IPMN is a benign, borderline, low-grade dysplasia or invasive cancer derived from pancreatic ductal epithelium. Tumor cells are tall columnar mucous-rich epithelial cells, with or without papillary protrusions, extensively invading the main pancreatic duct (MPD) and/or branch pancreatic duct (BPD), causing cystic dilation. With the continuous development of diagnostic standardization and imaging techniques, IPMNs are becoming increasingly routinely diagnosed in clinical practice. According to the communication with the pancreatic duct, they are morphologically divided into MD-IPMN, BD-IPMN and MT-IPMN. Approximately 40% to 65% of IPMNs occur in the branch pancreatic ducts, while they are found in the MPD accounting for about 15% to 35% cases. The probability of simultaneous occurrence in the both pancreatic ducts is only 15%-20% (28).

According to the different histology and mucin expression of IPMN, it can be divided into four epithelial subtypes as below: gastric-type, intestinal-type, pancreatobiliary-type and oncocytic-type, each of which has various kinds of risks of malignant progression. Oncocytic- and gastric-type IPMNs are often low-grade neoplasms, while intestinal- and pancreatobiliary-type IPMNs have a disposition to high-grade neoplasms and are usually related to invasive cancer (IC) (11). The prognosis of IPMNs is superior to pancreatic ductal adenocarcinoma (PDAC) after surgical resection (29). It has been reported that the incidence of cancers derived from IPMN is between 6% and 46% (30–33), including IPMN with high-grade dysplasia (HGD) and IC, and its prognosis is as poor as that of PDAC (34). Assessing the rate of malignancy in IPMN has become a clinical challenge. The risk factors for malignant tumors include weight loss, patient age, relationship with mural nodules, increased jaundice/bilirubin levels, and elevated CEA levels. However, there is no established standard that can safely and accurately distinguish malignant and nonmalignant lesions (35). Therefore, the key to the treatment of IPMN is to accurately predict the risk of malignancy. At the same time, it is also important to evaluate the probability of surgical resection benefit. Imaging takes a significant role in the evaluation and detection of IPMNs (36).

The aims of imaging examination of IPMN are as follows: first, to detect IPMN and exclude other PCLs; second, to distinguish the relationship between lesions and the pancreatic duct, which is conducive to typing; and third, to determine the key risk factor of malignancy and estimate the resectability of clinical surgery. Various imaging methods are used to achieve these goals.

MCNs are cystic tumors derived from the pancreatic epithelium that have the potential for malignancy. They are relatively rare pancreatic cystic tumors in the clinic, accounting for 29% of all PCNs (37). Compared with IPMNs, MCNs do neither communication with the MPD nor BPD system. They are often solitary and are covered by a thin fibrous cyst wall. The cyst is lined with tall columnar mucous cells that secrete mucin and often form papillae, and the subepithelial stroma is often ovarian-like stroma with abundant cells (38). There is no significant difference in incidence between sexes in their 60s and 70s, and the probability of occurrence in the body/tail of the pancreas is greater than that in the pancreatic head (67.3%–99.4%) (39), with 89.5% in the present series.

The malignant probability of MCNs varies between 6% and 36%, which is still significant (12). The features predictive of malignancy include irregular or focal thickening of the cyst wall, a large volume, and solid content inside or outside the cyst (40, 41). The size (> 8.5 cm) and volume of the MCNs on CT/MR imaging are the only features associated with HGD/IC, and the average growth rate is very slow, about 4 millimeters (0.16 in) every year (42). The mucinous transitional epithelium is the origin of almost whole malignancies arising from MCNs. MCNs can be divided into three major categories according to the grade of dysplasia as well as IPMNs: low- or intermediate-GD, HGD and IC (13, 43). Resection is advocated for whole types of MCNs according to current the guidelines and clinical consensus unless there are contraindications to surgery (44). For MCNs with IC, the prognosis is closely related to the extent of lesions invasion, tumor stage and R0 resection rate. The two- and five-years survival rates of resectable MCNs with IC are 67% and 50%, respectively (7). Therefore, early detection and identification of MCNs with invasive cancer by imaging methods are of great significance.

SPNs are an rare pancreatic tumor and, as their name implies, have a solid pseudopapillary structure formed by epithelial cells of a single shape in a loose arrangement. They are prone to hemorrhage and cystic transformation. These tumors account for only 0.2~2.7% of all exocrine pancreatic neoplasms (45). According to the morphological and structural characteristics of the lesion, different reports use the nomenclature solid-cystic neoplasm, papillary-cystic neoplasm, solid-cystic acinar neoplasm, solid-papillary neoplasm, but the actual use of “solid pseudopapillary neoplasm” (SPN) is similarly only a descriptive name that represents morphological features yet retains the openness of histogenesis. Regardless of the presence or absence of metastatic disease, SPNs are generally considered to be low-grade tumors with an indolent growth pattern. The origin of these tumor cells in the pancreas is uncertain. There are two classical theories about the origin of SPN: the one suggests that it origins from pluripotent pancreatic cell, then the other proposes a female genital bud origin (46).

Due to the wide application of high-quality and -resolution imaging examinations, mainly US, CT, and MRI, it has been reported high frequency in the past few decades. In a recent review of all SPN description that published in the English journal up to 2014, Law et al. (47) confirmed a total of 2744 cases of this neoplasm. Yao et al. systematically reviewed 2,450 SPN cases in a Chinese population before January 2020, which was published in both the Chinese and English literature (48). They concluded that SPN is an indolent neoplasm and seldom seen that mainly occurs in young females. The clinical manifestations are abdominal masses and abdominal pain, most of them present as pancreatic head and tail space occupying, and the prognosis is excellent after complete resection. Generally, they are indolent, but a few have malignant potential. Regrettably, the prognostic factors that predict malignant potential have been hard to identify (49). Most patients present with local lesions, and only 9-15% have metastases or local infiltration.

At present, the main treatment is still surgical resection, and its prognosis is different from that of pancreatic cancer. According to reports, the five-year survival rate can be as high as 94-97% (45, 50). Rare SPNs can occur at any age and in both genders, especially young females. Although the survival rate is typically high, histological images cannot accurately predict its biological behavior. Lesions without obvious malignant signs but only suspicious morphological signs can also cause late recurrence, metastasis and even death. The exact histogenesis is still unclear, and it may originate from primordial cells. More research on SPT is needs for further clarification.

SCNs are benign tumors of the pancreatic exocrine glands that account for 16%~33.3% of whole cystic neoplasms of the pancreas. It is a slow-growing benign lesion with an extremely low probability of malignant transformation (14, 15, 51). The concept of SCNs as benign disease entities without the risk of malignant transformation was revised after George et al. revealed the first case of a malignant pancreatic SCN in 1989 (52). The malignancy of SCNs, serous cystadenocarcinomas, are limited to 25~30 cases report in the global literature, representing <1% of all SCNs (15), including the largest sample size, which found three patients with serous cystadenocarcinomas among 2622 cases (53). Therefore, SCNs of the pancreas have extremely low malignant potential but are not totally benign.

Patients are often discovered with SCN in their late 50th or early 60th. SCN usually develops in the body/tail of the pancreas. Despite these neoplasms are mostly benign, they often grow slowly and may have large diameters (13). SCNs are representatively honeycombed microcystic tumors consisting of uniform, cuboidal, glycogen-rich epithelial cells. Thus far, there are four known variants of serous cystadenoma, namely, macrocystic serous cystadenoma, solid serous adenoma, VHL-related SCN, and mixed serous neuroendocrine neoplasm, in which the serous epithelial components of these variants are identical to those of serous cystadenoma.

Pancreatic serous cystadenomas are benign lesions and could be regulated by surveillance, which does not commonly mandate surgical resection unless they exhibit aggressiveness or unspecific characteristics that hinder accurate diagnosis. CT is the preferred first-line examination modality for characterizing SCNs and differentiating them from their mimickers (54).

Among the three types of IPMN, the branched type is the most common, followed by the mixed type, and the MPD type is relatively rare. The latter is further divided into two sub-types, the segmental type and the diffuse type. The communication between cystic lesions and the MPD is one of the diagnostic points of branch-type IPMN. Branch-type IPMN specific imaging findings are as follows: tubular or earthworm-like shadows in low-density cystic lesions, cystic walls and septate microenhancing nodules. The dilated MPD is not limited to the distal end of the lesion. The imaging signs that affect the diagnosis are as follows: the lesions are oval or dumbbell-shaped, the lesions’ density is close to that of water, and there are non-separations or tiny nodules in the lesions. Thin-slice scanning combined with three-phase enhancement, coronal or sagittal image reconstruction, and careful observation of cyst wall and intralesional structures can help improve the diagnostic accuracy ( Figure 2 ).

The MD-type IPMN-specific imaging manifestations are as below: moderate or greater dilatation of the MPD, continuous expansion of the pancreatic duct without bead-like changes, enhanced mural nodules on the cyst wall, slight atrophy of the pancreatic parenchyma, and markedly dilated MPD asymmetry with mildly atrophied pancreatic parenchyma ( Figure 3 ). Imaging signs that affect the diagnosis are as follows: in the early stage of the disease, dilatation of the MPD is limited or mild. In the early stage of the disease, this is easily confused with the slight dilatation of the pancreatic duct caused by anatomical variations; diffuse IPMN is easily confused with chronic pancreatitis, and localized IPMN is easily confused with pancreatic fusiform pseudocysts.

Mixed-type IPMN often comprises both branched IPMN and MPD IPMN, but it is not a simple combination of the two. In MT-IPMN, the expansion of the MPD can be localized or diffuse, and localized expansion can manifest as multiple discrete segmental expansions, but no beaded changes occur. There is no strict boundary between the limited MPD dilatation and branched IPMN with mild MPD dilatation. Some studies believe that mixed IPMN is caused by the further development of branched IPMN, which is the key to distinguishing between branched IPMN and the mixed type. It is unclear whether there are tiny nodules in the MPD that are adjacent to the expansion, and if there are tiny nodules in the expanded MPD, it is of a mixed-type ( Figure 4 ). Mixed-type IPMN can be accompanied by one or more branch types. Therefore, the imaging manifestations of mixed-type IPMN are more complex and can differ, but simultaneous expansion of the BPT and MPD is also the easiest subtype to diagnose.

Magnetic resonance T₁WI showed that the liquid in the pancreatic duct of the IPMN dilated pancreatic duct had a low signal that was slightly higher than water; T₂WI showed a high signal that was slightly lower than water. Some lesions showed hyperintensity on T₁WI and hyperintensity on T₂WI. The spatial resolution of MRI is limited, and the ability to show small mural nodules is not as good as CT, but MR helps to show larger nodules. The nodules show a lower signal on T₁WI, which is between normal pancreatic tissue and dilated pancreatic duct fluid. T₂WI is helpful to show streak-like hypointense separation or dilated pancreatic duct wall within the hyperintensity of branched IPMN, showing isohypointense mural nodules in the dilated pancreatic duct. The DWI signal of pancreatic IPMN varies greatly. Some DWI is iso-intense, and some is hyperintense. DWI helps to detect metastatic lymph nodes. MRCP helps to determine the relationship between the lesion and the MPD and shows the MPD protruding into the duodenum. Thin-layer MRCP is more helpful to determine the relationship between the MPD and the duodenum. During dynamic enhancement, the enhancement of the pancreatic duct wall of IPMN is similar to that of CT, the enhancement of small nodules is not as good as that of CT, and the enhancement of large nodules is close to or better than CT.

Mucinous cystadenoma with invasive carcinoma often manifests as multilocular cystic lesions with uneven wall thickness, wall nodules and calcifications in the lesion. After enhancement, the intracapsular septum thickens, and the wall nodules are obviously strengthened. Sometimes it is not easy to distinguish between benign and malignant tumors by imaging. If the following signs appear in the cyst, mucinous cystadenoma with invasive carcinoma is often indicated: ①There are more solid components in the cyst. ②There are obviously enhanced mural nodules. ③ Irregular thickening of the cyst wall and the presence of multiple daughter cyst near the large cyst. ④Local pancreatic lymphadenopathy and intrahepatic metastasis are observed and adjacent large blood vessels have been invaded. ⑤The tumor is large, with a diameter >8 cm.

Mucinous cystadenoma usually manifests as a clear boundary with hypo-intensity on T₁WI and hyper-intensity on T₂WI, but sometimes it has different manifestations due to the composition of the cyst fluid. The advantage of MRI is that it can accurately reflect the composition of the cyst fluid of mucinous cystadenoma. Sometimes the signal on T₁WI is uneven and has a high signal, which is pathologically related to mucin in the cyst fluid or intracystic hemorrhage. In addition, MRI shows better separation between the wall and wall nodules in the lesion capsule than CT. The cyst cavity of mucinous cystadenoma is generally not connected to the pancreatic duct, which helps to distinguish it from intraductal papilloma (communicating with the pancreatic duct). MRCP examination is helpful for determining whether the mucinous cystadenoma is connected to the pancreatic duct. MR examination of mucinous cystadenoma with invasive carcinoma can not only show that the tumor is cystic but also clearly depicts the tumor cyst wall, septum and mural nodules ( Figure 5 ).

Uneven thickening of the intratumoral septum and cyst wall or the appearance of mural nodules, invasion of the common bile duct (CBD) or pancreatic duct (PD) and surrounding blood vessels are all helpful for the diagnosis of mucinous cystadenoma with invasive carcinoma. MR examination is helpful for the differentiation of benign and malignant pancreatic mucinous cystadenoma. Since the blood supply of the tumor is mainly concentrated in the cyst wall and septum, enhancement may appear after an enhanced scan.

CT scans of serous microcystic adenoma show a clear boundary, lobulation, and a mass formed by several small water-like cysts. The diameter of a single capsule is usually less than 2 cm. Star-shaped central scars can be seen in the lesions, and calcifications usually occur in the central scars. The enhanced scan shows progressive medium-strength enhancement of the central scars. It is often difficult to identify the tumor septum when it is thin, and enhancement can help with visualization when it is thicker. Star-shaped central scarring with or without calcification is considered to be a specific manifestation of serous microcystic adenoma. CT showing the intratumoral septum, central scar and size of the cyst is key to the imaging diagnosis of microserous microcystic adenoma. Sometimes serous microcystic adenomas are composed of numerous tiny vesicles, which show a honeycomb or spongy appearance, and most of the single cysts are one to several millimeters long ( Figure 6 ). Clear edges and a cystic space with soft tissue structure can be seen on plain CT scan, and moderate enhancement is observed during the enhanced scan. At this time, it should be distinguished from SPN. When the lesion wall is thick, the internal components are complicated, and there is bleeding, calcification, liquefaction and necrosis, a solid pseudopapillary tumor is indicated.

MR shows serous cystadenoma basically similar to the CT appearance. However, the ability of MR to display calcification is weaker than that of CT, and MR can better display intracapsular hemorrhage, separation, and capsule wall. it can reveals more soft tissue information than CT scans. Soft tissue display can provide more information than CT. In addition, the display of the pancreatic duct and bile duct is more valuable for the differential diagnosis of diseases. Serous microcystic adenoma showed typical polycystic or honeycomb changes on MRI. The contents of the sac are a clear protein-containing liquid, generally showing a liquid signal shadow of long T₁ and long T₂. Sometimes the cyst cavity shows a slightly low signal shadow on T₂WI, which is caused by the local fluid concentration in the cyst cavity and the high protein content. Small cysts can only sometimes be a few millimeters. Intensified scanning of the cyst wall and separation are often mildly continuously enhanced. The lobulated contour, as if the collapsed wall dumped toward the center, is a feature of microcystic serous cystadenoma caused by the traction of the central star-shaped scar; the central scar sometimes has calcification, which appears to be sunlight-radiating on CT, with certain characteristics, but MRI has obvious shortcomings in showing scar tissue calcification.

MRI can almost show the cyst wall and the space within it, even for tumors with small diameters. The intratumoral septum can be seen more clearly on T₂WI. Serous tumor microcystic adenoma of the pancreas generally does not cause dilation of the CBD and PD. Although serous microcystic adenoma of the pancreatic head is adjacent to the CBD, there are few signs of abnormal compression or obstruction in the bile duct system. This may be due to the tumor’s soft body and slow growth, which does not compress the bile duct, also proving that the tumor is not aggressive. However, a small number of cases of serous microcystic adenoma with mild dilation of the pancreatic duct or CBD have also been reported. Mild pancreatic duct widening may be caused by mild compression changes or mild inflammatory changes in the pancreas. The relationship between the cyst cavity and the pancreaticobiliary duct is of great significance to the diagnosis and differential diagnosis of this disease.

Serous microcystic adenomas caused by pancreatic tail duct dilatation should be differentiated from BD-IPMN. MRCP can easily determine the relationship between the neoplasm and the pancreatic duct. Microcystic cystadenoma with typical changes and other pancreatic cystic tumors are not difficult to distinguish. Serous oligocystic adenoma tumors show typical unicystic or multicystic changes on MRI. Cystic lesions are larger than serous microcystic adenomas. The lesions are clearly separated from the surrounding pancreas, and the edges are smooth. The characteristics of the signal of the cyst contents and the relationship with the pancreaticobiliary duct are roughly similar to those of serous microcystic adenoma. There is no sign of communication between the cyst cavity of the lesion and the pancreatic duct, and the adjacent CBD often shows no obvious compression or obstruction ( Figure 7 ).

Solid serous cystadenoma contains a large number of fibrous interstitial blood vessels, and there is no cyst in the tumor according to general pathology. Only tiny cysts can be seen under the microscope, with abundant interstitial blood vessels ( Figure 8 ). This type of cystadenoma is rare. Because the capsule is very small, it is difficult to display watery signals in the lesion on T₂WI, which often leads to the misdiagnosis of pancreatic neuroendocrine tumors.

Regarding the typical CT appearance of a SPN, the solid part of the pancreas is slightly low-density, and cystic necrosis is shown as a lower-density area. The pathological basis that causes uneven density and mixed signals is tumor cystic transformation, hemorrhage, and the calcification of lesions( Figure 9 ). The distribution of cystic and solid components are also different; they can exist alternately, solid components can be located around the tumor, or multiple cysts of different sizes can be located at the edge of the tumor. Pathologically, the tumor cells in the pseudopapillary area form branched pseudopapillae with slender fibrous blood vessels as the axis. The cells are arranged in nests or lumps and in multiple layers. They are far away from the tumor cells around the blood vessels and are prone to degeneration. Necrosis, liquefaction and cystic changes can occur. Histologically, bleeding is prone to occur due to the large number of fragile, thin-walled blood vessels and the lack of a strong stent structure. Hemorrhage is one of the characteristics of this tumor.

Bleeding can occur in the cystic part or the solid part, showing gel-like or cystic tissue; the cystic and solid components of CT are scattered and patchy and show a high density. Calcification in the lesion is more common and can manifest in various ways: small spots, diffuse calcification, incomplete arc-shaped calcification of the envelope ( Figure 9 ), and sometimes complete arc-shaped calcification. If the lesion is mainly cystic, most of the cyst is not strengthened, and a few solid parts inside are obviously strengthened, which are distributed in the low-density liquid tissue in the form of sheets, forming the so-called “floating cloud sign” ( Figure 10 ). The surrounding envelope is obviously enhanced. In the case of a cystic solid structure, the solid part of the arterial phase is mostly papillary or wall nodular enhancement. For the solid structure, the solid part of the arterial phase is slightly enhanced, and the parenchymal and delayed phases are further strengthened, showing progressive filling, but both are lower than the degree of pancreatic parenchymal enhancement.

For larger lesions, where cystic and solid lesions are often the main focus, the cystic component is not enhanced after enhancement, and the solid and cystic structures are clearly demarcated. Note that the pancreatic tissue has a “cup-mouth” boundary ( Figure 10 ). Although the lesion is sometimes large, the MPD or CBD is generally not dilated. In a few cases, it may be slightly dilated, usually due to tumor compression of the adjacent duct. MRCP shows expansion of the MPD more intuitively than MDCT. A larger SPT can cause compression of adjacent blood vessels in the portal vein, splenic vein, inferior vena cava, and renal vein.

The MRI scan of SPN showed tumors with mixed signals on T₁WI and T₂WI and slightly hyper-intensity on DWI. The basis for the confounding of tumor signals is the tumor’s cystic degeneration, necrosis, hemorrhage and calcification. MRI is more sensitive to the detection of tumor hemorrhage than CT. Usually, hemorrhage MRI shows a hyper-intensity on T₁WI and a hyper- or hypo-intensity on T₂WI. Due to the coexistence of blood and other liquid components, signs of stratification can be seen, showing liquid levels. MRCP helps to show dilatation of the pancreaticobiliary duct.

Recently, EUS has been recognized as an essential diagnostic tool for PCN management. When the results of the radiological diagnosis of malignant tumors are certain and/or when the PCNs are considered to have clinical or radiological characteristics, EUS is nominated as a second-line examination method after CT/MRI. Because the stomach and pancreas are adjacent to each other in the body, the EUS transducer can be placed close to the pancreas, and the gland can be clearly imaged. In this way, the pancreatic cyst wall and its contents can be evaluated in detail, and internal septae and solid areas within the cysts can be differentiated.

One study (77) showed that EUS is the best diagnostic method for differentiating nonneoplastic cysts and PCNs and to characterize PCNs, being superior to both CT and MRI. However, another large multicenter study reported the opposite result: the accuracy of applying EUS to diagnose mucinous versus nonmucinous cysts was only 51% (16). It is speculated that one of the limitations of EUS may be due to different interpretations among endoscopists. Another study revealed that the accuracy of detecting neoplasms with malignant potential ranged from 40% to 93% among 8 different endoscopists invited to interpret the same EUS procedure (78).

Making an accurate diagnosis with cross-sectional imaging and EUS alone is challenging, so EUS-FNA (fine needle aspiration) is frequently employed to obtain a cyst aspirate. Cyst fluid cytology suffers from poor sensitivity, which is specific. During EUS, various analyses (cytology, biochemistry and molecular) of pancreatic cyst fluid acquired from FNA can observably increase the accuracy of diagnosis (1, 16, 79). In terms of the rate of correct diagnosis, EUS-FNA increased the accuracy by 36% after CT and the accuracy by 54% after MRI (80). The risk of infection, hemorrhage, and pancreatitis of EUS-FNA increases compared to noninvasive imaging, while most studies have shown that its benefits outweigh the risks (81, 82).

EUS-FNA is a commonly used method of diagnosing IPMNs. However, the interpretation of cytological features relies on clinical and radiological findings. The presence of large amounts of thick mucin in the correct clinical setting can only suggest a diagnosis of IPMN. In contrast, it is difficult to distinguish the presence of limited mucin and low-grade mucinous epithelium from the presence of normal gastric epithelium (4). Compared with low-grade IPMN, the following features are more supportive of differentiating HGD IPMN: background of necrosis, aberrant chromatin patterns (hypochromasia or hyperchromasia), the presence of large vacuole single cells, significant nuclear irregularities, increased nuclear-plasmic ratio, and small cell sizes (≤12 μm duodenal cell) (83). Furthermore, The high-grade atypia of IPMNs tend to be larger (≥30 mm), have enhanced mural nodules (≥5 mm) (84) or have solid contents and dilated MPD (≥5 mm) (85).

Differentiating between IPMN and MCN is also difficult for cytopathologists. Due to involvement with the pancreatic duct observed with IPMN but not MCN, correlation with radiologic findings can significantly facilitate diagnosis compared to others. At the same time, mucinous cystic tumors develop almost entirely in females, and presence of an ovarian stroma is pathognomonic. Aspirates of low-grade MCNs (mucinous cystic adenomas) that account for more than 75% of MCNs will show honeycomb sheets of bland mucin-containing epithelium but often lack the presence of complex papillary architecture compared to high-grade MCNs (4). The mucin-containing nucleus has a smooth contour, fine chromatin, and inconspicuous nucleoli (86).

EUS with FNA for SCNs has both low specificity and sensitivity, SCNs usually contain hemosiderin-laden macrophages and paucicellular cells with clear or hemorrhagic backgrounds. The highly vascularized fibrous septa of the SCN leads to the hemorrhagic nature of these specimens. SCNs do not involvement with the PD system and have low CEA levels, with cysts are often filled of clear-yellow serous fluid with low viscosity, compared with IPMNs. The cells of SCNs are bland. The nucleus is round, and the contour is smooth. Chromatin is evenly distributed in the nucleus, and the nucleoli are inconspicuous. When the background contains mucin, associate with CEA levels and radiologic imaging would be cautiously exclude MCN. Macroscopically, SCNs are usually arranged around star-shaped scars, which show cysts with a distinctive spongy or honeycomb appearance.

The contents of hypercellular smears shown by EUS-guided FNA of SPTs include slender papillary fragments with fibrous vascular stalks and perivascular myxoid matrix. They are arranged by monomorphic cubic cells into cohesive groups and isolated cells. The neoplastic cells are round to oval, and the cytoplasmic boundary was unclear. The nucleus is grooved or bean-shaped, while the chromatin is fine-grained, and occasionally invisible or small nucleoli can be seen. Macroscopically, SPNs are large, with round to oval shapes, clear margin and fibrous pseudocapsules. SPNs are complex neoplasms with varies components (e.g., solid, cystic, hemorrhagic, and necrotic). Cystic degeneration is a common phenomenon that occurs during progression. Moreover, the larger the neoplasm is, the more obvious the cystic component (45, 87).

Studies have shown that pancreatic cyst fluid analysis of tumor markers and molecular markers can help characterize PCNs. At present, perfect biomarker testing for detecting pancreatic tumor has not yet been developed. The most commonly used blood test to monitor and detect pancreatic cancer is the serum marker CA 19-9. However, its sensitivity is limited, especially for small malignancies (88). When CEA levels > 192 ng/mL, the sensitivity and specificity of the diagnosis of mucus lesions are 73% and 84%, respectively (16). Negative and positive predictive values for mucin etiology are both 94% when CEA levels <5 ng/mL or >800 ng/mL (17). The utilization of CEA using pancreatic cyst fluid to diagnose malignant cysts is less effective, as a previous meta-analysis suggested that both the diagnostic sensitivity and the specificity were 63% (89). Chemical analysis of liquid CEA and amylase levels may be helpful, but this approach cannot differentiate between MCNs and IPMNs. Elevated CEA can be used as a marker to distinguish between mucinous and non-mucinous cysts rather than benign or malignant cysts. When the critical value of CEA was ≥192~200 ng/mL, the accuracy increased to 80% for the diagnosis of mucinous cysts, showing high specificity but low sensitivity (16).

The current view is that serous cystic adenomas originate from centroacinar cells, where staining for cytokeratins and calretinin is positive but staining for CEA, mucin, estrogen receptors, and progesterone receptors is negative. SCN and ductal adenocarcinoma and neuroendocrine tumors can be distinguished by inhibin and calcarein, which were found to be helpful immunostaining markers in recent studies (90, 91).

To compensate for the limitations posed by cytology and tumor markers, specific molecular markers for diagnosing PCNs and predicting malignant tumors are currently being developed. A molecular DNA analysis method for pancreatic cyst fluid is currently on the market. However, a molecular analysis method for cyst fluid is still in development. KRAS mutations support the diagnosis of mucous cysts more accurately, but KRAS does not always indicate a malignant cyst. It may be helpful to use GNAS mutations to differentiate between obvious mucinous cysts and indolent cysts that can be managed conservatively (92). Genetic analysis showed that the dual mutations in KRAS and GNAS were highly specific for IPMN. Compared with MCNs and SCN (lack GNAS codon 201 mutations), several research found that mutations in GNAS codon 201 are present in some IPMNs (41%-66%) and can even reach 74% to 100% in enteric IPMN (18–20).

With the aim of distinguishing MCNs from other PCNs, such as IPMN and SPN, some research have revealed that MCN is a kind of cystic neoplasm without the GNAS mutation and generally without the CTNNB1 mutation (21). KRAS mutations have been reported in MCNs (50%-75%) (19, 21). In serous cystadenomas, the absence of CTNNB1 mutation can be used to distinguish them from SPNs. In addition, KRAS and GNAS mutations are often expressed in IPMNs and MCNs rather than SCNs. There are a few studies of protein expression in SCAs. VEGF is a protein inhibited by a kind of tumor suppressor that is usually encoded by the VHL gene. In cysts or pancreatic duct fluid, VEGF-A levels can aid in diagnosis. The sensitivity and specificity for diagnosing SCAs were 100% and 97%, respectively, when VEGF-A levels were > 8500 pg/mL and 100% and 90%, respectively, when VEGF-C levels were > 200 pg/mL (22). Combining both VEGF-A and VEGF-C provides 100% sensitivity and specificity for the diagnosis of SCA. In addition, VEGF-A (using a critical value of >5000 pg/mL) combined with CEA (<10 ng/mL) can detect SCAs with a sensitivity of 95.5% and a specificity of 100% (93).

Few research have focused on the glycoproteomics of SCAs. The research have shown that SCAs express MUC1 and MUC6 instead of MUC5AC, which provides proof that SCAs originate from pancreatic central cells and intralebar ducts (23, 24). An extracellular matrix protein implicated in pancreatic cancer called periostin was found to increase 8-fold in SCA cyst fluid compared to mucinous lesions (94). On the other side of the shield, serous cystadenomas are related to von Hippel Lindau (VHL) syndrome, while mutations in the VHL gene are present in all SCAs in patients with VHL syndrome. VHL loss-of-function mutations may also be reflected in the development of sporadic SCAs (95). The macrocyst (oligocyst) variant is a rare type of SCN with fewer but more numerous cysts and without a stellate central scar (96); solid variant, which is devoid of cysts; and mixed serous-neuroendocrine variant (91).

Studies have been performed to study several proteins related to the above genes in SPN tissue: B-catenin, androgen receptor, lymphoid enhancer-binding factor 1 (LEF1), and transcription factor for immunoglobulin heavy-chain enhancer 3 (TFE3) (25). Among them, B-catenin has a sensitivity of 98.9% and a specificity of 97% for the diagnosis of SPT, which is the most sensitive indicator of diagnosis. The combination of LEF1 and TFE3S increases the sensitivity to 100% but decreases the specificity to 91.9%. Another investigation explored the use of B-catenin to diagnose SPT, reporting a 100% sensitivity and 87% specificity (26). The combined application of B-catenin, TEF3, and SOX11 can be used to distinguish SPN, with a sensitivity and specificity of 97%. These tissue findings are also relevant for EUS-FNA biopsy samples (27). The absence of KRAS, GNAS, or RNF43 can distinguish SPTs from other PCNs. Because of the good prognosis of SPNs, complete resection of these inert neoplasms can be cured.