Cystic tumors of the pancreas: Current perspectives on diagnosis and management

Abstract

Pancreatic cystic lesions are being increasingly detected, mainly due to the widespread use of cross-sectional imaging. The reported prevalence ranges from 13% to 18% in asymptomatic individuals. These lesions display a broad histologic spectrum, from benign pseudocysts to premalignant mucinous cystic neoplasms and invasive carcinomas. Although many classification and management strategies exist, the natural history of numerous pancreatic cystic lesions remains incompletely understood, contributing to significant clinical uncertainty. Current diagnostic tools, including computed tomography, magnetic resonance imaging, endoscopic ultrasound, and cyst fluid analysis, are constrained by either suboptimal sensitivity or high costs. Cytology, while specific when positive, suffers from low sensitivity. Biochemical markers such as carcinoembryonic antigen, amylase, and glucose can help in cyst differentiation, whereas molecular testing (e.g., KRAS, GNAS, RNF43 mutations) provides additional diagnostic and prognostic value. However, the application of molecular diagnostics is still restricted in routine practice due to costs, access issues, and a lack of standardization. This diagnostic uncertainty leads to both overtreatment and undertreatment. Some patients undergo unnecessary surgeries for benign lesions, which exposes them to procedural risks and long-term consequences. Others may experience delays in interventions for high-risk cysts and missing opportunities for cancer prevention. Additionally, prolonged, and often unnecessary surveillance burdens patients and healthcare systems psychologically and financially. In this minireview, we present a comprehensive overview of the classification, diagnostic approach, and management of pancreatic cystic lesions, incorporating recent evidence and current international guidelines (Fukuoka, American Gastroenterological Association, European). We also highlight the limitations of existing strategies and emerging tools such as radiomics, next-generation sequencing, and novel biomarkers. Additionally, we emphasize the urgent need for cost-effective, accurate, and accessible diagnostic pathways. A more refined risk stratification approach is essential to optimize outcomes, reduce healthcare waste, and improve the quality of life for patients with pancreatic cystic lesions.

Key Words: Pancreatic cystic lesions; Intraductal papillary mucinous neoplasm; Radiomics; Molecular diagnostics; Risk stratification; Artificial intelligence; Endoscopic ultrasound; Next-generation sequencing; Pancreatoscopy; Multidisciplinary management

Core Tip: Pancreatic cystic lesions are increasingly detected in clinical practice, yet their diagnosis and management remain challenging due to overlapping imaging features, variable malignant potential, and limitations in current diagnostic tools. This minireview provides an up-to-date overview of classification systems, diagnostic strategies, including imaging, cyst fluid analysis, and molecular testing, as well as international guideline recommendations. It also highlights emerging technologies such as radiomics and next-generation sequencing. Emphasis is placed on the urgent need for cost-effective, accurate, and patient-focused risk stratification models to guide personalized surveillance and therapeutic decisions.

INTRODUCTION

Pancreatic cystic lesions (PCLs) have become an increasingly common clinical finding, primarily due to the widespread use of cross-sectional imaging modalities such as computed tomography (CT) and magnetic resonance imaging (MRI). Studies estimate their prevalence to be between 13% and 18% in asymptomatic individuals, particularly in older populations[1]. While many of these lesions are benign or indolent, others carry significant malignant potential, accounting for approximately 15% of pancreatic ductal adenocarcinoma cases[2].

PCLs encompass a broad histopathologic spectrum, ranging from benign pseudocysts to premalignant and malignant neoplasms such as intraductal papillary mucinous neoplasms (IPMNs) and mucinous cystic neoplasms (MCNs). Differentiating between these entities is crucial to avoid both overtreatments, including unnecessary pancreatic surgery, and undertreatment, which risks delayed diagnosis of potentially curable pancreatic cancer.

Despite the availability of various diagnostic tools, including advanced imaging, endoscopic ultrasound (EUS), and cyst fluid analysis, accurate risk stratification remains suboptimal, with a high rate of surgery for non-malignant cysts[3]. Current guidelines[4-6] offer varying recommendations regarding surveillance and intervention, often resulting in inconsistent clinical practices. Furthermore, the increasing psychological and economic burden of prolonged surveillance[7] underscores the need for the development of more precise diagnostic and management strategies.

This minireview aims to provide a comprehensive and up-to-date overview of the classification, diagnostic approach, and clinical management of PCLs. We also explore emerging diagnostic technologies and discuss the need for more refined, accessible, and cost-effective risk-stratification pathways.

CLASSIFICATION OF PCLS

PCLs involve a broad spectrum of entities ranging from benign non-neoplastic collections to premalignant and malignant neoplasms. Accurate classification is essential for risk stratification, clinical decision-making, and optimizing patient outcomes. Below is an overview of the major categories of PCLs, categorized by etiology and malignant potential.

Non-neoplastic cystic lesions

Pseudocysts: Pancreatic pseudocysts are the most common cystic lesions of the pancreas[8]. They result from encapsulated collections of fluid, rich in pancreatic enzymes, typically following acute pancreatitis, chronic pancreatitis, trauma, or pancreatic ductal disruption. Pseudocysts are surrounded by fibrous tissue rather than an epithelial lining, distinguishing them from true cystic neoplasms[9]. Although they can appear similar to mucinous cystic tumors, a history of pancreatitis or trauma is often key to diagnosis. Most resolve spontaneously or with supportive and endoscopic therapy. Surgical management is rarely indicated unless complications arise.

Benign neoplastic lesions

Serous cystadenoma: Serous cystadenomas (SCAs) are benign epithelial tumors composed of numerous small cysts lined by glycogen-rich cuboidal epithelium and supported by a fibrous stroma[9]. They account for approximately 30% of all cystic neoplasms of the pancreas and are most frequently observed in older women, commonly located in the pancreatic head[10]. SCAs do not communicate with the pancreatic duct and rarely undergo malignant transformation. Most are asymptomatic and require no intervention unless symptomatic due to mass effect. They may be associated with von Hippel-Lindau disease[11].

Premalignant and malignant cystic neoplasms

MCN: MCNs are mucin-producing cystic tumors that almost exclusively occur in middle-aged women and are typically located in the body or tail of the pancreas[12]. They are defined histologically by the presence of ovarian-type stroma and do not communicate with the pancreatic ductal system[9]. Although many are benign, up to one-third may harbor high-grade dysplasia or invasive carcinoma. All MCNs are considered surgical lesions due to their potential for malignancy. The risk of high-grade dysplasia or invasive adenocarcinoma within an MCN has varied in the literature from 10% to 39%[13-15].

IPMN: IPMNs are epithelial neoplasms characterized by intraductal proliferation of mucin-producing cells. IPMNs are divided into “main duct type”, which arise from the main pancreatic duct, and “branch duct type”, which arise in one of the branches of the main pancreatic duct and are considered the most common pancreatic cyst[16]. The risk of malignancy transformation in main duct type and branch duct type IPMNs is approximately 38% to 68% and 22%, respectively[17,18]. IPMNs are more common in older men and may occur multifocally. They are considered precursor lesions to pancreatic adenocarcinoma and require individualized management based on risk features and patient factors[19].

Solid pseudopapillary neoplasm: Solid pseudopapillary neoplasms are rare, low-grade malignant tumors that primarily affect young women, particularly of Asian or African descent[20]. They are composed of poorly cohesive epithelial cells forming pseudopapillary structures[9]. Although indolent, they can grow large and may cause mass effect. Surgical resection is curative in most cases, and recurrence is rare.

Rare and miscellaneous cystic lesions

True epithelial cysts: True congenital epithelial cysts are rare and typically asymptomatic[21]. They are lined by cuboidal epithelium and lack neoplastic potential[9]. They are often incidental findings and may be associated with systemic syndromes like von Hippel-Lindau disease[22].

Ductal adenocarcinoma with cystic degeneration: Although pancreatic ductal adenocarcinoma is typically solid, cystic degeneration may occur due to central necrosis or obstructed branch ducts. These lesions may mimic benign cysts but are usually associated with features of invasive malignancy, such as ductal obstruction, vascular encasement, and rapid progression. Cystic presentation of adenocarcinoma should always raise suspicion in the appropriate clinical context[23].

Other rare entities: Several uncommon tumors and lesions may present with cystic morphology, including lymphangiomas, cystic teratomas, sarcomas, paragangliomas, and metastatic lesions (e.g., from renal cell carcinoma). These are exceedingly rare and typically diagnosed via histologic or genetic analysis following surgical resection[24].

Cystic neuroendocrine tumors (cNETs) are rare PCLs that can be associated with multiple endocrine neoplasia type 1 (MEN1)[21]. These tumors account for a small percentage of pancreatic cysts and may present as incidental findings or with symptoms like abdominal pain and weight loss[25].

On CT and MRI, cNETs typically appear as well-defined masses, sometimes with calcifications. EUS helps detect cystic lesions with solid components, which can be indicative of cNETs[26].

Management: Functional cNETs (e.g., insulinomas) often require surgical resection due to hormone secretion, while non-functional cNETs are typically resected if they exceed 3 cm or show signs of malignancy[27]. Patients with MEN1 require close monitoring due to the increased risk of multifocal tumors and malignancy[28].

DIAGNOSTIC MODALITIES

Understanding the nature and behavior of each type of lesion is crucial for risk stratification and informed therapeutic decision-making, particularly as we transition toward more personalized and conservative management strategies. However, current diagnostic tools differ significantly in their accuracy, accessibility, and cost-effectiveness[29]. A multimodal approach is often necessary, integrating imaging, endoscopic techniques, cyst fluid analysis, and, when available, molecular diagnostics.

Cross-sectional imaging

MRI with cholangiopancreatography is the preferred noninvasive modality for the initial evaluation of PCLs, as it provides high-resolution visualization of cyst morphology, ductal communication, internal septations, and mural nodules[5,30]. Its lack of ionizing radiation also makes it particularly suitable for long-term surveillance. CT, on the other hand, is widely available and offers excellent spatial resolution, making it particularly helpful in detecting calcifications, mural nodules, and cyst wall thickening, features that may indicate malignant potential. However, CT is less sensitive than MRI in identifying more subtle findings, such as communication with the pancreatic duct[31].

EUS

EUS plays a pivotal and evolving role in the diagnostic workup and risk stratification of pancreatic cystic neoplasms, especially those with intermediate risk. It enables high-resolution visualization of cyst morphology and vascular structures, and is particularly helpful in identifying mural nodules, septations, and assessing vascular invasion. With the rising incidence of pancreatic cysts and the recognition that only a small proportion progress to malignancy, EUS has emerged as a critical tool to prevent unnecessary surgeries and tailor surveillance. EUS-guided fine-needle aspiration (FNA) enables the analysis of cyst fluid for key biomarkers, such as carcinoembryonic antigen (CEA), glucose, and amylase[32], as well as for cytology and DNA mutational analysis (e.g., KRAS, GNAS)[33], thereby enhancing diagnostic precision. Advanced imaging techniques, such as contrast-enhanced EUS, help distinguish true mural nodules from mucin or debris[34]. Harima[35] reported that contrast-enhanced harmonic EUS (CH-EUS) demonstrated significantly superior diagnostic accuracy for detecting mural nodules (98%) compared to CT (92%) and conventional EUS (72%), with statistically significant differences (CT vs CH-EUS, P < 0.05; conventional EUS vs CH-EUS, P < 0.01). Meanwhile, EUS elastography, although still under investigation, may aid in differentiating benign from malignant lesions based on tissue stiffness[36]. A 2025 study by Wang et al[37] highlighted that while EUS rarely alters the diagnosis in low-risk pancreatic cysts, it influences management in approximately 30% of cases by refining surgical decisions and surveillance strategies. Although agreement between cross-sectional imaging and EUS was high, 97% for high-risk stigmata and 94% for worrisome features, EUS often led to reclassification of lesions, particularly those near key thresholds (e.g., mural nodules < 5 mm or cysts near 3 cm). These findings support a more selective use of EUS, reserving it primarily for indeterminate or intermediate-risk lesions, where it offers the most significant clinical utility in confirming or refuting malignancy potential.

Cyst fluid analysis

Biochemical analysis of cyst fluid plays a supportive role in differentiating PCLs, although no single marker offers definitive diagnostic accuracy. Cyst fluid CEA remains one of the most widely utilized biochemical markers for differentiating mucinous from non-mucinous pancreatic cysts. Traditionally, a threshold of ≥ 192 ng/mL has been used to suggest mucinous etiology. However, recent data from a large retrospective cohort of 1169 cases[38], including 394 histologically confirmed lesions and 237 with confirmative molecular profiles (e.g., KRAS, GNAS, or RNF43 mutations), challenge the diagnostic performance of this cutoff. While median CEA levels were significantly higher in mucinous cysts compared to non-mucinous counterparts (323.9 ng/mL vs 204.6 ng/mL, P < 0.001), the optimal cut-off for differentiating cyst type was found to be much lower (20 ng/mL), yielding high sensitivity (89%) but modest specificity (64%). At the conventional threshold of 192 ng/mL, sensitivity decreased to 56% while specificity improved to 78%. To achieve a specificity of 85%, as initially reported in earlier studies, a higher threshold of 250 ng/mL was required, although this resulted in more false negatives. These findings underscore that while CEA remains a valuable tool in assessing PCLs, it is less specific than previously believed, and its interpretation should be contextualized alongside imaging, clinical features, and complementary molecular or cytological analyses.

In addition to CEA, other cyst fluid biomarkers, such as amylase and glucose, contribute to diagnostic insight but also have significant limitations. Elevated amylase levels are indicative of communication with the pancreatic duct, as seen in pseudocysts and IPMNs, but do not aid in distinguishing between mucinous and non-mucinous neoplasms[39]. Conversely, low cyst fluid glucose levels (< 50 mg/dL) have recently gained attention as a low-cost, accessible, and relatively sensitive marker for identifying mucinous cysts. Several studies have shown that glucose may outperform CEA in terms of diagnostic accuracy in specific clinical settings, particularly due to its binary interpretability and availability[40]. However, none of these biomarkers, whether used alone or together, provides definitive diagnostic certainty. Their values must be interpreted in the context of the clinical presentation, radiologic features, and, when available, cytological or molecular findings to guide appropriate management and avoid misclassification.

Cytology

Cyst fluid cytology is highly specific when positive but suffers from poor sensitivity due to the paucity of cellular material in most cysts. A positive cytologic diagnosis of high-grade dysplasia or malignancy is rare but clinically decisive[41].

Molecular diagnostics

Molecular analysis of cyst fluid represents a significant advancement in the diagnostic evaluation of PCLs, particularly in differentiating cyst subtypes and predicting malignant potential. Specific genetic alterations have been linked to distinct lesion types and varying degrees of dysplasia. KRAS and GNAS mutations are commonly found in IPMNs, while mutations in RNF43, TP53, and PIK3CA are more frequently associated with high-grade dysplasia or invasive carcinoma[42].

Despite their promise, the use of molecular diagnostics in routine practice remains limited. Significant barriers include the high cost of testing, restricted availability in non-tertiary settings, and the lack of standardization in panel selection, testing platforms, and interpretation of results. Furthermore, variability in the sensitivity and specificity of molecular findings across studies complicates their application in real-world settings.

Emerging molecular techniques such as next-generation sequencing (NGS) and digital droplet polymerase chain reaction (ddPCR) offer highly sensitive and precise methods for detecting these mutations in cystic fluid. NGS provides a comprehensive profile of multiple genetic alterations, aiding in the classification of PCLs, while ddPCR is highly effective in detecting specific mutations, even in limited cyst fluid samples[43].

The clinical implications of molecular diagnostics are significant, as identifying high-risk mutations, such as KRAS or TP53, can guide treatment decisions, including whether surgical resection is necessary. However, the high costs of these molecular tests, along with challenges in standardization and limited availability, present obstacles to their widespread clinical implementation.

CURRENT INTERNATIONAL GUIDELINES

Given the diversity of PCLs and their variable malignant potential, several professional societies have developed clinical guidelines to support their evaluation, risk stratification, and management. However, notable differences exist among these guidelines, resulting in substantial variability in clinical decision-making across institutions and specialties. The three most widely adopted frameworks include the revised international consensus guidelines (Fukuoka 2017), the American Gastroenterological Association (AGA) guidelines 2015, and the European evidence-based guidelines on pancreatic cystic neoplasms 2018. Each offers a distinct approach to classification, surveillance, and surgical referral criteria.

Fukuoka guidelines

The Fukuoka guidelines[5] focus primarily on IPMNs and MCNs. Management decisions are stratified according to specific clinical and radiological features: (1) High-risk stigmata, including obstructive jaundice, an enhancing mural nodule ≥ 5 mm, or a main pancreatic duct diameter ≥ 10 mm, warrant immediate surgical referral; (2) Worrisome features, such as cyst size ≥ 3 cm, mural nodules < 5 mm, thickened or enhancing cyst walls, main duct dilation of 5-9 mm, or a history of pancreatitis, warrant further evaluation with EUS; and (3) Surveillance recommendations are individualized based on cyst characteristics, with lifelong monitoring suggested for patients with IPMNs, given their potential for malignant transformation.

AGA guidelines

The AGA guidelines (2015)[44] adopt a more conservative and cost-conscious approach, aiming to reduce unnecessary surgical interventions. Surveillance is recommended only for patients exhibiting two or more high-risk features, such as a cyst size of 3 cm or greater, the presence of a solid component, or main pancreatic duct dilation. Surgical referral is advised if additional concerning features are detected on imaging or EUS-FNA, including positive cytology or elevated cyst fluid CEA. In cases where a cyst remains stable and without worrisome features for five years, discontinuation of surveillance is recommended. Although this strategy minimizes overtreatment, it has been criticized for its lower sensitivity in detecting high-grade dysplasia or early malignancy when compared to more aggressive guidelines such as Fukuoka[45].

European guidelines

The European evidence-based guidelines (2018)[46] offer a broader scope by addressing all types of PCLs, including both mucinous and non-mucinous entities. Key recommendations include: (1) Risk stratification incorporating both imaging characteristics and cyst fluid biomarkers; (2) Strong emphasis on multidisciplinary team discussion for cases with indeterminate or borderline features; (3) Surveillance protocols that are tailored based on cyst type, size, patient age, and comorbidities, enabling more individualized care; and (4) Preferential use of MRI over CT for routine follow-up, given its superior soft-tissue contrast and lack of radiation exposure, and selective application of EUS-FNA and molecular analysis to further characterize unclear lesions.

Together, these three guideline frameworks reflect varying clinical philosophies, ranging from aggressive surgical intervention to risk-adapted, conservative surveillance. This highlights the need for more unified, evidence-based approaches that strike a balance between diagnostic accuracy, patient safety, and healthcare resource utilization (Table 1).

Table 1 Comparative overview of major international guidelines for the management of pancreatic cystic lesions.

	Fukuoka (2017)	AGA (2015)	European (2018)
Focus	IPMN and MCN	All PCLs	All PCLs
High-risk stigmata	Yes	Not emphasized	Yes
Role of EUS-FNA	Essential for worrisome	Only if ≥ 2 risk factors	Strongly recommended
Surveillance duration	Lifelong	Stop after 5 years stable	Tailored, often lifelong
Role of MDT	Optional	Not emphasized	Strongly recommended
Molecular testing	Not addressed	Not included	Recommended in selected cases

AGA: American Gastroenterological Association; EUS: Endoscopic ultrasound; FNA: Fine-needle aspiration; MDT: Multidisciplinary team; IPMN: Intraductal papillary mucinous neoplasm; MCN: Mucinous cystic neoplasm; PCLs: Pancreatic cystic lesions.

CHALLENGES IN CLINICAL MANAGEMENT

While all guidelines aim to balance risk with cost and patient burden, their differing thresholds for intervention and surveillance create inconsistencies and uncertainties in clinical practice. The absence of unified criteria highlights the need for more standardized, evidence-based pathways that incorporate both clinical features and emerging diagnostic technologies.

Overdiagnosis and overtreatment

The increasing use of high-resolution imaging has led to a significant rise in the incidental detection of PCLs, many of which are indolent or benign[47]. However, given the current limitations in reliably excluding malignancy preoperatively, a significant number of patients undergo surgical resection for cysts that are ultimately diagnosed as non-neoplastic or low-risk mucinous neoplasms. Pancreatic surgery is associated with considerable morbidity, including the risk of postoperative pancreatic fistula, delayed gastric emptying, and long-term endocrine or exocrine insufficiency[48]. Consequently, unnecessary surgical interventions may adversely affect patient quality of life, particularly in elderly or medically complex individuals, and contribute to increased healthcare costs and resource utilization[49].

Undertreatment and missed malignancies

On the other hand, reliance on a conservative management strategy based solely on imaging characteristics or cyst size may result in the under-recognition of high-risk or malignant PCLs. This is particularly concerning in cases lacking overt clinical or radiological indicators of malignancy. Certain branch-duct IPMNs and MCNs may harbor high-grade dysplasia while appearing morphologically indolent on imaging[50]. The inability to detect these subtle malignant features can lead to delayed surgical intervention, thereby forfeiting the opportunity for potentially curative resection and increasing the risk of disease progression.

Burden of surveillance

Surveillance protocols, particularly those outlined in the Fukuoka and European guidelines, frequently recommend long-term or, in some cases, lifelong follow-up with serial imaging, such as MRI or EUS, at defined intervals[5,46]. While concerns have been raised regarding the potential psychological impact of repeated surveillance, current evidence suggests that the overall psychological burden for most patients remains modest[51]. Economically, the cumulative cost of repeated imaging imposes a substantial burden on both healthcare systems and individual patients, particularly when advanced modalities such as EUS are used[7].

Inconsistent interpretation of guidelines

The presence of various clinical guidelines, each offering different thresholds for surveillance, intervention, and the use of ancillary diagnostics, has led to considerable variation in the management of pancreatic cystic disease lesions, often affecting even physician adherence. Interpretations of malignancy risk often diverge across specialties, with gastroenterologists, surgeons, and radiologists emphasizing different clinical and radiologic parameters. Furthermore, variations in access to advanced diagnostics, such as EUS, molecular testing, and high-resolution imaging, often influence decision-making in ways that existing guidelines do not fully account for.

Diagnostic uncertainty and decision fatigue

Despite the availability of multimodal diagnostic strategies, a substantial proportion of PCLs remain indeterminate. Unclear findings, such as sub-centimeter mural nodules, borderline CEA levels, or inconclusive cytology, often fall within a diagnostic gray zone that lacks precise management algorithms[52]. In such cases, clinicians may default to a cautious “watch-and-worry” approach, characterized by ongoing surveillance without a well-defined endpoint or strong clinical rationale. The absence of diagnostic certainty can also lead to a conservative bias in clinical decision-making, with physicians opting for prolonged follow-up or even premature intervention out of concern for missing a potentially malignant lesion[53]. This tendency, while well-intentioned, may inadvertently contribute to overtreatment and increased patient burden, particularly in cases where the actual risk of malignant progression is low.

EMERGING TOOLS AND FUTURE DIRECTIONS

Given the limitations of conventional diagnostic tools and current guidelines, considerable effort has been devoted to developing novel strategies to enhance the risk stratification and management of PCLs[54].

Radiomics and artificial intelligence

Radiomics is a novel imaging analysis technique that extracts quantitative features, such as texture and pixel intensity, from cross-sectional scans, enabling characterization of PCLs beyond what is visible to the human eye[55]. When combined with machine learning, radiomics has shown promise in differentiating cyst types (e.g., IPMNs, MCNs, SCAs) and stratifying the risk of malignancy. Early studies report high diagnostic accuracy, with some models achieving over 90% sensitivity and specificity in identifying high-grade dysplasia or carcinoma in IPMNs[56]. Integration of radiomics with clinical and morphological features further enhances performance. However, its use remains limited to research settings, primarily based on single-center or retrospective studies. There is also the issue of the lack of standardized protocols for image acquisition and analysis, as well as the absence of phase III trials, which limits their clinical applicability and generalizability. Despite these limitations, radiomics holds significant potential as a non-invasive tool for refining diagnosis and guiding management in patients with PCLs.

EUS-guided needle-based confocal laser endomicroscopy

EUS-guided needle-based confocal laser endomicroscopy (nCLE) offers a real-time “optical biopsy” by advancing a fluorescein-primed mini-probe through a 19-G needle into the cyst cavity[57]. Characteristic image signatures, papillary fronds (IPMN), horizontal epithelial bands (MCN), and a superficial vascular network, have been validated with sensitivities and specificities approaching 95%-100% for distinguishing mucinous from non-mucinous lesions[58]. Independent cohorts confirm a diagnostic yield of approximately 84%, markedly superior to the traditional composite of cross-sectional imaging, cytology, and CEA[59]. Limitations, such as small prospective datasets, operator-dependent image interpretation, and high procedural costs, currently hinder the widespread adoption of nCLE. As a result, it has not yet been incorporated into standard clinical guideline algorithms. Larger, cost-effectiveness studies are required before routine adoption in pancreatic cyst management.

EUS-guided through-the-needle biopsy

EUS-guided through-the-needle biopsy (EUS-TTNB) has garnered interest in recent years as a valuable tool for determining cyst type and assessing the risk of neoplastic cysts stratification[60]. Several systematic reviews and meta-analyses have revealed a higher diagnostic performance of EUS-TTNB compared with EUS-FNA[61-63]. However, safety concerns[64], questions regarding patient selection[65], and doubts about the procedure’s clinical impact[64] have prevented its widespread adoption among endoscopists. A meta-analysis compared EUS-TTNB with nCLE[66]. The diagnostic yield (defined as the likelihood of obtaining a diagnosis) was higher with nCLE (85% vs 74%, P < 0.0001). Still, the sensitivity and the specificity were similar (pooled sensitivity: 80% vs 86% and pooled specificity: 80% vs 83% for EUS-TTNB and nCLE, respectively), as were technical success and adverse event rates[66]. The EUS-TTNB should be used when it provides added diagnostic value and can influence the decision-making process[65].

Novel biomarkers

Recent biochemical advancements aim to enhance the diagnostic performance of cyst fluid analysis in PCLs. Glucose has emerged as a promising, low-cost biomarker; levels below 50 mg/dL have demonstrated high sensitivity and specificity for identifying mucinous cysts, potentially serving as a practical alternative to CEA[67,40]. Additionally, novel molecular markers, including microRNAs, DNA methylation profiles, proteomic signatures, and exosomal components, are under active investigation and may enhance both diagnostic accuracy and risk stratification[68]. Furthermore, inflammatory and immune-related markers offer potential insights into the tumor microenvironment and its role in neoplastic progression[69]. While these emerging biomarkers are encouraging, their clinical utility (except for the use of glucose levels, which are widely incorporated into the clinical setting) remains limited pending validation in large, prospective studies.

Integrated clinical algorithms

Multiple research groups are working to develop comprehensive diagnostic models that integrate clinical data, imaging characteristics, EUS findings, cyst fluid biomarkers, and molecular profiles into unified risk stratification tools. Notable examples include the PANC-CYST and PACYFIC scoring systems, as well as artificial intelligence-driven risk calculators incorporated into radiology platforms or electronic health records[70,71]. These integrated approaches aim to move beyond fixed guideline criteria, enabling more personalized management strategies, such as tailored surveillance intervals and individualized surgical decision-making, based on quantified risk of malignancy.

Minimally invasive therapies

Experimental techniques, such as EUS-guided cyst ablation using ethanol, paclitaxel, or radiofrequency ablation, are being explored as alternatives to surgery in select patients with low-to-intermediate-risk cysts. EUS ablation of pancreatic cysts yields acceptable rates of complete resolution and a low incidence of severe adverse events, with the use of chemoablative agents resulting in higher performance rates[72]. These approaches remain investigational and should be pursued within clinical trials.

THE NEED FOR BETTER RISK STRATIFICATION

A central challenge in managing PCLs is the inability to reliably distinguish between benign, premalignant, and malignant lesions using currently available diagnostic modalities. This diagnostic uncertainty often results in either overtreatment or missed opportunities for cancer prevention. Therefore, the development of robust, accessible, and personalized risk stratification strategies is crucial.

Limitations of size-based and morphologic criteria

Current clinical guidelines primarily use cyst size, the presence of mural nodules, and ductal dilation as surrogate indicators of malignant potential in PCLs. However, these morphological parameters lack sufficient specificity to reliably guide management decisions[73]. Larger cysts are not necessarily high-risk, as many are benign, while smaller lesions may harbor high-grade dysplasia or even early invasive carcinoma. Moreover, imaging interpretation is often limited by ambiguity; features such as subtle mural nodules or thickened cyst walls can be challenging to distinguish from mucin or debris, even with high-resolution modalities like MRI or EUS[74]. Consequently, relying solely on size-based thresholds may lead to misclassification of risk, resulting in both overtreatment and missed opportunities for timely intervention[75].

Personalized surveillance and intervention strategies

Risk stratification in the management of PCLs should go beyond the simple decision of surgical intervention and provide a framework for personalized care throughout the patient’s treatment. It should guide the frequency and duration of surveillance, allowing lower-risk patients to be monitored with less intensive imaging. Additionally, it should help identify candidates for advanced diagnostics, such as molecular testing or EUS, particularly in cases with indeterminate or intermediate-risk features[76]. Surgical referral should be reserved for those at high risk of malignancy with an acceptable operative profile[77]. Sharib et al[29] highlight the significant costs of surveillance, especially with frequent imaging, but a comprehensive comparison of cost-effectiveness ratios, such as between early surgery and intensive monitoring, is necessary. This comparison will enable clinicians to make more informed decisions, balancing unnecessary procedures with the benefits of early intervention for high-risk lesions. A tiered, individualized approach can optimize clinical outcomes, reduce unnecessary interventions, and lessen the economic and psychological burden on both patients and healthcare systems.

Risk-stratified individualized strategy for PCLs

The management of PCLs should be guided by a flexible, risk-stratified approach that combines the high-risk features from the Fukuoka guidelines with cost-effective principles outlined by the AGA. This strategy aims to optimize patient care while minimizing unnecessary procedures and healthcare costs. The management plan is divided into three risk groups: High-risk, intermediate-risk, and low-risk, with decisions based on cyst characteristics, imaging findings, and molecular diagnostics. A multidisciplinary team should be involved in high-risk cases for optimal decision-making. A flowchart illustrating this strategy is presented as Figure 1.

Figure 1 Risk-stratified management of pancreatic cystic lesions. MRI: Magnetic resonance imaging; MRCP: Magnetic resonance cholangiopancreatography; EUS: Endoscopic ultrasound; MPD: Main pancreatic duct; FNA: Fine-needle aspiration; IPMN: Intraductal papillary mucinous neoplasm; cNET: Cystic neuroendocrine tumor; MEN1: Multiple endocrine neoplasia type 1.

Postoperative surveillance

Surveillance of the remnant pancreas following resection for pancreatic cystic neoplasms remains a clinically essential yet understudied area. While recurrence rates vary across studies, with reported repeat pancreatectomy rates ranging from 1.4% to 11%[78], most recurrences occur within the first three years postoperatively, underscoring the importance of intensive early surveillance[79]. However, late recurrences also occur, justifying the need for prolonged, and in some cases, lifelong follow-up. Surveillance protocols vary widely, particularly in the context of resected IPMNs. Some experts recommend 3-6 months intervals in the initial two years, followed by annual imaging, whereas others propose less intensive schedules. Follow-up may be discontinued when patients become surgically unfit or decline further intervention[80]. For MCNs, ongoing surveillance is generally reserved for cases with invasive carcinoma[79]. High-risk histologic features, such as high-grade dysplasia or multifocality, warrant more vigilant monitoring, and invasive IPMNs should be followed using protocols similar to those for resected pancreatic ductal adenocarcinoma. Imaging, primarily MRI, CT, and EUS, remains central to surveillance, with a focus on high-risk stigmata, such as solid masses or main duct dilation. The role of serum tumor markers, including carbohydrate antigen 19-9 and CEA, is less well-defined in the postoperative setting, although they are often monitored at 3-month to 6-month intervals. Elevated markers may precede imaging-detectable recurrence, though their sensitivity remains suboptimal[81]. Advances in molecular diagnostics, including circulating tumor DNA, microRNAs, and other multi-omics approaches, hold promise for improving recurrence detection, but data specific to PCLs are still limited[82]. Notably, a recent algorithm that integrates cyst fluid biomarkers and imaging data has shown high diagnostic accuracy for both mucinous and non-mucinous cysts, demonstrating its potential utility in guiding postoperative surveillance[71]. Nevertheless, further prospective studies are needed to validate the use of such biomarker-driven strategies and to define optimal surveillance intervals, modalities, and endpoints for patients with resected PCLs.

Pancreatic exocrine insufficiency and metabolic dysfunction following pancreatic surgery

Pancreatic exocrine insufficiency (PEI) is a common complication following pancreatic surgery, irrespective of the underlying pathology. A systematic review and meta-analysis by Beger et al[83] reported that 44.9% of patients developed new-onset postoperative PEI after a mean follow-up of 32 months postoperatively. However, current data primarily derive from heterogeneous surgical cohorts and do not specifically reflect outcomes in patients with PCLs.

Metabolic dysfunction, including steatotic liver disease (SLD) and new-onset diabetes (NOD), represents significant long-term complications following pancreatic surgery, with incidence rates closely linked to the type and extent of resection. Evidence suggests that up to one-third of patients undergoing pancreatectomy for pancreatic cystic neoplasms may develop NOD[84], while approximately one in four may develop SLD[84]. The pathophysiology of these conditions is multifactorial and closely related to the type and extent of the surgical resection. Specifically, pancreaticoduodenectomy, which includes both pancreatic and duodenal resection, has a higher incidence of metabolic disturbances due to its impact on both endocrine and exocrine functions. This surgical procedure leads to a loss of pancreatic tissue that contains insulin-producing β-cells and significantly reduces insulin secretion, contributing to endocrine insufficiency[85].

The regional distribution of β-cells across the pancreas plays a significant role in metabolic outcomes after surgery. β-cells are more densely concentrated in the tail of the pancreas, and resection of this part can result in more profound endocrine dysfunction[86]. The loss of β-cell mass and function following resection can lead to NOD, particularly when the remaining pancreatic tissue is insufficient to compensate for insulin production. Interestingly, islet cell plasticity and the trans differentiation of exocrine cells into insulin-producing cells have been observed in some patients, which may offer partial functional recovery post-surgery; however, the extent of this regenerative process is limited and highly variable.

Additionally, in some instances, resection of malignant or pre-malignant cystic lesions has been reported to result in the resolution of diabetes, suggesting the potential involvement of a paraneoplastic mechanism. This phenomenon is thought to be related to the suppression of insulin secretion caused by a cystic lesion or its secretory products[87]. However, more research is needed to better understand this interaction and its potential for therapeutic strategies.

Similar to PEI, metabolic disturbances such as NOD and SLD can significantly impact nutritional status, postoperative recovery, and long-term survival. Patients with SLD are at risk for malabsorption and fat-soluble vitamin deficiencies. At the same time, those with NOD often face challenges with glycemic control, which can complicate the management of other post-surgical complications[88].

CONCLUSION

PCLs represent a growing clinical challenge due to their increasing incidental detection and broad spectrum of malignant potential. While existing classification systems and guidelines provide a foundational framework for management, they are hindered by diagnostic uncertainty, variability in recommendations, and the inherent limitations of current tools. Imaging, EUS, and cyst fluid analysis remain central to the diagnostic workup but often fall short in accurately predicting malignancy. Additionally, they often fail to include other non-mucinous malignant cystic entities such as cystic ductal adenocarcinoma and cNETs. Emerging modalities, such as radiomics, molecular diagnostics, and novel biomarkers, show great promise but are not yet fully integrated into standard clinical practice due to issues of cost, access, and standardization. Imperfect diagnostics, guideline variability, and a lack of standardized, risk-based decision-making frameworks hinder the clinical management of PCLs. As a result, both overtreatment and undertreatment coexist, leading to substantial patient and healthcare system burden. A more refined, evidence-driven approach is urgently needed. Emerging technologies, including radiomics, molecular diagnostics, and novel biomarkers, hold great potential to transform the management of PCLs. As these tools become more validated and accessible, they may allow for more individualized, accurate, and cost-effective care. However, integration into routine practice will require robust clinical validation, standardization, and collaboration across specialties. Moving forward, multidisciplinary collaboration, large-scale prospective validation of risk models, and the development of scalable, cost-effective diagnostic pathways will be essential. By refining our approach to PCLs, we can optimize patient outcomes, reduce harm, and divide up healthcare resources more effectively.