Pediatric Surgery International

, Volume 29, Issue 3, pp 305–310

Non-operative treatment versus percutaneous drainage of pancreatic pseudocysts in children


    • Department of SurgeryUniversity of Utah
  • D. C. Barnhart
    • Division of Pediatric SurgeryPrimary Children’s Medical Center, University of Utah
    • Department of SurgeryUniversity of Utah
  • J. Madden
    • Department of SurgeryUniversity of Utah
  • E. Leeflang
    • Department of SurgeryUniversity of California San Diego
  • W. D. Jackson
    • Pediatric GastroenterologyPrimary Children’s Medical Center, University of Utah
  • G. P. Feola
    • Pediatric Interventional RadiologyPrimary Children’s Medical Center, University of Utah
  • R. L. Meyers
    • Division of Pediatric SurgeryPrimary Children’s Medical Center, University of Utah
    • Department of SurgeryUniversity of Utah
  • E. R. Scaife
    • Division of Pediatric SurgeryPrimary Children’s Medical Center, University of Utah
    • Department of SurgeryUniversity of Utah
  • M. D. Rollins
    • Division of Pediatric SurgeryPrimary Children’s Medical Center, University of Utah
    • Department of SurgeryUniversity of Utah
Original Article

DOI: 10.1007/s00383-012-3236-x

Cite this article as:
Russell, K.W., Barnhart, D.C., Madden, J. et al. Pediatr Surg Int (2013) 29: 305. doi:10.1007/s00383-012-3236-x



The objective of this study was to characterize the clinical course and outcomes of children with pancreatic pseudocysts that were initially treated non-operatively or with percutaneous drainage.


A retrospective review of children with pancreatic pseudocysts over a 12-year period was completed. Categorical variables were compared using Fischer’s exact method and the Student’s t test was used to compare continuous variables. Analysis was done using logistic and linear regression models.


Thirty-six children met the criteria for pancreatic pseudocyst and 33 children were treated either non-operatively or with percutaneous drainage. Of the 22 children managed non-operatively, 17 required no additional intervention (77 %) and five required surgery. Operative procedures were: Frey procedure (3), distal pancreatectomy (1), and cystgastrostomy (1). Eight of the 11 children treated with initial percutaneous drainage required no additional treatment (72 %). The other three children underwent distal pancreatectomy. Success of non-operative management or percutaneous drainage was not dependent on size or complexity of the pseudocyst Logistic regression did not identify any patient demographic (gender, age, and weight), etiologic (trauma, non-traumatic pancreatitis) or pseudocyst characteristic (size, septations) that predicted failure of non-operative therapy.


In children, pancreatic pseudocysts can frequently be managed without surgery regardless of size or complexity of the pseudocyst. When an intervention is needed, percutaneous drainage can be performed successfully, avoiding the need for major surgical intervention in the majority of patients.


Pancreatic pseudocystPediatricsNon-operative managementPercutaneous drainage


Pancreatic pseudocysts are cystic cavities that originate from the pancreas and lack an epithelial lining. In children, the development of a pancreatic pseudocyst is a relatively uncommon complication of acute pancreatitis, but may occur more frequently following pancreatic injury. After pancreatic trauma in children, pseudocysts occur up to 44 % of the time and suggest major ductal disruption [1]. When pseudocysts develop, interventional options include operative internal drainage, endoscopic internal drainage and external drainage Baron [17]. Non-operative management consists of pain control, nausea control, and nutrition either parenterally or with nasojejunal tube feeds. Small case series have also promoted the use of somatostatin to aid in pseudocyst resolution, but no large trial has validated any benefit [2].

The aim of this study was to characterize the course of children with pancreatic pseudocysts treated non-operatively or with percutaneous drainage. This characterization included the overall success rate of each strategy, factors predicting success or failure of therapy, associated complications, duration of treatment as well as overall resource utilization.


Study population

After obtaining IRB approval, a retrospective review of all children with the diagnosis of pancreatic pseudocyst (ICD-9 code 577.2) at a regional tertiary care children’s hospital from 1998 to 2010 was performed. Patients were included only if the peripancreatic fluid collection persisted at least 4 weeks beyond the time of initial diagnosis of pancreatitis or pancreatic injury. The 4-week time point was chosen so that only patients with true pseudocysts were included and patients with peripancreatic fluid collections alone were excluded Bradley [18]. The paper and electronic medical records of all patients meeting pseudocyst criteria were reviewed to obtain demographic data, pseudocyst characteristics, therapeutic measures and outcomes. Patients that were initially treated with surgery were removed from the study population. Remaining patients were grouped and evaluated based on their initial treatment: non-operative management or percutaneous drainage.

Once strict pseudocyst criteria were met, size was measured by largest diameter in centimeters. Additionally, a pseudocyst was considered complex if it had any septations. The decision to pursue operative treatment, external drainage or non-operative management was based on the treating surgeon’s preference. No clear pattern of initial treatment choice could be identified in this review. Non-operative treatment primarily consisted of pancreatic rest with total parenteral nutrition, nasojejunal tube feeds, or more commonly, a combination of both. Patients treated with percutaneous drainage had similar nutritional management. Patients were considered to have failed therapy if they remained symptomatic with pain, nausea, early satiety or failure to thrive and had a persistent pseudocyst demonstrated on imaging.

Statistical analyses

Comparison between groups was done using Fischer’s exact method for categorical variables and Student’s t test for continuous variables. Logistic regression was used to model treatment success and linear regression was used to model resource utilization. Resource utilization items considered in the model included need for parenteral nutrition, need for nasoenteric tube feeds, total number of CT scans, total number of ultrasounds, days to full oral intake, total hospital days, and number of hospital admissions. A probability (p value) of <0.05 was considered significant. Results are expressed as medians with ranges (minimum–maximum). Analyses were done using SAS 9.2.


Thirty-six children met inclusion criteria for having a pancreatic pseudocyst. A diagnosis of non-traumatic pancreatitis preceded the development of pseudocyst in 27 patients (75 %) and trauma incited pseudocyst development in nine patients (25 %). Etiology of pancreatitis and pseudocyst development in the non-trauma group included idiopathic (20), congenital anatomic malformation (2), gallstone disease (2), familial (1), cystic fibrosis (1) and bulimia (1). Patients with post-traumatic pseudocysts were injured by non-accidental trauma (2), handle bar injury (2), personal watercraft crash (1), ATV rollover (1), fall from bed (1), motor vehicle accident (1), and blunt abdominal trauma from an assault (1). This information is summarized in Table 1. The patients were grouped into non-traumatic pancreatitis and trauma for the purpose of analyses. Median age, weight, and gender were similar between the two groups (Table 2).
Table 1

Etiologies of pancreatic pseudocysts





Traumatic injury


Congenital anatomic malformation






Cystic fibrosis




Table 2

Demographics based on etiology




p value

Age (years)

13.0 (2–17)

9.0 (1–23)


Gender (female)

15 %

38 %


Weight (kg)

61.3 (11–78)

25.8 (7.1–86)


Data are displayed as medians with ranges

Of the 36 children, three were immediately treated operatively without a trial of non-operative management and were thus excluded from further analyses. Of the remaining 33 children, 22 were treated with non-operative management and 11 underwent percutaneous drainage. There were no endoscopic internal drainage procedures performed at our facility during this study period. The patient demographics were compared between the treatment arms and there was no significant difference found in age, gender, or weight. In the non-operative group, the median pseudocyst size was 5.4 cm (1–19.1) and 41 % were simple. In the percutaneous drainage group, the median pseudocyst size was 8.7 cm (5.7–11.4) and 18 % were simple (Table 3). These results were not statistically different.
Table 3

Characteristics based on treatment group


Non-operative (n = 22)

Percutaneous drainage (n = 11)

p value

Gender (female)

32 %

36 %


Weight (kg)

25.2 (7.1–86)

43.2 (9.7–81)


Pseudocyst size (cm)

5.4 (1–19.1)

8.7 (5.7–11.4)


Simple pseudocyst (non-septated)

41 %

18 %


Data are displayed as medians with ranges

Of the 22 children treated with non-operative management, 17 required no additional intervention (77 %). The other five children eventually required operations including Frey procedure (3), distal pancreatectomy (1), and cystgastrostomy (1).

Eight of the 11 children treated with percutaneous drainage required no additional treatment (72 %). The other three children underwent distal pancreatectomy. Success of non-operative management or percutaneous drainage was not dependent on pseudocyst size (p = 0.57) (Fig. 1) or complexity (p = 0.60) (Fig. 2). Logistic regression did not identify any patient demographic, etiologic or pseudocyst characteristic that predicted failure of non-operative therapy.
Fig. 1

Success of treatment in varying sizes of pseudocysts
Fig. 2

Success of management of simple and complex pseudocysts

Eight of the 33 children in this review remained symptomatic and required surgery despite initial non-operative management (5) and percutaneous drainage (3). Symptoms included pain, nausea, early satiety and failure to thrive. In the non-operative group, three patients underwent Frey procedure. In two of these patients the etiology was non-traumatic pancreatitis and though the pseudocysts were successfully managed non-operatively in both, they developed chronic pancreatitis with pancreatic duct dilatation requiring surgery. The third patient had a traumatic pancreatic laceration with a persistent pseudocyst, dilated duct and symptoms 82 days after initiation of non-operative management. One patient underwent distal pancreatectomy for pseudocyst with an ERCP demonstrating main pancreatic duct obstruction. Another patient underwent cystgastrostomy for a persistent and symptomatic 8 cm pseudocyst 102 days after initiation of non-operative therapy. In the percutaneous drainage group, three patients eventually required distal pancreatectomy. One child was symptomatic with multiple persistent pseudocysts despite drainage procedures. A second child developed pancreatic infection and necrosis that did not resolve with percutaneous drainage. The third child underwent distal pancreatectomy for management of a pancreatic pseudocyst with ERCP documentation of a fistulous tract to the main pancreatic duct.

There were two complications associated with non-operative management and two complications associated with IR drainage. One patient suffered hemorrhage into a pseudocyst occurring 21 days after initiation of non-operative therapy. This required embolization of the superior pancreaticoduodenal artery. A second patient had pseudocyst rupture, which occurred on day 19 of non-operative management. This was treated with percutaneous drainage of the fluid collection. However, the pseudocyst recurred 2 months later and was managed with a distal pancreatectomy. Two children treated initially with drain placement developed infection of the pseudocyst. In one patient the infection was controlled with percutaneous drainage and IV antibiotics. In the other patient an attempt at percutaneous drainage was unsuccessful and the patient went on to develop persistent fevers and pancreatic necrosis requiring debridement. There were no deaths related to pancreatic disease in this study population.

All patients achieved full oral feeds during the study period. The median number of days to full oral feeds was 29.5 (4–647) days. Pancreatic insufficiency, defined as use of pancreatic enzymes or insulin, was documented in 36 % of patients. At a median follow-up of 141 days (24 days to 6 years 288 days), 19 % of patients continued to receive pancreatic enzyme supplementation and 6 % of patients required insulin. No statistical difference was observed between groups regarding pancreatic insufficiency.

Resource utilization was assessed in each of the treatment groups (Table 4). This was based on the number of imaging tests performed during admission, days from admission to full oral feeds, total hospital days and total hospital admissions related to their diagnosis. The differences between groups were not statistically significant for any of the resources. Linear regression did not identify any patient demographic, etiologic or pseudocyst characteristic that predicted duration of therapy or hospitalization.
Table 4

Resource utilization


Non-operative (n = 22)

Percutaneous drainage (n = 11)

Parenteral nutrition

63 %

82 %

Nasoenteric tube feeds

86 %

64 %

CT (total number)

3.5 (1–7)

4 (1–8)

Ultrasound (total number)

3 (1–12)

1 (0–8)

Days to full oral intake



Hospital days (all admissions)

25 (8–113)

19 (10–111)

Hospital admissions

1 (1–8)

2 (1–6)

Data are displayed as medians with ranges


The rarity with which children develop pancreatitis and pancreatic pseudocyst make both of these events challenging to study. Most episodes of pancreatitis in children resolve without sequelae. The pediatric treatment strategies for pancreatic pseudocysts have historically been adopted from the adult literature Warshaw [19]. This approach may be complicated by the fact that children often have different etiologies resulting in pancreatitis than adults Benifla [12], Ford [14]. Traditionally, surgical treatment has been considered the gold standard for the management of pseudocysts; however this can come with a morbidity as high as 35 % and a mortality as high as 10 % in children [3].

Our series confirms the overall success of managing pancreatic pseudocysts in children non-operatively. This was true for children with pseudocysts due to pancreatitis and trauma. Of the 33 children in the series, only eight required operative treatment. All others were successfully managed with either non-operative treatment or percutaneous drainage, avoiding the need for operation in 76 % of our patients. Although the relatively small size of this series limits its ability to identify predictors of treatment failure, the series does demonstrate success of treatment over the full spectrum of pseudocysts. Even in children with septated pseudocysts, operative therapy was not typically required. Similarly, pseudocysts as large as 20 cm were successfully managed non-operatively. Importantly, there was no size or degree of pseudocyst complexity that clearly predicted failure of non-operative treatment. We believe that this series supports an initial non-operative approach (with or without percutaneous drainage) in almost all children.

There is a broad spectrum of current treatment practices across both the adult and pediatric populations. In adults, Vitas and Sarr demonstrated that approximately 58 % of pseudocysts resolve if conservative management is utilized [4]. Existing literature is sparse, but some recent reviews of children have reported success with non-operative management of pseudocysts following blunt abdominal trauma Bass [13], Jaffe [15], Burnweit [16], Wales [20]. In a recent multi-institutional review, pancreatic injuries managed non-operatively had a high rate of pseudocyst formation, but half of the pseudocysts were able to be managed non-operatively and the overall rate of surgical intervention was less compared to the group managed with immediate surgery [5]. Kouchi et al. examined the effectiveness of non-operative treatment of blunt pancreatic injury in 19 patients. Pancreatic injuries included contusion, laceration and injury to the main duct. Pseudocysts developed in ten patients with only two patients eventually requiring surgical intervention. Five of these patients that developed pseudocysts had main pancreatic ductal injury, and three were successfully managed non-operatively [6]. Similarly, Shilyansky et al [7]. reported successful non-operative management of ten pancreatic pseudocysts which developed in their experience of 28 patients with pancreatic injury managed non-operatively. Despite growing evidence that non-surgical treatment is warranted for traumatic pancreatic injury, some institutions still recommend immediate surgery. Canty et al. recommends ERCP for all patients with pancreatic injury and suspected ductal involvement, followed by stenting or distal pancreatectomy for patients with ductal disruption. They ultimately concluded that aggressive initial treatment of pancreatic injuries resulted in less pseudocyst development [8]. Recently, endoscopic cyst gastrostomy has been proposed as the treatment of choice for pancreatic pseudocysts in children [9, 10]. At our institution, advanced endoscopic techniques such as ERCP and endoscopic cyst gastrostomy are not readily available. Although these techniques should be considered in the management of pancreatic pseudocysts, we cannot comment on their effectiveness based on our experience.

In 2006, Teh et al. reviewed 24 pediatric patients with pseudocysts and found that 29 % of the pseudocysts resolved without any surgical or endoscopic intervention. They found that non-traumatic etiologies were more likely to require intervention at a rate of 92 %, compared to 45 % in those that were trauma related. Intervention in this study was either surgery or endoscopic sphincterotomy with stenting [11]. This finding that non-traumatic pseudocysts were more likely to require intervention was not confirmed in our series.

There are several limitations to this study. First, it is a retrospective review of an uncommon problem including a limited number of patients. This small sample size limits the power to detect characteristics that may predict treatment failure. The initial treatment plan of operative intervention, non-operative management or percutaneous drainage of the pancreatic pseudocyst was based on the preference of the treating pediatric surgeon. The variables that led to this decision were not clear upon review of the patient records. Additionally, treatment practices such as parenteral and enteral feeds, somatostatin use, and hospitalization end-point were also based on the preference of the treating surgeon and individualized according to the family social situation. It is possible that confounding factors existed among these patients that were unrecognized in our review.


This review represents one of the largest series of children with pancreatic pseudocysts in the literature. This experience shows that the majority of patients who experience pancreatitis or pancreatic injury complicated by the development of a pseudocyst may be successfully managed without operative intervention regardless of the size or characteristic of the pseudocyst. Furthermore, we rarely observed complications related to non-operative therapy in our patients. Although resource utilization with non-operative treatment may be significant, one could argue that this outweighs the potential morbidity of operative intervention and allows for preservation of pancreatic parenchyma.

Copyright information

© Springer-Verlag Berlin Heidelberg 2012