Gastric heterotopic pancreas and stromal tumors smaller than 3 cm in diameter: clinical and computed tomography findings

An ensemble machine learning model assists in the diagnosis of gastric ectopic pancreas and gastric stromal tumors

OBJECTIVE

To develop an ensemble machine learning (eML) model using multiphase computed tomography (MPCT) for distinguishing between gastric ectopic pancreas (GEP) and gastric stromal tumors (GIST) in lesions < 3 cm.

METHODS

In this study, we retrospectively collected MPCT images from 138 patients between April 2017 and June 2023 across two centers. Cohort 1 comprised 94 patients divided into a training cohort and an internal validation cohort, while the 44 patients from Cohort 2 constituted the external validation cohort. Deep learning (DL) models were constructed based on the lesion region, and radiomics features were extracted to develop radiomics models, which were later integrated into the fusion model. Model performance was assessed through the analysis of the area under the receiver operating characteristic curve (AUROC). The diagnostic efficacy of the optimal model was compared with that of a radiologist. Additionally, the radiologist with the assistance of the eML model provides a secondary diagnosis, to assess the potential clinical value of the model.

RESULTS

After evaluation using an external validation cohort, the radiomics model demonstrated the highest performance in the venous phase, achieving AUROC of 0.87. The DL model showed optimal performance in the non-contrast phase, with AUROC of 0.81. The eML achieved the best performance across all models, with AUROC of 0.90. The use of eML-assisted analysis resulted in a significant improvement in the junior radiologist's accuracy, rising from 0.77 to 0.93 (p < 0.05). However, the senior radiologist's accuracy, while improving from 0.86 to 0.95, did not exhibit a statistically significant difference.

CONCLUSION

eML model based on MPCT can effectively distinguish between GEPs and GISTs < 3 cm.

CRITICAL RELEVANCE STATEMENT

The multiphase CT-based fusion model, incorporating radiomics and DL technology, proves effective in distinguishing between GEP and gastric stromal tumors, serving as a valuable tool to enhance diagnoses and offering references for clinical decision-making.

KEY POINTS

No studies yet differentiated these tumors via radiomics or DL. Radiomics and DL methodologies unveil potentially distinct phenotypes within lesions. Quantitative analysis on CT for GIST and ectopic pancreas. Ensemble learning aids accurate diagnoses, assisting treatment decisions.

Development of a multi-phase CT-based radiomics model to differentiate heterotopic pancreas from gastrointestinal stromal tumor

BACKGROUND

To investigate whether CT-based radiomics can effectively differentiate between heterotopic pancreas (HP) and gastrointestinal stromal tumor (GIST), and whether different resampling methods can affect the model's performance.

METHODS

Multi-phase CT radiological data were retrospectively collected from 94 patients. Of these, 40 with HP and 54 with GISTs were enrolled between April 2017 and November 2021. One experienced radiologist manually delineated the volume of interest and then resampled the voxel size of the images to 0.5 × 0.5 × 0.5 mm3, 1 × 1 × 1 mm3, and 2 × 2 × 2 mm3, respectively. Radiomics features were extracted using PyRadiomics, resulting in 1218 features from each phase image. The datasets were randomly divided into training set (n = 66) and validation set (n = 28) at a 7:3 ratio. After applying multiple feature selection methods, the optimal features were screened. Radial basis kernel function-based support vector machine (RBF-SVM) was used as the classifier, and model performance was evaluated using the area under the receiver operating curve (AUC) analysis, as well as accuracy, sensitivity, and specificity.

RESULTS

The combined phase model performed better than the other phase models, and the resampling method of 0.5 × 0.5 × 0.5 mm3 achieved the highest performance with an AUC of 0.953 (0.881-1), accuracy of 0.929, sensitivity of 0.938, and specificity of 0.917 in the validation set. The Delong test showed no significant difference in AUCs among the three resampling methods, with p > 0.05.

CONCLUSIONS

Radiomics can effectively differentiate between HP and GISTs on CT images, and the diagnostic performance of radiomics is minimally affected by different resampling methods.

Comparing endoscopic ultrasonography and double contrast-enhanced ultrasonography in the preoperative diagnosis of gastric stromal tumor

BACKGROUND

This study was designed to perform a comparative analysis between endoscopic ultrasonography (EUS) and double contrast-enhanced ultrasonography (DCEUS) for the preoperative diagnosis of gastric stromal tumors (GSTs).

METHODS

A retrospective study was conducted involving 139 patients with histologically confirmed GSTs. All patients preoperatively underwent DCEUS and EUS. The pathology reports were treated as the baseline and were retrospectively compared with the findings of EUS and DCEUS.

RESULTS

Of the 139 lesions, 120 and 113 were correctly identified by DCEUS and EUS, respectively, with an accuracy of 86.3% and 81.3%. The results revealed an insignificant difference between these two methods (p = 0.189).

CONCLUSIONS

DCEUS can display not only the locations, sizes, shapes, borders, internal echoes, but also show the blood perfusion patterns of GSTs. It is a highly accurate, noninvasive, and convenient method to be used at the pre-treatment stage.

A CT-based nomogram established for differentiating gastrointestinal heterotopic pancreas from gastrointestinal stromal tumor: compared with a machine-learning model

OBJECTIVE

To identify CT features and establish a nomogram, compared with a machine learning-based model for distinguishing gastrointestinal heterotopic pancreas (HP) from gastrointestinal stromal tumor (GIST).

MATERIALS AND METHODS

This retrospective study included 148 patients with pathologically confirmed HP (n = 48) and GIST (n = 100) in the stomach or small intestine that were less than 3 cm in size. Clinical information and CT characteristics were collected. A nomogram on account of lasso regression and multivariate logistic regression, and a RandomForest (RF) model based on significant variables in univariate analyses were established. Receiver operating characteristic (ROC) curve, mean area under the curve (AUC), calibration curve and decision curve analysis (DCA) were carried out to evaluate and compare the diagnostic ability of models.

RESULTS

The nomogram identified five CT features as independent predictors of HP diagnosis: age, location, LD/SD ratio, duct-like structure, and HU lesion/pancreas A. Five features were included in RF model and ranked according to their relevance to the differential diagnosis: LD/SD ratio, HU lesion/pancreas A, location, peritumoral hypodensity line and age. The nomogram and RF model yielded AUC of 0.951 (95% CI: 0.842-0.993) and 0.894 (95% CI: 0.766-0.966), respectively. The DeLong test found no statistically significant difference in diagnostic performance (p > 0.05), but DCA revealed that the nomogram surpassed the RF model in clinical usefulness.

CONCLUSION

Two diagnostic prediction models based on a nomogram as well as RF method were reliable and easy-to-use for distinguishing between HP and GIST, which might also assist treatment planning.

A computed tomography-based preoperative risk scoring system to distinguish lymphoepithelioma-like gastric carcinoma from non-lymphoepithelioma-like gastric carcinoma

Purpose

The aim of this study was to develop a preoperative risk scoring model for distinguishing lymphoepithelioma-like gastric carcinoma (LELGC) from non-LELGC based on contrast-enhanced computed tomography (CT) images.

Methods

Clinicopathological features and CT findings of patients with LELGC and non-LELGC in our hospital from January 2016 to July 2022 were retrospectively analyzed and compared. A preoperative risk stratification model and a risk scoring system were developed using logistic regression.

Results

Twenty patients with LELGC and 40 patients with non-LELGC were included in the training cohort. Significant differences were observed in Epstein–Barr virus (EBV) infection and vascular invasion between the two groups (p < 0.05). Significant differences were observed in the distribution of location, enhancement pattern, homogeneous enhancement, CT-defined lymph node status, and attenuations in the non-contrast, arterial, and venous phases (all p < 0.05). Enhancement pattern, CT-defined lymph node status, and attenuation in venous phase were independent predictors of LELGC. The optimal cutoff score of distinguishing LELGC from non-LELGC was 3.5. The area under the receiver operating characteristic curve, sensitivity, specificity, and accuracy of risk identification model in the training cohort were 0.904, 87.5%, 80.0%, and 85.0%, respectively. The area under the receiver operating characteristic curve, sensitivity, specificity, and accuracy of risk identification model in the validation cohort were 0.705 (95% CI 0.434–0.957), 75.0%, 63.6%, and 66.7%, respectively.

Conclusion

A preoperative risk identification model based on CT imaging data could be helpful for distinguishing LELGC from non-LELGC.

Computed tomography features and clinicopathological characteristics of gastric glomus tumor

Background

Gastric glomus tumor (GGT) is a rare neoplasm that is difficult to distinguish from other gastric submucosal tumors due to a lack of diagnostic experience. The goal of this study was to better understand GGT by looking at its clinicopathological features, computed tomography (CT) features, and differential diagnosis.

Methods

The clinical data and CT findings of 21 pathologically confirmed GGT patients were examined. The clinical characteristics and CT findings of benign GGT were compared to gastric stromal tumors (GST) (n = 30) and heterotopic pancreas (n = 30).

Results

The 21 cases included six males and fifteen females ranging in age from 42 to 64 years. The lesions were found in the gastric body in four cases and the antrum in seventeen. GGT was diagnosed as benign in 20 cases and malignant in one. In benign cases, the glomus cells were small, uniform, showed perivascular hemangiopericytoma‑like or solid nest‑like structures. Obvious mitotic figures were observed in the malignant case. SMA staining was positive in the tumor cells. A quasi-round or round solid mass protruded into the gastric cavity in 20 benign cases, with a clear and smooth edge. The long to short diameter ratio was 1.01 ± 0.15. All of the benign cases had obvious enhancement, with homogeneous enhancement in ten cases and heterogeneous enhancement in ten cases, as well as central filling enhancement in 12 cases. The ratio of CT value of lesion to abdominal aorta in arterial phase and venous phase were (0.41 ± 0.11) and (0.81 ± 0.20), which were significantly higher than GST and heterotopic pancreas. The irregular mass broke through the gastric wall and invaded liver with poorly defined boundary and internal necrosis, heterogeneous persistent moderate enhancement with thickening blood supply arteries was seen in one malignant case with a long diameter of 150 mm and a thick diameter of 30 mm.

Conclusions

CT enhancement usually shows persistent obvious enhancement, especially in arterial phase, which provides important value for the diagnosis. CT findings can help in the differential diagnosis of GGT and other submucosal tumors.

Computed Tomography Features and Clinical Prognostic Characteristics of Hepatoid Adenocarcinoma of the Stomach

PURPOSE

Hepatoid adenocarcinoma of the stomach (HAS) is a highly malignant and aggressive tumor. The purpose of this study was to describe the clinical, computed tomography (CT), and prognostic features of HAS to increase the awareness of this entity and determine its distinguishing features from non-HAS tumors.

METHODS

The CT features and clinical data of 47 patients in our hospital with pathologically documented HAS were retrospectively analyzed, and the relevant differences between pure HAS (pHAS) and mixed HAS (mHAS) were determined. In addition, 141 patients with non-HAS tumors in the same T stage in the same period were selected as the control group. The data were compared between the two groups, and factors affecting the prognosis of HAS were analyzed. In addition, we included 9 patients with HAS and 27 patients with non-HAS tumors from another center for external validation.

RESULTS

The patients in the HAS group were predominantly men (n = 33), and the tumor location was mostly the cardia or fundus (n = 27). Between the HAS and non-HAS groups, there were observed differences in terms of: sex, serum alpha-fetoprotein (AFP), carbohydrate antigen (CA)-125, and CA-724 levels; longest tumor diameter; degree of differentiation; vascular invasion; N stage, M stage, and tumor-node-metastasis (TNM) stage; thickest tumor diameter; plain CT attenuation; arterial-phase CT attenuation; CT attenuation between the venous and arterial phases; enhancement modes; and degrees of enhancement (all P < 0.05). In the data from another center for external validation, there were observed differences in terms of: age, degree of differentiation, vascular invasion, thickest tumor diameter, the ratio of arterial CT attenuation to CT attenuation of the abdominal aorta at the same level (RA), CT attenuation difference between the venous phase and arterial phase (HUv-a) (all P < 0.05). The results of the multivariate analysis revealed that the independent factors for differentiation were serum AFP level (P = 0.001), M stage (P = 0.038), and tumor enhancement on CT (P = 0.014). Among patients in the HAS group, 72.34% had pHAS and 27.66% had mHAS. The thickest tumor diameter and the longest short diameter of the metastatic lymph nodes of the mHAS group were on average 6.39 cm and 1.45 cm, respectively, which were larger than those in the pHAS group. The median progression-free survival time was 18.25 months in the HAS group, which was shorter than that in the non-HAS group (72.96 months; P = 0.001). The median overall survival time in the HAS group was 24.80 months, which was shorter than that in the non-HAS group (67.96 months; P = 0.001). The factors affecting the prognosis of HAS were M stage (P = 0.001), overall TNM stage (P = 0.048), presence of vascular cancer emboli (P = 0.040), and pHAS type (P = 0.046). Multifactorial analysis revealed that M stage (P = 0.027) and pHAS type (P = 0.009) were independent risk factors affecting the prognosis of HAS.

CONCLUSION

Although HAS is a rare clinical entity, it should be considered in the differential diagnosis of gastric tumors. Patients with HAS often have advanced-stage disease at presentation and a worse prognosis than patients with non-HAS tumors. CT findings, combined with laboratory results, can support the diagnosis of HAS. However, the final diagnosis needs to be confirmed with a histopathologic examination. If the postoperative pathologic findings reveal the mHAS type, a rapid clinical intervention and a detailed follow-up with CT are essential.

Insight into gastrointestinal heterotopic pancreas: imaging evaluation and differential diagnosis

Heterotopic pancreas (HP) is an uncommon congenital abnormality in the developmental process of the pancreas, with gastrointestinal heterotopic pancreas (GHP) being the most common HP. The clinical manifestations of GHP may have variable patterns of presentation, dictated by both the anatomic location and the functional ability of the lesion. The most common imaging modality in detecting GHP is computed tomography (CT), while gastrointestinal barium fluoroscopy, endoscopic ultrasonography, and magnetic resonance imaging (MRI) are also applied. The density and enhancement patterns of GHP are consistent with histological classifications. GHP with a predominantly acinar tissue component manifests homogeneous and marked enhancement on CT images, whereas a predominantly ductal GHP presents heterogeneous and mild enhancement. On MRI, the appearance and signal intensity of GHP were paralleled to the normal pancreas on all sequences and were characterized by T1-weighted high signal and early marked enhancement. This article provides a comprehensive review of the histopathology, clinical manifestations, imaging features of various modalities, and differential diagnosis of GHP. It is hoped that this review will improve clinicians' knowledge of GHP and aid in accurate preoperative diagnosis, thereby reducing the misdiagnosis rate.

Endoscopic submucosal dissection for gastric ectopic pancreas: a single-center experience

BACKGROUND AND OBJECTIVE

Endoscopic submucosal dissection (ESD) is a minimal invasive technology and could allow "en bloc" resection for superficial gastric tumors. The aim of this study is to evaluate the safety and feasibility of ESD for gastric ectopic pancreas (EP).

METHODS

A total of 93 patients diagnosed with ectopic pancreas who underwent ESD between January 2011 and June 2017 were enrolled. The demographic, clinical, and endoscopic data were collected and analyzed.

RESULTS

The average maximal diameter of lesions was 1.01 (range 0.4-3.0) cm with mean age of patients which was 39.75 (range 15-66) years. Overall, all of procedures en bloc was successful. The median operative time was 76.87 (range 30-160) min. A total of 12 patients experienced complications. In seven patients, bleeding occurred during the operation and was treated using hot biopsy forceps or metal clip. Five cases suffered from pneumoperitoneum which was managed well. The mean length of postoperative hospital stay was 5.7 (range 2-17) days. There was no relapse in any cases during the follow-up.

CONCLUSION

ESD appears to be a safe and feasible approach for curative treatment in gastric ectopic pancreas. Larger studies are needed to identify the role and the outcomes of ESD in another center.

Heterotopic pancreas involved muscularis propria mimicking mesenchymal tumor in stomach

Heterotopic pancreas is defined as pancreatic tissue found outside the usual anatomic location of the pancreas. It is often an incidental finding and can be found at various sites in the gastrointestinal tract. It is usually asymptomatic, however, symptoms such as abdominal pain, nausea, vomiting or even gastrointestinal bleeding could be possible.

In this report, a 38-year-old woman with epigastric fullness and endoscopic impression of the gastric submucosal tumor is described. Preoperative surveys including endoscopic ultrasound and computed tomography suggested gastric mesenchymal tumor such as leiomyoma. A surgical operation was arranged with a diagnosis of the heterotopic pancreas as confirmed by pathological assessment.