Yan-Qin Long, Wen-Li Xu, Lu-Xiu Li, Hui-Qin He, Jing-Jie Wang, Guo-Dong Shan, Ning Dai, and Hong-Tan Chen. J Neurogastroenterol Motil 2024;30:184-93. https://doi.org/10.5056/jnm23084

Pancreatic cystic neoplasms (PCNs) are a complex group of lesions with a spectrum of malignancy. Accurate differentiation of PCN types is crucial for patient management, as misdiagnosis can result in unnecessary surgeries or treatment delays, affecting the quality of life. The significance of developing a non-invasive, accurate diagnostic model is underscored by the need to improve patient outcomes and reduce the impact of these conditions. We developed a machine learning model capable of accurately identifying different types of PCNs in a non-invasive manner, by using a dataset comprising 449 MRI and 568 CT scans from adult patients, spanning from 2009 to 2022. The study’s results indicate that our multimodal machine learning algorithm, which integrates both clinical and imaging data, significantly outperforms single-source data algorithms. Specifically, it demonstrated state-of-the-art performance in classifying PCN types, achieving an average accuracy of 91.2%, precision of 91.7%, sensitivity of 88.9%, and specificity of 96.5%. Remarkably, for patients with mucinous cystic neoplasms (MCNs), regardless of undergoing MRI or CT imaging, the model achieved a 100% prediction accuracy rate. It indicates that our non-invasive multimodal machine learning model offers strong support for the early screening of MCNs, and represents a significant advancement in PCN diagnosis for improving clinical practice and patient outcomes. We also achieved the best results on an additional pancreatic cancer dataset, which further proves the generality of our model.

Background and Objectives: Bleeding from isolated gastric varix type I (IGV1) is less common but more dangerous and fatal than other varices. This study compared the safety and efficacy of endoscopic clipping and cyanoacrylate injection (EC-CYA) with EUS–guided coil and cyanoacrylate injection (EUS-coil/CYA) in the secondary prophylaxis of IGV1. Methods: A retrospective study was conducted on 131 patients with cirrhosis and a history of IGV1 bleeding who underwent either EC-CYA ( n = 55) or EUS-coil/CYA ( n = 76) at three medical centers. Primary outcomes were gastric variceal rebleeding and ectopic embolism. Secondary outcomes were technical success, variceal obliteration rate, cyanoacrylate volume, cyanoacrylate embolization efficiency, other causes of rebleeding, other adverse events, and mortality. Results: In the EC-CYA group, the gastric variceal rebleeding rate (14.5% vs . 8.7%, P = 0.307), technical success rate (96.4% vs . 94.7%, P > 0.999), other causes of rebleeding rate (12.7% vs . 5.8%, P = 0.303), other adverse events rate (5.5% vs . 5.8%, P > 0.999), and mortality (12.7% vs . 8.7%, P = 0.467) showed no significant differences compared with the EUS-coil/CYA group. No ectopic embolism was found in either group. The EC-CYA group had a higher cyanoacrylate volume (3.14 ± 1.45 mL vs . 1.99 ± 0.72 mL, P < 0.001) and a higher variceal obliteration rate at first follow-up (82.23% ± 27.09% vs . 66.59% ± 30.28%, P = 0.032). However, the EUS-coil/CYA group was superior to the EC-CYA group in terms of cyanoacrylate embolization efficiency (2.85 ± 2.05 cm ² /mL vs . 1.98 ± 1.46 cm ² /mL, P = 0.044). Conclusions: Both EC-CYA and EUS-coil/CYA are safe and effective in preventing rebleeding of IGV1. The EUS-coil/CYA has better embolization efficiency, whereas EC-CYA has better operational convenience.

Endoscopic ultrasound-guided lauromacrogol injection for treatment of colorectal cavernous hemangioma: Two case reports

Gastroduodenal artery pseudoaneurysm (GDAP) is usually treated by transcatheter arterial embolization (TAE) [1]. However, when the pseudoaneurysm combines with an abnormal course of the artery, TAE may fail. Herein, we report a case in which endoscopic ultrasound-guided coil and glue injection (EUS-CGI) was useful for treatment of a GDAP. A 58-year-old man with a GDAP underwent TAE prior to presentation to our hospital. The TAE failed due to the variant course of a communicating artery ([Fig. 1]). An ultrasound endoscope was introduced, and a 22-gauge puncture needle (EchoTip, Ultra Endoscopic Ultrasound Needle; Cook Medical, Bloomington, Indiana, USA) was selected to penetrate the GDAP. A preloaded 0.018-inch fibered interlock detachable coil (F-IDC; 2D 12 mm, 30 cm; Boston Scientific, Marlborough, Massachusetts, USA) and two 0.18-inch pushable coils (2D 12 mm, 5 cm; Boston Science, Marlborough, Massachusetts, USA) were slowly pushed out of the needle tip into the entrance of the GDAP. Thereafter, 2 ml saline followed by 0.5 ml glue and 2 ml saline was injected rapidly. Subsequent color Doppler imaging showed that the blood flow had disappeared ([Video 1]). Gastric arterial CT angiography (CTA) was performed for follow-up and revealed the GDPA to be partially filled with contrast ([Fig. 2]). The patient was therefore readmitted. Under EUS, the GDAP showed a hyperechoic mass with posterior acoustic shadow. Flows could be seen in the residual cavity, and pulsed wave Doppler imaging showed arterial blood flow, indicating recanalization. The residual cavity was punctured with a 22-gauge needle. Two 0.18-inch pushable coils were slowly placed. Thereafter, 2 ml saline followed by 0.5 ml glue and 2 ml saline was injected rapidly. Doppler imaging after the intervention showed the flow to have disappeared ([Video 1]). Gastric arterial CTA was performed 2 days later and showed no obvious contrast filling ([Fig. 3]). EUS-CGI as a treatment for GDAP makes use of the natural lumen of the digestive tract. It is not only a safe and effective treatment for GDAP, but also is repeatable, accurate, and avoids radiation exposure. Endoscopy_UCTN_Code_TTT_1AS_2AG Article published online: 19 December 2024 © 2024. The Author(s). This is an open access article published by Thieme under the terms of the Creative Commons Attribution License, permitting unrestricted use, distribution, and reproduction so long as the original work is properly cited. (https://creativecommons.org/licenses/by/4.0/). Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany