Long-term survival following resection of a solid pseudopapillary tumour of the pancreas with arterial and venous encasement

Abstract

Solid pseudopapillary neoplasms (SPNs) of the pancreas (also known as Franz tumours) are rare, representing 1%–2% of pancreatic tumours. They are of exocrine gland origin and are more prevalent in women under the age of 30. They are classified as low-grade malignancies and are usually treated with surgery.

We present the case of a young woman who was found to have a pancreatic head lesion obstructing the superior mesenteric vein and encasing the superior mesenteric artery. Diagnosis of SPN was confirmed on histology. Initially deemed unresectable due to vascular encasement, the patient received neoadjuvant chemotherapy to downstage the tumour. This was unsuccessful. Given the patient’s young age and lack of other options, surgery was performed. Instead of vascular resection, the tumour was split and peeled off the superior mesenteric artery giving an R1 resection. The patient has achieved long-term recurrence free survival at 10 years.

Background

Solid pseudopapillary neoplasms (SPNs) are uncommon pancreatic tumours that often present as an incidental radiological finding. The lack of symptoms mean they are often detected at larger sizes (mean 6–8 cm).¹ Uncertainty has surrounded the malignant potential of these tumours, and this is reflected in the evolving terminology. Initially, SPNs were described as a group of borderline tumours. In 2010, the WHO categorised the tumours as malignant in nature.²

The recommendation for management of SPNs is complete surgical resection. Survival rates approach 96% at 10 years after complete resection.^{1 3} There is little reported on cases involving the vasculature of the pancreas and the outcomes of these tumours following resection with vascular involvement or positive surgical margins.

This case report is important as it demonstrates that vascular involvement may not be contraindication to surgical management and good results can still be achieved with appropriate operative planning and expertise.

Case presentation

A 28-year-old woman presented to the emergency department with severe right upper quadrant pain and elevated liver function tests. She was not jaundiced. Her pain resolved with analgesia and she underwent outpatient imaging. An ultrasound confirmed cholelithiasis to be the likely source of her pain. An incidental solid/cystic mass measuring 74 mm × 63 mm × 67 mm was visualised in the head of the pancreas. This was causing mild dilatation of the extrahepatic and intrahepatic biliary tree.

A computer tomography scan (CT) of the abdomen demonstrated a heterogeneous mass in the head of the pancreas, encasing the superior mesenteric artery (SMA) over a distance of 5 cm and occluding the superior mesenteric vein (SMV) (see figures 1 and 2). There was no CT evidence of cirrhosis of the liver. Multiple enlarged lymph nodes at the base of the mesentery and aorta were seen. The CT chest was clear of metastatic disease. Referral was made to our specialist hepatobiliary unit for ongoing management.

Figure 1.

Axial CT of 7 cm solid pseudopapillary neoplasm with complete encasement of the superior mesenteric artery.

Figure 2.

Coronal CT of solid pseudopapillary neoplasm with obstruction of the superior mesenteric vein and encasement of the superior mesenteric artery.

The patient underwent endoscopic ultrasound (EUS), confirming the radiological findings of a hypoechoic, heterogeneous pancreatic head lesion measuring 60 × 60 mm. The portal vein was patent, filling via collaterals. Established varices were present. The SMV and spleno-portal confluence were obliterated, and SMA was encased by the tumour. Fine-needle aspiration (FNA) demonstrated loosely cohesive epitheloid cells displaying moderate anisonucleosis with round to oval nuclei.

Immunohistochemistry was positive for vimentin, CD10 and beta catenin. Immunostaining for E-cadherin was negative. These findings were consistent with solid pseudopapillary neoplasm of the pancreas.

Investigations

Liver function tests and tumour markers were as follows: alanine aminotransferase 110 U/L, aspartate aminotransferase 43 U/L, alkaline phosphatase 142 U/L, gamma-glutamyl transferase 49 U/L, lipase 320 U/L, carcinoembryonic antigen <0.5 μg/L, total serum bilirubin 5 μmol/L.

Differential diagnosis

There is yet to be defined radiological diagnostic criteria for SPNs. Common radiological features include large masses with heterogeneous, cystic or solid components with or without calcification.⁴ They can be similar in appearance to other cystic tumours of the pancreas including serous cystadenoma, mucinous cystadenoma, cystic neuroendocrine tumours or pancreatic pseudocysts. This diagnosis of SPN must be considered in female patients under the age of 40 presenting with a cystic lesion in the pancreas.

Treatment

After multidisciplinary discussion, due to the extent of vascular involvement and the likelihood of an R1 resection, a decision was made to consider neoadjuvant chemotherapy in an attempt to downstage the tumour. There were only anecdotal reports in the literature regarding suitable chemotherapeutic agents to treat SPN.⁵ The patient was treated with three cycles of gemcitabine chemotherapy and tolerated this well. She developed further derangement of the liver function tests and a reduced dose was delivered.

A CT scan after three cycles of chemotherapy demonstrated no reduction of tumour volume, but no progression. Chemotherapy was therefore discontinued and surgical resection considered. The plan was to do a Whipple’s operation. Given that the SMA was entirely encased over a long distance, cadaveric artery was obtained from the tissue bank with the aim to resect the affected segment of SMA and replace it with an interposition graft. There was no plan to reconstruct the SMV as there was already well-established collaterals present, decompressing the gut through the inferior mesenteric and splenic veins.

During the operation, control of the SMA was obtained both proximal and distal to the tumour. Because there appeared to be a clear plane of cleavage between the tumour and the artery, the decision was made to split the tumour and preserve the native SMA rather than risk performing an interposition graft. We accepted that it was not optimal to breach the tumour; however, we decided this was a balance of risks.

As planned, the SMV was not reconstructed as there was excellent venous return via the inferior mesenteric vein.

The patient spent 4 weeks in hospital requiring total parenteral nutrition due to delayed gastric emptying. She developed chylous ascites that needed peritoneal drainage. She suffered with diarrhoea for 6 months, likely a combination of pancreatic insufficiency and the ramification of neural skeletonisation of the SMA. This was controlled with pancreatic enzymes and loperamide.

Macroscopic analysis of the sample revealed a solid haemorrhagic mass encompassing most of the pancreatic head measuring 70 mm × 65 mm × 65 mm. Microscopic sections revealed a solid pseudopapillary neoplasm with invasion into pancreatic parenchyma and peripancreatic fat. Distal to the lesion, perineural invasion was evident with 1 of 19 lymph nodes positive. Immunohistochemistry was positive for vimentin, CD10 and beta catenin. Immunostaining for E-cadherin was negative. This was an R1 resection with broad involvement of the uncinate process and the neck of the pancreas margin.

Outcome and follow-up

The patient returned to full-time work after 6 months and has been seen at regular intervals with repeat CT scans. She did not continue with chemotherapy postoperatively. Despite positive margins, the patient remains well and recurrence free at the 10-year review.

Discussion

SPNs were first recognised in 1959 by Virginia K Franz in three patients with tumours displaying similar characteristics.⁶ They are a rare pancreatic lesion, accounting for 1%–2% of pancreatic neoplasms.⁷ They occur with a female predominance of 10:1, between the ages of 20 and 40. Some researchers believe that hormonal status may play a role in pathogenesis.⁵

There seems to be no predilection for a particular site in the pancreas.⁸ The location of the tumour influences the type of surgical treatment, with head of pancreas tumours requiring the higher morbidity Whipple’s operation.⁶

Due to the pancreas’ posterior position in the abdomen, SPNs may grow to a considerable size before they cause symptoms. Up to 50% of tumours present as an incidental finding on imaging. Papavtamidis reports the most common symptom is abdominal pain, followed by an epigastric mass, nausea, anorexia and vomiting. Some patients may present with compression of surrounding structures leading to jaundice. Late presentations include metastatic disease or haemoperitoneum from ruptured vasculature.⁹

The most common sites of metastatic disease reported in the literature include: liver, portal/splenic/SMVs, spleen, diaphragm, omentum, peritoneum, duodenum and stomach.^{6 8}

Since their discovery, there has been debate regarding the history and malignant potential of SPNs. Initially, SPNs were identified as adenocarcinomas or cystadenomas, among other pancreatic lesions. In 1996, the WHO classified SPNs as a pancreatic cystic neoplasm, behaving in a borderline malignant fashion. However, this definition cast doubt on the metastatic potential of the tumour.¹ In 2010, the lesions were reclassified as low-grade malignant lesions due to their demonstrated ability to display malignant characteristics, namely perineural and deep tissue invasion.^{7 8}

There is no consensus on malignant features of SPNs to help guide treatment options. There is some agreement that malignant features include larger tumour size >5 cm and histological characteristics (nuclear atypia, mitotic activity).^{7 10} Kim et al observed that tumour size >5 cm was associated with a high-grade behaviour.⁷ Given the morbidity of pancreatic surgery, future studies will be important to determine if lesions with more favourable characteristics may be treated with a watch-and-wait approach.

There are no gold standard imaging characteristics of these lesions that differentiate them from other pancreatic tumours. Diagnosis should be strongly considered when a cystic lesion of the pancreas occurs in a younger woman. MRI will demonstrate an encapsulated mass with cystic and solid components, making it a sensitive scan to rule out other pathologies such as adenocarcinoma.

EUS and FNA are useful in providing a preoperative histopathological diagnosis.¹¹ SPN is positive for markers CD10, beta catenin (86.1% expression) and vimentin (90.2% expression).⁴

Aggressive surgical treatment remains the gold standard treatment for SPN. The operation will vary depending on tumour location. The validity of extensive lymph node dissection not clear due to the relatively low prevalence of lymphatic metastasis.^{7 10}

Additional therapy for metastatic disease is limited to case reports and is determined by the location of the metastases. There are few reports on the use of adjuvant chemotherapy and even less so reporting use of neoadjuvant chemotherapy as a downstaging tool. Of note, these articles have focused on the use of combination chemotherapy, gemcitabine and capecitabine.¹² It is unlikely we would offer neoadjuvant chemotherapy again given the lack of tumour response unless more data relating to neoadjuvant chemotherapy becomes available.

Overall, the prognosis of SPNs with or without metastatic disease is favourable. Surgical resection provides excellent outcomes for localised disease. We have demonstrated a favourable long-term outcome after aggressive surgical resection for a locally advanced tumour with vascular involvement and nodal disease. We took a calculated risk by cleaving the tumour to preserve the SMA and this accounted for our positive margin. The tumour had a pushing rather than invading edge, and it was this feature that made cleavage possible, avoiding the ischaemic and thrombotic risks associated with en bloc tumour resection and SMA grafting. In hindsight, this positive margin has not affected local recurrence and with the indolent biology of this disease it was a risk worth taking.

Patient’s perspective.

The best way I could describe my whole experience of dealing with the Franz tumour, from diagnosis to the treatment and subsequent surgery and recovery, would be tumultuous and surreal.

Sometimes it felt like I wasn’t even directly part of the process, that I was viewing it as a spectator, instead of the patient. In hindsight, I realise that that was most likely a coping mechanism I used to try to deal with everything going on at the time. I remember trying my best to laugh and joke my way through it all—all the tests, all the medications and procedures, and being very aware throughout that my chances of survival were not great by any means. My surgeon was clear and open about there not being a huge amount of data or information about this particular tumour, and that a lot of the suggested treatment options were a kind of new territory, so to speak.

It was the biggest risk I had ever taken in my life up to that point, going ahead with the surgery, but it turned out well in the end. I still have ongoing issues that will likely never go away, but I can now manage them relatively easily. I feel very fortunate that I had the medical team that I did have, and that I had such a strong support network around me.

Learning points.

• Solid pseudopapillary neoplasms (SPNs) may demonstrate a spectrum of malignant potential.

• In this case, neoadjuvant chemotherapy did not downstage the disease prior to surgery.

• Locally advanced tumours with vascular encasement may be considered resectable.

• Further clinical research should aim to understand the natural history of SPNs to optimise treatment for patients.

• Despite the incidence of SPNs appearing to increase, little is known about their true natural history and whether a regimen of careful observation may be considered for small tumours rather than immediately resorting to aggressive surgical management. It is important that ongoing publication of these cases occurs.

Ethics statements

Patient consent for publication

Consent obtained directly from patient(s).