Comparison of CT and MRI findings of different types of pancreatic Serous cystic neoplasm: combined with histopathology

doi:10.21203/rs.3.rs-4846988/v1

Download PDF

Research Article

Comparison of CT and MRI findings of different types of pancreatic Serous cystic neoplasm: combined with histopathology

https://doi.org/10.21203/rs.3.rs-4846988/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objectives

A comparative analysis of the imaging findings across different types Serous Cystic Neoplasm（SCN）was conducted, the objective is to delineate distinct image characteristics, thus enhancing diagnostic accuracy.

Methods

The retrospective analysis involved 65 patients diagnosed with Serous Cystic Neoplasm (SCN) confirmed by pathology. Patients were categorised into macrocystic, microcystic, mixed, and solid types based on morphology. A comparative analysis of the imaging findings across different SCN types was conducted, elucidating their respective imaging characteristics. Additionally, microscopic pathological observations were juxtaposed to discern disparities among the various SCN types.

Results

Age, lesion location, clinical symptoms, and underlying diseases did not exhibit significant differences among the various types of Serous Cystic Neoplasm (SCN) (P > 0.05). However, the volume of lesions in the mixed type was notably larger than that in the microcystic type (P < 0.05). Additionally, the CT values of the mixed and microcystic types were higher than those of the macrocystic type in precontrast scans (P < 0.05). The enhancement degree of the macrocystic type significantly differed from that of the mixed and microcystic types (P < 0.05). Notably, central fibrous scars were evident in mixed and microcystic SCN interiors, contrasting with the absence of such scars in macrocystic types (P < 0.05). While the proportion of calcification and pancreatic duct dilatation was higher in mixed and microcystic types. Electron microscopic revealed that macrocystic types featured large cyst cavities with thin fiber separation and few blood vessels, whereas microcystic types exhibited small cavities with a high number of fibers and visible scar tissue. Mixed types represented a combination of macrocystic and microcystic features. Solid types displayed very small cavities, high proportions of fibrous stroma, and rich blood vessels.

Conclusion

Differences in CT and MR findings among different types of SCN suggest a correlation with their pathological basis.

Pancreatic neoplasms

Serous cystic neoplasm

Tomography

X-ray computed

Magnetic resonance imaging

Pancreatic Serous Cystic Neoplasm (SCN) is a relatively uncommon benign tumor, often incidentally detected due to advancements in imaging technology[1]. While various guidelines offer different management approaches for SCN[2–4], asymptomatic cases typically do not require surgical intervention. A multicenter study involving 2262 patients reported a low mortality rate of 0.1% associated with SCN, contrasting with a surgical mortality rate of 0.6%[5]. Despite its benign nature, SCN may undergo unnecessary surgical procedures due to misdiagnosis as other tumor types[6]. Therefore, achieving an accurate diagnosis of SCN is crucial, although the diversity of imaging presentations often leads to a high misdiagnosis rate[1].

SCN can be categorized into four morphological types[7]: macrocystic, microcystic, mixed, and solid. Microcystic SCN comprises multiple small cysts, each < 1cm in size, separated by thin fibrous septa, sometimes resembling a honeycomb pattern. Macrocystic SCN consists of one or more cyst lumens ≥ 1cm in size. Mixed type SCN represents a combination of macrocystic and microcystic features, while solid type SCN lacks visible cystic cavities on cross-sectional imaging. The distinct imaging features of different SCN types, ranging from simple cystic to completely solid, pose challenges for radiologists in diagnosis[8, 9]. Currently, there is limited attention given to these radiological differences, and no comparative studies have been conducted.

In this retrospective analysis, we examined imaging findings and pathological data from 65 surgically diagnosed SCN cases. Focusing on the characteristic features identified in previous studies, we compared imaging disparities among various SCN types and analyzed their pathological characteristics. Histopathological examination was utilized to elucidate the underlying reasons for imaging differences observed in different SCN types, thereby enhancing our understanding of the disease. A comprehensive understanding of imaging distinctions among different SCN types is crucial for improving diagnostic accuracy, aiding clinical decision-making, and preventing unnecessary surgical interventions[9].

This study obtained approval from the Ethics Committee of the XXX Hospital, with the informed consent requirement waived. We conducted a retrospective analysis of patients diagnosed with Serous Cystic Neoplasm (SCN) between March 2017 and July 2023. Inclusion criteria involved histological confirmation of SCN through either surgical or endoscopic biopsy, completion of MR or CT plain scan and dynamic enhanced scan within 30 days before surgery at our hospital, clear visibility of the lesion in the images, and no prior treatment before examination. A total of 65 patients were enrolled, with 53 undergoing CT and 48 undergoing MR imaging. Among them, 38 cases were pathologically confirmed after surgery, while 27 cases were confirmed through endoscopic fine needle puncture biopsy. We collected demographic data (age, sex), clinical information (symptoms, tumor markers), imaging results (CT and MR images), and pathological data for comparative analysis.

Image inspection parameter setting and data acquisition

All 53 CT patients underwent upper abdominal precontrast CT scan and double-phase enhanced pancreas scan prior to surgery, following a fasting period of 4–6 hours. The scanning was conducted using various models including the 256-layer Brilliance iCT scanner (Philips, Eindhoven, Netherlands), 64-layer Discovery CT750CT scanner (GE Healthcare, Milwaukee, Wisconsin, USA), and 64-layer LightSpeedVCT scanner (GE Healthcare, Milwaukee, Wisconsin, USA). The scanning range extended from the top of the diaphragm to the lower edge of both kidneys. Automatic tube current modulation technology was utilized with a tube voltage of 120kV, and the reconstructed layer thickness and layer distance were set at 5mm. A nonionic contrast agent, iodihexyl (350mgI/ml, OSU, Yangzijiang Pharmaceutical Group, China), was administered through the median elbow vein at a rate of 3ml/s using a high-pressure syringe (Ulrich, Ulm, Germany), with a total contrast agent volume of 1.5ml/kg and normal saline flush of 20ml. The enhanced arterial phase was delayed for 35s, and the portal phase was delayed for 65s.

All 48 patients underwent MR plain scan and MRCP scan, with 28 of them also undergoing MR enhanced scan. MRI was conducted using the Signa Explorer 1.5T (GE Healthcare, Milwaukee, Wisconsin, USA) superconducting MRI scanner with body phased array surface coils for 31 cases and the Signa Architect 3.0T (GE Healthcare, Milwaukee, Wisconsin, USA) magnetic resonance machine with body phase-controlled surface coils for 17 cases. The scanning sequences included conventional axial and coronal sequences: T1 weighted image (T1 WI) fat suppression (TR 350 ms, TE 2.7 ms) and T2 weighted image (T2 WI) fat suppression (TR 2182ms, TE 107ms), covering from the upper margin of the liver to the lower margin of the kidney. Dynamic enhanced scanning utilized the LAVA sequence (layer thickness 4 mm, TR 2.9 ms, TE 1.3 ms). A high-pressure syringe was used for contrast injection at a flow rate of 2.0 mL/s, with 20 mL of Gd-DTPA (Bayer Medical, Leverkusen, Germany) injected at a dose of 0.1 mmol/kg, followed by approximately 10 mL of normal saline. MRCP employed respiratory trigger with TR 2521 ms, TE 760 ms, and a matrix of 300×225.

Image analysis and measurement

CT and MR images of all patients were imported into the PACS system for evaluation. Two radiologists, with 10 and 8 years of imaging diagnosis experience, respectively, completed the measurement evaluation independently and blindly to the clinical and pathological results. Quantitative parameters were measured independently by both radiologists. In case of any inconsistency in the qualitative assessment, the senior radiologist participated in the evaluation. Subsequently, the quantitative parameters measured by both physicians were tested for consistency.

Qualitative parameters assessed from CT images included tumor location, shape, boundary, intracranial calcification, fibrous central scar, pancreatic duct dilation, relationship between lesions and peripancreatic blood vessels (including no relationship, extrusion, blood supply, invasion), peripancreatic lymph node enlargement (defined as peripancreatic lymph node diameter ≥ 1.0cm), liver metastasis, and distant metastasis. Additionally, the density of the lesion in precontrast scan, arterial stage, and portal pulse stage were evaluated using pancreatic parenchyma as reference.

Quantitative CT parameters included lesion size, precontrast scan CT value of lesion and surrounding normal pancreatic parenchyma (CTun), enhanced arterial phase CT value (CTap), and portal phase CT value (CTpp). The largest dimension of the tumor was selected for measurement of lesion size. Lesion length and diameter were measured three times, and the average value was taken. For CTap measurement, the ROI was placed in the solid area with significant tumor enhancement, covering approximately 1/2 to 2/3 of its area. Each ROI was measured three times and then averaged. Absolute enhancement values of arterial stage were calculated as ΔCTap = CTap - CTun, and absolute enhancement values of portal vein stage were calculated as ΔCTpp = CTpp - CTun. Arterial phase relative enhancement index (API) was calculated as lesion enhanced value in arterial phase divided by peripheral pancreas enhanced value in arterial phase; portal vein phase relative enhancement index (PPI) was calculated similarly. The degree of enhancement was categorized as no enhancement (< 10HU), mild enhancement (10-30HU), moderate enhancement (30-50HU), and obvious enhancement (> 50HU).

MR images were used to observe and measure lesion size, location, shape, boundary, T1WI and T2WI signal intensity (relative to hepatic parenchyma), fibrous central scar, pancreatic duct dilation, and the relationship between the lesion and pancreatic duct observed via MRCP. Additionally, the relationship between the lesion and surrounding structures and the presence of metastasis were evaluated.

Pathological examination method

The tumor tissues underwent fixation using a 4% neutral formaldehyde solution followed by paraffin embedding, with a section thickness of 4µm, and staining with hematoxylin and eosin (HE). The optical microscope was used to observe the morphology and cytoplasmic staining of tumor cells. The tumors were grouped based on different imaging classifications and compared with the imaging findings.

Statistical method

SPSS23.0(IBM Corporation, Chicago, IL, USA)software was used for data statistics.Qualitative variables are expressed in frequency (%).The measurement data were tested for normality and homogeneity of variance, and the homogeneity of variance was expressed as mean ± standard deviation in accordance with normal distribution. ANOVA was used for multi-group comparison in accordance with normal distribution.The non-normal distribution data is expressed as the median (quartile) and analyzed using non-parametric tests.Qualitative data were compared using χ2 or Fisher's test.P < 0.05 was considered statistically significant.The intra-group correlation coefficient (ICC) was used to assess the consistency of quantitative data measured by 2 physicians.ICC > 0.9 indicates good consistency.

1. General data of different types of SCN

A total of 65 patients diagnosed with SCN were analysed, comprising 18 cases of macrocystic type, 17 cases of mixed type, 29 cases of microcystic type, and 1 case of solid type. The patients' general characteristics are summarised in Table 1. There were no statistically significant differences observed in age, lesion location, clinical symptoms, or underlying diseases among the different types of SCN (P > 0.05). However, the lesion volume of the mixed type was found to be larger than that of the microcystic type (P < 0.05), whereas no significant difference was noted between the macrocystic type and mixed and microcystic types (P > 0.05).

2. The consistency of CT quantitative parameter measurement.

The consistency of CT quantitative parameters measured by the two physicians was evaluated. The intraclass correlation coefficient (ICC) values for plain precontrast CT value, arterial phase absolute enhancement value, venous phase absolute enhancement value, arterial phase enhancement index, and venous phase enhancement index were 0.979, 0.966, 0.975, 0.982, and 0.967, respectively, indicating good agreement between the two surveyors.

3. CT findings and comparison of different types of SCN

A total of 53 patients with SCN underwent CT examination, and the findings and comparative analysis of different types of SCN are presented in Table 2. CT values of the mixed type and microcystic type were significantly higher than those of the macrocystic type on precontrast scans (P < 0.05). While there were no significant differences in arterial phase absolute enhancement value, portal venous phase absolute enhancement value, arterial phase enhancement index, and portal venous phase enhancement index between the mixed type and microcystic type (P > 0.05). Notably, the macrocystic type demonstrated no evident enhancement in both arterial and portal venous phases, which differed significantly from the mixed type and microcystic type (P < 0.05). Furthermore, the mixed type and microcystic type exhibited varying degrees of enhancement, with no significant difference in the proportion of different enhancement degrees between the two types (P > 0.05).

Most CT scans of SCN depicted a lobulated external contour (66%), with the mixed type showing the highest proportion of this contour type (81.8%). Additionally, the macrocystic type demonstrated a well-defined cystic mass on CT, while a small portion of the margin was unclear in the mixed and microcystic types (P < 0.05). Moreover, fibrous central scars were prevalent in the mixed type and microcystic type, contrasting with the absence of stellate fibrous scars in the macrocystic type (P < 0.05). Although the proportion of calcification in the mixed type and microcystic type was higher than that in the macrocystic type, the difference was not statistically significant (P > 0.05). Furthermore, vascular displacement was predominant in most macrocystic types, while the mixed type and microcystic type commonly exhibited thickened blood vessels (P < 0.05).

The macrocystic type typically consisted of single or multiple large cysts (Fig. 1, a-c), with 70.6% exhibiting a lobulated contour. Precontrast CT scans revealed watery density (Fig. 1, a), with clear margins and no significant enhancement in subsequent enhancement stages (Fig. 1, b-c). Additionally, no communication with the pancreatic duct was observed, although squeezed blood vessels around the lesion were sometimes visible (Fig. 1, c).

The mixed type comprised multiple cystic cavities of varying sizes (Fig. 2, a-b), 81.8% of which displayed a lobulated contour (Fig. 2, a-b). Precontrast CT scans depicted either liquid or solid soft tissue density (Fig. 2, a-b), with significantly lower enhancement observed in the low-density liquid area compared to the parenchymal area (Fig. 2, b). Arterial enhancement was notable in the peripheral parenchymal area, accompanied by delayed enhancement of scar tissue and calcification within the scar (Fig. 2, a).

The microcystic type comprised multiple microcystic cavities (Fig. 3, Fig. 4), with 58.6% exhibiting a lobulated contour. Precontrast CT scans displayed densities between fluid and pancreatic parenchyma density (Fig. 3, a, Fig. 4, a), with varying degrees of enhancement observed during enhanced examination (Fig. 3, b-c, Fig. 4, b-d). Arterial stage enhancement was most significant in the peripheral parenchymal area, diminishing in the portal vein area (Fig. 4, b-c), with delayed enhancement noted in the lesion's scar (Fig. 4, d).

The solid type was characterised by a relatively small tumour volume, with precontrast CT scans showing slightly lower density (Fig. 5, a) and a circular shape (Fig. 5, b). Notably, significant enhancement was observed in the arterial stage (Fig. 5, b), diminishing in the portal vein stage (Fig. 5, c), which sometimes caused pancreatic duct.

4. MRI findings and comparison of different types of SCN

A complete MRI examination was conducted on 48 patients with SCN, and the MRI findings and analysis of various SCN types are outlined below (Table 3).

All 48 patients exhibited a low signal on T1-weighted images (T1WI) and a high signal on T2-weighted images (T2WI), with distinct boundaries and no evidence of communication with the pancreatic duct. The majority displayed a lobular contour. There were no statistically significant differences observed among the macrocystic, mixed, and microcystic types (P > 0.05). However, a higher incidence of stellate fibre scars was noted in the mixed and microcystic types compared to the absence of scars in the large cystic type, and this disparity was statistically significant (P < 0.05). Nevertheless, there were no statistically significant differences between the mixed type and microcystic type (P > 0.05).

The macrocystic type typically consists of either a single or several large cysts (Fig. 1, d, e), exhibiting a low signal on T1WI and a significantly high signal on T2WI (Fig. 1, d, e). Approximately 77.8% of cases exhibited a lobulated contour with clear edges, thin internal separation, and no communication with the pancreatic duct, and there was no obvious enhancement.

In the mixed type, multiple lumens of varying sizes are observed (Fig. 2, C-E). T2WI demonstrates a high signal (Fig. 2, c), with visible internal low-signal fibre compartments or scars (Fig. 2, c). Around 93.7% of cases present a lobular contour (Fig. 2, d), with clear edges and no communication with the pancreatic duct. Significant enhancement is observed in areas with smaller cystic cavities, with delayed enhancement in scar tissue (Figure 2, e).

The microcystic type comprises multiple microcystic lumens (Fig. 3, D-E, and 4E-G). T2WI reveals a high signal (Fig. 3, d, and 4e), with internal low-signal fibre compartments or scars visible (Fig. 4E-G). Approximately 86.4% of cases display a lobed outline with clear edges and no communication with the pancreatic duct.

The solid type exhibits a high signal on T2WI (Fig. 5, d) and a rounded appearance, with a clear edge. It demonstrates significant enhancement in the arterial stage (Fig. 5, e), which decreases in the portal vein stage (Fig. 5, f). This enhancement may compress the pancreatic duct, resulting in duct dilation (Fig. 5, g).

5. Microscopic pathology of different types of SCN

Under the microscope, sections stained with hematoxylin and eosin (HE) revealed that all tumours exhibited cystic structures with a pale red stain. Pathologically, all cases exhibited cystic structures with a single layer of cubic epithelial cells covering the inner wall. Different types of SCN displayed variations in cystic structures concerning size, number, arrangement, thickness of fibrous separation, presence of fibre scar, proportion of fibre interstitial, and abundance of small blood vessels.

Macrocystic type

Due to the large cyst cavity, only the capsule wall (Fig. 1, f) or fibrous septum is visible under the microscope, with a thin capsule wall or septum. The capsule wall is lined with a single layer of cubic epithelial cells and thin fibrous collagen beneath it. The capsule lining favours cubic epithelium over flat epithelium. The nuclei are relatively large, cytoplasm transparent, and small cysts are occasionally visible on the wall of a single cyst under the microscope.

Mixed type

The size of the capsule varies (Fig. 2, f), and microscopic features vary greatly in different areas, including large and microcystic manifestations.

Microcystic type

Under the microscope, the incised surface appears spongy and consists of numerous small sacs (Fig. 3, f), with diameters mostly ranging from a few hundred micrometres to a few millimetres. The sac walls are lined with a single layer of flat or cubic epithelial cells, with bright red cytoplasm, centrally located nuclei, round to oval in shape, consistent in size, with inconspicuous nucleoli, lacking nuclear division, and non-atypical cell shapes. The tumour's fibrous tissue is relatively abundant, with common fibrous scars (Fig. 4, h-i). Tiny cystic cavities and small blood vessel hyperplasia are visible only under the microscope in the fibrous septum. The smaller the cystic cavity, the more abundant the fibrous interstitium and small blood vessels. Fibrous scars consist of relatively dense fibrous interstitium, where hyalinoid tissue is evident, and small clusters of sacs are observed within and around fibrous scars. Unlike true scar tissue, these fibrous collagens are rich in small vascular structures.

Solid type

Under an electron microscope, very small cavities are visible (Fig. 5, h-i), often ranging from tens to hundreds of micrometres, lined with typical cubic epithelial cells. These cells are closely arranged into nests, sheets, and trabeculae, separated by thick fibrous bands, rich in interstitial components and small blood vessels. These vessels are mature, and the vascular wall structure is intact.

6. CT and MRI misdiagnosis of different types of SCN

Among the 53 patients with SCN who underwent CT examination, 22 (41.5%) were misdiagnosed. Macrocystic type is prone to being misdiagnosed as mucinous cystic neoplasm (MCN) and other cystic lesions. Mixed type can be mistaken for neuroendocrine neoplasm (PNET). It is crucial to differentiate microcystic type from various pancreatic cystic solid tumours. Solid type is often misdiagnosed as a PNET. The misdiagnosis rates for different types of SCN on CT are detailed in Table S1.

Of the 48 patients who underwent MR examination, 15 (31.3%) were misdiagnosed. Macrocystic type is commonly misidentified as MCN, whereas mixed, microcystic, and solid types are frequently misdiagnosed as PNETs. The misdiagnosis rates for different types of SCN on MR imaging are outlined in S2.

Pancreatic SCN is a rare exocrine gland tumour originating from the central cell of the pancreatic acinar, accounting for approximately 1%-2% of all pancreatic exocrine tumours[7]. The majority of SCN lesions are benign, with only rare cases progressing to serous cystadenocarcinoma[5]. Non-surgical observation of imaging characteristics in typical and elderly frail patients can facilitate accurate diagnosis, which holds significant value[10]. The considerable differences in imaging among various SCN types contribute substantially to misdiagnosis[11, 12]. Therefore, a comprehensive comparison and analysis of imaging findings combined with pathological insights are essential to delineate their imaging characteristics and deepen understanding.

SCN typically manifests in middle-aged and elderly women, with an average age of around 60 years[13]. Most patients are asymptomatic, although some may experience abdominal discomfort or pain[7], consistent with our study findings. Our results revealed no significant differences in age, location, or clinical symptoms among different types of SCN patients. Notably, the mixed type was observed to be larger than the microcystic type upon initial detection, possibly due to variations in growth rates and detection rates, necessitating further investigation with larger samples.

We observed significant variations in CT precontrast scan density, enhancement degree, and enhancement mode among SCN types. Precontrast scan density correlated with cystic cavity size and fibrous interstitial proportion. Macrocystic type exhibited low density, akin to water sample density, owing to its large sac cavity and thin fibre separation. Conversely, other SCN types, with relatively smaller lumens and more abundant fibrous compartments, demonstrated densities resembling soft tissue or mixed cystic solid lesions. However, SCN densities typically remained lower than pancreatic parenchyma due to luminal fluid content[14].

Combining imaging with pathology, we noted that tumour enhancement degree correlates with stromal component proportion, with abundant tumour stromal blood vessels serving as the basis for enhancement. Given the varied stromal component proportions across SCN types, enhancement degrees varied accordingly. Solid type showed the highest stromal component proportion and the most pronounced enhancement, while macrocystic type exhibited the lowest. Microcystic and mixed types typically displayed moderate to significant enhancement, reflecting their rich tumour stromal vasculature[15]. Moreover, solid type featured mature small vessels, explaining its conspicuous enhancement, with intact vascular endothelial cells minimising bleeding incidence[16].

Differences in lumen size, fibrous interstitial ratio, and vascular richness were evident across SCN types, with variability observed even within the same type. Consequently, mixed and microcystic types may exhibit mild, moderate, or significant enhancement, limiting SCN diagnosis solely based on enhancement degree. Moreover, different tumour parenchymal regions displayed varying enhancement patterns, such as delayed enhancement or wash-in and wash-out patterns[15], reflecting differing vascular densities. Identifying fibre scar presence within lesions holds diagnostic significance[17], with dense fibre tissue showing delayed enhancement on endoscopy.

Our study highlighted the prevalence and characteristics of fibrous scars in microcystic and mixed types, contrasting with macrocystic type's typical absence of fibrous scars. Notably, fibrous central scar exhibited low T2WI signal intensity, distinguishing it from SCN with high signal intensity fluid in cystic cavities. Therefore, MRI surpasses CT in revealing mass contour, fibre separation, and central scars[18]. Intralesional calcification, although a specific SCN sign, remains rare, with CT superior to MRI in its detection[1]. Thus, combined CT and MR imaging may enhance SCN diagnostic accuracy.

Peripheral blood vessel relationships varied among SCN types, with macrocystic type typically lacking significant peripheral blood vessels. Conversely, mixed and microcystic types exhibited notable peripheral blood vessel abundance, correlating with lesion enhancement. "Peripheral vascular signs," more prevalent in mixed and microcystic types[19], likely result from abundant peripheral blood vessels and benign tumour characteristics.

While SCN may present cystic or solid features with varying enhancement degrees, common imaging features include cystoid structure, lobulated contour, clear edges, high T2WI signal, and no communication with pancreatic duct. Most SCN comprise multiple cystic lumens, with a lobulated contour prevalent across all types. Lobulated contour may appear as the "extracapsular cyst sign," aiding in SCN diagnosis[20]. SCN typically do not communicate with the pancreatic duct, although large or proximal lesions may compress it[21]. Tumour lumens contain clear watery fluid, with high T2WI signal dominance despite potential signal reduction due to increased fibre content. SCN[17], as benign tumours, exert pressure on surrounding tissues without invasion or metastasis[5].

Our analysis of misdiagnosis revealed SCN's propensity for misidentification as various pancreatic tumours, commonly mucinous cystic neoplasm (MCN) and pancreatic endocrine neoplasm (PENT). Other misdiagnoses included solid pseudopapillary neoplasm (SPN) and intraductal papillary mucinous neoplasm (IPMN). SCN, particularly those with single or few cystic characteristics, require differentiation from MCN, which typically presents in female patients without pancreatic head involvement, featuring thick cyst walls and septa, potential cyst wall calcification, and intracystic papillary prominences or nodules[22, 23].

SCN with abundant blood supply often resemble PENT, particularly when lesions are small and CT depicts primarily solid findings, resulting in high misdiagnosis rates[16]. CT values of SCN on unenhanced scans are notably lower than those of PENT, with richly vascularised SCN displaying higher relative washout rates[24]. Various imaging features, including tumour size, T2-weighted signal intensity, apparent diffusion coefficient values, and fibrous envelope presence, suggest solid type SCN[25]. Notably, MRI offers advantages in revealing central scars and fibre separation within lesions. Clinical history of endocrine disease, liver metastasis, and lymph node enlargement further aids PENT-SCN differentiation[9].

SPN, prevalent in young women, typically presents as a mixed cystic mass with complex internal components, complete envelope, peripheral calcifications, "floating cloud sign," and progressive enhancement on enhanced scans, distinguishing it from SCN[10]. IPMN, more common in middle-aged to elderly men, typically arises in the pancreatic head, featuring club-like cystic shapes, dilated main or branch pancreatic ducts, and strip-like compartments with papillary wall nodules showing enhancement, alongside cystic cavity-main pancreatic duct communication[9].

Despite certain imaging differentials, SCN with atypical features pose diagnostic challenges. Radiomics and deep learning hold promise for improving SCN differential diagnosis[26–28]. Recent studies demonstrate the comparable performance of radiomic-based approaches to experienced radiologists[29], albeit requiring further validation. Endoscopic ultrasonography offers diagnostic value via cyst fluid aspiration and fine needle puncture biopsy[30], although its limited availability restricts widespread use. CT and MRI remain the primary diagnostic modalities for SCN. Thus, comprehending pancreatic cystic tumour differentials and understanding SCN types enhances diagnostic accuracy, averting unnecessary surgeries.

This study has limitations, including the relatively low SCN incidence and small sample size despite incorporating data from two research centres. Additionally, retrospective study design and incomplete CT and MR examinations in some patients, particularly lacking delayed scanning, may have impacted fibrous scar imaging. Moreover, utilising different CT and MR scanners might have introduced minor judgement biases.

In conclusion, SCN types exhibit distinct imaging findings attributed to variations in cystic cavity size, number and arrangement, fibrous interstitial proportion, and vascular abundance. Nonetheless, common imaging characteristics among SCN types exist. A comprehensive understanding of these features and key differential points aids in accurate SCN.

Author Contribution

Guangmang Li and Aichun Lei carried out the studies, participated in collecting data, and drafted the manuscript. Peng Li and Bo Li performed the statistical analysis and participated in its design. Peng Wang, and Guanghai Ji participated in acquisition, analysis, or interpretation of data and draft the manuscript. All authors read and approved the final manuscript.

Data Availability

All data generated or analysed during this study are included in this published article.

Ning K, Salamone A, Manos L, Lafaro KJ, Afghani E (2023) Serous Cystadenoma: A Review on Diagnosis and Management. J Clin Med 12 https://doi.org/10.3390/jcm12237306
Megibow AJ, Baker ME, Morgan DE, Kamel IR, Sahani DV, Newman E, Brugge WR, Berland LL, Pandharipande PV (2017) Management of Incidental Pancreatic Cysts: A White Paper of the ACR Incidental Findings Committee. J Am Coll Radiol 14:911–23. https://doi.org/10.1016/j.jacr.2017.03.010
(2018) European evidence-based guidelines on pancreatic cystic neoplasms. Gut 67:789–804. https://doi.org/10.1136/gutjnl-2018-316027
Elta GH, Enestvedt BK, Sauer BG, Lennon AM (2018) ACG Clinical Guideline: Diagnosis and Management of Pancreatic Cysts. Am J Gastroenterol 113:464–79. https://doi.org/10.1038/ajg.2018.14
Jais B, Rebours V, Malleo G et al (2016) Serous cystic neoplasm of the pancreas: a multinational study of 2622 patients under the auspices of the International Association of Pancreatology and European Pancreatic Club (European Study Group on Cystic Tumors of the Pancreas). Gut 65:305–12. https://doi.org/10.1136/gutjnl-2015-309638
Chen Y, Guo C, Zhang Q, Shen Y, Li Y, Li X, Bai X, Liang T (2018) Patients with pancreatic cystic neoplasms can benefit from management of multidisciplinary team: Experience from a Chinese academic center. Pancreatology 18:799–804. https://doi.org/10.1016/j.pan.2018.07.010
Kimura W, Moriya T, Hirai I et al (2012) Multicenter study of serous cystic neoplasm of the Japan pancreas society. Pancreas 41:380–7. https://doi.org/10.1097/MPA.0b013e31822a27db
Sun HY, Kim SH, Kim MA, Lee JY, Han JK, Choi BI (2010) CT imaging spectrum of pancreatic serous tumors: based on new pathologic classification. Eur J Radiol 75:e45-55. https://doi.org/10.1016/j.ejrad.2009.11.017
Chu LC, Singhi AD, Haroun RR, Hruban RH, Fishman EK (2017) The many faces of pancreatic serous cystadenoma: Radiologic and pathologic correlation. Diagn Interv Imaging 98:191–202. https://doi.org/10.1016/j.diii.2016.08.005
Lee LS (2021) Updates in diagnosis and management of pancreatic cysts. World J Gastroenterol 27:5700–14. https://doi.org/10.3748/wjg.v27.i34.5700
Chu LC, Singhi AD, Hruban RH, Fishman EK (2014) Characterization of pancreatic serous cystadenoma on dual-phase multidetector computed tomography. J Comput Assist Tomogr 38:258 – 63. https://doi.org/10.1097/RCT.10.1097/RCT.0b013e3182ab1556
Takeshita K, Kutomi K, Takada K, Kohtake H, Furui S, Takada T, Fukushima J (2005) Unusual imaging appearances of pancreatic serous cystadenoma: correlation with surgery and pathologic analysis. Abdom Imaging 30:610–5. https://doi.org/10.1007/s00261-004-0295-z
Park C, Kim HJ, Kim SY, Lee SS, Byun JH, Kim SC, Kim MH (2019) Growth rate of serous pancreatic neoplasms in vivo: a retrospective, observational study. Acta Radiol 60:433–40. https://doi.org/10.1177/0284185118787350
Kim HJ, Lee DH, Ko YT, Lim JW, Kim HC, Kim KW (2008) CT of serous cystadenoma of the pancreas and mimicking masses. AJR Am J Roentgenol 190:406–12. https://doi.org/10.2214/ajr.07.2808
Sahara S, Kawai N, Sato M et al (2012) Differentiation of pancreatic serous cystadenoma from endocrine tumor and intraductal papillary mucinous neoplasm based on washout pattern on multiphase CT. J Comput Assist Tomogr 36:231–6. https://doi.org/10.1097/RCT.0b013e3182483bb7
Hayashi K, Fujimitsu R, Ida M, Sakamoto K, Higashihara H, Hamada Y, Yoshimitsu K (2012) CT differentiation of solid serous cystadenoma vs endocrine tumor of the pancreas. Eur J Radiol 81:e203-8. https://doi.org/10.1016/j.ejrad.2011.01.111
Morana G, Ciet P, Venturini S (2021) Cystic pancreatic lesions: MR imaging findings and management. Insights Imaging 12:115. https://doi.org/10.1186/s13244-021-01060-z
Manfredi R, Ventriglia A, Mantovani W, Mehrabi S, Boninsegna E, Zamboni G, Salvia R, Pozzi Mucelli R (2015) Mucinous cystic neoplasms and serous cystadenomas arising in the body-tail of the pancreas: MR imaging characterization. Eur Radiol 25:940–9. https://doi.org/10.1007/s00330-014-3493-2
Wang GX, Wang ZP, Chen HL, Zhang D, Wen L (2020) Discrimination of serous cystadenoma from mucinous cystic neoplasm and branch duct intraductal papillary mucinous neoplasm in the pancreas with CT. Abdom Radiol (NY) 45:2772–8. https://doi.org/10.1007/s00261-020-02664-7
Chen HY, Zhao J, Lu YF, Lv SY, Chen JY, Pan Y, Shi D, Xu XF, Yu RS (2018) The "extracapsular cystic" sign in pancreatic serous cystic neoplasms: A clinicopathologic study of 177 patients with cystic pancreatic lesions. Eur J Radiol 106:167–72. https://doi.org/10.1016/j.ejrad.2018.08.002
Liu QY, Zhou J, Zeng YR, Lin XF, Min J (2016) Giant Serous Cystadenoma of the Pancreas (⩾10 cm): The Clinical Features and CT Findings. Gastroenterol Res Pract 2016:8454823. https://doi.org/10.1155/2016/8454823
Yadav RK, Jiang X, Chen J (2020) Differentiating benign from malignant pancreatic cysts on computed tomography. Eur J Radiol Open 7:100278. https://doi.org/10.1016/j.ejro.2020.100278
Yang J, Guo X, Zhang H, Zhang W, Song J, Xu H, Ma X (2019) Differential diagnosis of pancreatic serous cystadenoma and mucinous cystadenoma: utility of textural features in combination with morphological characteristics. BMC Cancer 19:1223. https://doi.org/10.1186/s12885-019-6421-7
Fang X, Jiang H, Cao K, Li J, Liu F, Wang L, Lu J, Shao C, Bian Y (2022) Distinguishing pancreatic solid serous cystadenomas from nonfunctional pancreatic neuroendocrine tumors by computed tomography: A propensity score analysis. Medicine (Baltimore) 101:e30523. https://doi.org/10.1097/md.0000000000030523
Chen JY, Chen HY, Pan Y, Shi D, Yu RS (2019) Computed tomography and magnetic resonance imaging features of solid serous cystadenomas of the pancreas. Oncol Lett 18:898–906. https://doi.org/10.3892/ol.2019.10346
Shen X, Yang F, Yang P et al (2020) A Contrast-Enhanced Computed Tomography Based Radiomics Approach for Preoperative Differentiation of Pancreatic Cystic Neoplasm Subtypes: A Feasibility Study. Front Oncol 10:248. https://doi.org/10.3389/fonc.2020.00248
Chen HY, Deng XY, Pan Y et al (2021) Pancreatic Serous Cystic Neoplasms and Mucinous Cystic Neoplasms: Differential Diagnosis by Combining Imaging Features and Enhanced CT Texture Analysis. Front Oncol 11:745001. https://doi.org/10.3389/fonc.2021.745001
Liang W, Tian W, Wang Y et al (2022) Classification prediction of pancreatic cystic neoplasms based on radiomics deep learning models. BMC Cancer 22:1237. https://doi.org/10.1186/s12885-022-10273-4
Chu LC, Park S, Soleimani S et al (2022) Classification of pancreatic cystic neoplasms using radiomic feature analysis is equivalent to an experienced academic radiologist: a step toward computer-augmented diagnostics for radiologists. Abdom Radiol (NY) 47:4139–50. https://doi.org/10.1007/s00261-022-03663-6
Huang P, Staerkel G, Sneige N, Gong Y (2006) Fine-needle aspiration of pancreatic serous cystadenoma: cytologic features and diagnostic pitfalls. Cancer 108:239–49. https://doi.org/10.1002/cncr.21911

Table 1 Comparison of clinical information on different types of SCN

	Macrocystic type(n＝18)	Mixed type(n＝17)	Microcapsule type(n＝29)	Solid type(n＝1)	P	Total (n ＝65)
Age(years)	54.6±15.3	60.1±9.93	58.±12.1	62	0.432	57.7±12.7
Gender,n(%)
Male (example)	6(33.3)	4(23.5)	9(31.0)	1	0.830	20(30.8)
Female (example)	12(66.7)	13(76.5)	20(69.0)	0	0.830	45(69.2)
Initial detection size (mm)	34.3±13.9	42.9±15.1	29.6±10.8	12.5	0.019	34.1±14.2
Location,n(%)
Head /neck	6(33.3)	8(47.1)	9(31.0)	0	0.744	21(32.3)
Body/tail	12(66.7)	9(52.9)	20(69.0)	1	0.744	44(67.7)
Clinical symptoms,n(%)
Abdominal pain	7(38.9)	3(17.6)	8(30.7)	0	0.381	18(27.7)
Physical examination or occasional	11(61.1)	14(82.4)	18(69.3)	1	0.320	44(67.7)
Hyperglycemia	0(0)	0(0)	2(6.9)	0	0.497	2(3.1)
Nausea and vomiting	0(0)	0(0)	1(3.1)	0	1.000	1(1.5)
CA199, n(%)	1(5.6)	1(5.9)	0(0)	0	0.295	2(3.1)
Basic diseases,n(%)
Hypertension	6(33.3)	2(11.8)	5(17.2)	0	0.314	13(20.0)
Diabetes	1(5.6)	0(0)	2(6.9)	0	0.787	3(4.6)
Elevated uric acid	1(5.6)	0(0)	1(3.1)	0	1.000	2(3.1)
Cardiac insufficiency	0(0)	0(0)	1(3.1)	0	1.000	1(1.5)

The solid type was not included in the statistical difference analysis. SCN, pancreatic cystic neoplasm

Table 2 Comparison of CT imaging findings of different types of SCN

Imaging findings	Macrocystic type(n＝17)	Mixed type(n＝11)	Microcapsule type(n＝24)	Solid type(n＝1)	P	Total (n ＝53)
Precontrast scan CT value (HU)	8(3.75,10.25)	24(14,27)	18.5(15,23)	25	0.000	15(10,23)
Arterial phase absolute enhanced value (HU)	—	47(30.5，57)	34(19.75，47.25)	81	0.122	—
Portal vein phase absolute enhanced value (HU)	—	50(34.5，54)	38.5(26.5，57)	40	0.657	—
API	—	0.792(0.576，1.048)	0.544(0.429，0.672)	1.333	0.073	—
PPI	—	0.770(0.461，0.965)	0.612(0.540，0.838)	0.923	0.722	—
Enhancement degree (arterial phase), n (%)
No enhancement	17(100.0)	0(0)	0(0)	0	0.000	17(32.1)
Mild enhancement	0(0)	3(27.3)	7(29.2)	0	0.029	10(18.9)
Moderate enhancement	0(0)	2(18.2)	9(37.5)	0	0.008	11(20.7)
Obvious enhancement	0(0)	6(54.5)	8(33.3)	1	0.002	15(28.3)
Enhancement degree (portal vein phase), n (%)
No enhancement	17(100.0)	0(0)	1(4.2)	0	0.000	18(34.0)
Mild enhancement	0(0)	3(27.3)	10(41.7)	0	0.004	13(24.5)
Moderate enhancement	0(0)	3(27.3)	8(33.3)	1	0.018	12(22.6)
Obvious enhancement	0(0)	5(45.6)	5(20.8)	0	0.007	10(18.9)
External contour, n (%)
Smooth	4(23.5)	0(0)	5(20.8)	1	0.281	10(18.9)
Lobulated	12(70.6)	9(81.8)	14(58.6)	0	0.373	35(66.0)
Ambiguity	1(5.9)	2(18.2)	5(20.8)	0	0.473	8(15.1)
Boundary, n (%)
Clear	17(100.0)	8(72.7)	16(66.7)	1	0.018	42(79.2)
Unclear	0(0)	3(27.3)	8(33.3)	0	0.018	11(20.8)
Fibrous central scar,n(%)	0(0)	4(36.4)	3(12.5)	0	0.021	7(13.2)
Internal calcification,n(%)	3(17.6)	4(36.4)	8(33.3)	0	0.488	15(28.3)
Pancreatic duct dilation,n(%)	1(5.9)	3(27.3)	4(16.7)	1	0.331	9(17.0)
Relationship with blood vessels,n(%)
No relationship	4(23.5)	0(0)	1(4.2)	0	0.097	5(9.4)
Squeeze	12(70.6)	3(27.3)	5(20.8)	0	0.004	20(37.7)
Blood-supply	1(5.9)	8(72.7)	18(75.0)	1	0.000	28(52.9)
Distant organ and lymph node metastasis,n(%)	0(0)	0(0)	0(0)	0	NS	0(0)

The solid type was not included in the statistical difference analysis. SCN, pancreatic cystic neoplas; API, Arterial phase enhancement index; PPI, Portal vein phase enhancement index.

Table 3 Comparison of MRI imaging findings of different types of SCN

Imaging findings	Macrocystic type(n＝9)	Mixed type(n＝16)	Microcapsule type(n＝22)	Solid type(n＝1)	P	Total (n ＝48)
T1WI signal strength,n(%)
High signal	0(0)	0(0)	0(0)	0	NS	0(0)
Low signal	9(100.0)	16(100.0)	22(100.0)	1	NS	48(100.0)
T2WI signal strength,n(%)
High signal	9(100.0)	16(100.0)	22(100.0)	1	NS	48(100.0)
Low signal	0(0)	0(0)	0(0)	0	NS	0(0)
External contour,n(%)
Smooth	2(22.2)	1(6.3)	3(13.6)	1	0.546	7(14.6)
Lobulated	7(77.8)	15(93.7)	19(86.4)	0	0.546	41(85.4)
Boundary,n(%)
Clear	9(100.0)	16(100.0)	22(100.0)	1	NS	48(100.0)
Unclear	0(0)	0(0)	0(0)	0	NS	0(0)
Internal calcification,n(%)	0(0)	12(75.0)	12(54.5)	0	0.001	24(50.0)
Pancreatic duct dilation,n(%)	1(11.1)	3(18.7)	3(13.6)	1	0.874	8(16.7)
Relationship with pancreatic duct
No relationship	8(88.9)	13(81.3)	19(86.4)	0	0.874	40(83.3)
Squeeze	1(11.1)	3(18.7)	3(13.6)	1	0.874	8(26.7)
Communicating	0(0)	0(0)	0(0)	0	NS	0(0)

The solid type was not included in the statistical difference analysis. SCN, pancreatic cystic neoplasm; T1WI, T1‑weighted imaging; T2WI, T2‑weighted imaging.

No competing interests reported.

SupplementaryMatrrials.docx

Download PDF

Version 1

posted

You are reading this latest preprint version

Comparison of CT and MRI findings of different types of pancreatic Serous cystic neoplasm: combined with histopathology

Status:

Version 1

Abstract

Figures

Introduction

Materials and methods

Image inspection parameter setting and data acquisition

Image analysis and measurement

Pathological examination method

Statistical method

Results

Discussion

Declarations

Author Contribution

Data Availability

References

Tables

Additional Declarations

Supplementary Files

Status:

Version 1