Comparative analysis of the CT and MRI features of pancreatic colloid carcinoma

Abstract

Objective

To investigate the CT and MRI features of pancreatic colloid carcinoma (PCC) and to correlate imaging findings with histopathological characteristics.

Materials and methods

The CT and MRI images and clinical data of 15 patients with PCC confirmed by surgical resection and histopathological examination from August 2009 to October 2023 were retrospectively analyzed.

Results

All 15 PCC lesions presented as single masses, occurring in 8 males and 7 females (mean age 66 ± 8.4 years). The majority of lesions (73%) were located in the pancreatic head (11 cases), with 2 cases in the body, 1 in the tail, and 1 involving both body and tail. PCC typically manifested as cystic-solid masses with internal septations and calcification (8 cases, 53.3%). Associated findings included pancreatic ductal dilatation (13 cases), pancreatic atrophy (13 cases), and biliary dilatation (12 cases). On dynamic contrast-enhanced imaging, lesions demonstrated mild enhancement in the arterial phase and progressive sponge-like or reticular enhancement in the delayed phase. Compared with the surrounding parenchyma, the degree of enhancement in the arterial phase and portal venous phase was significantly lower (8.2 ± 6.7 vs. 19.3 ± 13.6, 9.67 ± 8.93 vs. 28.3 ± 17.2 HU, respectively; P < 0.05). The apparent diffusion coefficient (ADC) values showed no significant difference between PCC lesions and surrounding parenchyma (2.248 ± 0.454 × 10⁻³ mm²/s vs. 1.829 ± 0.514 × 10⁻³ mm²/s; P > 0.05). Pathologically, all tumors were characterized by neoplastic epithelium suspended in extracellular mucin pools comprising more than 80% of the tumor volume, meeting current WHO diagnostic criteria for PCC.

Conclusion

The combination of imaging features in PCC—including presentation as a cystic or cystic-solid mass with progressive delayed sponge-like or reticular enhancement, a salt-and-pepper appearance on T2-weighted images, and frequent calcification—may facilitate its differentiation from other pancreatic cystic neoplasms when considered together. Progressive delayed reticular enhancement reflects fibrous stroma segregating the mucin pools, as confirmed by histopathological correlation.

Introduction

Pancreatic colloid carcinoma (PCC), also known as mucinous non-cystic carcinoma, is a rare subtype of pancreatic ductal adenocarcinoma (PDAC). According to the current World Health Organization (WHO) Classification of Tumours of the Digestive System, PCC is histologically characterized by scant clusters or strands of neoplastic epithelial cells floating in pools of stromal extracellular mucin, with the mucinous component exceeding 80% of the tumor volume [1, 2]. PCC has a relatively low incidence rate of 1–3% but carries a more favorable prognosis than conventional PDAC. Previous studies have demonstrated that the 5-year survival rate of PCC patients is approximately 57%, significantly higher than that of PDAC at 12% [2, 3]. Therefore, accurate preoperative identification of PCC is crucial for clinical management and treatment planning. Given the paucity of studies on the radiological characteristics of PCC, we analyzed the CT and MRI findings of 15 PCC patients and correlated them with histopathological results, aiming to improve the accuracy of preoperative diagnosis and differential diagnosis.

Materials and methods

Basic clinical data

This retrospective study was approved by our institutional review board with a waiver of the requirement for written informed consent. From August 2009 to October 2023, clinical data, CT, and MRI findings of 15 patients with PCC confirmed by surgery and histopathological examination at Zhongshan Hospital, Fudan University, were analyzed. There were 8 males (53.3%) and 7 females (46.7%) with a mean age of 66 ± 8.4 years, and all cases presented as single lesions. Twelve patients underwent plain and contrast-enhanced CT examination, 10 patients underwent plain and contrast-enhanced MRI examination, and 7 patients underwent both CT and MRI examinations before surgery. Five cases had a history of pancreatitis: 2 cases presented with upper abdominal discomfort, 1 case with asthenia, poor appetite, and jaundice, and 2 cases with mild abdominal pain and diabetes mellitus. The remaining cases were either incidentally discovered or diagnosed during routine physical examinations. Laboratory tests revealed that tumor markers (carcinoembryonic antigen, alpha-fetoprotein, and carbohydrate antigen 19 − 9) were within normal ranges. Preoperative radiological diagnoses included cystic lesions with indeterminate or low-grade malignant potential (8 cases), serous cystadenoma (2 cases), mucinous cystadenocarcinoma (3 cases), solid pseudopapillary tumor (1 case), and pancreatic cancer (1 case). Endoscopic ultrasound-guided fine-needle aspiration (EUS-FNA) was not performed in this cohort, as all patients had imaging findings warranting surgical resection, including features concerning for malignancy, mass effect with biliary or pancreatic duct obstruction, or lesions with indeterminate malignant potential requiring definitive histological confirmation.

CT and MRI Examinations

CT examination

Siemens Sensation 16-slice spiral CT or Siemens Definition AS + 128-slice spiral CT was used. Technical scan parameters were as follows: tube voltage 120 kV, tube current 180 mA, slice thickness and spacing 5 mm, scanning range from the diaphragmatic dome to the level of the bilateral anterior superior iliac spine. Iohexol was injected through the antecubital vein at a flow rate of 2.5–3.0 mL/s with a dosage of 1.2 mL/kg. Arterial phase and portal venous phase images were acquired at 30 s and 70–80 s, respectively.

MRI examination

Imaging was performed on either 1.5 T (Siemens MAGNETOM Avanto, Erlangen, Germany; United Imaging Healthcare uMR 560, Shanghai, China) or 3.0 T MRI scanners (United Imaging Healthcare uMR 770, Shanghai, China; Siemens Verio, Erlangen, Germany). The routine MRI protocol included axial T2-weighted imaging (T2WI) with fat saturation (fs), axial in-phase and out-of-phase T1-weighted imaging (T1WI), unenhanced axial fs T1WI, and dynamic triple-phase contrast-enhanced MRI. Diffusion-weighted imaging (DWI) was performed using a breath-hold single-shot echo-planar imaging pulse sequence. For 1.5 T scanners, b values of 0 and 500 s/mm² were used, while for 3.0 T scanners, b values of 0 and 800 s/mm² were employed. The selection of b values was optimized for each field strength to achieve adequate image quality while minimizing susceptibility artifacts commonly encountered in pancreatic imaging.

For 1.5 T systems, the parameters were: T2WI (TR = 3500 ms, TE = 84 ms, flip angle = 140°, matrix = 194 × 256), T1WI [TR = 6.8 ms, TE = 2.35 ms (in-phase)/4.75 ms (opposed-phase), flip angle = 70°, matrix = 144 × 256], and DWI (SE-EPI sequence, b = 0 and 500 s/mm², TR = 2400 ms, TE = 66 ms, matrix = 112 × 128), all with slice thickness = 7.0 mm and spacing = 2.1 mm. For 3.0 T systems, the parameters were: T2WI fat-suppressed fast spin-echo sequence (TR = 4500–7000 ms, TE = 75–90 ms, slice thickness = 5–5.5 mm, gap = 1 mm, matrix = 320 × 320, field of view = 35 cm), T1WI gradient-echo in-phase (TR/TE = 2.6–4.1 ms/1.2–2.6 ms) and out-of-phase sequence (TR/TE = 2.7–4.1 ms/2.5–5.1 ms), and DWI sequence (TR/TE = 4300–7000 ms/50–80 ms, b values = 0 and 800 s/mm²).

Contrast-enhanced MRI was performed with gadopentetate dimeglumine (Magnevist^®, Bayer Schering Pharma, Berlin, Germany) via a power injector (Spectris Solaris^® EP MR, MEDRAD Inc., Indianola, PA, USA) at an infusion rate of 1.5–2 mL/s and a dosage of 0.1 mmol/kg. Following contrast injection, a three-dimensional fs T1W gradient-echo sequence (3D-VIBE-T1-FS, TR = 5.04 ms, TE = 2.31 ms, slice thickness 4 mm, no spacing, matrix 256 × 192, FOV 32 × 38 cm) was used to acquire images in the arterial phase (AP), portal venous phase (PVP), and delayed phase (DP) at 20–25 s, 60–80 s, and 150–180 s, respectively.

Image analysis

Images were independently analyzed by two experienced senior radiologists, with final interpretation reached by consensus. CT and MRI images were evaluated for lesion size, location, morphology, density or signal intensity, calcification, pancreatic duct dilatation, pancreatic atrophy, invasion of adjacent tissues, and lymph node or liver metastasis. With the surrounding normal pancreatic parenchyma serving as the reference, CT values of the lesion and surrounding parenchyma were measured on plain scan, arterial phase, and portal venous phase images. The degree of lesion enhancement during the arterial phase (EE_ap = enhanced CT value − plain scan CT value) and portal venous phase (EE_pp = enhanced CT value − plain scan CT value) was calculated, with calcification carefully avoided in the region of interest (ROI). MRI signal intensity was assessed relative to the surrounding pancreatic parenchyma and stratified as hypointense, isointense, or hyperintense. Apparent diffusion coefficient (ADC) values of the lesion and surrounding pancreatic parenchyma were measured.

Pathological analysis

All cases were confirmed by surgical resection and histopathological examination performed by board-certified pathologists at the Department of Pathology, Zhongshan Hospital, Fudan University. Pathological examination involved routine tissue sectioning, hematoxylin and eosin (HE) staining, and immunohistochemical analysis. The diagnosis of PCC was established according to the current WHO Classification of Tumours of the Digestive System (5th edition, 2019) [4], which requires that extracellular mucin comprise more than 80% of the tumor volume [1, 2]. Macroscopically, the cross-sections were soft and cystic, containing grayish-yellow gelatinous material. Microscopically, all tumors demonstrated the defining histological features of colloid carcinoma: clusters and strands of neoplastic epithelial cells floating within abundant extracellular mucin pools, with fibrous stroma separating the mucin into irregularly shaped pools. Thirteen cases were associated with intraductal papillary mucinous neoplasm (IPMN). Immunohistochemistry analysis demonstrated that MUC1 was negative while MUC2 was positive in all 15 cases.

Statistical analysis

SPSS version 19.0 software was used for statistical analysis. The Kolmogorov–Smirnov test was used to assess the normality of measurement data. Normally distributed data are expressed as mean ± standard deviation, while data with skewed distribution are expressed as median (interquartile range). Differences in CT density and ADC values between PCC lesions and surrounding normal pancreatic parenchyma were compared using paired t-tests, with P < 0.05 considered statistically significant.

Results

Tumor Size, Location, Morphology, and Boundaries

All 15 patients had single lesions ranging from 3.2 cm to 9.8 cm (largest dimensions: 7.1 × 7.9 cm), demonstrating an expansive growth pattern. Eleven lesions (73.3%) were located in the pancreatic head, 2 (13.3%) in the body, 1 (6.7%) in the tail, and 1 (6.7%) in both body and tail. Thirteen cases (86.7%) exhibited a quasi-circular or cystic-solid morphology with internal septations creating a multilocular cystic appearance with a soap-bubble pattern, while 2 cases (13.3%) were irregular in shape. Eleven cases (73.3%) had well-defined boundaries and 4 cases (26.7%) had indistinct boundaries with the surrounding pancreatic parenchyma. Eight cases (53.3%) demonstrated point-like, strip-like, or rod-like calcification (Figs. 1, 2, 3, 4 and 5; Table 1).

Fig. 1.

CT images of PCC in the pancreatic head. A Plain CT scan showing a cystic-solid mass with focal linear calcification (arrows). B, C Mild peripheral and septal progressive enhancement on arterial and portal venous phase images (arrows)

Fig. 2.

MRI of PCC (the same patient as Fig. 1). A Fat-saturated T2WI showing a lesion with mixed signal in the pancreatic head (white arrow) and pancreatic ductal dilatation (black arrowhead). B, C In-phase and out-of-phase T1WI showing low signal with no signal drop. D Hyperintense signal on DWI (b = 800 s/mm²). E Heterogeneous mild high signal on ADC map. F Mild enhancement in the arterial phase and progressive moderate enhancement in the portal venous phase

Fig. 3.

MRI of PCC in the pancreatic head. A Fat-saturated T2WI showing a well-defined mildly hyperintense lesion (white arrow, 5.1 × 5.4 cm). B, C In-phase and out-of-phase T1WI showing slightly low signal. D MRCP showing mild hyperintense signal with biliary and pancreatic ductal dilatation. E Coronal T2WI showing mild hyperintense signal with proximal common bile duct dilatation. F Microscopic appearance showing characteristic clusters of neoplastic epithelial cells suspended in abundant extracellular mucin pools, separated by fibrous stroma (HE, 100×)

Fig. 4.

MRI of PCC in the pancreatic body. A, B T1WI/T2WI showing a well-defined lesion (3.5 × 2.8 cm) with hypointense T1 and hyperintense T2 signal; tiny nodules visible on T2WI-fs (black arrowhead). C, D High signal on DWI (b = 800 s/mm²) without significantly decreased ADC values (white arrow). E–H Mild enhancement with rim enhancement (white arrow) and enhanced tiny papillary structures (white arrowhead) on axial and coronal images

Fig. 5.

MRI of PCC in the pancreatic head. A, B T1WI/T2WI showing an irregular cystic-solid lesion (3.8 × 6.7 cm) with internal septations (white arrowhead), heterogeneous hypointense T1 and hyperintense T2 signal (white arrow). C, D Rim enhancement (white arrow), septal enhancement (white arrowhead), and reticular enhancement (hollow white arrow) in the arterial phase and portal venous phase

Table 1.

Clinical and associated imaging data of the 15 PCC patients

Parameters	Number (n)	Percentage (%)
Lesion Size	3.2–9.8 (cm)	NA
Lesion Location
Head	11	73.3
Body	2	13.3
Tail	1	6.7
Body and tail	1	6.7
Morphology
Irregular	2	13.3
Quasi-circular/cystic-solid	13	86.7
Boundary
Distinct	11	73.3
Indistinct	4	26.7
Calcification
Present	8	53.3
Absent	7	46.7
Pancreatic Duct Dilatation
Present	13	86.7
Absent	2	13.3
Biliary Duct Dilatation
Present	12	80.0
Absent	3	20.0
Pancreatic Atrophy
Present	13	86.7
Absent	2	13.3
Adjacent Tissue Invasion
Present	1	6.7
Absent	14	93.3

CT and MRI characteristics

Twelve cases demonstrated cystic or solid low-density lesions on plain CT scan with mild progressive enhancement on contrast-enhanced scans (Fig. 1). Compared with the surrounding pancreatic parenchyma, the mean CT values of lesions were lower on plain CT scans (13.6 ± 6.0 vs. 32.1 ± 5.6 HU), arterial phase (21.8 ± 16.0 vs. 51.3 ± 16.0 HU), and portal venous phase (23.3 ± 14.0 vs. 60.4 ± 18.5 HU). The degree of enhancement was significantly lower during both the arterial phase (8.2 ± 6.7 vs. 19.3 ± 13.6 HU) and portal venous phase (9.67 ± 8.93 vs. 28.3 ± 17.2 HU), with statistically significant differences (P < 0.05) (Table 2).

Table 2.

Comparison of CT values between PCC lesions and surrounding pancreatic parenchyma (n = 12)

	Plain Scan (HU)	AP (HU)	PP (HU)	EE_ap (HU)	EE_pp (HU)
PCC	13.6 ± 6.0	21.8 ± 16.0	23.3 ± 14.0	8.2 ± 6.7	9.67 ± 8.93
Pancreatic Parenchyma	32.1 ± 5.6	51.3 ± 16.0	60.4 ± 18.5	19.3 ± 13.6	28.3 ± 17.2
T-value	7.718	5.911	5.539	2.518	3.329
P-value	0.000	0.000	0.000	0.020	0.003

AP arterial phase, PP portal venous phase, EE extent of enhancement

On MRI, 10 cases showed hypointense signal on T1WI and heterogeneously hyperintense signal on T2WI (Figs. 2 and 3, and 5), while 1 case was isointense on both T1WI and T2WI. All lesions demonstrated hyperintense signal on DWI (Figs. 2D and 4C) with generally mild-to-moderate high signal intensity on ADC maps (Figs. 2E and 4D). On magnetic resonance cholangiopancreatography (MRCP), lesions showed heterogeneously high signal intensity (Fig. 3D). In the arterial phase, the peripheral margins and septations of the lesions showed mild enhancement. Lesions demonstrated progressive delayed peripheral and internal sponge-like or reticular enhancement in the portal venous and delayed phases, while the cystic components showed no enhancement (Figs. 3, 4 and 5). In 2 cases (13.3%), moderate delayed enhancement was observed owing to a greater proportion of solid components. The ADC values of PCC lesions were not significantly different from those of the surrounding pancreatic parenchyma (2.248 ± 0.454 × 10⁻³ mm²/s vs. 1.829 ± 0.514 × 10⁻³ mm²/s; P = 0.069) (Tables 3 and 4).

Table 3.

MRI signal intensities and ADC values of PCC lesions (n = 10)

Signal Intensity	T2WI	T1WI	Arterial Phase	Portal Venous Phase	Delayed Phase
Hypointense	0	12	15	13	6
Isointense	2	3	0	2	9
Hyperintense	13	0	0	0	0

Table 4.

The comparison of the ADC values of the PCC lesion and surrounding pancreatic parenchyma (n = 10)

	ADC value (x10-3mm2/s)	T test	P value
PCC	2.248 ± 0.454	-1.935	0.069
Surrounding Pancreatic Parenchyma	1.829 ± 0.514	-1.935	0.069

The signal intensity between the most obvious enhanced part of the solid area of the PCC lesion and the surrounding pancreatic parenchyma

Associated findings

Thirteen cases (86.7%) demonstrated pancreatic ductal dilatation and parenchymal atrophy, and biliary dilatation was present in 12 cases (80.0%). Fourteen cases (93.3%) showed no adjacent tissue invasion (Fig. 3D and E), while only 1 case (6.7%) demonstrated duodenal and superior mesenteric vessel involvement with abdominal lymph node metastasis (Table 1).

Pathological features

All cases were confirmed by surgical resection and histopathological examination according to current WHO criteria [4]. Macroscopically, the cross-sections were soft and cystic, containing grayish-yellow gelatinous material. Microscopically, all tumors demonstrated the defining histological features of colloid carcinoma, with clusters and strands of neoplastic epithelial cells floating within abundant extracellular mucin pools, separated by fibrous stroma (Figs. 3F and 6).

Fig. 6.

Microscopic images of PCC showing clusters of neoplastic epithelial cells suspended in abundant extracellular mucin pools (A, B) (HE, 200×)

Discussion

The main clinical manifestations of PCC include upper abdominal pain and jaundice, with more than 50% of cases having a history of chronic pancreatitis [3]. In our cohort, 5 patients had a history of pancreatitis, while the remaining cases were diagnosed during routine examinations. The demographic profile showed a higher incidence in elderly individuals, with a mean age of 66 years. Our study included 8 males and 7 females (M/F ratio = 8/7); we speculate that this near-equal distribution may be related to the small sample size. The primary treatment for PCC is surgical resection.

Macroscopically, PCC is characterized by soft, gelatinous, grayish-white masses with relatively distinct boundaries, rich in mucin and containing mucinous epithelium. Most cases are associated with IPMN; in our study, 13 cases (86.7%) showed IPMN association. Both PCC and IPMN are typically rich in mucin 2 (MUC2) glycoprotein [5]. However, IPMN is characterized as a cystic tumor with mucin accumulated within the cysts, whereas PCC predominantly comprises solid nodules composed mainly of extracellular mucin. In PCC, fibrous stroma segregates the mucin into irregularly shaped and sized pools, with extracellular mucin comprising more than 80% of the tumor volume [1–4], infiltrating the pancreatic parenchyma. MUC2 was positive in all 15 cases in our study.

In our study, all PCC lesions presented as single masses, predominantly localized in the pancreatic head or uncinate process, consistent with findings reported by Fouladi et al. [6]. While individual imaging features such as cystic or cystic-solid morphology are not specific to PCC, the combination of characteristic features may aid in its differentiation from other pancreatic cystic neoplasms. Due to abundant extracellular mucin, the CT features of tumor lesions were soft, round, or quasi-circular cystic or cystic-solid masses with mild enhancement and sometimes indistinct boundaries. During dynamic contrast enhancement, the central portion of lesions showed minimally enhancing mucin pools, while progressive sponge-like or reticular enhancement was observed peripherally and internally [7]. This reticular enhancement pattern is consistent with the histopathological finding of fibrous stroma separating the mucin pools, as clearly demonstrated in Fig. 3F. The mucin nodules with relatively clear boundaries, characterized by stromal segregation of mucin, demonstrated multilocular vesicles, soap-bubble, or honeycomb patterns, showing hypointense or isointense signal on T1WI and markedly hyperintense signal on T2WI. Within lesions, the fibrous stroma segregating the mucin pools or solid soft tissue density components created a mesh-like low signal intensity pattern on T2WI.

The tumors demonstrated diffusely scattered small nodular T2 hyperintense foci of various sizes with intervening mesh-like linear hypointense septations and internal hypointense foci, manifesting as a “salt-and-pepper appearance.” We believe this salt-and-pepper pattern on T2WI reflects the histopathologic composition of the tumors: small nodular T2 hyperintense areas correspond to mucin pools, while the mesh-like internal hypointense signal and tiny hypointense foci reflect fibrous stroma and neoplastic cells [8, 9]. Calcification is common, predominantly characterized by point-like, strip-like, or rod-like morphologies, as reported by Ren et al. [10].

PCC demonstrated significantly high signal intensity on DWI sequences; however, ADC values were only slightly higher than those of the surrounding pancreatic parenchyma. Lesions also exhibited high signal intensity on MRCP, attributed to a T2-shine-through effect caused by the abundant mucinous components. No statistically significant difference in ADC values was found between PCC and pancreatic parenchyma. We hypothesize that the abundant gelatinous mucin may have obscured the small neoplastic cellular components, resulting in manually measured ADC values that do not fully reflect the true cellular nature of the lesion. On careful inspection, tiny papillary nodules demonstrate mild-to-moderate restricted diffusion on ADC maps. Following contrast administration, lesions demonstrated mild progressive enhancement or progressive peripheral and internal sponge-like reticular enhancement during the portal venous phase. Minimal enhancement was observed within the mucin pools. Enhancement patterns on CT and MRI were similar, though MRI provided superior delineation of enhancement characteristics within lesions, consistent with findings reported by Gao et al. [9]. This enhancement pattern is attributed to tumor tissue proliferation and fibrous stroma segregation within the tumor. In patients with pancreatic or biliary duct dilatation or pancreatic atrophy, these findings may be related to secondary chronic pancreatitis caused by the tumor [11].

PCC should be differentiated from the following tumors. First, mucinous cystadenoma or cystadenocarcinoma occurs almost exclusively in premenopausal women (> 90%), is predominantly localized in the pancreatic tail (> 95%), and demonstrates peripheral calcification and irregular wall thickening, presenting as thick-walled cystic tumors with wall thickness exceeding 2 mm [12–14]. The septations are thick, with marked enhancement of the cystic wall and septations on contrast-enhanced imaging. Compared with mucinous cystic neoplasms, the enhancement degree of colloid carcinoma is lower. Second, IPMN is caused by cells with malignant potential in the main pancreatic duct or branch ducts and can be classified as main duct type, branch duct type, or mixed type. Main duct and mixed type IPMN demonstrate higher malignant potential due to main pancreatic duct involvement, characterized by main pancreatic duct dilatation greater than 1 cm, cyst size of mural nodules 3 cm or larger, cystic wall thickening, main pancreatic duct stenosis with distal pancreatic atrophy, and enlarged adjacent lymph nodes [14–16]. Branch duct type IPMN connects to the pancreatic duct without main duct dilatation, demonstrating grape-like clustered or lobulated cystic morphology, with the connection between the pancreatic duct and cyst being a typical feature. Intestinal-type IPMN is prone to malignant progression into PCC, while the gastric subtype has more aggressive biological characteristics and may be associated with tubular adenocarcinoma [17]. High-grade dysplasia may be present in IPMN [18], but tumors do not demonstrate main pancreatic duct dilatation. Third, serous cystadenoma is a benign entity [19, 20] most common in elderly women, manifesting as single or multiple microcysts with possible calcification in the cyst wall or septations. Cyst fluid is relatively clear, with marked enhancement of cyst walls and septations but no enhancement of cyst fluid on CT [21, 22]. In colloid carcinoma, the cyst wall is irregular, and neoplastic cells float in mucin pools; enhancement occurs not only in cyst walls and septations but also heterogeneously within the cyst fluid contents on CT.

Our study has several limitations. First, PCC is a rare disease with low incidence; therefore, our sample size of 15 cases is limited, which may affect the generalizability of the results. Second, as a retrospective study, scan parameters and contrast injection rates could not be completely standardized, potentially affecting quantitative measurements. Third, inter-observer variability in determining imaging findings was not formally assessed due to the small cohort. Fourth, although all pathological diagnoses were independently confirmed by board-certified pathologists according to current WHO criteria [4], no pathologist was included as a co-author, which may limit the depth of pathological interpretation. Further prospective investigations with larger cohorts and multidisciplinary collaboration are warranted. Notwithstanding these limitations, we believe that the radiological findings and radiological–pathological correlation presented herein will prove useful for the differential diagnosis of PCC.

In conclusion, although PCC, IPMN, and ductal adenocarcinoma share similar demographics including gender, age, and tumor location, with higher frequency in elderly males and predominant localization in the pancreatic head [23], the combination of CT and MRI features of PCC may facilitate its identification. Key imaging features include presentation as an expansive cystic or cystic-solid mass with mesh-like solid tumor components, point-like and strip-like calcification, progressive sponge-like internal contrast enhancement, and a salt-and-pepper appearance on T2WI, accompanied by pancreatic atrophy and pancreatic duct dilatation. Progressive delayed reticular enhancement corresponds to fibrous stroma segregating the mucin pools, as confirmed by histopathological correlation.

Authors’ contributions

Mingliang Wang submitted ethics approval. Mingliang Wang and Qingyang Wu collected the data and drafted the original manuscript. Shasha He and Mengsu Zeng served as supervising investigators. Qingyang Wu, Qihua Yin, and Hongxia Zhang prepared the original elements of the protocol and supervised data collection. All authors have read and agreed to the published version of the manuscript.

Funding

None.

Data availability

Further details and additional raw data are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

This study was reviewed and approved by Zhongshan Hospital, Fudan University, and all patients involved signed the informed consent form.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Competing interests

The authors declare no competing interests.