Search Results (39)

Background: Pancreatic cystic lesions (PCLs), including intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs), pose a diagnostic challenge due to their variable malignant potential. Current guidelines, such as Fukuoka and American Gastroenterological Association (AGA), have moderate predictive accuracy and may lead to overtreatment or missed malignancies. Artificial intelligence (AI), incorporating machine learning (ML) and deep learning (DL), offers the potential to improve risk stratification, diagnosis, and management of PCLs by integrating clinical, radiological, and molecular data. This is the first systematic review to evaluate the application, performance, and clinical utility of AI models in the diagnosis, classification, prognosis, and management of pancreatic cysts. Methods: A systematic review was conducted in accordance with PRISMA guidelines and registered on PROSPERO (CRD420251008593). Databases searched included PubMed, EMBASE, Scopus, and Cochrane Library up to March 2025. The inclusion criteria encompassed original studies employing AI, ML, or DL in human subjects with pancreatic cysts, evaluating diagnostic, classification, or prognostic outcomes. Data were extracted on the study design, imaging modality, model type, sample size, performance metrics (accuracy, sensitivity, specificity, and area under the curve (AUC)), and validation methods. Study quality and bias were assessed using the PROBAST and adherence to TRIPOD reporting guidelines. Results: From 847 records, 31 studies met the inclusion criteria. Most were retrospective observational (n = 27, 87%) and focused on preoperative diagnostic applications (n = 30, 97%), with only one addressing prognosis. Imaging modalities included Computed Tomography (CT) (48%), endoscopic ultrasound (EUS) (26%), and Magnetic Resonance Imaging (MRI) (9.7%). Neural networks, particularly convolutional neural networks (CNNs), were the most common AI models (n = 16), followed by logistic regression (n = 4) and support vector machines (n = 3). The median reported AUC across studies was 0.912, with 55% of models achieving AUC ≥ 0.80. The models outperformed clinicians or existing guidelines in 11 studies. IPMN stratification and subtype classification were common focuses, with CNN-based EUS models achieving accuracies of up to 99.6%. Only 10 studies (32%) performed external validation. The risk of bias was high in 93.5% of studies, and TRIPOD adherence averaged 48%. Conclusions: AI demonstrates strong potential in improving the diagnosis and risk stratification of pancreatic cysts, with several models outperforming current clinical guidelines and human readers. However, widespread clinical adoption is hindered by high risk of bias, lack of external validation, and limited interpretability of complex models. Future work should prioritise multicentre prospective studies, standardised model reporting, and development of interpretable, externally validated tools to support clinical integration. Full article

Pancreatic cystic lesions (PCLs) are increasingly identified via computerized tomography (CT) and magnetic resonance (MR), with a prevalence of 2–45%. Distinguishing mucinous PCLs (M-PCLs), which include intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs) that can progress to pancreatic ductal adenocarcinoma, from non-mucinous PCLs (NM-PCLs) is essential. Carcinoembryonic antigen (CEA) remains widely used but often demonstrates limited sensitivity and specificity. In contrast, endoscopic ultrasound-guided measurement of intracystic glucose more accurately differentiates PCL subtypes, as tumor-related metabolic changes lower cyst fluid glucose in mucinous lesions. Numerous prospective and retrospective studies suggest a glucose cut-off between 30 and 50 mg/dL, yielding a sensitivity of 88–95% and specificity of 76–91%, frequently outperforming CEA. Additional benefits include immediate point-of-care assessment via standard glucometers and minimal interference from blood contamination. DNA-based biomarkers, including KRAS and GNAS mutations, enhance specificity (up to 99%) but exhibit moderate sensitivity (61–71%) and necessitate specialized, expensive platforms. Molecular analyses can be crucial in high-risk lesions, yet their uptake is constrained by technical challenges. In practice, combining glucose assessment with targeted molecular assays refines risk stratification and informs the choice between surgical resection or active surveillance. Future investigations should establish standardized glucose thresholds, improve the cost-effectiveness of genetic testing, and integrate advanced biomarkers into routine protocols. Ultimately, these strategies aim to optimize patient management, limit unnecessary interventions for benign lesions, and ensure timely therapy for lesions at risk of malignant transformation. Full article

The increasing use of high-resolution cross-sectional imaging has significantly enhanced the detection of pancreatic cystic lesions (PCLs), including pseudocysts and neoplastic entities such as IPMN, MCN, and SCN. However, accurate categorization of PCLs remains a challenge. This study aims to improve PCL evaluation by developing and validating a radiomics-based software tool leveraging machine learning (ML) for lesion classification. The model categorizes PCLs into mucinous and non-mucinous types using a custom dataset of 261 CT examinations, with 156 images for training and 105 for external validation. Three experienced radiologists manually delineated the images, extracting 38 radiological and 214 radiomic features using the Pyradiomics module in Python 3.13.2. Feature selection was performed using Least Absolute Shrinkage and Selection Operator (LASSO) regression, followed by classification with an Adaptive Boosting (AdaBoost) model trained on the optimized feature set. The proposed model achieved an accuracy of 89.3% in the internal validation cohort and demonstrated robust performance in the external validation cohort, with 90.2% sensitivity, 80% specificity, and 88.2% overall accuracy. Comparative analysis with existing radiomics-based studies showed that the proposed model either outperforms or performs on par with the current state-of-the-art methods, particularly in external validation scenarios. These findings highlight the potential of radiomics-driven machine learning approaches in enhancing PCL diagnosis across diverse patient populations. Full article

Endoscopic ultrasound (EUS)-guided radiofrequency ablation (RFA) is a promising minimally invasive technique for the treatment of pancreatic lesions. This review first focuses on the technical aspects in EUS-RFA: the procedure typically employs EUS probes with integrated radiofrequency electrodes, enabling accurate targeting and ablation of pancreatic lesions. Different types of RFA devices, monopolar and bipolar energy delivery systems, are discussed, along with considerations for optimal ablation, including energy settings, procedure time, and pre- and post-procedural management. This paper presents a comprehensive literature review of EUS-RFA applied to both solid and cystic pancreatic lesions, including functioning and non-functioning pancreatic neuroendocrine tumors (pNETs), pancreatic cystic lesions (PCLs), pancreatic ductal adenocarcinoma (PDAC), and pancreatic metastases (PMs), discussing current evidence on safety, efficacy, clinical outcomes, and adverse events (AEs). EUS-RFA is an emerging technique with expanding potential for the treatment of both benign and malignant conditions; however, further studies are needed to better define patient selection criteria, assess long-term benefits, and establish definitive indications for its use. Full article

Pancreatic cystic lesions (PCLs) represent a spectrum of non-neoplasms and neoplasms with varying malignant potential, posing significant challenges in diagnosis and management. While some PCLs are precursors to pancreatic cancer, others remain benign, necessitating accurate differentiation for optimal patient care. Conventional approaches to PCL management rely heavily on radiographic imaging, and endoscopic ultrasound (EUS) guided fine-needle aspiration (FNA), coupled with clinical and biochemical data. However, the observer-dependent nature of image interpretation and the complex morphology of PCLs can lead to diagnostic uncertainty and variability in patient management strategies. This review critically evaluates current PCL diagnosis and surveillance practices, showing features of the different lesions and highlighting the potential limitations of conventional methods. We then explore the potential of artificial intelligence (AI) to transform PCL management. AI-driven strategies, including deep learning algorithms for automated pancreas and lesion segmentation, and radiomics for analyzing heterogeneity, can improve diagnostic accuracy and risk stratification. These advanced techniques can provide more objective and reproducible assessments, aiding clinicians in decision-making regarding follow-up intervals and surgical interventions. Early results suggest that AI-driven methods can significantly improve patient outcomes by enabling earlier detection of high-risk lesions and reducing unnecessary procedures for benign cysts. Finally, this review emphasizes that AI-driven approaches could potentially reshape the landscape of PCL management, ultimately leading to improved pancreatic cancer prevention. Full article

Background: Pancreatic cystic lesions (PCLs) are frequently detected incidentally and vary from benign to malignant. Accurate differentiation between mucinous (M-PCLs) and non-mucinous PCLs (NM-PCLs) is essential for appropriate management. This study aims to validate the accuracy of on-site glucose measurement using a glucometer with a cut-off of 50 mg/dL for distinguishing M-PCLs from NM-PCLs. Methods: In this prospective multicenter study, conducted at three European academic hospitals, patients who underwent endoscopic ultrasound-guided fine-needle aspiration for PCLs between 2019 and 2020 were included. On-site glucose measurement was performed using a conventional glucometer. Data on demographics, clinical features, EUS findings, and histopathology were collected. Results: Fifty patients were enrolled, with 37 having glucose levels < 50 mg/dL and 13 ≥ 50 mg/dL. M-PCLs were more common in the <50 mg/dL group (81%) compared to the ≥50 mg/dL group (23%, p < 0.001). The median CEA was higher in the <50 mg/dL group (146 ng/mL) than in the ≥50 mg/dL group (3 ng/mL, p = 0.047). On-site glucose testing < 50 mg/dl demonstrated a sensitivity of 93.2%, a specificity of 76.5%, and an accuracy of 89% for detecting M-PCLs with an AUC of 0.74 and an OR of 14.29 (p < 0.001). In comparison, CEA > 192 ng/mL had a sensitivity of 55.6%, a specificity of 87.5%, and an accuracy of 75.8% for M-PCLs, with an AUC of 0.65 and an OR of 4.44. Conclusions: On-site glucose measurement using a glucometer with a cut-off of <50 mg/dL is a highly accurate, rapid, and cost-effective method for differentiating M-PCLs from NM-PCLs. Our results validate the glucose cut-off in a multicentric prospective cohort supporting its integration into standard diagnostic protocols for PCLs. Full article

Objectives: To assess the feasibility of 5.0 T magnetic resonance imaging (MRI) in characterizing pancreatic cystic lesions (PCLs), compared with 3.0 T MRI and multidetector computed tomography (MDCT). Methods: Thirty-five patients with PCLs underwent 5.0 T MR alongside 3.0 T MR or MDCT. Two observers measured subjective and objective image quality scores. The consistency of two observers between 5.0 T and 3.0 T was calculated by intraclass correlation coefficients. The characteristics of PCLs and their specific diagnosis, as well as benignity/malignancy, were evaluated across MDCT, 3.0 T, and 5.0 T MRI. Results: The 5.0 T MR demonstrated significantly higher subjective image quality and SNR on T1WI compared to that in 3.0 T MR (p < 0.05). The 5.0 T MRI identified more cyst lesions than the 3.0 T MRI (40 and 32) and MDCT (82 and 56). The sensitivity, specificity, and accuracy for differentiating benign from malignant lesions with 5.0 T MRI (75%, 100%, and 91.4%, respectively) surpassed those of 3.0 T MRI and MDCT. The accuracy of the specific diagnosis of PCLs at 5.0 T MRI (80%) was superior to 3.0 T MRI and MDCT. Conclusions: 5.0 T MRI exhibits certain superiority in delineating details of PCLs and in clinical diagnostic accuracy, outperforming MDCT and 3.0 T MRI while maintaining sufficient image quality. Full article

Pancreatic cystic lesions (PCLs) have been increasingly identified in recent years, encompassing a wide spectrum ranging from benign non-evolutive to malignant invasive lesions. Despite various clinical, laboratory, imaging, endoscopic ultrasound, and cytohistological features that may aid clinicians in the complex differential diagnosis of PCLs, only a few pathognomic signs distinctive to specific PCLs have been identified. Although rarely encountered, their proper recognition is crucial for the appropriate management of PCLs. The aim of our review is to extensively discuss and illustrate pathognomic signs in the setting of PCLs. Full article

Pancreatic cystic lesions (PCLs) are increasingly diagnosed owing to the wide use of cross-sectional imaging techniques. Accurate identification of PCL categories is critical for determining the indications for surgical intervention or surveillance. The classification and management of PCLs rely on a comprehensive and interdisciplinary evaluation, integrating clinical data, imaging findings, and cyst fluid markers. EUS (endoscopic ultrasound) has become the widely used diagnostic tool for the differentiation of pancreatic cystic lesions, offering detailed evaluation of even small pancreatic lesions with high sensitivity and specificity. Additionally, endoscopic ultrasound–fine-needle aspiration enhances diagnostic capabilities through cytological analysis and the assessment of fluid viscosity, tumor glycoprotein concentration, amylase levels, and molecular scrutiny. These detailed insights play a pivotal role in improving the clinical prognosis and management of pancreatic neoplasms. This review will focus mainly on the latest recommendations for the differentiation, management, and treatment of pancreatic cystic lesions, highlighting their clinical significance. Full article

Pancreatic cystic lesions (PCLs) pose a diagnostic challenge due to their increasing incidence and the limitations of cross-sectional imaging and endoscopic-ultrasound-guided fine-needle aspiration (EUS-FNA). EUS-guided through the needle biopsy (EUS-TTNB) has emerged as a promising tool for improving the accuracy of cyst type determination and neoplastic risk stratification. EUS-TTNB demonstrates superior diagnostic performance over EUS-FNA, providing critical preoperative information that can significantly influence patient management and reduce unnecessary surgeries. However, the procedure has risks, with an overall adverse event rate of approximately 9%. Preventive measures and further prospective studies are essential to optimize its safety and efficacy. This review highlights the potential of EUS-TTNB to enhance the diagnostic and management approaches for patients with PCLs. It examines the current state of EUS-TTNB, including available devices, indications, procedural techniques, specimen handling, diagnostic yield, clinical impact, and associated adverse events. Full article

Pancreatic cystic lesions (PCL) are frequently encountered in clinical practice and some are referred to surgery due to their neoplastic risk or malignant transformation. The management of PCL involves complex decision-making, with postoperative surveillance being a key component for long-term outcomes, due to the potential for recurrence and postoperative morbidity. Unfortunately, the follow-up of resected patients is far from being optimal and there is a lack of consensus on recommendations with regard to timing and methods of surveillance. Here, we summarize the current knowledge on the postoperative surveillance of neoplastic pancreatic cysts, focusing on the mechanisms and risk factors for recurrence, the recurrence rates according to the initial indication for surgery, the final result of the surgical specimen and neoplastic risk in the remaining pancreas, as well as the postsurgical morbidity comprising pancreatic exocrine insufficiency, metabolic dysfunction and diabetes after resection, according to the type of surgery performed. We analyze postsurgical recurrence rates and morbidity profiles, as influenced by different surgical techniques, to better delineate at-risk patients, and highlight the need for tailored surveillance strategies adapted to preoperative and operative factors with an impact on outcomes. Full article

The malignant progression of pancreatic cystic lesions (PCLs) remains understudied with a knowledge gap, yet its exploration is pivotal for effectively stratifying patient risk and detecting cancer at its earliest stages. Within this review, we delve into the latest discoveries on the molecular level, revealing insights into the IPMN molecular landscape and revised progression model, associated histologic subtypes, and the role of inflammation in the pathogenesis and malignant progression of IPMN. Low-grade PCLs, particularly IPMNs, can develop into high-grade lesions or invasive carcinoma, underscoring the need for long-term surveillance of these lesions if they are not resected. Although KRAS and GNAS remain the primary oncogenic drivers of neoplastic development in IPMNs, additional genes that are important in tumorigenesis have been recently identified by whole exome sequencing. A more complete understanding of the genes involved in the molecular progression of IPMN is critical for effective monitoring to minimize the risk of malignant progression. Complicating these strategies, IPMNs are also frequently multifocal and multiclonal, as demonstrated by comparative molecular analysis. Algorithms for preoperative cyst sampling and improved radiomic techniques are emerging to model this spatial and temporal genetic heterogeneity better. Here, we review the molecular pathology of PCLs, focusing on changes associated with malignant progression. Developing models of molecular risk stratification in PCLs which can complement radiologic and clinical features, facilitate the early detection of pancreatic cancer, and enable the development of more personalized surveillance and management strategies are summarized. Full article

Undetermined pancreatic cystic lesion (PCL) differentiation benefits from endoscopic ultrasound (EUS) based on morphology and cyst fluid analysis, but room for new biomarkers exists. Our aim was to assess the intracystic and serum diagnostic value of neutrophil gelatinase-associated lipocalin (Ngal) and interleukin 1 beta (IL-1β) for differentiation of PCLs. This prospective study included patients from one tertiary hospital, evaluated between April 2018 and May 2020. EUS fine-needle aspiration or pancreatic pseudocysts drainage was the source of PCL intracystic liquid. The final diagnosis was based on surgery or EUS results (morphology, cytology, glucose, and CEA—carcinoembryogenic antigen). The intracystic samples were tested for Ngal, IL-1β, glucose, and CEA, and serum for Ngal and IL-1β. We evaluated 63 cysts, 33 pseudocysts, and 30 non-inflammatory cysts. The diagnostic sensitivity and specificity for mucinous PCL was 70.8% and 92.3% for intracystic Ngal (cut-off: 500–800 ng/dL), without correlation with serum Ngal, no matter the inclusion of infected pseudocysts. After exclusion of infected pseudocysts, the sensitivity and specificity for glucose were 87% and 75%, respectively, and for CEA, they were 87.1%, and 96.8%, respectively. Intracystic Ngal shows promise in differentiating mucinous PCLs, but researchers need to conduct further studies to confirm its effectiveness. Intracystic IL-1β and serum Ngal made no diagnostic contribution. Full article

Due to their aspecific macroscopic appearance, uncommon pancreatic cystic lesions (PCLs) are often misdiagnosed as mucinous lesions and improperly resected. We aimed to evaluate the endoscopic ultrasound (EUS)-guided through-the-needle biopsy (TTNB) capacity of the preoperative diagnosis of uncommon PCLs. Overall, 136 patients with PCLs who underwent EUS-TTNB between 2016 and 2022 were retrospectively identified. Common histotypes (e.g., IPMN, serous cystadenoma, and mucinous cystadenoma) were excluded and 26 (19.1%) patients (15 female, mean age 52.9 ± 10.4) were analyzed. The EUS findings, adverse events (AEs), and TTNB outcomes in uncommon PCLs were evaluated. The cysts histotype was accurately diagnosed by TTNB in 24/26 (92.3%) cases (seven cystic neuroendocrine tumors, four squamoid cysts, three acinar cells cystadenomas, two lymphoepithelial cysts, two mucinous non-neoplastic cysts, two bronchogenic cysts, two cystic lymphangiomas, one solid-pseudopapillary neoplasm, and one schwannoma). In the remaining two cases, lymphangioma was eventually diagnosed after resection. Surgery was performed in 15/26 (57.7%) patients. The mean follow-up of non-surgical patients was 32.5 months. One severe acute case of pancreatitis (3.8%) that required surgery occurred after EUS-TTNB. Uncommon pancreatic/peripancreatic lesions represent the 19.1% of PCLs in our series, with mainly benign histotypes. TTNB demonstrated a high diagnostic performance with a low rate of AEs in this setting, representing a reliable tool with which to avoid useless surgery. Full article