Cystic Pancreatic Lesions Diagnosed on Endoscopic Ultrasound: Frequency, Morphology, and Factors Associated with Malignant Nature of these Lesions

Abstract

Background

Cystic pancreatic lesions (CPLs) constitute an important part of pancreatic diseases. This study aimed to identify the frequency of the various CPLs evaluated through endoscopic ultrasound (EUS) and factors associated with malignant histology of the cyst.

Materials and methods

In this observational study, all patients with CPLs referred for EUS were included. All extra-pancreatic fluid collections were excluded. The frequency, morphology, and demographics of patients with CPLs were recorded. Factors associated with malignant histology of the cyst were statistically analyzed.

Results

Of the total 56 patients included in the study, females were more common (34; 60.7%) than males, and the mean age was 46.5 (±14.6) years. The majority of the CPLs were pure cystic lesions 41 (73.2%), while the remaining were solid-cystic lesions 15 (26.8%). The majority of the CPLs were located in the head 30 (53.6%), followed by the body 14 (25%) of the pancreas. The commonest CPLs were: Serous cystadenoma (SCN) 11 (19.6%), intraductal papillary neoplasm (IPMN) 10 (17.9%), mucinous cystic neoplasm (MCN) 9 (16.1%), associated with a specific disease 7 (12.5%), malignancy 7 (12.5%), pancreatitis 6 (10.7%), simple cyst 5 (8.9%) and solid pseudopapillary tumor 1 (1.8%). Malignancy was detected among seven patients with CPL: Cystic degeneration of pancreatic ductal adenocarcinoma 3, neuroendocrine cancer 3, and IPMN with malignant transformation 1. Factors associated with the malignant nature of cyst were age >55 (p = 0.029), a solid-cystic morphology (p = 0.001), and size >4 cm (p = 0.002) on univariate analysis; while solid-cystic morphology (p = 0.036) on multivariate analysis.

Conclusion

The commonest CPLs evaluated through EUS were SCN, followed by IPMN, MCN, and those associated with a specific disease. The chance of a cystic pancreatic lesion being malignant is higher if it has a solid-cystic morphology.

Introduction

Cystic pancreatic lesions (CPLs) constitute an important part of pancreatic diseases. The various CPLs include the non-neoplastic pancreatic cysts (true cysts, retention cysts, lymphoepithelial cysts), pancreatic cystic neoplasms [mucinous cystic neoplasm (MCN), intraductal papillary mucinous neoplasm, serous cystic neoplasms, solid pseudopapillary tumors], and cystic degeneration in solid pancreatic tumors (pancreatic ductal adenocarcinoma, acinar cell carcinoma, and endocrine neoplasms like neuroendocrine tumors/carcinomas).¹ The importance of these CPLs lies in their ability to become malignant in the future. The serous cystic lesions have a benign course, while mucinous lesions and intraductal papillary mucinous lesions predispose to future chances of malignant transformation. Thus, careful surveillance of premalignant cysts, either radiologic or endoscopic, is an important part of their management.²

The symptoms caused by these cysts can range from none or mild symptoms (vague upper abdominal pain) to severe symptoms (pancreatitis, obstructive jaundice, gastric outlet obstruction), depending upon their size and location within the pancreas, requiring endoscopic or surgical management to decompress the biliary and upper gastrointestinal tract systems.³ The treatment of the benign lesions is largely conservative unless they cause severe symptoms due to enlarging size, while the premalignant CPLs often require multidisciplinary meetings for surgical resection when the criterion for their removal is met.^4,5

Various studies have been performed to learn the clinical behavior of the various CPLs utilizing different imaging techniques like transabdominal ultrasound, computed axial tomography (CAT) scan, and magnetic resonance imaging (MRI) of the abdomen.^6,7 Several advancements have also been made in the field of endoscopic ultrasound (EUS), whereby newer techniques are being utilized to study the cyst structure, including confocal laser endomicroscopy (CLE) and contrast-enhanced EUS.⁸ Our study aimed to utilize EUS as an imaging modality in identifying the frequency of various types of CPLs in our part of the world and also to determine clinical parameters associated with the malignant nature of the cyst.

Materials and Methods

This observational study was performed in the department of gastroenterology of Sindh Institute of Urology and Transplantation, Karachi, from June 2018 to May 2025. All patients with CPL identified on abdominal imaging, such as ultrasound, CT scan, or MRI of the abdomen, and who were referred for evaluation with EUS were included. All patients who had extra-pancreatic fluid collections associated with acute or chronic pancreatitis were excluded. However, those patients with the development of an intra-pancreatic cystic lesion after pancreatitis were included. Permission was obtained from the institutional board and ethical committees of our unit. The study was performed in accordance with the Declaration of Helsinki, 2024. Informed written consent of the patient to participate in this study was obtained.

All relevant clinical findings in the history, general and systemic physical examination, blood investigations, and abdominal imaging were recorded in a preformed proforma. The patient was asked to report for the EUS after 8 hours of fasting on both out- and an in-patient basis. The EUS equipment used was any of the following: Pentax EG38-J10UT, Pentax EG-3870UTK, or Olympus TGF-UC180J. The morphology of patients with CPLs was recorded. Endoscopic ultrasound was performed to determine the type of cystic lesion based on the morphological characteristics of the cyst, including location, number, size, locularity, echogenicity, septation, and communication with the pancreatic duct. The size and contour of the pancreatic duct and the presence or absence of a mural or mucous nodule were noted. The type of the CPL was determined in accordance with the sonographic appearance and cystic fluid contents (Table 1).⁹

Table 1.

Determination of the type of CPL with EUS (based on the sonographic morphologic features and cystic fluid contents)

Clinical characteristics	PP	SCN	MCN	IPMN	SPT	Simple cyst
Sonographic features	Unilocular Variable size and wall thickness Echogenic material Features of acute/chronic pancreatitis	Microcystic – Multiple small cysts – Honeycomb appearance – Star-shaped central scar with calcification Macrocystic – Multiple large cysts	Macrocystic Multilocular Septations Nodularity or papillary projections	MD-IPMN – Dilated main PD – Mural nodule – Ductal communication BD-IPMN – One/more cysts communicating with the side branch of PD appearing as “bunch of grapes.”	Well circumscribed Heterogeneous Solid and cystic components Fibrous capsule Calcifications	Round, well-defined Anechoic without internal echoes or solid components
Communication with PD	Sometimes	No	No	Yes	No	No
Fluid amylase	High	Low	Variable	High	Low	Low
Fluid CEA	Low	Low	High	High	Variable	Low

IPMN, intraductal papillary mucinous neoplasm; MCN, mucinous cystic neoplasm; PP, pancreatic pseudocyst; SCN, serous cystic neoplasm; SPT, solid pseudopapillary tumor

Whenever indicated, the cystic pancreatic lesion was subjected to aspiration of the fluid with fine needle aspiration (FNA) equipment (Boston Scientific; 19G and 22G). The fluid was then sent to the laboratory to determine the levels of the various markers, including fluid amylase, fluid carcinoembryonic antigen (CEA), culture sensitivity, and cytology. If aspiration of the cystic fluid failed due to the viscous nature of the fluid, the procedure was abandoned. When evaluating a solid cum cystic pancreatic lesion with suspicion of development of malignancy, a biopsy was obtained using a fine needle biopsy (FNB) of size 22G. The adequacy of the tissue was determined with macroscopic on-site evaluation (MOSE) by the endosonographer. Procurement of at least two cores of tissue of lengths at least one inch each was considered satisfactory MOSE. The tissue was then sent for histopathologic and tumor marker analysis to rule out the nature and grade of malignancy. The gold standard for the diagnosis of the malignant nature of the cyst was evidence of malignancy on histopathologic analysis.

The data were analyzed using the Statistical Package for the Social Sciences (SPSS) version 24 (IBM Corp, Armonk, NY, USA). Values were expressed as means for continuous variables, depending on the distribution, or as percentages for categorical variables. Using histopathologic evidence as the gold standard for a diagnosis of malignancy in the cystic pancreatic lesion, univariate analysis was performed initially to identify the clinical and sonographic categorical parameters associated with the malignant nature of the cyst. To determine this association, the χ² test or Fisher's exact test was used as indicated. The identified factors were then utilized to perform multivariate regression analysis. A p-value of less than 0.05 was considered statistically significant.

Results

Of the total 56 patients included in the study, there were more females (34/56; 60.7%) than males (22/56: 39.3%). The mean age was 46.5 (±14.6) years. The majority of the CPLs were pure cystic lesions 41 (73.2%), while the remaining were solid cum cystic lesions 15 (26.8%) (Table 2). Among the various pure cystic lesions, the commonest types were serous cystadenoma (SCN) and intraductal papillary neoplasms (IPMNs), followed by MCNs; while among the solid cum cystic lesions, the most common type was malignant cystic neoplasm, followed by those associated with a systemic disease (Fig. 1). The gender and age distribution and the part of the pancreas affected among patients with CPL are shown in Table 3. Among the gender involvement, females were more commonly affected by SCN, IPMNs, and MCNs, while males were more commonly affected when CPLs were associated with any disease [e.g., von Hippel-Lindau disease (vHL), adult polycystic kidney disease (APKD)]. Among the age-groups affected, malignant lesions were more common among elderly people aged 55 years. The majority of the cystic lesions were located in the head 30 (53.6%), followed by the body 14 (25%), tail 5 (8.9%), and neck 2 (3.5%) of the pancreas. Five patients had multiple cysts that were diffusely distributed throughout the pancreas. The malignant cystic neoplasms were mostly located in the head region, followed by the body of the pancreas. Those CPLs associated with a systemic disease had a generalized distribution throughout the pancreas, while simple cysts were equally common in the tail or head.

Table 2.

Baseline characteristics of patients with CPL (n = 56)

Characteristics	Number (n)	Percentage (%)
Gender
Females	34	60.7
Males	22	39.3
Age
<55 yr	38	67.9
>55 yr	18	32.1
Morphology
Pure cystic	41	73.2
Solid cum cystic	15	26.8
Diagnosis
SCN	11	19.6
IPMN	10	17.9
MCN	9	16.1
Disease-associateda	7	12.5
Malignancy	7	12.5
Pancreatitisb	6	10.7
Simple cyst	5	8.9
SPT	1	1.8
Distribution
Head (predominantly)	30	53.6
Body (predominantly)	14	25.0
Tail	5	8.9
Neck	2	3.6
Diffuse/multiple	5	8.9

^aCysts associated with specific disease, i.e., von Hippel-Lindau disease, adult polycystic kidney disease; ^bIntra-pancreatic fluid collections associated with acute or chronic pancreatitis

Figs 1A and B.

Frequency of the subtypes of CPL among the pure cystic lesions (A) (n = 41) and solid cum cystic lesions (B) (n = 15). SCN, serous cystic neoplasm; IPMN, intraductal papillary mucinous neoplasm; MCN, mucinous cystic neoplasm; Pancr, pancreatitis-related intra-pancreatic fluid collection; DA, disease associated (e.g., von Hippel-Lindau disease, adult polycystic kidney disease); Malig, malignant

Table 3.

Gender and age distribution and pancreatic part affected among the various CPL (n = 56)

Distribution features	SCN (n = 11)	IPMN (n = 10)	MCN (n = 9)	DA (n = 7)	Malignant (n = 7)	Pancreatitis (n = 6)	Simple (n = 5)	SPT (n = 1)
Gender
Males (n = 22)	5	3	2	5	3	3	1	–
Females (n = 34)	6	7	7	2	4	3	4	1
Age
>55 (n = 18)	5	3	2	2	5	1	–	–
<55 (n = 38)	6	7	7	5	2	5	5	1
Part of the pancreas
Head (n = 30)	6	6	4	3	6	2	2	1
Body (n = 14)	5	2	4	–	1	2	–	–
Diffuse (n = 5)	–	1	–	4	–	–	–	–
Tail (n = 5)	–	1	–	–	–	1	3	–
Neck (n = 2)	–	–	1	–	–	1	–	–

The commonest CPLs were: SCN 11 (19.6%), IPMN 10 (17.9%), MCN 9 (16.1%), associated with a specific disease 7 (12.5%), malignancy 7 (12.5%), pancreatitis 6 (10.7%), simple cyst 5 (8.9%) and solid pseudopapillary tumor 1 (1.8%) (Table 2). Among CPLs that were associated with a systemic disease, four were associated with vHL disease, two with APKD, and one with multiple cysts of unknown cause. Among the seven patients with CPLs that were found to have malignancy on EUS-guided biopsy, three had pancreatic ductal adenocarcinoma with cystic degeneration, three had neuroendocrine cancer with cystic degeneration, and one had intraductal papillary mucinous neoplasm with malignancy-related histologic features.

Univariate analysis that was performed to determine factors associated with the malignant nature of the CPL showed that patient age >55 years (p = 0.029), a solid cum cystic cyst morphology (p = 0.001), and size of CPL > 4 cm (p = 0.002) showed statistically significant association. Singularity (vs multiplicity) of the cystic lesion showed a tendency toward association with malignancy, but was not statistically significant (p = 0.082). Multivariate analysis demonstrated that only solid cum cystic morphology (p = 0.036) was associated with the malignant nature of CPL (Table 4).

Table 4.

Factors associated with the malignant nature of the CPL (n = 56)

Clinical characteristics	Malignant	Non-malignant	Univariate analysis			Multivariate analysis
			OR	CI	p-value	p-value
Age (years)
>55	5	13	6.92	1.19–40.16	0.029	0.174
<55	2	36
Gender
Male	3	19	1.18	0.24–5.88	1.000
Female	4	30
Comorbid
Yes	2	19	0.63	0.11–3.59	0.700
No	5	30
Body mass index (kg/m2)
>25	2	21	0.53	0.09–3.02	0.688
<25	5	28
Location of CPL
Head	6	22	0.136	0.02–1.21	0.101
Other	1	27
Solid component
Solid cum cyst	6	9	0.038	0.04–0.35	0.001	0.036
Pure cyst	1	40
Size of CPL
>4cm	7	18	0.720	0.56–0.92	0.002	0.998
<4cm	0	31
Multiplicity of cysts
Multiple	0	19	0.811	0.69–0.95	0.082
Single	7	30

Discussion

In our study, the commonest CPL evaluated through EUS was SCN, followed by IPMN and MCN. Various studies have been done in this regard worldwide, whereby peri-pancreatic pseudocyst has been considered the commonest cystic pancreatic lesion.¹⁰ However, in our study, we only included intra-pancreatic cystic fluid collections due to pancreatitis and excluded all the extra-pancreatic ones; therefore, these lesions formed only a small part of our overall CPLs. We noticed that among those with pure cystic lesions, serous cystadenomas were the commonest variety of CPLs, followed by IPMNs and MCNs. This is in accordance with worldwide observation that SCNs form a total of one-third of all CPLs, followed by IPMNs and then MCNs, which constitute about half of those of IPMNs.¹¹ However, among the group consisting of solid cum cystic lesions, the commonest variety we observed was the malignant cysts: pancreatic ductal adenocarcinomas or the neuroendocrine cancers with cystic degeneration.

In our study, females were more commonly affected by SCNs, IPMNs, and MCNs, while males were affected by cysts related to systemic disease, like vHL or APKD. Also, we found that the majority of the CPLs were found to be located within the head of the pancreas, followed by those in the body. That the neoplastic cysts are more common among females and in the head region has also been shown by Falqueto et al.¹² from Brazil, who studied the demographic data of patients with CPLs. However, surprisingly, they found that the risk of malignancy was more common among males and ages above 60 years. Furthermore, in our study, the majority of the patients with CPLs that were associated with systemic disease were those with vHL syndrome and APKD, and were all young patients. This is probably because these are genetic diseases that tend to involve young individuals, and also because our gastroenterology department was an appendage of a larger kidney disease-related hospital, where most cases were referrals from urologists to rule out the development of pancreatic cancer.

In our study, the clinical parameter associated with malignant nature on multivariate analysis of the cyst was the presence of a solid cum cystic morphology. Also, it showed that most of the malignant cystic lesions were located in the head region of the pancreas. The fact that a pancreatic cyst having a solid component within it has a higher risk of developing or having malignancy has already been demonstrated by previous research work.⁷ Similarly, larger cyst sizes have been known to carry higher malignant potential compared to those less than 3 cm.¹³ Other studies show that a pancreatic cyst size >4 cm, thick irregular walls, peripheral calcifications, pancreatic duct dilatation, and surrounding lymphadenopathy are all associated with a high risk of malignancy within the cyst.^14,15 Besides, Pausawasdi and Scheiman¹⁶ from Thailand state that the clinical predictors for malignancy at surgery are male sex, age greater than 50 years, weight loss, and high cyst fluid CEA.¹⁶ In our study too, advanced age, i.e., >55 years, was found to be associated with the malignant nature of cyst on univariate analysis but not on multivariate analysis. This mutual difference in observation could be due to our small sample size.

Conclusion

The commonest CPL evaluated through EUS were serous cystic neoplasms, followed by intraductal papillary mucinous neoplasms and MCNs. Females were more commonly affected by the neoplastic cystic lesions, and the most common region affected was the head of the pancreas. Pure cysts were more commonly observed compared to solid cum cystic lesions. Those cystic lesions commonly found to be malignant on histopathology were those having large cystic degeneration of pancreatic ductal and neuroendocrine cancers, or malignant transformation of intraductal papillary mucinous neoplasm. The clinical features associated with malignancy among patients with CPL were essentially the presence of a large size and the development of a solid component within the cystic lesions.

Strengths and Limitations

The limitation of our study was the small sample size of 56 patients, and only seven of them had a malignant lesion. Another limitation of our study was that we evaluated only those CPL that were referred to us by various medical facilities that had encountered them on abdominal imaging and deemed their evaluation necessary. This may represent an under-estimation of the actual burden of the disease in the community. However, the strength of our study is that it has been performed in a resource-poor developing country where basic research work regarding CPL and their association with malignancy, in particular, is scarce.

Artificial Intelligence (AI) Disclosure

No AI is used in the study performed.

Authors’ Contributions

AAT: Conceptualized, did the procedure, collected the data, and wrote the first draft of the manuscript. NHL: Did the procedures, collected the data, and wrote the first draft. ZM: Collected the data and analysis and wrote the final draft. SA: Collected the data. MMS: Wrote the final draft. RTY: Wrote the final draft. SML: Did the procedure.

Orcid

Abbas A Tasneem https://orcid.org/0000-0002-9964-0977

Nasir H Luck https://orcid.org/0000-0002-4752-4157

Zain Majid https://orcid.org/0000-0002-6961-3011

Raja TY Khan https://orcid.org/0000-0002-5504-5084