Pancreaticopleural fistula in a Thai boy with SPINK1 c.101A>G substitution variant-related chronic pancreatitis: a case report and literature review

A 10-year-old boy presented at a community hospital due to intermittent epigastric pain and weight loss of 8 kg in a month. An initial chest radiograph showed massive left pleural effusion (Figure 1). He underwent thoracic drainage. Pleural fluid analysis showed exudative profile, and was positive for anti-nuclear antibodies (ANA). He was treated with intravenous antibiotics and retained chest tube drainage for 6 days, the left pleural effusion had decreased. However, 4 days later, he complained of chest tightness and dyspnea, and chest radiography showed massive right pleural effusion. He was again treated with right chest tube drainage. Contrast-enhanced chest CT revealed bilateral pleural effusion (Figure 2A). Therefore, the patient was referred to Songklanagarind Hospital, a tertiary care, and teaching hospital of the Faculty of Medicine, Prince of Songkla University, Hat Yai, to determine the cause of pleural effusion. We obtained written informed consent in writing from the patient's parent to publish the present case report and associated images, and the CARE reporting guidelines were used when writing the report [17]. The patient information for this case was anonymized as far as possible to avoid identification of the patient or their family.

Figure 1

Figure 2

The patient's height was 145 cm and weight was 35 kg. He had no known underlying disease and no family history of pancreatitis. Physical examination revealed decreased breath sound on the right hemithorax. He had elevated serum amylase of 879 U/L and lipase of 817 U/L. Fasting lipids, blood glucose, and calcium levels were within normal range. The pleural fluid analysis revealed an elevated amylase level of 4079 U/L, but negative for ANA. Polymerase chain reaction (PCR) testing for tuberculosis, and bacterial cultures were negative for pathogenic organisms. Serology for autoimmune and connective tissue disease including anti-ds DNA, anti-cardiolipin, anti-B2-glycoprotein, and anti-Sm (Sm, small nuclear ribonucleoproteins autoantibodies) was all negative. The chest CT from the community hospital was reviewed and we found features of chronic pancreatitis including irregular pancreatic duct dilatation with diffuse atrophic pancreatic parenchyma, and a small pancreatic pseudocyst (Figure 2B). A diagnosis of pancreatitis was established and the patient was conservatively treated with fasting, parenteral nutrition, antibiotics, and octreotide for 10 days. The thoracic drainage was eventually removed. Subsequent MRCP performed on day 21 revealed evidence of chronic pancreatitis seen as thinning of the pancreatic parenchyma with irregular dilated upstream pancreatic duct and a focal disruption of the duct at the pancreatic genu-body connecting to a large pancreatic pseudocyst which extended upward to the mediastinum (Figure 3A, B). Hence, we diagnosed PPF complicated by chronic pancreatitis. Then endoscopic ultrasound (EUS)-guided gastrocystostomy pseudocyst drainage was successfully performed by placing two 7-Fr, 5-cm double pigtail stents between the body of the stomach and the pancreatic pseudocyst. Subsequently, the patient was discharged from the hospital and scheduled for a followed-up out-patient appointment at 1 month later. However, he was lost to follow-up at his out-patient appointment. A chest-abdominal CT performed on day 90 showed interval increase in the size of the pancreatic pseudocyst. Three weeks later, the patient underwent an endoscopic retrograde cholangiopancreatography (ERCP), which confirmed the MRCP findings of ductal disruption with a disconnected pancreatic duct genu communicating with the pseudocyst (Figure 4A). However, the endoscopic interventionist was unable to pass the ERCP wire distally beyond the disrupted pancreatic duct. A month later, a pancreatic sphincterotomy with rendezvous pancreatic duct intervention was performed. Pancreatography via the punctured upstream pancreatic duct revealed a complete cut-off at the duct genu with failure to cannulate the duct downstream (Figure 4B). Thus, a 7-Fr, 7-cm double pigtail stent was placed across the fistula between the duct body and the stomach, creating pancreaticogastrostomy fistula. A rendezvous-assisted endoscopic retrograde pancreatography (ERP) via the matured fistula was performed 6 weeks later. A 10-Fr, 7-cm straight biliary stent was successfully inserted across the disrupted duct via the duodenal papilla. The patient made an uneventful recovery and was discharged from the hospital 5 days later. A follow-up CT 2 months later showed complete resolution of the pseudocyst and pleural effusions. Four months later, a follow-up ERP showed complete anatomical restoration of the pancreatic duct, and the duct stent was retrieved (Figure 4C). The patient was well and symptom-free for at least a further 10 months of follow up since successful placement of a stent in the duct. The total length of care at our institution was about 426 days. The episode of care was summarized in Figure 5. During this time, genetic testing was conducted, and we found a heterozygous missense substitution in SPINK1, c.101A>G (p. Asn34Ser), but otherwise negative for PRSS1 variants. Tests for other possible genetic variations related to chronic pediatric pancreatitis such as in CFTR, CTRC, or the gene for calcium sensing receptor (CASR) are not available in our institution. Family history investigations revealed no history of pancreatic diseases.

Figure 3

Figure 4

Figure 5

PPF is a rare cause of pleural effusion in children, mostly reported in the literature as case reports [1, 2, 8,9,10,11,12,13,14,15,16]. The pathophysiology of PPF involves the pancreatic fluid leaking directly from a focal pancreatic duct disruption or a pancreatic pseudocyst and moving upward into the thoracic cavity because of the pressure gradient between abdominal and thoracic cavities [18, 19]. This fluid can leak along the retroperitoneum, reaching toward the mediastinum or the diaphragm (Figure 6A, B) [18, 19]. Sometimes, the duct ruptures anteriorly and forms a pseudocyst in the lesser sac and the pancreatic juice then erodes the nearby diaphragm (Figure 6C) [19]. The causes of duct disruption can be direct trauma, or a late consequence of acute or chronic pancreatitis [2]. Focal pancreatic inflammation can result in a stricture, a protein plug, or stones in the duct, leading to increasing intraluminal pressure and disruption of the distal duct (Figure 6D) [18]. Diagnosis of PPF is established by the presence of an amylase-rich pleural effusion combined with direct visualization of the fistula, evidence of chronic pancreatitis or a pseudocyst in imaging [2, 20]. The diagnosis of PPF is often delayed because of the low incidence or lack of awareness of this abnormality [2, 20].

Figure 6

In addition to our case, we reviewed 12 other cases of PPF in children reported in the English literature 2001–2020 [1, 2, 8,9,10,11,12,13,14,15,161]. The patients’ ages ranged from 1.5 years to 13 years. All patients had presented with respiratory symptoms or vague abdominal pain with massive or intractable pleural effusion(s). The pleural effusion amylase content ranged from 950–157,000 IU/L. The details of the imaging findings in each case are shown in Table 1, and number of the abnormalities detected by imaging are provided in Table 2 (n = 13). Most PPFs could be identified by extension to the mediastinum (6 cases) or diaphragm (5 cases). A pseudocyst without demonstrable fluid tract had been seen in only 1 case of the 13 (including ours) [8]. In our patient, MRCP was the most useful imaging to depict all pancreatic duct pathologies including PPF, duct disruption, irregular duct dilatation, proximal duct stricture, and stones. These findings by MRCP were made in 6, 3, 7, 2, and 2 of 7 other cases, respectively (Table 2). Nevertheless, CT provided the first imaging clues leading to diagnosis of PPF in our patient, namely evidence of chronic pancreatitis and a pseudocyst. Although the fistula was detected by CT in 7 of 12 cases, the tomography demonstrated some features of chronic pancreatitis including irregular dilated pancreatic duct in 8 cases, pancreatic calcifications in 4 cases, and pancreatic pseudocysts in 7 cases (Table 2). CT is more available and more practical for children than MRI, which usually requires sedation and pediatric-imaging specialized operators. Another advantage of CT is that when a child in whom PPF is suspected is scheduled for a chest CT to evaluate pulmonary conditions, the scan length can inferiorly extend to cover the pancreas. The initial information regarding PPF and/or chronic pancreatitis received by pancreatic CT will help determine whether the patient is able to proceed to specific treatments of PPF, or to undergo further investigation with MRCP. When compared to MRCP, ERCP had lower detection rate of the fistula, at 5 cases in 9, but produced better visualization of duct stricture and stones (Table 2). The diagnostic rate of PPF by ERCP was highly variable depending on the time of examination and anatomical variations [1, 2, 9, 10, 12, 15, 16]. The reported causes of failed ERCP in children with PPF included pancreas divisum, associated downstream duct stricture, and duct stone [1, 10, 12]. Ultrasound has only a small role in PPF, with detection rates for PPF of 0 in 6 cases, irregular dilatation of the duct in 3 cases, and pancreatic calcifications in 2 cases (Table 2).

Management of the fistula should be tailored on the morphologies of the pancreatic duct [2]. If the duct looks normal or is mildly dilated without stenosis, conservative treatment may be adequate. In cases of downstream ductal stenosis, partial ductal disruption in the pancreatic head or body, complete ductal disruption, ductal obstruction proximal to fistula, or impacted duct stones, endoscopic or surgical procedures should be considered [2].

Chronic pancreatitis is the main etiology of PPF in children and genetic variations accounts for 73% of cases of chronic pancreatitis in children [2, 3]. One of the widely known pancreatitis-related variations is in SPINK1 [3]. The SPINK1 c.101A>G substitution variant, which is more frequently mentioned as the gene associated with chronic pancreatitis in patients with European ancestry than in patients with Asian ancestry, was, however, found in our Thai patient [3]. Other variants of SPINK1 that have been previously reported are a c.194 + 2T>C substitution variant in a Chinese boy with PPF and a IVS3 + 2T>C slice-site substitution variant in many Chinese children with chronic pancreatitis [2, 3]. Limitations of this report include lack of some specific genetic tests related to chronic pancreatitis at Songklanagarind Hospital. Therefore, multifactorial genetic factors contributed to chronic pancreatitis could not be determined in our study. In summary, the imaging and management guidelines for suspected cases of PPF are shown in Figure 7.

Figure 7

In children who present with massive pleural effusion with high levels of pleural fluid amylase, MRCP or pancreatic CT should be promptly conducted to enable the early diagnosis of PPF, which will reduce complications and shorten hospitalization in cases of pediatric PPF.