIST.
METHODS: The present systematic review and meta-analysis was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. PubMed, Embase, Cochrane, Web of Science and ClinicalTrials.gov databases were screened from January 1, 2003 to May 1, 2024.
RESULTS: A total of 4 studies were included, with a total of 507 patients enrolled. The objective response rate (ORR) of the RPT-treated advanced GIST was 17% (95%CI: 0.11-0.27), while the disease control rate (DCR) was 66% (95%CI: 0.59-0.73). The overall occurrence of adverse events with varying degrees was 97% (95%CI: 0.93-1), whereas that of grade ≥ 3 adverse reactions was 42% (95%CI: 0.28-0.63). The sensitivity analysis revealed that omitting some studies did not yield statistically notable variances in the aggregate data regarding the ORR, DCR, and the occurrence of adverse events of grade 3 or higher. The publication bias was absent because no significant asymmetry was observed in Begg\'s funnel plot in all studies.
CONCLUSIONS: RPT has favorable efficacy profiles in GIST patients, but the adverse reactions are obvious, and patient management needs to be strengthened to achieve better safety and tolerability.

Understanding the determinants of long-term liver metastasis (LM) outcomes in gastrointestinal stromal tumor (GIST) patients is crucial. We established the feature selection model of intratumoral microbiome at the surgery, achieving robust predictive accuracies of 0.953 and 0.897 AUCs in discovery (n = 74) and validation (n = 34) cohorts, respectively. Notably, despite the significant reduction in LM occurrence with adjuvant imatinib (AI) treatment, intratumoral microbiome exerted independently stronger effects on post-operative LM. Employing both 16S and full-length rRNA sequencing, we pinpoint intracellular Shewanella algae as a foremost LM risk factor in both AI- and non-AI-treated patients. Experimental validation confirmed S. algae\'s intratumoral presence in GIST, along with migration/invasion-promoting effects on GIST cells. Furthermore, S. algae promoted LM and impeded AI treatment in metastatic mouse models. Our findings advocate for incorporating intratumoral microbiome evaluation at surgery, and propose S. algae as a therapeutic target for LM suppression in GIST.

Imatinib-induced skin rash poses a significant challenge for patients with gastrointestinal stromal tumor, often resulting in treatment interruption or discontinuation and subsequent treatment failure. However, the underlying mechanism of imatinib-induced skin rashes in gastrointestinal stromal tumor patients remains unclear. A total of 51 patients (27 with rash and 24 without rash) were enrolled in our study. Blood samples were collected concomitantly with the onset of clinical manifestations of rashes, and simultaneously collecting clinical relevant information. The imatinib concentration and untargeted metabolomics were performed by ultra-high-performance liquid chromatography-tandem mass spectrometry. There were no significant differences in age, gender, imatinib concentration and white blood cells count between the rash group and the control group. However, the rash group exhibited a higher eosinophil count (P<0.05) and lower lymphocyte count (P<0.05) compared to the control group. Untargeted metabolomics analysis found that 105 metabolites were significantly differentially abundant. The univariate analysis highlighted erucamide, linoleoylcarnitine, and valine betaine as potential predictive markers (AUC≥0.80). Further enriched pathway analysis revealed primary metabolic pathways, including sphingolipid signaling pathway, sphingolipid metabolism, cysteine and methionine metabolism, biosynthesis of unsaturated fatty acids, arginine and proline metabolism, and biosynthesis of amino acids. These findings suggest that the selected differential metabolites could serve as a foundation for the prediction and management of imatinib-induced skin rash in gastrointestinal stromal tumor patients.

Gastrointestinal stromal tumor (GIST) is the most common mesenchymal tumor of the alimentary tract. The prognosis depends on the primary site, and small intestinal GISTs have a worse prognosis than gastric GISTs. Molecularly targeted drugs to inhibit tyrosine kinase activity of KIT were used for unresectable or recurrent GISTs. However, secondary resistance to the drugs is often acquired, and treatments based on other mechanisms are needed. Previously, we reported that cell adhesion molecule 1 (CADM1) was highly expressed in most of small intestinal GISTs but not in most of gastric GISTs. In the present study, we examined whether the antibody-drug conjugate (ADC) with anti-CADM1 antibody and monomethyl auristatin E (anti-CAD-ADC) shows anti-tumor effect on CADM1-expressing human GIST cells. The ADC adhibited in this study was previously used for CADM1-expressing human mesothelioma cells and showed anti-tumor effect for them in vitro. GIST-T1 cell line of gastric origin which scarcely expresses CADM1 and GIST-T1 cells transfected with CADM1 cDNA (GIST-T1-CAD cells) which highly expresses CADM1 and represents small intestinal GIST were used. In vitro, anti-CAD-ADC showed remarkable cytotoxic activity on GIST-T1-CAD cells, but control ADC did not. Both anti-CAD-ADC and control ADC did not show anti-tumor effect on original GIST-T1 cells. When GIST-T1-CAD cells were subcutaneously injected to the nude mice, intravenous administration of anti-CAD-ADC showed inhibitory effect for tumor enlargement. Tumor of GIST-T1 cells grew even after anti-CAD-ADC injection. When GIST-T1-CAD cells were injected into peritoneal cavity of the SCID mice, intraperitoneal administration of anti-CAD-ADC showed reduction of the peritoneal tumor. On the other hand, peritoneal tumor grew after control ADC administration. Tissue and organ damage due to administration of anti-CAD-ADC was not apparent by macroscopic and histological examinations in mice. These results indicate that anti-CAD-ADC could have apparent anti-tumor effect on CADM1-expressing human GIST cells both in in vitro and in vivo mouse models.

Currently, due to limited long-term evidence, there remains some controversy surrounding the recommended postoperative monitoring strategy for primary low-risk gastrointestinal stromal tumors (GISTs). This study recruited a total of 532 patients diagnosed with very low-risk and low-risk GISTs who underwent endoscopic resection from 2015 to 2021, including 460 very low-risk patients and 72 low-risk patients. Descriptive statistical analysis was used to evaluate the clinical and pathological characteristics of GIST patients, and Kaplan-Meier methods were employed for survival analysis. The results showed that the 5-year recurrence-free survival rates for very low-risk and low-risk patients were 98.5% and 95.9%, respectively. The 5-year disease-specific survival rates for both groups were 100%. Additionally, the 5-year overall survival rates were 99.7% for very low-risk patients and 100% for low-risk patients (P = 0.69). Therefore, it is suggested that routine follow-up monitoring, including endoscopic surveillance and imaging, may not be necessary for very low-risk and low-risk GISTs after endoscopic resection.

Gastrointestinal stromal tumors (GIST) are the most common mesenchymal-derived tumors of the GI tract. They can occur throughout the GI tract, and the survival time of some patients can be improved by first-line targeted therapy with imatinib. However, there are some limitations with imatinib treatment. Immunotherapy for GIST has attracted much attention in recent years, and as one of the most abundant cells in the GIST microenvironment, M2 macrophages play an important role in disease progression. They have unique anti-inflammatory and pro-tumorigenic effects and are one target for immunotherapy. This review summarizes the connection between different factors and the programmed death receptor-1/programmed death ligand-1 pathway and M2 macrophages to reactivate or enhance anti-tumor immunity and improve imatinib efficacy, and to provide new ideas for GIST immunotherapy.

Using Tyrosine Kinase Inhibitors (TKIs) for gastrointestinal stromal tumors (GIST) has significantly reduced the risk of recurrence and prolonged survival. Immunotherapy has demonstrated efficacy in multiple solid tumors, but its effectiveness in GIST remains uncertain. Although early clinical studies indicate good tolerability of immunotherapy in patients, the efficacy is not as desired. Therefore, identifying the subset of GIST patients who benefit from immunotherapy and coordinating the relationship between immunotherapy and TKI treatment are crucial issues to be explored. In this review, we aims to provide a retrospective analysis of relevant literature and find that GIST patients exhibit a rich presence of tumor-infiltrating immune cells, which play critical roles in the immune surveillance and evasion processes of tumors. This review incorporates a selection of 48 articles published between 2002 and 2023, sourced from PubMed, EBSCO, and Google Scholar databases.

OBJECTIVE: The objective was to assess and examine radiomics models derived from contrast-enhanced CT for their predictive capacity using the sub-regional radiomics regarding the Ki-67 proliferation index (PI) in patients with pathologically confirmed gastrointestinal stromal tumors (GIST).
METHODS: In this retrospective study, a total of 412 GIST patients across three institutions (223 from center 1, 106 from center 2, and 83 from center 3) was enrolled. Radiomic features were derived from various sub-regions of the tumor region of interest employing the K-means approach. The Least Absolute Shrinkage and Selection Operator (LASSO) regression was employed to identify features correlated with Ki-67 PI level in GIST patients. A support vector machine (SVM) model was then constructed to predict the high level of Ki-67 (Ki-67 index >8%), drawing on the radiomics features from each sub-region within the training cohort.
RESULTS: After features selection process, 6, 9, 9, 7 features were obtained to construct SVM models based on sub-region 1, 2, 3 and the entire tumor, respectively. Among different models, the model developed by the sub-region 1 achieved an area under the receiver operating characteristic curve (AUC) of 0.880 (95% confidence interval [CI]: 0.830 to 0.919), 0.852 (95% CI: 0.770-0.914), 0.799 (95% CI: 0.697-0.879) in the training, external test set 1, and 2, respectively.
CONCLUSIONS: The results of the present study suggested that SVM model based on the sub-regional radiomics features had the potential of predicting Ki-67 PI level in patients with GIST.