Role of endoscopic ultrasound-guided fine needle aspiration biopsies in diagnosing pancreatic neoplasms in the paediatric population: experience from a tertiary center and review of the literature

Endoscopic ultrasonography (EUS) is a well-established diagnostic method for the evaluation of a diverse range of pancreatic lesions in adult patients, however it is a relatively new technique in paediatric pancreatology.^1,2

Studies have also shown that EUS is a better diagnostic procedure than transabdominal ultrasound (TUS), CT scan or endoscopic retrograde cholangiopancreatography for smaller pancreatic lesions, mainly because of the proximity of the EUS probe, making possible a better evaluation of the lesion.^1,2 EUS enables a detailed evaluation of pancreatic parenchyma and the ductal system, especially in paediatric patients with pancreatic masses, suspected autoimmune pancreatitis or fluid collections.^3,4

In addition, it allows for the intervention procedures, namely sampling of lesions with fine-needle aspiration biopsy (FNAB) or core tissue biopsies, or performing drainage procedures in cystic lesions while using Doppler to assess and avoid the vasculature.³ However, the methods of cell or tissue sampling depend on the size and location of the pancreatic lesion and the professional expertise.

Pancreatic neoplasms are exceptionally rare in children, with malignant tumours having an estimated incidence of 0.02 per 100,000.³ It is estimated that solid pseudopapillary neoplasm (SPN) accounts for up to 71% of pancreatic tumours in children and adolescents, followed in order of frequency by pancreatic neuroendocrine tumour, serous cystadenoma and pancreatoblastoma, which is the most common malignant tumour in the patients under 10 years old.^3,5,6

The radiologic preoperative diagnosis usually starts with a TUS, which allows a diagnosis of most paediatric pancreatic tumours, except for functional neuroendocrine tumours like insulinomas and gastrinomas, which may cause symptoms even though they’re small.³ This is especially true in the case of SPN, which presents as a well demarcated lesion, with cystic or necrotic components and with occasional calcifications within solid tumour.^3,4

The Clinical Department of Gastroenterology at the University Clinical Centre Ljubljana and the Department of Cytopathology at the Institute of Pathology (IP), Faculty of Medicine (FM), University of Ljubljana (UL), Slovenia, have been closely collaborating on endoscopic ultrasound-guided fine needle aspiration biopsies (EUS FNAB) since the introduction of the method in 2010. The aim of present study was to analyse the cases of EUS FNAB of pancreatic neoplasms in children in a tertiary care centre and to incorporate a review of the existing literature.

The first case was a 8-year-old female patient, weighting 40 kg, who was referred to EUS FNAB from GHN due to a dense lesion in the pancreas found by TUS when searching for the cause of increased liver function tests (AST 1.19 µkat/L [normal up to 0.52 µkat/L], ALT 1.19 µkat/L [normal up to 0.52 µkat/L], normal gama-GT, alcaline phosphatase and bilirubin) performed due to nonspecific disease signs (fatigue, nausea, headache).

EUS showed a round, well circumscribed, isoechogenic lesion in the pancreatic neck with a diameter of 15 mm (Figures 1, 2A). The cell sample was highly cellular, with abundant eosinophilic stroma, around which the tumour cells were arranged in rounded clusters (Figure 2B–C). The individual cells had moderate, basophilic cytoplasm. The nuclei were round to elongated, with grooves, the chromatin was pale with incospicuous nucleoli (Figure 2D–E). The background of the specimen contained blood and a few siderophages. Only a few tumour cell nuclei were positive for the proliferation marker Ki67 (less than 1%), the rest of the immunocytochemical reactions were consistent with the final cytopathological diagnosis of SPN (Table 1) (Figure 2F). Additional immunohistochemical reactions revealed positive results for Vimentin, CD56, CD10, and NSE, while E-Kadherin showed a negative result.

FIGURE 1.

FIGURE 2.

Eighteen days after EUS FNAB, a partial resection of the pancreas with the lymph node dissection at a. hepatica communis and cholecystectomy followed. The frozen section and definitive specimens showed a solid pseudopapillary tumour confined to the pancreas, measuring 18×10 mm in the largest diameter, lobulated, homogeneous, removed 3 mm from the pancreatic surgical margin (Figure 2B). There was no perineural or lymphovascular invasion, and no tumour tissue was present in the resected lymph node. The patient reported no problems at the outpatient follow-up, the weight gain was adequate, postoperative US showed no additional abnormalities and a fecal elastase, which assesses exocrine pancreatic function, was normal. Thirty-nine months after the EUS FNAB she had acute pancreatitis, which resolved with conservative treatment. Magnetic resonance imaging of the pancreas and abdominal MR with contrast showed no recurrence. At the last follow-up 50 months after the initial diagnosis, she had no medical complaints.

The second case was a 7-year-old female patient, weighting 43 kg, who came to EUS FNAB from the GHN where she presented with abdominal pain. A pancreatic cystic lesion was noted on TUS exam, which could represent a tumour or pseudocyst. EUS showed a 40×40 mm hypoechogenic tumour in the head of the pancreas, pressing the portal system and causing dilatation of the ductus choledochus to 9 mm (Figure 3A). On ROSE the EUS FNAB samples were all non-diagnostic (only blood), except for one pass, where numerous lymphocytes, histiocytes and a necrotic background was retrieved. In the filtered and concentrated sample, there were few preserved tumour cells. They were monomorphus, small, with a round, regular nuclei, single nucleoli and focally spickled chromatin (Figure 3D). The cytoplasm was coarsely granular. In the background, poorly preserved branched capillaries, surrounded by predominantly necrotic cells were visible (Figure 3C). According to the immunochemical reactions in the rare preserved cells, the cytopathologic diagnosis was SPN (Table 1). Thirteen days after EUS FNAB, a partial resection of the pancreas with duodenum, excision of the lymph node adjacent to the a. hepatica communis and cholecystectomy followed. Histopathologic examination confirmed SPN with extensive necrosis of the pancreatic head, with the largest diameter of 30 mm, at a distance of 0.1 mm from the nearest retropancreatic surgical margin (Figure 3B). There were perineural as well as lymphatic and vascular invasions (Figure 3E). No involvement with tumour cells was present in the excised lymph node and 6 regional lymph nodes. The gallbladder and duodenum were un-remarkable. Given the extensive tumour necrosis and lymphovascular invasion, the report mentioned the possibility of a more aggressive clinical course. Postoperative follow-up was uneventful. TUS showed changes consistent with the post-operation state. The stool elastase was reduced. Despite the absence of clinical signs, pancreatic enzyme therapy was initiated to ensure adequate nutrient resorption. She has gained considerable weight and has grown appropriately. The girl was presented to the haemato-oncology multidisciplinary team, which advised MRI of the abdomen every 3 months for the first 2 years and then every 6 months for up to 5 years, which she regularly undergoes. At the last follow-up, 31 months after the initial diagnosis, she had no complaints, the laboratory results and the abdomen MRI showed no signs of disease recurrence.

FIGURE 3.

The third case, which ultimately did not involve a pancreatic tumor, pertained to a 17-year-old boy with trisomy 21, weighing 46 kg. He exhibited an enlarged and inhomogeneous pancreatic head on TUS and magnetic resonance cholangiopancreatography following treatment for biliary pancreatitis that resolved without the need for endoscopic retrograde cholangiopancreatography. EUS-FNAB revealed no tumor cells. Consequently, a pancreatic biopsy was recommended, but the patient’s mother declined. Subsequent TUS follow-ups showed no evidence of tumor growth or disease progression.

Given the patient’s concurrent mild exocrine pancreatic insufficiency (faecal elastase 88 µg/g, normal > 200 µg/g), pancreatic enzyme replacement therapy was initiated. However, a more comprehensive diagnosis of the pancreatic disease, such as autoimmune or hereditary pancreatitis, proved unattainable as the family withheld consent, and the patient failed to attend follow-up visits.

The use of EUS FNAB in the paediatric population is limited, even though it is a well-established diagnostic procedure for the assessment of suspicious pancreatic tumours in adult patients. In paediatric cases, the primary indication for EUS-guided pancreatic tissue sampling is pancreatic mass or suspected autoimmune pancreatitis. These conditions often manifest as inhomogeneous lesions of the pancreas, detectable through TUS or EUS.³ Our retrospective study spanned the whole period since the introduction of EUS FNAB in the tertiary heath care centre in 2010 until the end of the review period in December 2021. During that period, we had the first case of EUS FNAB in a child in 2019 followed by 5 more cases with complete cytopathological reports and corresponding archival slides. Among the total of 6 children with EUS FNAB of gastrointestinal lesions, two had pancreatic neoplasm, diagnosed as SPN by cytopathological examination.

According to the reviewed literature (Table 2), the adoption of EUS for children with pancreatobilliary diseases was delayed due to different reasons, among them the low incidence of tumorous lesions and even rarer malignancies.⁸ Other important reasons were also the size limitation of EUS equipment relative to pediatric anatomy, the low number of skilled pediatric endoscopists with EUS expertise, the need for sedation or general anesthesia, and a limited awareness among pediatric practitioners of EUS diagnostic and therapeutic possibilities.^3,8 In our patients, the procedure and equipment used was similar to the adult population, apart from the use of general anaesthesia. The endoscopists were very experienced in the management of adult patients, enabling the procedure to be safe and efficient also in the pediatric patients, which corresponds well with other studies.^{1,3,8,9,10,11,16,17,18} While EUS was shown previously as a safe diagnostic method in the pediatric population, covering the age range of our patients (age 8 and 9 years, respectively), there is limited experience with EUS especially in smaller children (< 15 kg) due to the fear of an esophageal rupture associated with the large diameter (adult) EUS scope. Therefore, for smaller children (< 15 kg) endobronchial ultrasound has been used instead and a few studies have shown that EUS can be performed safely and with a high diagnostic accuracy even in these children.^{1,3,8,9,10,11,16,17,18}

The initial diagnostic radiologic assessment of pancreatic lesions is usually conducted by a TUS, which was also the approach in our two paediatric patients. In one patient, a solid isoechogenic lesion suspicious for tumour was found in the neck of the pancreas. For radiologically solid lesions, the differential diagnoses include autoimmune pancreatitis, neuroendocrine tumours, microcystic serous adenoma, pancreatoblastoma and acinar cell carcinoma.¹⁹ Autoimmune pancreatitis can also closely mimic radiologic impression of a tumour, so tissue acquisition for light microscopic morphological evaluation is a crucial step in reaching a diagnosis.^1,3,16 However, a cytologic impression can also be false positive.¹³

In the second patient, a pancreatic cystic lesion was noted on TUS, which was interpreted as either a tumor or a pseudocyst. Cystic lesions of the pancreas may represent benign or malignant processes and are identified incidentally during abdominal cross-sectional imaging or TUS performed for other indications.^3,8 The main radiologic differential diagnoses in the case of predominantly cystic lesions in children are pseudocysts after a blunt pancreatic injury, retention cysts in cystic fibrosis patients, congenital anomalies (duplication cyst), a part of syndromes like von Hippel Lindau or a tumour pathology of a cystic appearance, such as SPN, neuroendocrine tumours, macrocystic serous adenoma, or mucinous cystic neoplasm.^3,8,19

SPN is a low-grade tumour that mainly affects younger women and, to a lesser extent, children. It accounts for up to 30% of pancreatic tumours in patients under 40 years of age.²⁰

It may be clinically silent as in case 1 or presents as an abdominal pain for its mass effect as in case 2, or rarely as jaundice.²⁰ Usually it does not have an endocrine function and the tumorous markers are negative.²⁰

The radiologic imaging in our cases covered the ends of the spectrum of the adult SPN presentation which is illustrated by its old name of “solid-cystic tumour” reflecting its radiological picture of a solid tumor with necrotic cystic areas, which are more common in larger tumours.²⁰ However, we present a pediatric case of one completely solid and one almost completely cystic lesion, without the predominant solid-cystic appearance. Another difference was also the location of the lesions, with one located in the pancreatic head and one in the pancreatic neck, as opposed to the usual body or tail presentation.²⁰

The demographics matched the incidence of pediatric SPN with exclusively female patients, the youngest being 8 years old, as is also the case in the literature.²⁰ Nevertheless there are anecdotical cases of tumors in pediatric males.^21,22,23

With EUS one gains the ability to better characterise pancreatic cystic and less common solid lesions also in children and, when performed in combination with FNAB, the discrimination between the various types of cysts and solid tumors can be made, that my direct treatment from a surgical resection to chemotherapy.¹³ In our cases, ROSE was used to assure diagnostic cell samples and a specific diagnosis of SPN could be made in both cases based on cytomorphology supplemented by relevant immunocytochemical stainings.

The morphological and immunocytochemical features of SPN in paediatric population don’t differ from adult population. In the cytopathological specimens, SPN comprises pseudopapillary clusters, acinar groups and single cells, delicate capillaries and globules of amorphous myxoid material which was well presented in case 1, but not in case 2 which was mainly necrotic.^24–25 Histomorphologically and cytomorphologically, the tumour is most commonly composed of unimorphic cells, with sparse to moderately abundant cytoplasm, which may be pale or oncocytic, with vacuoles and round to oval nuclei, with grooves and a fine chromatin structure, which also presented in our cases.^24–25 There is however an exception, where a few large pleomorphic atypical multinucleated giant cells could be intermingled with more typical ones, which was not found in our cases.²⁵

Main cytopathologic differential diagnosis of SPN in children depends on the age, comprising neuroendocrine neoplasms in patients older than 10 years and a pancreatoblastoma in younger patients.³ In comparison to neuroendocrine neoplasms, SPNs have hyaline-mucinous stroma, cytoplasmic vacuoles, nuclei with grooves and negative reactions to chromogranin and cytokeratin.^20,26,27 Pancreatoblastoma, which occurs in younger children (median age 5 years) combines several types of differentiation, namely acinar, endocrine and ductal with morulae, with corresponding immunoprofile. Cells are positive for cytokeratin AE1/AE3, BCL 10, neuroendocrine markers, EMA, and have PAS-D positive granules in the cytoplasm.^20,26,27

For the concordant final cytopathological diagnosis of SPN, a combination of cytomorphological and immunohistochemical findings is widely recommended.²⁶ To differentiate between neuroendocrine tumours, the recently proposed β-catenin, CD10, and PR were used, but not CD99, which is not specific for SPN.^19,26 Instead, the CyclinD1, chromogranin and an additional, SOX11 were utilised (Table 1). The clue marker β-Catenin showed a consistent reaction between cytopathological (cell block) and resection samples. The differential diagnosis of pancreatoblastoma was dismissed with a negative CKAE1/AE3 reaction.

A diagnostic challenge arises because of the positive nuclear reaction to β-catenin, which is the marker of a somatic point mutation of the β-catenin gene (CCTNB), a driving mutation in more than 90% of SPN, but also present in most pancreatoblastomas and in 15% of pancreatic neuroendocrine neoplasms of higher stage (III/IV) but not in lower stage neoplasms.^19,20 That is the reason why the adjunctive immunohistochemical staining to SOX11 is recommended, which is positive in SPNs and negative in normal pancreatic tissue and in neuroendocrine neoplasms. Additionally, it does not stain background cells and it does not have non-specific membranous or cytoplasmic staining as opposed to β-catenin.²⁴

In general, the histopathological features make it difficult to predict the future behaviour of SPN, and radical resection is recommended in all cases, usually followed by an excellent prognosis, as in our cases.²⁰ A more aggressive course is expected in tumours with high-grade malignant transformation, a higher number of mitoses and severe nuclear atypia.²⁰ Vascular, lymphatic and perineural invasion or infiltrative growth into surrounding structures are not yet associated with a higher likelihood of relapse, as is also consistent with our second case.²⁰ Overall, disease recurs in up to 15% of cases, most commonly in the local lymph nodes, liver and peritoneum, possibly years after the removal of the primary tumour and the metastasis.²⁸ Compared to the adult population, the paediatric population with SPN has a better survival rate at all stages, despite the same treatment.^18,20, ²³ The 5-year survival rate of paediatric patients after completely resected SPN is 95%.⁵ In our institution both cases are without recurrence with a 31 and 50 months follow-up.

The sensitivity of EUS FNAB in diagnosing SPN in the adult population is estimated to be above 80%.⁹ In our institution, EUS-guided fine-needle aspiration biopsy (EUS FNAB) in diagnosing solid pseudopapillary neoplasms (SPN) in pediatric patients was so far accurate, based on two documented cases. In cases with necrotic samples, as in case 2, cytopathology specific techniques like filtration of cell sample to discard necrotic debris, and concentration are very helpful in detecting any individual viable cells, that can further undergo immunochemical stains. Unfortunately, necrosis and degenerated samples may also show a significant number of non-specific positive or false negative immunochemical reactions, making the diagnosis less reliable. The techniques that enable selection of viable diagnostic cells are not available in histology, which is yet another advantage of EUS FNAB.

The main drawback of the present study is the small sample size, connected to the rarity of the neoplasm, that could be avoided only by a multi-centric study.

Solid pseudopapillary neoplasm (SPN) is a low-grade tumour that mainly affects young women and very rarely female pediatric patients. In the study, we documented two cases, successfully diagnosed by EUS FNAB, without any complications regarding the procedure. Main differential possibilities, neuroendocrine tumours and pancreatoblastomas, were both excluded by several immunochemical reactions, essential for the correct diagnosis. The caveat of one almost completely necrotic sample was overcome with an aid of filtration, the cytopathology specific technique, retrieving the sparse viable cells.

To sum it up, the diagnosis of SPN in the paediatric population requires a high level of suspicion and good collaboration of all specialties (paediatricians, gastroenterologists, radiologists, pathologists and surgeons) due to the wide spectrum of symptoms, the variety of radiological presentations and (cyto)pathological morphology. EUS FNAB proved to be a safe and efficient technique even at paediatric population.