How to distinguish pancreatic tumors and assess the necessity for biopsy

Abstract

Endoscopic ultrasonography (EUS) is a valuable and widely used tool for evaluating pancreatic tumors. Accurate decision-making during EUS procedures, particularly for differentiating between benign and malignant lesions based on imaging characteristics and assessing the need for tissue sampling, is crucial. This review provides a comprehensive overview of pancreatic tumor features observed during EUS and highlights the key criteria for distinguishing between malignant and benign conditions. Additionally, we discuss the indications for fine-needle aspiration or biopsy to obtain histopathological and genetic confirmation. Improving our understanding of these critical aspects can help improve diagnostic accuracy and guide clinicians in determining the most appropriate management strategies for patients with pancreatic tumors.

Key Words: Endoscopic ultrasound; Contrast; Elastography; Pancreatic tumor; Fine-needle biopsy

Core Tip: This article highlights the endoscopic ultrasonography, assisted by contrast enhanced sonography and elastography, offers superior accuracy in detecting and characterizing pancreatic tumors. By assessing morphology, vascularity, and tissue stiffness, it differentiates malignant from benign processes and identifies high-risk cystic lesions. Endoscopic ultrasonography-guided fine needle biopsy enables histological and molecular diagnosis, including next generation sequencing for actionable mutations, guiding surgical decisions, neoadjuvant therapy, and targeted treatments. This integrated approach optimizes pancreatic cancer detection, staging, and personalized management while avoiding unnecessary interventions.

INTRODUCTION

The incidence of pancreatic cancers, mostly adenocarcinomas, has been increasing, and the 5-year survival rate remains low, at only 12%-13%. One reason for this is the difficulty of early diagnosis, which precludes some patients from undergoing curative surgical resection and leaves them with palliative chemotherapy as the only option[1-3]. Therefore, diagnostic tools play a crucial role in the early detection of pancreatic cancer. Common diagnostic tools include computed tomography (CT) and abdominal ultrasonography (US); however, their effectiveness is limited by factors such as tumor location and size[4,5]. In recent years, endoscopic US (EUS) has gained attention owing to its superior diagnostic ability, particularly in overcoming the limitations of transabdominal US. It has proven to be a cost-effective tool for assessing pancreatobiliary disorders, especially in cases in which other diagnostic methods have been unsuccessful[6]. EUS has detection rates of more than 90%, particularly for lesions smaller than 2-3 cm; its sensitivity of 99% is higher than that of CT, at 55%[7,8]. EUS can visualize the detailed morphology and characteristics of pancreatic lesions, enhancing the diagnostic accuracy for pancreatic malignancies and allowing for better characterization of pancreatic cystic lesions. Furthermore, EUS enables the aspiration or biopsy of pancreatic lesions during the procedure, facilitating cytohistological diagnosis and distinguishing between benign and malignant tumors. Compared to US or CT-guided tissue sampling of pancreatic masses, EUS-guided fine-needle biopsy (FNB) demonstrates higher diagnostic accuracy, particularly for small lesions. Additionally, EUS can detect metastatic lymph nodes near the gastrointestinal tract or liver metastases, which may be missed by other imaging modalities in approximately 10% of patients[9-11].

In addition to the widespread use of traditional EUS and EUS-FNB for the pathological evaluation of pancreatic tumors, contrast-enhanced (CE) EUS and elastography have recently been introduced to provide more detailed characterization of pancreatic lesions. These advanced techniques have not only improved the diagnostic accuracy for pancreatic malignancies but might also enhanced the efficiency of fine-needle aspiration (FNA) and FNB procedures[12,13].

In this article, we aim to identify the key characteristics of pancreatic masses observed using traditional EUS, CE-EUS, and elastography to differentiate malignant tumors from benign and inflammatory lesions and to assess the necessity of FNB.

MORPHOLOGY

Although pancreatic tumors exhibit highly diverse morphological characteristics, they are generally categorized as either solid or cystic. Solid pancreatic masses may result from inflammation associated with chronic pancreatitis or malignant processes. Ductal pancreatic adenocarcinoma is the most common type, accounting for > 95% of all malignant solid pancreatic tumors, whereas neuroendocrine tumors represent only a small proportion of cases. Other tumor types, such as squamous cell carcinoma, primary pancreatic lymphoma, and metastatic pancreatic malignancy, are exceedingly rare[14,15].

Distinguishing between benign and malignant masses is challenging, as symptoms such as weight loss, jaundice, and diabetes progression can occur in both cases. Serum carbohydrate antigen 19-9 Levels and ductal dilation are also observed in both benign and malignant cases[16]. Signs that indicate the possibility of pancreatic tumor malignancy include lymph node involvement, particularly in the celiac axis, peripancreatic region, porta hepatis, gastrohepatic ligament, aortocaval, and posterior mediastinal stations. The affected nodes are typically round, well-defined, and hypoechoic. Metastases of the left lobe of the liver, which are usually hypoechoic and well-defined, can be detected on EUS; FNB is performed if the lesion is accessible. Ascites may indicate peritoneal or venous occlusion. Assessing vascular invasion is crucial because tumors can affect specific vessels based on their location, potentially leading to vessel adherence, occlusion, or collateral formation[17].

Pancreatic solid lesions

Solid pancreatic masses include benign and malignant lesions. Benign masses primarily consist of a focal chronic pancreatitis, whereas malignant tumors include ductal adenocarcinomas, neuroendocrine tumors, and other less common malignancies.

Focal chronic pancreatitis: CT is the standard diagnostic approach for solid pancreatic lesions. However, its diagnostic utility may be limited, particularly in patients with chronic pancreatitis, in whom distinguishing between focal pancreatitis and pancreatic cancer remains challenging[18]. Acute inflammatory exacerbations of chronic pancreatitis can cause focal edema, which may appear similar to a neoplastic mass on CT and lead to a diagnostic error rate of 16%-23%[19,20]. However, there are some differences between inflammation-related tumors and neoplasms: Neoplastic masses are typically heterogeneous and distinctly hypoechoic, lack normal parenchymal features and ducts, and - unlike benign masses - rarely contain calcifications.

IgG4-associated autoimmune pancreatitis (AIP) is a distinct subtype of pancreatitis characterized by lymphoplasmacytic infiltration, fibrosis, and a strong steroid response. It typically presents with diffuse swelling of the pancreas or a focal pancreatic mass[21]. EUS demonstrates the classic features of AIP, including a diffusely enlarged (sausage-shaped) gland that is hypoechoic, coarse, and patchy with a heterogeneous parenchyma, and an irregular, ectatic main pancreatic duct[22]. Moreover, a distinct transition zone, where the gland shifts abruptly from normal to typical AIP features, is often observed on EUS but not on CT or magnetic resonance imaging (MRI). In this region, the main pancreatic duct often tapers smoothly, suggesting a benign lesion rather than a malignancy[23].

Glasbrenner et al[24] proposed the criteria for EUS findings indicative of an inflammatory pancreatic mass: A heterogeneous echo pattern, calcification, peripancreatic echostranding, and cysts. Compared with these findings, malignant masses show signs of irregular outer mass margins, invasion into adjacent organs, and enlargement of the nearby lymph nodes. These proposed criteria increase the diagnostic accuracy of inflammatory and malignant masses. The success rate of EUS is 96%; its overall diagnostic performance in predicting malignant masses had a sensitivity of 78%, specificity of 93%, positive predictive value of 93%, negative predictive value of 78%, and diagnostic accuracy of 85%.

Solid pancreatic malignancy: Pancreatic adenocarcinomas typically appear as heterogeneous hypoechoic masses with irregular margins on EUS. However, the diagnostic specificity of these morphological features alone is only 53%, as these features are also observed in focal pancreatitis, neuroendocrine tumors, and metastases[25]. Despite this limitation, EUS demonstrates a high sensitivity of 89%-100%, making it highly effective for the early detection of small adenocarcinoma[26,27]. Compared with other imaging techniques, EUS provides a more accurate staging, thereby enhancing the management of pancreatic cancer. EUS is particularly valuable for evaluating peripancreatic vascular and lymph node involvement. Large-scale studies have reported that when EUS is used for staging, the accuracy of T staging ranges 78%-91%, whereas the accuracy of nodal staging varies between 41% and 86%[26,28,29].

In patients with a chronic pancreatic disease, it is difficult to distinguish between neoplasms and benign lesions using EUS. Nevertheless, one study found that EUS had a higher differential diagnostic rate (92.5%) than CT (82.5%) in patients with chronic pancreatitis. This may be because EUS has a greater sensitivity for detecting pancreatic duct dilation - a key feature of pancreatic cancer. Notably, secondary signs such as duct interruption, bile or pancreatic duct dilation, distal pancreatic atrophy, mass effect, and contour abnormalities were more easily detected with EUS than with CT[30].

On EUS, neuroendocrine tumors typically appear as well-defined hypoechoic lesions with intense vascularization. Additionally, 60%-75% of neuroendocrine tumors are less than 1.5 cm in diameter[31]. Lesions > 3 cm in size are more likely to exhibit malignant potential and often present with a heterogeneous appearance, including cystic features, solid components, or pure fluid[32].

Metastatic pancreatic lesions: Metastatic tumors account for only a small proportion of all pancreatic tumors and present with varied morphologies based on the originating cancer type. For example, most metastatic pancreatic tumors have a hypoechoic pattern; however, hyperechoic metastases from bladder cancer and anechoic metastases from melanoma have also been observed. Patients with a history of malignancy have a higher risk of developing pancreatic metastasis, regardless of whether they received treatment. EUS with FNB is essential for definitive histological diagnosis, particularly in patients with prior cancer, to distinguish between primary and metastatic tumors. This helps avoid unnecessary surgeries and enable timely treatment for better patient outcomes[33,34].

Pancreatic cystic lesions

Pancreatic cystic lesions account for 1%-5% of all pancreatic malignancies[35]. As the imaging characteristics of benign and malignant cystic lesions often significantly overlap, distinguishing these lesions remains challenging. Compared with other imaging modalities, EUS has demonstrated superior accuracy in local staging, prediction of vascular invasion, and assessment of tumor resectability, particularly for tumors < 2 cm in diameter[30,36,37].

Common neoplastic cysts include intraductal papillary mucinous neoplasms (IPMN), mucinous cystic neoplasms (MCN), and serous cystic neoplasms (SCN). The less common types include solid pseudopapillary tumors (SPNs), neuroendocrine cystic tumors, and cystic adenocarcinomas. IPMNs, MCNs, and SPNs can undergo malignant transformation, whereas cystic pancreatic neuroendocrine tumors inherently possess metastatic potential. In contrast, pseudocysts, lymphoepithelial cysts, and retention cysts have lower malignant potential[38]. Thus, determining the malignant potential of pancreatic cystic lesions can help the selection of the most appropriate treatment strategy and prevent unnecessary surgery.

IPMN: IPMN is characterized by mucin hyperproduction, dilation of the duct of Wirsung, and a protruding papilla. It is classified into three types: Main duct IPMN (MD-IPMN), branch duct IPMN (BD-IPMN), and mixed-type IPMN. A main pancreatic duct dilation > 1 cm is suggestive of MD-IPMN, whereas a pancreatic mucinous cyst communicating with the pancreatic duct without significant main duct dilation suggests BD-IPMN. This classification is clinically relevant in assessing malignancy potential, as MD-IPMNs and mixed-type IPMNs carry a higher malignancy risk (36%-100%) compared to BD-IPMNs (11%-30%)[39]. According to the American Gastroenterological Association guidelines, certain characteristics indicate either high-risk stigmata or worrisome features suggestive of malignancy. A main pancreatic duct diameter ≥ 1.0 cm, obstructive jaundice with a cyst in the pancreatic head, and solid enhancing nodule within the cyst are high-risk stigmata, whereas worrisome features include cysts larger than 3 cm, thick or contrast-enhancing cyst walls, and a main pancreatic duct diameter > 0.5 cm[40]. Hence, surgical intervention should be considered for patients whose pancreatic cystic lesions are consistent with main duct IPMN and who experience symptoms such as obstructive jaundice or abdominal pain.

MCN: MCNs account for approximately 10% of all pancreatic cystic lesions. They are usually solitary, located in the body or tail of the pancreas (90%-95%), and do not communicate with the ductal system[41,42]. MCNs are characterized by thick-walled septate mucinous cysts with a distinctive ovarian-type stroma. The imaging features include round morphology, sharp margins, thick walls, septa, and fluid content, with mural nodules and peripheral calcifications suggestive of malignancy[29,43]. Carcinoembryonic antigen (CEA) levels in the cystic fluid can help differentiate MCNs from non-mucinous cystic lesions. Additionally, CE-EUS and MRI can aid in assessing malignant potential, with CE-EUS providing a detailed visualization of the inner structures. Given their premalignant nature, surgical resection is recommended for all eligible candidates, as per international guidelines[44].

SCN: Compared with lesions characterized by mucinous content, SCNs are a type of serous cystic lesion. In contrast to mucinous cystic lesions such as IPMN and MCN, serous cystic lesions have a lower risk of malignancy[45,46]. SCNs are benign and account for approximately 10%-16% of all pancreatic cystic lesions. They are characterized by glycogen-rich, clear cuboidal epithelial cells and are generally have a very low malignant potential. Increased use of advanced imaging techniques has led to a higher detection rate of these neoplasms, often as incidental findings. While the incidence of malignancy in SCNs is low, with a comprehensive review reporting a malignancy rate of approximately 3%, there remains a low risk of malignant transformation. Notably, SCN is a broad term that includes both serous cystadenoma (SCA), which is benign, and a very rare serous cystadenocarcinoma, which is malignant[45,47].

SCAs are usually solitary lesions that do not communicate with the main pancreatic duct and exhibit a distinctive multilocular honeycomb architecture with thin walls and septa. They typically consist of multiple (more than six) well-defined microcystic structures (< 2 cm) with delicate septa. SCAs are most commonly located in the pancreatic body and tail[48,49]. They may contain central stellate calcifications or a central scar (15% of the cases). Pancreatic SCAs are predominantly benign neoplasms with a very low risk of malignancy[50]. While the exact rate of malignant transformation has not been definitively established, studies suggest that it is less than 3%. For instance, a comprehensive review reported a malignancy rate of approximately 3%, indicating that SCAs are generally benign but carry a small risk of becoming malignant lesions such as serous adenocarcinoma[45]. Given their low malignant potential, SCAs are often conservatively managed, especially in asymptomatic patients. Surgical intervention may be considered for lesions that are symptomatic, show significant growth, or cannot be certainly distinguished from other cystic pancreatic neoplasms with higher malignant potential. Hence, an FNA biopsy is not routinely required for SCAs if characteristic imaging findings are observed. However, in cases with uncertain diagnosis, symptomatic cysts, or rapid growth, FNA with cystic fluid analysis can provide additional diagnostic clarity[51,52].

CONTRAST-ENHANCED ENDOSCOPI US

CE-EUS was first introduced by Kato et al[53], who used fundamental EUS with carbon dioxide gas. CE-EUS is an advanced imaging technique that combines traditional EUS with ultrasound contrast agents to enhance the visualization of tissues, blood flow, and vascular structures. The ultrasound contrast agents, such as the commonly used Levovist^®, Sonovue^®, and Sonazoid^®, consist of microbubbles that enhance vascular imaging by creating pseudo-Doppler signals, allowing better differentiation between malignant and benign masses. This method improves detection based on the characteristics of different tumors during the differential phase of contrast injection[54,55].

The study conducted by Kitano et al[56] highlights the efficacy of CE-EUS in diagnosing small pancreatic carcinomas. Among 156 patients evaluated, EUS exhibited a significantly higher sensitivity for detecting tumors ≤ 2 cm compared with contrast-enhanced CE-CT (94.4% vs 50%, respectively). Based on CE-EUS and power Doppler EUS, tumors can be classified into three vascular patterns: Hypovascular (mostly ductal carcinomas), isovascular, and hypervascular (including endocrine and metastatic tumors). CE-EUS has demonstrated strong utility in distinguishing between hypovascular and hypervascular tumors, although it has limitations in differentiating among hypervascular tumors. Therefore, CE-EUS is a highly sensitive method for the detecting and classifying small pancreatic tumors based on their vascular characteristics.

CE-EUS can reliably differentiate pancreatitis from pancreatic cancer, with sensitivity, specificity, positive predictive value, and negative predictive value of 91%, 93%, 100%, and 88%, respectively, making it a reliable diagnostic tool[57,58]. It is particularly effective for detecting small pancreatic tumors, as it more accurately reveals hypoenhancing lesions that indicate malignancy than standard EUS. Furthermore, CE-EUS outperforms helical CT in visualizing blood flow in solid pancreatic tumors. Hence, CE-EUS is a valuable tool for the precise identification of pancreatic tumors[59].

Adenocarcinoma

Pancreatic ductal adenocarcinoma (PDAC) typically presents as hypo-enhancement in all phases of CE-US. This hypovascular pattern is associated with a desmoplastic reaction, leading to low vascular density. Studies report that CE-EUS has a high accuracy of 87.8% in characterizing PDAC. Additionally, a meta-analysis showed that the sensitivity of CE-EUS in diagnosing PDAC is approximately 86.47%, which is comparable to the sensitivity of multidetector CT[60].

Pancreatic neuroendocrine

Pancreatic neuroendocrine tumors (pNETs) are among the most common secondary solid tumors of the pancreas, although they are relatively rare compared with PDAC. pNETs typically appear as hyperenhancing lesions during the arterial phase of CE-EUS. They demonstrate either homogeneous or heterogeneous enhancement owing to their abundant arterialization. Heterogeneous enhancement often indicates areas of necrosis within a tumor, particularly in larger tumors. Compared with PDAC, pNETs exhibit a distinct vascular profile, enabling effective differentiation on CE-EUS. Owing to the differential characteristics of PDAC and pNETs, CE-EUS is recommended for distinguishing adenocarcinoma from pNETs according to the ENETS guidelines[61].

Focal chronic pancreatitis

Patients with chronic and IgG4-associated pancreatitis may occasionally develop mass-forming pancreatitis. Its primary feature is progressive interstitial fibrosis accompanied by inflammatory infiltration[62]. Differentiating mass-forming pancreatitis from neoplastic diseases is challenging they may present with similar symptoms and clinical signs[62]. On US, mass-forming pancreatitis typically appears as a hypoechoic mass in a localized region of the gland, most commonly the head, and often closely resembles pancreatic carcinoma. Sometimes, small calcifications observed in the tumor may indicate an inflammatory origin but have low specificity[63]. On CE-EUS, 94% of mass-forming pancreatitis cases exhibited isoenhancement of the pancreatic lesion, whereas 6% remained hypoechoic throughout the dynamic phase[64]. CE-EUS diagnosed mass-forming pancreatitis with 88.6% sensitivity, 97.8% specificity, and 96% overall accuracy. The presence of parenchymographic enhancement or hypoechogenicity in the dynamic phase significantly improves diagnostic confidence.

Pancreatic cystic lesions

One major update in recent guidelines is the inclusion of CE-EUS findings in the assessment of high-risk stigmata and worrisome features. EUS is particularly valuable for detecting signs suggestive of high-grade dysplasia or invasive carcinoma, such as the presence and size of mural nodules. Additionally, EUS provides an opportunity for cyst fluid sampling and biopsies of solid components, further enhancing the diagnostic accuracy[65]. In pancreatic cystic lesions, the presence of vascularized mural nodules and tumoral vegetation indicates the possibility of malignancy. CE-EUS can easily demonstrate vascularized mural nodules and tumor tissue from avascular mucin plug areas. It also has a diagnostic accuracy similar to that of MRI in detecting cystic septa and nodules[66].

ELASTOGRAPHY

EUS elastography for pancreatic tissue evaluation was first reported in 2006[67]. It utilizes an embedded sonoelastography module, which is an advanced technique for assessing tissue elasticity during conventional EUS examination. This method visualizes the tissue elasticity modulus by overlaying a color map (red-green-blue) onto a conventional grayscale image. Tissue characteristics can be evaluated qualitatively, with a predominance of blue indicating malignancy, or quantitatively, with a hue histogram value on a scale of 1-256, where values above 175 indicate malignancy[68]. A review reported that the pooled sensitivity ranges from 95% to 99%, whereas the specificity ranges from 67% to 76%[69].

Dawwas et al[70] recently reported that EUS elastography showed a high sensitivity (100%) but poor specificity (16.7%) in the evaluation of solid pancreatic lesions. These findings are inconsistent with the high specificity of EUS elastography reported in previous studies[71]. The authors suggested that EUS elastography should complement tissue sampling with EUS-FNA. They also recommended repeating EUS-FNA in cases of solid pancreatic lesions with negative FNA results but a high strain ratio and advised reassurance for patients when both EUS-FNA results were negative and the strain ratio was low[70]. Furthermore, the specificity of EUS-FNA can be improved by combining it with EUS elastography[12]. During the staging of small lymph node metastases, EUS elastography can identify the smallest metastatic changes in tissue hardness. However, there are no reports of EUS elastography for the T- and M-staging of pancreatic cancers.

Focal chronic pancreatitis

In patients with chronic pancreatitis, the fibrotic background decreases the sensitivity, specificity, and accuracy of pancreatic lesion diagnosis. Accurate diagnosis is crucial for managing benign and malignant lesions in chronic pancreatitis. The study reported by Săftoiu et al[72] analyzed a group of 68 patients, including those with a normal pancreas, chronic pancreatitis, pancreatic adenocarcinoma, and neuroendocrine tumors and reported that the sensitivity, specificity, and accuracy of distinguishing benign from malignant cases under EUS elastography are 91.4%, 87.9%, and 89.7%, respectively. However, previous studies reported that the sensitivity of EUS-FNB for diagnosing pancreatic adenocarcinoma in the context of chronic pancreatitis decreased from 89%-91% to 54%-73%[73,74]. In EUS elastography, inflammatory tissue tends to have a softer consistency, resulting in lower strain ratios in mass-forming chronic pancreatitis than in pancreatic adenocarcinoma. Several studies have explored the use of EUS elastography to distinguish benign and malignant solid pancreatic lesions[75]. Based on the results of the current study, EUS elastography may serve as a valuable supplemental diagnostic tool alongside EUS-FNB in cases with normal pancreatic parenchyma and chronic pancreatitis.

EUS-GUIDED FNB AND ASPIRATION

Before EUS became available, physicians diagnosed pancreatic masses by using abdominal US, CT-guided biopsy, endoscopic retrograde cholangiopancreatography-guided brush cytology, or surgery. The accuracy of endoscopic retrograde cholangiopancreatography brush cytology in detecting pancreatic cancer is low, and the procedure carries a risk of pancreatitis[76]. EUS not only provides more detailed images of small pancreatic masses (0.5-2 cm) but also can also guide FNA and FNB. With the assistance of on-site macroscopic evaluation, obtaining sufficient tissue cores from EUS-FNB has recently become the primary method for diagnosing malignancies[77]. These techniques are useful for confirming tissue diagnoses and staging tumors, including through biopsies of metastatic liver lesions and lymph nodes[78].

Solid pancreatic tumor

Current guidelines suggest that tissue confirmation via EUS-FNB is not recommended for patients with resectable symptomatic pancreatic tumors as a non-diagnostic biopsy should not delay surgical resection when clinical suspicion for pancreatic cancer is high. Some authors recommend using EUS for initially resectable tumors to avoid major surgery and to provide a second staging assessment; however, it remains essential to ensure that this does not cause significant delays in surgery[55,59,79]. When the pancreatic tumor is borderline resectable, it needs to receive neoadjuvant therapy; thus, EUS-guided biopsy is required. Genetic testing for inherited mutations is recommended in patients with a confirmed diagnosis before neoadjuvant therapy, particularly when the tumor is unresectable or metastatic. A positive biopsy result is required before systemic treatment can be administered. According to current evidence, EUS-FNB is preferred over CT-guided FNB in patients with resectable disease owing to its superior diagnostic yield, enhanced safety, and potentially lower risk of peritoneal seeding[80-82].

Next-generation sequencing (NGS) has been introduced for the molecular analysis of pancreatic tumors. NGS plays a crucial role in identifying diverse genetic alterations and offers significant insights into tumor biology, malignant potential, and therapeutic vulnerabilities of solid pancreatic tumors[83]. This technique facilitates the detection of key oncogenic mutations, such as KRAS, TP53, CDKN2A, and SM-AD4, which are commonly observed in PDAC (Table 1). Identifying actionable mutations, such as BRCA1/2 or PALB2, can inform targeted therapies use, such as poly ADP ribose polymerase inhibitors or platinum agents[84]. Additionally, detecting rare but targetable alterations such as neurotrophic tyrosine receptor kinase fusions or mismatch repair deficiency may make patients eligible for immunotherapy. Importantly, NGS can be successfully performed on small samples obtained by EUS-FNA or EUS-FNB, making it highly applicable in clinical settings where surgical specimens are not available[85].

Table 1 Main genetic alteration in pancreatic tumors.

Pancreatic tumor	Gene altered	Frequency (%)
PDAC	KRASTP53	95
	CDKN2A	35-40
	p16	95-100
	SMAD4	45-60
IPMN	KRAS	45-50
	GNAS	35-50
P-NET	MEN1	10-30
MCN	KRAS	20-50
	GNAS	25-50
SCA	VHL	40-50

PDAC: Pancreatic ductal adenocarcinoma; IPMN: Intraductal papillary mucinous neoplasm; P-NET: Pancreatic neuroendocrine tumor; MCN: Mucinous cystic neoplasm; SCA: Serous cystadenoma.

Obtaining high-quality samples for NGS requires skilled technique. FNB is preferred to FNA because of its capacity to yield more tissue and preserve architectural integrity. Sampling should target viable, non-necrotic regions of the tumor under EUS guidance, ideally with multiple passes to ensure sufficient material. When available during the procedure, rapid on-site evaluation can confirm the presence of tumor cells[86].

Pancreatic cystic lesions

Considering that EUS is an invasive procedure, it is not recommended as a first-line investigation for small benign cysts. In addition, EUS alone is often insufficient to fully characterize pancreatic cystic lesions or accurately assess their malignant potential. However, the true value of EUS lies not only in its ability to provide a detailed morphological assessment of the cyst but also in enabling FNA, which allows for cyst fluid analysis. This combination provides valuable diagnostic information, especially in cases of indeterminate cysts or those with high-risk features[87].

In accordance with current guidelines, resectable pancreatic cystic lesions with objective worrisome signs, including obstructive jaundice due to from lesions at the head portion, enhancing solid components within the cyst, and a main pancreatic duct larger than 1.0 cm in size. However, EUS-FNA, cytological interpretation, and cytological analysis play a role in the evaluation of some pancreatic cystic lesions with equivocal worrisome features, such as BD-IPMN, without worrisome signs[88]. Previous reports have stated that “high-grade epithelial atypia”, which refers to epithelial cells with cellular atypia, is not qualitatively indicative of malignancy[88,89]. In a previous study, such cells in the cyst fluid predicted malignancy in a mutinous cyst with 72% sensitivity and 80% positive predictive value[90]. In another study, 30% more tumors were detected in small IPMN than in IPMNs with worrisome features[88]. CEA cutoff greater than 192-200 ng/mL demonstrates approximately 80% diagnostic accuracy for mucinous cysts[91,92]. Cystic fluid amylase levels are not uniformly elevated in IPMNs, and MCNs may also show elevated amylase levels. Serous cysts typically contain low levels of CEA and amylase[38]. A recent study reported that the detection of KRAS mutations supported mutinous rather than malignant cysts more accurately[93,94].

The typical morphological patterns of SCN are polycystic, honeycomb, and oligocystic. Differentiating between a small oligocystic SCN and a BD-IPMN is challenging and may require EUS-FNA with cystic fluid CEA[95,96].

When cytology and CEA levels are inconclusive, gene sequencing can significantly improve diagnostic accuracy and malignancy risk stratification. Differentiating mucinous from non-mucinous cysts by detecting mutations such as KRAS and GNAS (suggestive of IPMN or MCN, respectively) vs VHL or CTNNB1 (associated with SCN or SPN, respectively)[97]. The presence of high-risk mutations (TP53, PIK3CA, PTEN, and SMAD4) correlated with advanced neoplasia and supported decisions regarding surgical interventions (Table 1). Furthermore, NGS can be applied to cystic fluid obtained by EUS-FNA, offering a minimally invasive method to enhance the diagnostic capacity. Additionally, integrating NGS with imaging and clinical features allows for the development of more accurate management strategies for patients with pancreatic cystic lesions[85].

NGS can be performed on the cyst fluid during sampling pancreatic cystic lesions obtained via EUS-FNA, particularly mucinous cysts such as IPMNs or MCNs. Mural nodules and solid components should be targeted preferentially when present, as they provide more reliable DNA for molecular testing. In cases of fluid-only cysts, maximizing the aspirated volume improves the chance of obtaining adequate DNA (typically > 20 ng). Samples should be handled carefully to avoid contamination and degradation[98,99].

Notably, cystic fluid analysis of mucinous cystic lesions, as well as cysts of any size with worrisome features, is not recommended. Aspiration should be avoided due to the risk of cyst leakage, which can lead to peritoneal dissemination and infection[100]. Currently, EUS-FNA with cytological and molecular analyses plays an investigational role; however, it is recommended for the evaluation of cystic lesions when high-grade dysplasia or invasive carcinoma is suspected.

CONCLUSION

EUS is a highly sensitive imaging tool for detecting and staging pancreatic tumors, particularly small lesions, and offers superior accuracy compared to CT and MRI. When combined with EUS-FNA and FNB, it enhances diagnostic precision by enabling tissue sampling. Advanced techniques, such as CE-EUS and EUS elastography, further improve the differentiation between benign and malignant tumors by assessing vascularization and tissue stiffness. These advancements have also enhanced the feasibility and accuracy of EUS-FNB. Endoscopists can analyze EUS images to distinguish tumors and determine whether further aspiration or biopsy is necessary to obtain tissue confirmation, especially in challenging cases, such as chronic pancreatitis mimicking malignancy or metastatic tumors. In addition, it would be valuable to assess how the integration of EUS imaging with molecular analyses such as NGS guides clinical decision-making, including the need for surgical intervention, adjustment of surveillance intervals, and consideration of targeted therapies. Establishing a stronger correlation between imaging and pathology may improve the diagnostic accuracy and support the advancement of personalized medicine in pancreatic disease management.