Prediction of risk of malignancy of gastrointestinal stromal tumors by endoscopic ultrasonography

Development and validation of a machine-learning model for preoperative risk of gastric gastrointestinal stromal tumors

BACKGROUND

Gastrointestinal stromal tumors (GISTs) have malignant potential, and treatment varies according to risk. However, no specific protocols exist for preoperative assessment of the malignant potential of gastric GISTs (gGISTs). This study aimed to use machine learning (ML) to develop and validate clinically relevant preoperative models to predict the malignant potential of gGISTs.

METHODS

This study screened patients diagnosed with gGISTs at the Affiliated Hospital of North Sichuan Medical College. Moreover, this study employed the Least Absolute Shrinkage and Selection Operator (LASSO) and logistic regression to identify risk factors. Subsequently, an ensemble of ML models was used to determine the optimal classifier. In addition, this study used SHapley Additive exPlanations (SHAP) for tailored risk profiling.

RESULTS

This study included 318 patients with gGISTs. Using LASSO regression and multifactorial logistic regression, this study analyzed the training dataset, revealing that the presence of endoscopic ultrasound (EUS) high-risk features, tumor border clarity, tumor diameter, and monocyte-to-lymphocyte ratio (MLR) were significant predictors of high malignancy risk in gGIST. As determined by our ML approach, the logistic classification model demonstrated optimal performance, with area under the receiver operating characteristic curves of 0.919 for the training set and 0.925 for the test set. Furthermore, decision curve analysis confirmed the clinical relevance of the model.

CONCLUSION

High-risk EUS features, ill-defined tumor margins, larger tumor diameters, and elevated MLR independently predicted increased malignant potential in gGIST. This study developed logistic regression models based on these factors, which were further interpreted using the SHAP methodology. This analytical approach facilitated personalized therapeutic decision-making among diverse patient populations.

Findings of Endoscopic US and CT of Esophageal Disease

Various diseases can affect the esophagus. Endoscopic ultrasound (EUS), which provides precise information about the layers of the esophageal wall, is the primary approach used to investigate esophageal diseases. However, CT is one of the most important imaging modalities for diagnosing esophageal diseases as it can elucidate mediastinal involvement, adjacent lymphadenopathy, and distant disease spread. These two modalities complement each other in the diagnosis of esophageal diseases. Although radiologists may be unfamiliar with EUS procedures and their interpretation, understanding them aids in the differential diagnosis of esophageal conditions. This pictorial essay illustrates the EUS and CT findings of various esophageal diseases originating in the esophageal wall.

Combining Endoscopic Submucosa Dissection and Endoscopic Full-Thickness Resection for Resection of a Gastric Schwannoma Mimicking Gastrointestinal Stroma Tumor

Subepithelial gastric tumors are a diagnostic challenge. Endoscopic ultrasound allows differentiation. Lesions originating from the submucosal layer can be resected using endoscopic submucosal dissection (ESD). Surgery or endoscopic full-thickness resection (EFTR) techniques are alternatives. We present a patient with an 11 × 8 mm submucosal tumor in the gastric antrum suggestive of a gastrointestinal stromal tumor, originating from the muscularis propria. Eventually endoscopic resection was performed, combining ESD and EFTR (hybrid ESD-EFTR). Contrary to expectations, histology revealed a gastric schwannoma. This case illustrates an efficient and safe endoscopic hybrid technique for the removal of submucosal gastric lesions.

Development and validation of a preoperative risk nomogram prediction model for gastric gastrointestinal stromal tumors

BACKGROUND AND STUDY AIMS

Gastrointestinal stromal tumors (GIST) carry a potential risk of malignancy, and the treatment of GIST varies for different risk levels. However, there is no systematic preoperative assessment protocol to predict the malignant potential of GIST. The aim of this study was to develop a reliable and clinically applicable preoperative nomogram prediction model to predict the malignant potential of gastric GIST.

PATIENTS AND METHODS

Patients with a pathological diagnosis of gastric GIST from January 2015 to December 2021 were screened retrospectively. Univariate and multivariate logistic analyses were used to identify independent risk factors for gastric GIST with high malignancy potential. Based on these independent risk factors, a nomogram model predicting the malignant potential of gastric GIST was developed and the model was validated in the validation group.

RESULTS

A total of 494 gastric GIST patients were included in this study and allocated to a development group (n = 345) and a validation group (n = 149). In the development group, multivariate logistic regression analysis revealed that tumor size, tumor ulceration, CT growth pattern and monocyte-to- lymphocyte ratio (MLR) were independent risk factors for gastric GIST with high malignancy potential. The AUC of the model were 0.932 (95% CI 0.890-0.974) and 0.922 (95% CI 0.868-0.977) in the development and validation groups, respectively. The best cutoff value for the development group was 0.184, and the sensitivity and specificity at this value were 0.895 and 0.875, respectively. The calibration curves indicated good agreement between predicted and actual observed outcomes, while the DCA indicated that the nomogram model had clinical application.

CONCLUSIONS

Tumor size, tumor ulceration, CT growth pattern and MLR are independent risk factors for high malignancy potential gastric GIST, and a nomogram model developed based on these factors has a high ability to predict the malignant potential of gastric GIST.

Prevalence, natural progression, and clinical practices of upper gastrointestinal subepithelial lesions in Korea: a multicenter study

BACKGROUND/AIMS

This study aimed to evaluate the prevalence and natural progression of subepithelial lesions (SELs) in the upper gastrointestinal (UGI) tract.

METHODS

The medical records of patients with UGI SELs who underwent endoscopic screening at eight university hospitals between January and December 2010 were retrospectively investigated. The follow-up evaluations were performed until December 2016.

RESULTS

UGI SELs were found in 1,044 of the 65,233 participants screened (endoscopic prevalence, 1.60%; the total number of lesions, 1,062; mean age, 55.1±11.2 years; men, 53.6%). The median follow-up period was 48 (range, 8-74) months. SELs were most frequently found in the stomach (63.8%) and had a mean size of 9.9±6.1 mm. Endoscopic ultrasonography (EUS) was performed in 293 patients (28.1%). The most common lesions were leiomyomas, followed by gastrointestinal stromal tumors (GISTs), and ectopic pancreas. The proportions of SELs with malignant potential according to size were 3% (<1 cm), 22% (1-2 cm), 27% (2-3 cm), and 38% (≥3 cm). In gastric SELs larger than 1 cm, resections were performed in 20 patients because of an increase in size, of which 12 were found to be GISTs.

CONCLUSION

The prevalence of UGI SELs was 1.60%. Further, 23% of gastric SELs ≥1 cm were precancerous lesions, most followed by EUS and clinical decisions without initial pathological confirmation.

Risk stratification for 1- to 2-cm gastric gastrointestinal stromal tumors: visual assessment of CT and EUS high-risk features versus CT radiomics analysis

OBJECTIVES

To investigate the ability of CT and endoscopic sonography (EUS) in predicting the malignant risk of 1-2-cm gastric gastrointestinal stromal tumors (gGISTs) and to clarify whether radiomics could be applied for risk stratification.

METHODS

A total of 151 pathologically confirmed 1-2-cm gGISTs from seven institutions were identified by contrast-enhanced CT scans between January 2010 and March 2021. A detailed description of EUS morphological features was available for 73 gGISTs. The association between EUS or CT high-risk features and pathological malignant potential was evaluated. gGISTs were randomly divided into three groups to build the radiomics model, including 74 in the training cohort, 37 in validation cohort, and 40 in testing cohort. The ROIs covering the whole tumor volume were delineated on the CT images of the portal venous phase. The Pearson test and least absolute shrinkage and selection operator (LASSO) algorithm were used for feature selection, and the ROC curves were used to evaluate the model performance.

RESULTS

The presence of EUS- and CT-based morphological high-risk features, including calcification, necrosis, intratumoral heterogeneity, irregular border, or surface ulceration, did not differ between very-low and intermediate risk 1-2-cm gGISTs (p > 0.05). The radiomics model consisting of five radiomics features showed favorable performance in discrimination of malignant 1-2-cm gGISTs, with the AUC of the training, validation, and testing cohort as 0.866, 0.812, and 0.766, respectively.

CONCLUSIONS

Instead of CT- and EUS-based morphological high-risk features, the CT radiomics model could potentially be applied for preoperative risk stratification of 1-2-cm gGISTs.

KEY POINTS

• The presence of EUS- and CT-based morphological high-risk factors, including calcification, necrosis, intratumoral heterogeneity, irregular border, or surface ulceration, did not correlate with the pathological malignant potential of 1-2-cm gGISTs. • The CT radiomics model could potentially be applied for preoperative risk stratification of 1-2-cm gGISTs.

ACG Clinical Guideline: Diagnosis and Management of Gastrointestinal Subepithelial Lesions

Subepithelial lesions (SEL) of the GI tract represent a mix of benign and potentially malignant entities including tumors, cysts, or extraluminal structures causing extrinsic compression of the gastrointestinal wall. SEL can occur anywhere along the GI tract and are frequently incidental findings encountered during endoscopy or cross-sectional imaging. This clinical guideline of the American College of Gastroenterology was developed using the Grading of Recommendations Assessment, Development, and Evaluation process and is intended to suggest preferable approaches to a typical patient with a SEL based on the currently available published literature. Among the recommendations, we suggest endoscopic ultrasound (EUS) with tissue acquisition to improve diagnostic accuracy in the identification of solid nonlipomatous SEL and EUS fine-needle biopsy alone or EUS fine-needle aspiration with rapid on-site evaluation sampling of solid SEL. There is insufficient evidence to recommend surveillance vs resection of gastric gastrointestinal stromal tumors (GIST) <2 cm in size. Owing to their malignant potential, we suggest resection of gastric GIST >2 cm and all nongastric GIST. When exercising clinical judgment, particularly when statements are conditional suggestions and/or treatments pose significant risks, health-care providers should incorporate this guideline with patient-specific preferences, medical comorbidities, and overall health status to arrive at a patient-centered approach.

A Gastroenterologist's Approach to the Diagnosis and Management of Gastrointestinal Stromal Tumors

Gastrointestinal stromal tumors (GISTs) are the most common mesenchymal tumors of the gastrointestinal (GI) tract. These tumors have been shown to harbor oncogenic mutations of the c-kit tyrosine kinase receptor or platelet-derived growth factor receptor alpha (PDGFRA). Immunohistochemical analysis of GISTs allows for the differentiation of these tumors from other mesenchymal tumors of the GI tract such as leiomyomas and leiomyosarcomas. All GISTs have the potential to behave in a malignant fashion. Tumor location, size, and mitotic index are factors used to predict the risk of malignant behavior. Endoscopy and endoscopic ultrasound play a critical role in the diagnosis of GISTs and can yield important information to further risk-stratify tumors and determine management. This article provides a gastroenterologist's perspective on the diagnosis and management of GISTs.

Risk factors for the failure of endoscopic resection of gastric submucosal tumors: a long-term retrospective case-control study

OBJECTIVE

Endoscopic resection (ER) is an effective treatment method for gastric submucosal tumors (G-SMTs), but endoscopic resection failure requires emergency surgery. The purpose of this study was to assess potential risk factors for endoscopic resection failure.

METHODS

A total of 1041 patients with G-SMT undergoing endoscopic resection were enrolled. Twenty-five patients in whom endoscopic resection failed, requiring a transition to surgery midway through the operation, were included in the failed group, and 1016 patients who received successful endoscopic resection were included in the successful endoscopic resection group. Baseline and lesion characteristics were recorded, and the differences in tumor characteristics and risk factors for resection failure of G-SMT were analyzed. Sensitivity analysis was performed to detect the stability of the indicator.

RESULTS

Of the 1041cases included, there were 25 cases (2.4%) of failed endoscopic resection. Binary logistic analysis showed that the independent risk factors included tumors originating from deep muscularis propria(OR = 14.42, 95% CI 4.47-46.52), size > 3 cm (OR = 7.75, 95% CI 2.64-22.70), exophytic growth pattern (OR = 4.98, 95% CI 1.62-15.29), endoscopist with less experience (OR = 5.99, 95% CI 1.07-12.19), and irregular borders (OR = 4.13, 95% CI 1.40-12.19). The stable risk factors were tumors size, tumor origin and growth pattern according to sensitivity analysis.

CONCLUSIONS

Tumors originating from the deep muscularis propria, tumor size > 3 cm, endoscopists with less experience, an exophytic growth pattern, and irregular boundaries were found to be independent risk factors for endoscopic resection failure. To reduce the risk of endoscopic resection failure, physicians should carefully evaluate G-SMT characteristics preoperative.

Gastric subepithelial tumor: long-term natural history and risk factors for progression

BACKGROUND

Subepithelial tumors are often detected incidentally during upper gastrointestinal endoscopy. However, their natural history and clinical management have yet to be clearly established. The aim of this study was to evaluate the natural history and risk factors for progression of gastric SETs.

METHODS

The study retrospectively reviewed the medical records of patients who were diagnosed with gastric SET using upper gastrointestinal endoscopy between January 2005 and December 2017. Tumor progression was defined by a ≥ 25% increase in diameter.

RESULTS

Among 3237 patients, 1859 underwent serial upper gastrointestinal endoscopy for more than six months. Endoscopic ultrasonography was further performed in 733 (39.43%) patients. Resection was performed in 73 (3.93%) patients. Tumor progressed in 138 (7.42%) patients over a mean follow-up period of 59.41 months (range, 5-215 months). In progressed tumors, the mean initial size was 15.01 mm (range, 2-50 mm) and the mean size increment was 12.86 mm (range, 3-50 mm). Large initial tumor size (OR: 1.03, 95% CI: 1.01-1.05), surface ulcer or erosion (OR: 2.47, 95% CI: 1.21-5.06), lobulated shape (OR: 3.76, 95% CI: 2.00-7.06), and middle-third location (OR: 1.65, 95% CI: 1.08-2.52) were significant risk factors for tumor progression. Large SETs had higher rates of progression and tended to grow faster than smaller tumors (r = 0.44, p < 0.001).

CONCLUSIONS

The majority of gastric SETs did not increase in size during the long-term follow-up. Serial endoscopy may be sufficient as a follow-up tool for small SETs with intact overlying mucosa without lobulated shape.

A preoperative risk prediction model for high malignancy potential gastrointestinal stromal tumors of the stomach

BACKGROUND

Gastric gastrointestinal stromal tumors (GISTs) exhibit various degrees of aggression and malignant potential. However, no systematic preoperative evaluation strategy to predict the malignancy potential of gastric GISTs has yet been developed. This study aimed to develop a reliable and easy-to-use preoperative risk-scoring model for predicting high malignancy potential (HMP) gastric GISTs.

METHODS

The data of 542 patients with pathologically confirmed gastric GISTs who underwent resection were reviewed. Multivariate logistic regression analysis was used to identify significant predictors of HMP. The risk-scoring system (RSS) was based on the predictive factors for HMP, and its performance was validated using a split-sample approach.

RESULTS

A total of 239 of 542 (44.1%) surgically resected gastric GISTs had HMP. Multivariate analysis demonstrated that tumor size, location, and surface changes were independent risk factors for HMP. Based on the accordant regression coefficients, the presence of surface ulceration was assigned 1 point. Tumor sizes of 4-6 cm and > 6 cm were assigned 2 and 5 points, respectively. Two points were assigned to cardia or fundus locations. A score of 3 points was the optimal cut-off value for HMP prediction. HMP were found in 19.8% and 82.7% of the low and high-risk groups of the RSS, respectively. The area under the receiver-operating characteristic curve for predicting HMP was 0.81 (95% confidence interval (CI) 0.75-0.86). Discrimination was good after validation (0.75, 95% CI 0.69-0.81).

CONCLUSION

This simple RSS could be useful for predicting the malignancy potential of gastric GISTs and may aid preoperative clinical decision making to ensure optimal treatment.

Preoperative imaging of gastric GISTs underestimates pathologic tumor size: A retrospective, single institution analysis

BACKGROUND

How well imaging size agrees with pathologic size of gastric gastrointestinal stromal tumors (GISTs) is unknown. GIST risk stratification is based on pathologic size, location, and mitotic rate. To inform decision making, the size discrepancy between imaging and pathology for gastric GISTs was investigated.

METHODS

Imaging and pathology reports were reviewed for 113 patients. Bland-Altman analyses and intraclass correlation (ICC) assessed agreement of imaging and pathology. Changes in clinical risk category due to size discrepancy were identified.

RESULTS

Computed tomography (CT) (n = 110) and endoscopic ultrasound (EUS) (n = 50) underestimated pathologic size for gastric GISTs by 0.42 cm, 95% confidence interval (CI): (0.11, 0.73), p = 0.008 and 0.54 cm, 95% CI: (0.25, 0.82), p < 0.001, respectively. ICCs were 0.94 and 0.88 for CT and EUS, respectively. For GISTs ≤ 3 cm, size underestimation was 0.24 cm for CT (n = 28), 95% CI: (0.01, 0.47), p = 0.039 and 0.56 cm for EUS (n = 26), 95% CI: (0.27, 0.84), p < 0.0001. ICCs were 0.72 and 0.55 for CT and EUS, respectively. Spearman's correlation was ≥0.84 for all groups. For GISTs ≤ 3 cm, 6/28 (21.4% p = 0.01) on CT and 7/26 (26.9% p = 0.005) on EUS upgraded risk category using pathologic size versus imaging size. No GISTs ≤ 3 cm downgraded risk categories. Size underestimation persisted for GISTs ≤ 2 cm on EUS (0.39 cm, 95% CI: [0.06, 0.72], p = 0.02, post hoc analysis).

CONCLUSION

Imaging, particularly EUS, underestimates gastric GIST size. Caution should be exercised using imaging alone to risk-stratify gastric GISTs, and to decide between surveillance versus surgery.

Correlation of Endoscopic Ultrasonography Features with the Mitotic Index in 2- to 5-cm Gastric Gastrointestinal Stromal Tumors

BACKGROUND

Predicting the malignancy potential of gastrointestinal stromal tumor (GIST) before resection could improve patient management strategies as gastric GISTs with a low malignancy potential can be safely treated endoscopically, but surgical resection is required for those tumors with a high malignancy potential. This study aimed to evaluate endoscopic ultrasound (EUS) features of 2- to 5-cm gastric GISTs that might be used to predict their mitotic index using surgical specimens as the gold standard.

PATIENTS AND METHODS

Forty-nine patients (30 females and 19 males; mean age 55.1 ± 12.7 years) who underwent EUS examinations, followed by surgical resections of 2- to 5-cm gastric GISTs, were retrospectively reviewed.

RESULTS

The mean tumor size was 3.44 ± 0.97 (range 2.1-5.0) cm. A univariate analysis revealed no significant differences in age, sex, and tumor location in the low mitotic index and high mitotic index groups (all p > 0.05). In terms of EUS features, there were no significant differences in the mitotic indexes with respect to the shape, surface lobulation, border regularity, echogenicity, homogeneity, growth patterns, presence of mucosal ulceration, hyperechogenic foci, anechoic spaces, and hypoechoic halos (all p > 0.05). However, the tumor size was larger in the high mitotic index group than that in the low mitotic index group (3.97 ± 1.05 vs. 3.27 ± 0.9 cm, p = 0.03).

CONCLUSION

Conventional EUS features are not reliable for predicting the mitotic index of 2- to 5-cm gastric GISTs. Further modalities for predicting the mitotic index are needed to prevent unnecessary surgical resections in patients with a low risk of malignancy.

Gastric Calcifying Fibrous Tumor: An Easy Misdiagnosis as Gastrointestinal Stromal Tumor-A Systemic Review

Background and Objectives: Calcifying fibrous tumor (CFT) in the stomach is extremely rare and is easily misdiagnosed as a gastrointestinal stromal tumor (GIST). This study aims to determine the best method to differentiate between gastric CFT and GIST after a systemic review and meta-analysis. Materials and Methods: A systematic search of articles using electronic databases (MEDLINE, EMBASE, and LILACS) was conducted and resulted in 162 articles with 272 CFT cases published from January 1988 to September 2019. Results: Of these cases, 272 patients, 60 patients with gastric CFT (32 men and 28 women, mean age 49.2 years) were analyzed. The mean tumor size was 2.4 cm in patients with gastric CFT. Both endoscopic ultrasound (EUS) and computed tomography (CT) findings revealed well-defined (100% vs. 77.8%), heterogeneous (100% vs. 77.8%), iso-hypoechoic (71.4% vs. 33.3%), and calcified (85.7% vs. 77.8%) lesions, respectively. The majority of patients (53.3%) were symptomatic, with the most common symptom being abdominal discomfort (55.6%). None of the patients with gastric CFT showed recurrence after treatment, and most patients received nonendoscopic treatment (56%, n = 28/50). Both age and tumor size were statistically significant in patients with gastric CFT than GIST (49.2 vs. 65.0 years and 2.4 vs. 6.0 cm; both p < 0.001). The ratio of children among patients with CFT (5%) and GIST (0.05%) was also significantly different (p = 0.037). The calcification rates of gastric CFT had significantly higher calcification rates than GIST on images of EUS and CT (85.7% vs. 3.6% and 77.8% vs. 3.6%; both p < 0.001). Conclusions: Compared with patients with GIST, patients with gastric CFT were younger, had smaller tumor size, and were symptomatic. Furthermore, gastric CFT was well-defined, heterogeneous in the third layer, and had high calcification rates on the images.

Current Status of Endoscopic Ultrasonography in Gastrointestinal Subepithelial Tumors

Gastrointestinal subepithelial tumors (GSTs) are usually detected incidentally on endoscopic or radiologic examinations. In conventional endoscopy, a GST usually presents as a protuberant lesion with an intact mucosal surface. As the lesion is located beneath the mucosal layer of the gastrointestinal tract, conventional biopsy typically does not reveal the pathologic diagnosis. First, a GST should be differentiated from an extrinsic compression through the positional change of the patient during conventional endoscopic examination. In cases of GSTs originating from the gastrointestinal wall, endoscopic ultrasonography (EUS) can be beneficial for narrowing the differential diagnosis through delineation of echo findings and by determining the layer of origin. EUS findings can also help determine the management strategies for GSTs by making a differential diagnosis according to malignant potential.

The roles of CT and EUS in the preoperative evaluation of gastric gastrointestinal stromal tumors larger than 2 cm

OBJECTIVE

This study aimed to investigate the endoscopic ultrasound (EUS) and computed tomography (CT) features of gastric gastrointestinal stromal tumors (GISTs) for assessing potential malignancy and prognosis.

METHODS

Fifty consecutive patients with primary gastric GISTs larger than 2 cm were retrospectively enrolled in this study. The association of CT and EUS features with malignancy was analyzed using univariate and stepwise logistic regression method. The agreement between EUS/CT lesion size and pathologic tumor size was analyzed by calculating the intraclass correlation coefficient (ICC) value, and the association of imaging features with mitotic counts was further analyzed using univariate analysis. The Kaplan-Meier method and Cox proportional hazards models were used to assess the value of imaging features for predicting the prognosis of GIST patients.

RESULTS

Tumor size > 5 cm and an exophytic/mixed growth pattern on CT as well as tumor size > 5 cm and the presence of cystic spaces on EUS were independent predictors of highly malignant GISTs (all p < 0.05). The ICC values of CT/EUS lesion size relative to pathologic tumor size showed very good reliability (0.853 for EUS and 0.831 for CT). Only tumor shape and growth pattern on CT were significant for predicting mitotic index (both p < 0.05). Direct organ invasion on CT (p = 0.036; hazard ratio [HR] = 11.891) and serosal invasion on EUS (p = 0.015; HR = 8.223) were independent adverse prognostic factors.

CONCLUSIONS

CT features may be more useful than EUS features for predicting tumor mitotic index. In addition, preoperative imaging features can help predict the prognosis of gastric GISTs.

KEY POINTS

• Both CT and EUS features can be used for risk stratification of gastric GISTs larger than 2 cm. • CT features performed better than EUS features for predicting tumor mitotic index. • Preoperative imaging features can help predict the prognosis of gastric GISTs.

Current clinical management of gastrointestinal stromal tumor

Gastrointestinal stromal tumors (GISTs) are the most common malignant subepithelial lesions (SELs) of the gastrointestinal tract. They originate from the interstitial cells of Cajal located within the muscle layer and are characterized by over-expression of the tyrosine kinase receptor KIT. Pathologically, diagnosis of a GIST relies on morphology and immunohistochemistry [KIT and/or discovered on gastrointestinal stromal tumor 1 (DOG1) is generally positive]. The prognosis of this disease is associated with the tumor size and mitotic index. The standard treatment of a GIST without metastasis is surgical resection. A GIST with metastasis is usually only treated by tyrosine kinase inhibitors without radical cure; thus, early diagnosis is the only way to improve its prognosis. However, a GIST is usually detected as a SEL during endoscopy, and many benign and malignant conditions may manifest as SELs. Conventional endoscopic biopsy is difficult for tumors without ulceration. Most SELs have therefore been managed without a histological diagnosis. However, a favorable prognosis of a GIST is associated with early histological diagnosis and R0 resection. Endoscopic ultrasonography (EUS) and EUS-guided fine needle aspiration (EUS-FNA) are critical for an accurate diagnosis of SELs. EUS-FNA is safe and effective in enabling an early histological diagnosis and adequate treatment. This review outlines the current evidence for the diagnosis and management of GISTs, with an emphasis on early management of small SELs.

Pattern Classification for Gastrointestinal Stromal Tumors by Integration of Radiomics and Deep Convolutional Features

Predicting malignant potential is one of the most critical components of a computer-aided diagnosis system for gastrointestinal stromal tumors (GISTs). These tumors have been studied only on the basis of subjective computed tomography findings. Among various methodologies, radiomics, and deep learning algorithms, specifically convolutional neural networks (CNNs), have recently been confirmed to achieve significant success by outperforming the state-of-the-art performance in medical image pattern classification and have rapidly become leading methodologies in this field. However, the existing methods generally use radiomics or deep convolutional features independently for pattern classification, which tend to take into account only global or local features, respectively. In this paper, we introduce and evaluate a hybrid structure that includes different features selected with radiomics model and CNNs and integrates these features to deal with GISTs classification. The Radiomics model and CNNs are constructed for global radiomics and local convolutional feature selection, respectively. Subsequently, we utilize distinct radiomics and deep convolutional features to perform pattern classification for GISTs. Specifically, we propose a new pooling strategy to assemble the deep convolutional features of 54 three-dimensional patches from the same case and integrate these features with the radiomics features for independent case, followed by random forest classifier. Our method can be extensively evaluated using multiple clinical datasets. The classification performance (area under the curve (AUC): 0.882; 95% confidence interval (CI): 0.816-0.947) consistently outperforms those of independent radiomics (AUC: 0.807; 95% CI: 0.724-0.892) and CNNs (AUC: 0.826; 95% CI: 0.795-0.856) approaches.

A Bleeding Duodenal GIST Masquerading as Refractory Peptic Ulcer Disease

Gastrointestinal stromal tumors (GIST) are the most common mesenchymal tumors of the gastrointestinal tract; however, the occurrence of a GIST in the duodenum is rare. Our case demonstrates the importance of considering GIST in the evaluation of refractory duodenal ulcers, as well as the utilization of endoscopic ultrasound in the evaluation of these lesions.

Small EUS-suspected gastrointestinal stromal tumors of the stomach: An overview for the current state of management

Gastrointestinal stromal tumors (GIST) are the most common mesenchymal tumors found in the gastrointestinal (GI) tract, with the stomach being the most common site. They represent a distinct group of GI tumors originating from the interstitial cells of Cajal and are characterized by gain-of-function mutations of KIT. KIT oncoprotein serves as both diagnostic and therapeutic targets. Prognosis is related to size, mitotic activity, and site of the tumor. Asymptomatic, small endoscopic ultrasonography (EUS)-suspected GISTs are increasingly encountered with the wide availability of endoscopic/endosonographic examination. The majority of small GISTs are biologically indolent, albeit possibly harboring c-KIT gene mutations. An ongoing controversy exists regarding the management and surveillance policy for small gastric GISTs. A number of reports on the management of GISTs have been published, not confidently addressing the issue of gastric GISTs of small size. This work provides an overview on the current state of management considerations, specifically focusing on small EUS-suspected gastric GISTs, which are increasingly encountered by clinicians.

Synchronous occurrence of gastrointestinal stromal tumor and acute myeloid leukemia: A case report and review of the literature

Gastrointestinal stromal tumors (GISTs) originate from the mesenchymal tissue of the gastrointestinal tract. The pathogenesis of GIST is associated with the mutational activation of the receptor tyrosine kinase cluster of differentiation (CD)117 or platelet-derived growth factor receptor-α. Overall, ~60% of GISTs occur in the stomach. Clinically, GISTs may coexist with various types of cancer, including liver cancer, pancreatic tumors and lymphoma, either synchronously or metachronously. The present study reports the case of a patient with the synchronous occurrence of a CD117-positive GIST and acute myeloid leukemia. A 69-year-old man was hospitalized for heart palpitations and dizziness, and was diagnosed with acute myeloid leukemia (AML) by bone marrow aspiration and flow cytometry analysis. An abdominal computed tomograpy and gastroscopy revealed the presence of GIST. The patient received chemotherapy in combination with imatinib (400 mg/day), and the mass was removed 2 months later. To the best of our knowledge, the present study is the first reported case of the synchronous development of a CD117-positive GIST and AML. Additional studies are required in order to understand the association between GIST and hematological malignancies.

Current endoscopic ultrasound-guided approach to incidental subepithelial lesions: optimal or optional?

Subepithelial lesions (SEL) are identified during endoscopic procedures on a regular basis. They can occur anywhere in the gastrointestinal (GI) tract and are located beneath the normal epithelial layer, which explains why a tissue diagnosis is difficult to obtain with routine biopsies. Endoscopic ultrasound (EUS) is used to further characterize these lesions. EUS can distinguish intramural lesion from extramural compression. Furthermore, it allows allocation of intramural lesions to a specific layer of the GI wall and offers additional information as to whether a lesion could be benign or malignant. EUS also assists in choosing the optimal means of tissue acquisition. The choice of tissue acquisition is based on a number of factors, such as tumor size, EUS features, and location within the GI tract or within a specific layer of the GI wall. Furthermore, local expertise and patient factors should be considered when deciding whether tissue acquisition, surgical intervention or follow up is recommended. In this review we offer an EUS-guided approach to the evaluation of incidental SEL based on current evidence and point out areas of uncertainty, which explain why the proposed algorithmic approach may be optional rather than optimal.

Refractory left pleural effusion in an older patient: Atypical presentation of colorectal carcinoma

The current study reports the case of a 60-year-old male that presented with a four-month history of recurrent chest pain. Chest X-ray examination revealed a left-sided pleural effusion. A closed thoracic drainage procedure was performed, but the chest pain relapsed shortly afterwards. The pleural fluid was exudative, with no tumor cells detected. A computed tomography scan subsequently revealed a large mass located in the splenic curve of the colon, with involvement of the greater curvature of the stomach. Endoscopic biopsies confirmed the diagnosis of adenocarcinoma. A gastrointestinal stromal tumor was excluded by endoscopic ultrasonography and biopsy. A subtotal colectomy with partial excision of the stomach, diaphragm and left liver lobe was successfully performed, followed by the administration of six cycles of adjuvant chemotherapy. At present, the patient is asymptomatic and there is no evidence of tumor recurrence following a 12-month follow-up. The present study summarizes the characteristics of refractory left pleural effusion by colorectal malignancies.