REVIEW
Operative Versus Nonoperative Management of Blunt Pancreatic Trauma in Children A Systematic Review Ezra Y. Koh, MD,*† Daan van Poll, MD, PhD,* J. Carel Goslings, MD, PhD,† Olivier R. Busch, MD, PhD,† Erik A. Rauws, MD, PhD,‡ Matthijs W. Oomen, MD, PhD,* and Marc G. Besselink, MD, PhD†
Abstract: The aim of this study was to compare operative versus nonoperative management of blunt pancreatic trauma in children. A systematic literature search was performed. Studies including children with blunt pancreatic injuries classified according to the American Association for the Surgery of Trauma classification were included. The primary outcome was pseudocyst formation. After screening 526 studies, 23 studies with 928 patients were included. Sufficient data were available for 674 patients (73%). Of 309 patients with grade I or II injuries, 258 (83%) were initially managed nonoperatively with a 96% success rate. Of 365 patients with grade III, IV, or V injuries, nonoperative management was initially chosen for 167 patients (46%) with an 89% success rate. Pseudocysts occurred in 18% of patients managed nonoperatively versus 4% of patients managed operatively (P < 0.01), of whom 65% were treated nonoperatively. Hospitalization was 20.5 days after nonoperative versus 15.1 days after operative management (nonparametric t test, P = 0.41). Blunt pancreatic trauma in children can be managed nonoperatively in the majority of patients with grade I or II injuries and in about half of the patients with grade III to V injuries. Although pseudocysts are more common after nonoperative management, two thirds can be managed nonoperatively. Key Words: pancreatic injury, abdominal trauma, children (Pancreas 2017;46: 1091–1097)
P
ancreatic injuries are diagnosed in less than 5% of children admitted for blunt abdominal trauma.1,2 Despite this low incidence, these injuries pose a significant challenge for clinicians because of the associated morbidity and lack of evidence on treatment strategies.3 Recently, nonoperative management has become the standard of care for pediatric solid organ injuries with success rates exceeding 90%.4,5 Pancreatic trauma, however, has lower success rates of nonoperative management.1 Current guidelines stratify management of these injuries depending on severity according to the American Association for the Surgery of Trauma (AAST) Organ Injury Scale, see Table 1.6 Grade I and II injuries indicate parenchymal injuries without pancreatic duct involvement, which can be considered minor, whereas grade III, IV, and V injuries include pancreatic duct lacerations/ transections and are considered major injuries. For adults with blunt pancreatic trauma, observation and/or drainage is recommended for From the *Department of Pediatric Surgery, Emma Children's Hospital, and Departments of †Surgery, ‡Gastroenterology and Hepatology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. Received for publication January 30, 2017; accepted August 9, 2017. Address correspondence to: Marc G. Besselink, MD, MSc, PhD, Department of Surgery, Academic Medical Center, University of Amsterdam, PO Box 22660, 1100 AZ Amsterdam, The Netherlands (e‐mail: [email protected]). Matthijs W. Oomen and Marc G. Besselink share senior authorship. The authors declare no conflict of interest. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/MPA.0000000000000916
Pancreas • Volume 46, Number 9, October 2017
minor injuries, whereas resection with drainage is recommended for major injuries.7–9 It is currently unclear whether this same approach should be applied for children. Proponents of operative management argue that surgery is associated with faster recovery and a minimal risk of pseudocysts.10 Advocates of nonoperative management argue that complete recovery is often feasible without surgery.11 A recent systematic review did not compare outcomes of these strategies because of the absence of randomized trials; therefore, to our knowledge, this is the first review comparing both approaches in blunt pancreatic trauma in children.12
MATERIALS AND METHODS Study Selection A systematic literature search, shown in Figure 1, was performed on September 16, 2016, using PubMed, Embase, and the Cochrane Library. This review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines.13 The following search strategy was conducted: “(((trauma[tiab]) OR blunt injury [MeSH]) AND (pancreas[Title/Abstract] OR pancreatic[Title/Abstract])) AND children [MeSH Terms].”
Eligibility Criteria Included were studies reporting on at least 5 children with nonoperative or operative management of blunt pancreatic trauma using the AAST pancreatic injury scale classification (Table 1). Excluded were studies in languages other than English and that included adult patients. Operative management was defined as immediate surgery upon establishing the diagnosis of pancreatic injury. Nonoperative management was defined as observation, endoscopic procedures, interventional radiology procedures, and surgery for other organ injuries.
Methodological Quality Assessment Individual studies were graded for level of evidence using criteria from the Oxford Center for Evidence-Based Medicine.14 Critical appraisal was performed using the Newcastle-Ottowa Scale, and studies with a score lower than 6 were considered low quality.
Outcomes The primary outcome was the success rate of nonoperative management, defined as patients in which a nonoperative strategy was chosen after diagnosis of pancreatic injury that did not require surgery. Secondary outcomes included complications, pseudocysts, days on total parenteral nutrition (TPN), length of stay, long-term outcomes, and mortality. www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1091
Pancreas • Volume 46, Number 9, October 2017
Koh et al
FIGURE 1. Selection of articles for review.
Statistical Analysis All statistical analyses were performed using SPSS Statistics 21 (IBM, Armonk, NY). Statistical significance was defined as P < 0.05.
RESULTS Search Results After review (Fig. 1), 23 studies including a total of 928 patients with blunt pancreatic trauma were included. Sufficient data for further analysis were available for 674 patients (73%). Study characteristics are displayed in Table 2.
Methodological Quality All eligible studies were retrospective case series, level 4 evidence. Of all studies, 18 (78%) of 23 were of low methodological quality, see Table 3. None of the studies used case matching. Most studies did not have a thorough follow-up.
Patient Characteristics Overall, 57% of patients were male. Bicycle accidents caused 24% (160 of 650 patients) of injuries in studies that reported the mechanisms of injuries.6,11,15–23,25–28,30,32–35 In studies
that provided individual patient data (51 patients from 5 studies), the incidence of grade III to V injuries did not differ between age groups (63%, 70%, and 50% of patients; 1-way analysis of variance, P = 0.17).21,22,27,32,34
Grade I or II Pancreatic Injuries There were 309 patients with grade I or II injuries in whom the treatment strategy could clearly be extracted. Of these patients, 258 (83%) were initially managed nonoperatively, whereas 51 patients (17%) were initially managed operatively, with the surgical technique reported in 39 patients. In total, surgical drainage was performed in 19 patients, distal pancreatectomy in 12, cystogastrostomy in 6, and gastrostomy in combination with a feeding jejunostomy in 1 patient. Primary repair was performed in 1 patient. Nonoperative management was successful in 248 (96%) of the 258 patients. Arkovitz et al15 reported 6 patients with grade II injuries who failed nonoperative management. Of those 6 patients, 2 required a distal pancreatectomy and 3 required a cystogastrostomy. The procedure performed for 1 patient could not be extracted from the study. Snajdauf et al21 reported 4 patients with grade II injuries who developed symptomatic pseudocysts and who were treated with cystogastrostomy at an average of 3.8 months after admission.
FIGURE 2. Pseudocyst occurrence in grade III to V injuries.
1092
www.pancreasjournal.com
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
Management of Blunt Pancreatic Trauma in Children
TABLE 1. The AAST Pancreatic Injury Classification7 Grade* I II III IV V
Type of Injury
Description of Injury
Hematoma Laceration Hematoma Laceration Laceration Laceration Laceration
Minor contusion without duct injury Superficial laceration without duct injury Major contusion without duct injury or tissue loss Major laceration without duct injury or tissue loss Distal transection or parenchymal injury with duct injury Proximal† transection or parenchymal injury involving ampulla Massive disruption of pancreatic head
*Advance 1 grade for multiple injuries up to grade III. † Proximal pancreas is to the patients' right of the superior mesenteric vein.
One of these patients had previous computed tomography– guided drainage.
Grade III, IV, or V Pancreatic Injuries There were a total of 365 patients with grade III, IV, or V injuries in whom the treatment strategy could clearly be extracted from the article. Nonoperative management was initially chosen for 167 patients (46%), whereas 198 patients (55%) had initial operative management. The operative technique could be determined for 139 patients. Distal pancreatectomy was performed in 99 patients, operative drainage in 19 patients,
pancreaticojejunostomy in 17 patients, cystogastrostomy in 3 patients, and feeding jejunostomy in 1 patient. Nonoperative management was successful in 148 patients (89%). There were 5 patients who required surgery after failed percutaneous drainage.20 Distal pancreatectomy was performed in 4 patients, and pancreaticojejunostomy was performed in 1 patient. Endoscopic stenting was unsuccessful in 1 patient who had a distal pancreatectomy 2 days later for a persistent leak.34 Six patients underwent surgery for progressive symptoms, distal pancreatectomy was performed in 4 patients, and pancreaticojejunostomy was performed in 2 patients.16,21,29 Eight patients required cystogastrostomy for pseudocysts that occurred after nonoperative management, one of which was after failed endoscopic
TABLE 2. Study Characteristics Reference 15
Arkovitz et al Firstenberg et al16 Canty et al17 Wales et al18 Jacombs et al19 Stringer20 Snajdauf et al21 Houben et al22 Mattix et al6 Vane et al23 Juric et al24 Sutherland et al25 Wood et al26 Klin et al27 Paul et al28 Borkon et al29 Cigdem et al30 Cuenca et al31 Yamana et al32 Beres et al33 Abbo et al11 Iqbal et al10 Garvey et al34 Dai et al35
Year
Country
Study Design
Level of Evidence
Sample Size
Male, %
Age*
ISS
1997 1999 2001 2001 2004 2005 2007 2007 2007 2009 2009 2010 2010 2011 2011 2011 2011 2012 2012 2013 2013 2014 2015 2015
United States United States United States Canada Australia United Kingdom Czech Republic United Kingdom United States United States Croatia Australia United States Israel United States United States Turkey United States Japan Canada France United States United States China
Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series
4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
26 9 18 9 65 9 13 15 173 14 7 91 43 10 43 25 31 79 6 39 36 167 7 33
85 67 44 67 60 66 69 73 63 79 57 72 66 90 70 48 90 66 29 69 72 64 42 —
7.4 8.7 — 8.0 6.5 — 8.5 10.0 7.7 6.9 8.3 8.7 7.3 9.7 — 10.5 9.1 8.6 7.6 9.2 8.7 9.1 7.0 —
— — — 25 — — — — — — — — 10 — 12 20.7 — 9 7.1 19.2 16.5 15.1 — —
*Reported as mean or median. ISS indicates injury severity score.
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1093
Pancreas • Volume 46, Number 9, October 2017
Koh et al
TABLE 3. Methodological Quality Selection Reference
Outcome
Representativeness Selection Ascertainment
Arkovitz et al15 Firstenberg et al16 Canty et al17 Wales et al18 Jacombs et al19 Stringer20 Snajdauf et al21 Houben et al22 Mattix et al6 Vane et al23 Juric et al24 Sutherland et al25 Wood et al26 Klin et al27 Paul et al28 Borkon et al29 Cigdem et al30 Cuenca et al31 Yamana et al32 Beres et al33 Abbo et al11 Iqbal et al10 Garvey et al34 Dai et al35
Conflicted FU Adequacy Total Interest Comparability Assessment Length of FU Points
1 1
1 1
1 1
0 0
0 0
1 1
0 0
0 0
4 4
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
0 0 0 0 0 0 0 0 0 0
0 N/A 0 0 1 1 N/A 0 N/A N/A
1 1 1 1 1 1 1 1 1 1
0 1 0 0 1 1 N/A 0 1 0
0 1 0 0 1 1 N/A 0 1 0
4 6 4 4 7 7 4 4 6 4
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
0 0 0 0 0 0 0 0 0 0 0 0
1 0 0 N/A 0 0 0 1 N/A 0 1 N/A
1 1 1 1 1 1 1 1 1 1 1 1
0 0 0 0 1 0 0 0 0 0 0 0
0 0 0 0 1 0 0 0 0 0 0 0
5 4 4 4 6 4 4 5 4 1 5 4
FU indicates follow-up; N/A, not applicable.
drainage.10,17,22 Overall, success of nonoperative management was 96% for grade I or II injuries versus 89% for grade III, IV, or V injuries (Fisher exact test, P < 0.01).
Indications for Operative Management In 16 studies, the indications for operative management were discussed, of which 8 studies provided individual numbers. A total of 19 patients with grade I or II injuries who underwent operative management were identified. Of these patients, 18 had operative exploration for pancreatic indications. In the remaining patients, the decision to pursue operative management was due to a positive diagnostic peritoneal lavage, after which during exploration a grade I pancreatic contusion was identified and a gastrostomy and feeding jejunostomy were placed. Of the 37 patients with grade III to V injuries, all had surgery for identified pancreatic injuries. None of the studies provided criteria for operative or nonoperative management.
Endoscopic Retrograde Cholangiopancreatography Endoscopic retrograde cholangiopancreatography (ERCP) was performed in 49 patients. In 3 separate studies, stenting was performed in 14 patients (7 grade IV injuries, 6 grade III injuries, 1 grade II injury).17,22,34 Sphincterotomy was performed in 2 patients (both grade III injuries).34 In 1 patient, stenting and sphincterotomy were performed simultaneously (grade III).
1094
www.pancreasjournal.com
Houben et al22 reported 2 technical failures because of difficult anatomy. Two patients required distal pancreatectomy after ERCP (1 after stent and 1 after sphincterotomy, 1 grade III and 1 grade IV injury). In total, 3 patients required repeat ERCP for stent replacement.
Complications Five patients (2%) had surgical complications that required repeat surgery (2 cases of bowel obstruction, 2 cases of ongoing leakage, and 1 case of wound dehiscence). Eleven patients (4%) had postoperative complications that did not require surgical intervention (3 cases of line sepsis, 2 cases of pancreatitis, 2 cases of pancreatic fistula, and 1 case each of small bowel obstruction, high drain output, intraabdominal abscess, and Clostridium infection).20,27–29 During nonoperative management, a total of 12 patients (3%) developed complications (6 cases of sepsis, 2 cases of abscesses, and 1 case each of infected fluid collection, pancreatitis, and a Mallory-Weiss tear).19,24,30,33,34 All patients recovered successfully without surgical intervention.
Pseudocysts In total, 140 patients (15%) developed pseudocysts with sufficient available data for further analysis in 85 patients (61%). There were 76 patients (18%) who developed pseudocysts after nonoperative management, compared with 9 patients (4%) who © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
developed pseudocysts after operative management Fisher’s exact test, P < 0.01; number needed to treat = 7.14). We were able to identify a total of 22 patients (9%) with minor pancreatic injury who developed pseudocysts after nonoperative management and 1 patient (2%) who developed a pseudocyst after operative management (Fisher’s exact test, P = 0.14; number needed to treat = 14.3). Pseudocysts developed in 54 patients (32%) with major injuries after nonoperative management versus 8 patients (4%) after operative treatment (Fisher’s exact test, P < 0.01; number needed to treat = 3.6) (Fig. 2). Percutaneous drainage of the pseudocyst was performed in 26 patients (31%), whereas endoscopic drainage was performed in 9 patients (11%). In 8 patients (23%), percutaneous or endoscopic drainage was unsuccessful. A total of 29 patients (35%) had surgical cystogastrostomy.
Total Parenteral Nutrition Eight studies reported on the use of TPN. On average, TPN was administered in 62% of patients for a period of 15.1 days. The TPN was administered in 65% of patients treated nonoperatively for an average of 19.4 days and 68% of patients treated surgically for an average of 10 days. The median TPN duration between nonoperative and operative management was similar (nonparametric t test, P = 0.10)
Length of Stay Thirteen studies reported on length of stay, which overall was mean 21.7 (standard deviation, 12.2) days. The median (interquartile range) hospitalization was 20.5 (11.3–37.6) days after nonoperative management in comparison with 15.1 (12–20) days after operative management (nonparametric t-test, P = 0.414).
Mortality Mortality occurred in 49 patients (5%), often as a result of associated injuries. There were 4 patients who died of traumatic brain injury.15,17,29 Another 10 patients died of either traumatic brain injury or multiorgan failure.31 Four patients died because of pancreatic injuries.18,23 One patient had a grade V injury, although the severity of injury in the other patients is unknown. Of the remaining 31 patients, the exact cause of death could not be extracted. However, both studies reported that these deaths were not attributable to pancreatic injury.18,25
Follow-Up In total, 3 studies with 37 patients reported follow-up results. One study reported excellent follow-up results of patients treated nonoperatively; after 47 months follow-up, none of the 9 patients had exocrine or endocrine insufficiency.19 Another study reported uneventful follow-up in 15 patients treated nonoperatively, 1 of which eventually required surgery, without any pancreatic insufficiency after a median follow-up of 2 years.22 Snajdauf et al21 also reported excellent follow-up results of operative management in a series in which distal pancreatectomy was performed in 6 patients, pancreaticoduodenectomy in 4 patients, and cystogastrostomy in 3 patients. In a follow-up ranging from 1 to 12 years, none of the 13 patients had either exocrine or endocrine insufficiency.21
DISCUSSION This systematic review comparing nonoperative and operative management demonstrates that the vast majority of children with grade I or II blunt pancreatic injuries and about half of patients with grade III – V injuries can be managed nonoperatively with excellent success rates. Although pseudocysts are more © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Management of Blunt Pancreatic Trauma in Children
common (18%) in patients managed nonoperatively, these can be treated nonoperatively in two thirds of patients. Nonoperative management does not lead to prolonged hospital stay. A previous Cochrane systematic review failed to include any studies because the authors were only interested in randomized trials, which are nonexistent in this topic.12 Therefore the present review brings together the best available evidence, but at the same time acknowledging that these are all retrospective studies with a significant risk of selection bias. None of the included studies reported on specific selection criteria for operative management. Little is known about the epidemiology of blunt pancreatic trauma in children. Our results suggest that pancreatic trauma is more common in males and that almost a quarter of these injuries are caused by bicycle accidents. No relationship between age and severity of injury could be identified.21,22,27,32,34 Low body weight has been associated with these injuries.22 Data from the included studies showed an almost equal distribution of minor and major pancreatic injuries, but it is important to consider that minor injuries may be underrepresented in the literature because these patients may not always seek treatment. According to current guidelines, identifying ductal injury is crucial to determine which type of intervention is indicated. Computed tomography and magnetic resonance imaging both have shown high accuracy in assessing pancreatic injury.36 The diagnostic accuracy of computed tomography is reported to be 75%.37 Endoscopic retrograde cholangiopancreatography is an accurate method of investigating the pancreatic duct but is less attractive because it is invasive with associated risks such as pancreatitis. Magnetic resonance cholangiopancreatography has been reported as an accurate, noninvasive method of determining whether ductal injury is present.38,39 Because of the low incidence of blunt pancreatic trauma in children, large, high-quality studies, such as randomized trials, are not available and will probably never be performed. This review suggests that many patients can be managed nonoperatively. However, this comes at a cost of a higher rate of pseudocysts after nonoperative management (18% vs 4% in this review). Nevertheless, there are many concerns about operative management as well. Patients may develop endo- or exocrine insufficiency after distal pancreatectomy. Surprisingly, this was not reported by studies included in this systematic review. Studies have shown that after elective pancreatic surgery in children, postoperative diabetes occurs in less than 10% of children, whereas exocrine dysfunction occurs in up to 60% of children.40–43 Overall, pseudocysts developed in 142 children (15%) with blunt pancreatic trauma. Currently, there are several options for managing these pseudocysts. Observation is an acceptable initial step, because pseudocysts may spontaneously resolve.44 In case of persistent symptoms or infection, interventional radiology, endoscopy, or surgery can be used to drain the pseudocyst internally. Both radiologic and endoscopic drainages show good results, and technical success is over 90%.44–47 When comparing both techniques, endoscopic drainage seems to have a greater procedural risk but has a lower risk of reinterventions, residual collections, and the need for subsequent surgery due to internal drainage of pseudocysts.48 Laparoscopic techniques also show favorable results, although selection bias may be present.49 This review has several limitations. First, all included studies were retrospective and hence selection bias has played a clear role in the decision. Second, the heterogeneous nature of the studies led to frequent missing data. Despite contacting corresponding authors, we were unable to attain all missing data. Therefore, we were unable to compare outcomes according to overall injury severity using the injury severity score and among different www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1095
Pancreas • Volume 46, Number 9, October 2017
Koh et al
age groups. Lastly, not all studies had sufficient follow-up to determine whether pseudocysts developed or not.
CONCLUSIONS In conclusion, blunt pancreatic trauma in children can be managed nonoperatively in most patients with high success rates. Higher injury severity is associated with an increased need for surgical intervention, although selection criteria are lacking and should be developed. Pseudocysts are more common after nonoperative management but do not require surgery in two thirds of patients. The exact role of endoscopy and interventional radiology in these patients remains unclear. Further studies are warranted to validate the risks and benefits of endoscopy, interventional radiology, or surgery for these patients to establish definitive guidelines.
14. Howick J, Chalmers I, Glasziou P, et al. OCEBM Levels of Evidence Working Group. The Oxford 2011 Levels of Evidence. 2011. Available at: http://www.cebm.net/ocebm-levels-of-evidence/. Accessed September 29, 2016. 15. Arkovitz MS, Johnson N, Garcia VF. Pancreatic trauma in children: mechanisms of injury. J Trauma. 1997;42:49–53. 16. Firstenberg MS, Volsko TA, Sivit C, et al. Selective management of pediatric pancreatic injuries. J Pediatr Surg. 1999;34:1142–1147. 17. Canty TG Sr, Weinman D. Management of major pancreatic duct injuries in children. J Trauma. 2001;50:1001–1007. 18. Jacombs AS, Wines M, Holland A, et al. Pancreatic trauma in children. J Pediatr Surg. 2004;39:96–99. 19. Wales PW, Shuckett B, Kim PC. Long-term outcome after nonoperative management of complete traumatic pancreatic transection in children. J Pediatr Surg. 2001;36:823–827. 20. Stringer MD. Pancreatic trauma in children. Br J Surg. 2005;92:467–470.
ACKNOWLEDGMENTS The authors thank Dr Mary Brindle (Department of Surgery, Alberta Children's Hospital, Calgary, Canada) for providing additional data from her original study. REFERENCES 1. Holmes JH 4th, Wiebe DJ, Tataria M, et al. The failure of nonoperative management in pediatric solid organ injury: a multi-institutional experience. J Trauma. 2005;59:1309–1313.
21. Snajdauf J, Rygl M, Kalousová J, et al. Surgical management of major pancreatic injury in children. Eur J Pediatr Surg. 2007;17:317–321. 22. Houben CH, Ade-Ajayi N, Patel S, et al. Traumatic pancreatic duct injury in children: minimally invasive approach to management. J Pediatr Surg. 2007;42:629–635. 23. Vane DW, Kiankhooy A, Sartorelli KH, et al. Initial resection of potentially viable tissue is not optimal treatment for grades II-IV pancreatic injuries. World J Surg. 2009;33:221–227. 24. Jurić I, Pogorelić Z, Biocić M, et al. Management of blunt pancreatic trauma in children. Surg Today. 2009;39:115–119.
2. Bradley EL 3rd, Young PR Jr, Chang MC, et al. Diagnosis and initial management of blunt pancreatic trauma: guidelines from a multiinstitutional review. Ann Surg. 1998;227:861–869.
25. Sutherland I, Ledder O, Crameri J, et al. Pancreatic trauma in children. Pediatr Surg Int. 2010;26:1201–1206.
3. Lin BC, Chen RJ, Fang JF, et al. Management of blunt major pancreatic injury. J Trauma. 2004;56:774–778.
26. Wood JH, Partrick DA, Bruny JL, et al. Operative vs nonoperative management of blunt pancreatic trauma in children. J Pediatr Surg. 2010; 45:401–406.
4. Dervan LA, King MA, Cuschieri J, et al. Pediatric solid organ injury operative interventions and outcomes at Harborview Medical Center, before and after introduction of a solid organ injury pathway for pediatrics. J Trauma Acute Care Surg. 2015;79:215–220. 5. Partrick DA, Moore EE, Bensard DD, et al. Operative management of injured children at an adult level I trauma center. J Trauma. 2000;48: 894–901. 6. Mattix KD, Tataria M, Holmes J, et al. Pediatric pancreatic trauma: predictors of nonoperative management failure and associated outcomes. J Pediatr Surg. 2007;42:340–344. 7. Moore EE, Cogbill TH, Malangoni MA, et al. Organ injury scaling, II: pancreas, duodenum, small bowel, colon, and rectum. J Trauma. 1990;30: 1427–1429. 8. Biffl WL, Moore EE, Croce M, et al. Western Trauma Association critical decisions in trauma: management of pancreatic injuries. J Trauma Acute Care Surg. 2013;75:941–946. 9. Ho VP, Patel NJ, Bokhari F, et al. Management of adult pancreatic injuries: a practice management guideline from the Eastern Association for the Surgery of Trauma. J Trauma Acute Care Surg. 2017;82:185–199. 10. Iqbal CW, St Peter SD, Tsao K, et al. Operative vs nonoperative management for blunt pancreatic transection in children: multi-institutional outcomes. J Am Coll Surg. 2014;218:157–162. 11. Abbo O, Lemandat A, Reina N, et al. Conservative management of blunt pancreatic trauma in children: a single center experience. Eur J Pediatr Surg. 2013;23:470–473.
27. Klin B, Abu-Kishk I, Jeroukhimov I, et al. Blunt pancreatic trauma in children. Surg Today. 2011;41:946–954. 28. Paul MD, Mooney DP. The management of pancreatic injuries in children: operate or observe. J Pediatr Surg. 2011;46:1140–1143. 29. Borkon MJ, Morrow SE, Koehler EA, et al. Operative intervention for complete pancreatic transection in children sustaining blunt abdominal trauma: revisiting an organ salvage technique. Am Surg. 2011;77:612–620. 30. Cigdem MK, Senturk S, Onen A, et al. Nonoperative management of pancreatic injuries in pediatric patients. Surg Today. 2011;41:655–659. 31. Cuenca AG, Islam S. Pediatric pancreatic trauma: trending toward nonoperative management? Am Surg. 2012;78:1204–1210. 32. Yamana I, Kawamoto S, Kamitani T, et al. Pancreatic injury in children: review of 7 cases and the pertinent literature. Hepatogastroenterology. 2012;59:574–577. 33. Beres AL, Wales PW, Christison-Lagay ER, et al. Non-operative management of high-grade pancreatic trauma: is it worth the wait? J Pediatr Surg. 2013;48:1060–1064. 34. Garvey EM, Haakinson DJ, McOmber M, et al. Role of ERCP in pediatric blunt abdominal trauma: a case series at a level one pediatric trauma center. J Pediatr Surg. 2015;50:335–338. 35. Dai LN, Chen CD, Lin XK, et al. Abdominal injuries involving bicycle handlebars in 219 children: results of 8-year follow-up. Eur J Trauma Emerg Surg. 2015;41:551–555.
12. Haugaard MV, Wettergren A, Hillingsø JG, et al. Non-operative versus operative treatment for blunt pancreatic trauma in children. Cochrane Database Syst Rev. 2014;2:CD009746.
36. Panda A, Kumar A, Gamanagatti S, et al. Evaluation of diagnostic utility of multidetector computed tomography and magnetic resonance imaging in blunt pancreatic trauma: a prospective study. Acta Radiol. 2015;56: 387–396.
13. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269, W264.
37. Moschetta M, Telegrafo M, Malagnino V, et al. Pancreatic trauma: the role of computed tomography for guiding therapeutic approach. World J Radiol. 2015;7:415–420.
1096
www.pancreasjournal.com
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
38. Fulcher AS, Turner MA, Yelon JA, et al. Magnetic resonance cholangiopancreatography (MRCP) in the assessment of pancreatic duct trauma and its sequelae: preliminary findings. J Trauma. 2000;48: 1001–1007.
Management of Blunt Pancreatic Trauma in Children
44. Russell KW, Barnhart DC, Madden J, et al. Non-operative treatment versus percutaneous drainage of pancreatic pseudocysts in children. Pediatr Surg Int. 2013;29:305–310.
39. Nirula R, Velmahos GC, Demetriades D. Magnetic resonance cholangiopancreatography in pancreatic trauma: a new diagnostic modality? J Trauma. 1999;47:585–587.
45. Akshintala VS, Saxena P, Zaheer A, et al. A comparative evaluation of outcomes of endoscopic versus percutaneous drainage for symptomatic pancreatic pseudocysts. Gastrointest Endosc. 2014;79:921–928; quiz 983.e2, 983.e5.
40. Marchegiani G, Crippa S, Malleo G, et al. Surgical treatment of pancreatic tumors in childhood and adolescence: uncommon neoplasms with favorable outcome. Pancreatology. 2011;11:383–389.
46. Coelho DE, Ardengh JC, Carbalo MT, et al. Clinicopathologic characteristics and endoscopic treatment of post-traumatic pancreatic pseudocysts. Pancreas. 2011;40:469–473.
41. Muller CO, Guérin F, Goldzmidt D, et al. Pancreatic resections for solid or cystic pancreatic masses in children. J Pediatr Gastroenterol Nutr. 2012;54: 369–373.
47. Thomson DA, Krige JE, Thomson SR, et al. The role of endoscopic retrograde pancreatography in pancreatic trauma: a critical appraisal of 48 patients treated at a tertiary institution. J Trauma Acute Care Surg. 2014; 76:1362–1366.
42. Park HH, Kim HY, Jung SE, et al. Long-term functional outcomes of PPPD in children—nutritional status, pancreatic function, GI function and QOL. J Pediatr Surg. 2016;51:398–402. 43. Sugito K, Furuya T, Kaneda H, et al. Long-term follow-up of nutritional status, pancreatic function, and morphological changes of the pancreatic remnant after pancreatic tumor resection in children. Pancreas. 2012;41: 554–559.
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
48. Keane MG, Sze SF, Cieplik N, et al. Endoscopic versus percutaneous drainage of symptomatic pancreatic fluid collections: a 14-year experience from a tertiary hepatobiliary centre. Surg Endosc. 2016;30:3730–3740. 49. Palanivelu C, Senthilkumar K, Madhankumar MV, et al. Management of pancreatic pseudocyst in the era of laparoscopic surgery—experience from a tertiary centre. Surg Endosc. 2007;21:2262–2267.
www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1097
Operative Versus Nonoperative Management of Blunt Pancreatic Trauma in Children A Systematic Review Ezra Y. Koh, MD,*† Daan van Poll, MD, PhD,* J. Carel Goslings, MD, PhD,† Olivier R. Busch, MD, PhD,† Erik A. Rauws, MD, PhD,‡ Matthijs W. Oomen, MD, PhD,* and Marc G. Besselink, MD, PhD†
Abstract: The aim of this study was to compare operative versus nonoperative management of blunt pancreatic trauma in children. A systematic literature search was performed. Studies including children with blunt pancreatic injuries classified according to the American Association for the Surgery of Trauma classification were included. The primary outcome was pseudocyst formation. After screening 526 studies, 23 studies with 928 patients were included. Sufficient data were available for 674 patients (73%). Of 309 patients with grade I or II injuries, 258 (83%) were initially managed nonoperatively with a 96% success rate. Of 365 patients with grade III, IV, or V injuries, nonoperative management was initially chosen for 167 patients (46%) with an 89% success rate. Pseudocysts occurred in 18% of patients managed nonoperatively versus 4% of patients managed operatively (P < 0.01), of whom 65% were treated nonoperatively. Hospitalization was 20.5 days after nonoperative versus 15.1 days after operative management (nonparametric t test, P = 0.41). Blunt pancreatic trauma in children can be managed nonoperatively in the majority of patients with grade I or II injuries and in about half of the patients with grade III to V injuries. Although pseudocysts are more common after nonoperative management, two thirds can be managed nonoperatively. Key Words: pancreatic injury, abdominal trauma, children (Pancreas 2017;46: 1091–1097)
P
ancreatic injuries are diagnosed in less than 5% of children admitted for blunt abdominal trauma.1,2 Despite this low incidence, these injuries pose a significant challenge for clinicians because of the associated morbidity and lack of evidence on treatment strategies.3 Recently, nonoperative management has become the standard of care for pediatric solid organ injuries with success rates exceeding 90%.4,5 Pancreatic trauma, however, has lower success rates of nonoperative management.1 Current guidelines stratify management of these injuries depending on severity according to the American Association for the Surgery of Trauma (AAST) Organ Injury Scale, see Table 1.6 Grade I and II injuries indicate parenchymal injuries without pancreatic duct involvement, which can be considered minor, whereas grade III, IV, and V injuries include pancreatic duct lacerations/ transections and are considered major injuries. For adults with blunt pancreatic trauma, observation and/or drainage is recommended for From the *Department of Pediatric Surgery, Emma Children's Hospital, and Departments of †Surgery, ‡Gastroenterology and Hepatology, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands. Received for publication January 30, 2017; accepted August 9, 2017. Address correspondence to: Marc G. Besselink, MD, MSc, PhD, Department of Surgery, Academic Medical Center, University of Amsterdam, PO Box 22660, 1100 AZ Amsterdam, The Netherlands (e‐mail: [email protected]). Matthijs W. Oomen and Marc G. Besselink share senior authorship. The authors declare no conflict of interest. Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved. DOI: 10.1097/MPA.0000000000000916
Pancreas • Volume 46, Number 9, October 2017
minor injuries, whereas resection with drainage is recommended for major injuries.7–9 It is currently unclear whether this same approach should be applied for children. Proponents of operative management argue that surgery is associated with faster recovery and a minimal risk of pseudocysts.10 Advocates of nonoperative management argue that complete recovery is often feasible without surgery.11 A recent systematic review did not compare outcomes of these strategies because of the absence of randomized trials; therefore, to our knowledge, this is the first review comparing both approaches in blunt pancreatic trauma in children.12
MATERIALS AND METHODS Study Selection A systematic literature search, shown in Figure 1, was performed on September 16, 2016, using PubMed, Embase, and the Cochrane Library. This review was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses guidelines.13 The following search strategy was conducted: “(((trauma[tiab]) OR blunt injury [MeSH]) AND (pancreas[Title/Abstract] OR pancreatic[Title/Abstract])) AND children [MeSH Terms].”
Eligibility Criteria Included were studies reporting on at least 5 children with nonoperative or operative management of blunt pancreatic trauma using the AAST pancreatic injury scale classification (Table 1). Excluded were studies in languages other than English and that included adult patients. Operative management was defined as immediate surgery upon establishing the diagnosis of pancreatic injury. Nonoperative management was defined as observation, endoscopic procedures, interventional radiology procedures, and surgery for other organ injuries.
Methodological Quality Assessment Individual studies were graded for level of evidence using criteria from the Oxford Center for Evidence-Based Medicine.14 Critical appraisal was performed using the Newcastle-Ottowa Scale, and studies with a score lower than 6 were considered low quality.
Outcomes The primary outcome was the success rate of nonoperative management, defined as patients in which a nonoperative strategy was chosen after diagnosis of pancreatic injury that did not require surgery. Secondary outcomes included complications, pseudocysts, days on total parenteral nutrition (TPN), length of stay, long-term outcomes, and mortality. www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1091
Pancreas • Volume 46, Number 9, October 2017
Koh et al
FIGURE 1. Selection of articles for review.
Statistical Analysis All statistical analyses were performed using SPSS Statistics 21 (IBM, Armonk, NY). Statistical significance was defined as P < 0.05.
RESULTS Search Results After review (Fig. 1), 23 studies including a total of 928 patients with blunt pancreatic trauma were included. Sufficient data for further analysis were available for 674 patients (73%). Study characteristics are displayed in Table 2.
Methodological Quality All eligible studies were retrospective case series, level 4 evidence. Of all studies, 18 (78%) of 23 were of low methodological quality, see Table 3. None of the studies used case matching. Most studies did not have a thorough follow-up.
Patient Characteristics Overall, 57% of patients were male. Bicycle accidents caused 24% (160 of 650 patients) of injuries in studies that reported the mechanisms of injuries.6,11,15–23,25–28,30,32–35 In studies
that provided individual patient data (51 patients from 5 studies), the incidence of grade III to V injuries did not differ between age groups (63%, 70%, and 50% of patients; 1-way analysis of variance, P = 0.17).21,22,27,32,34
Grade I or II Pancreatic Injuries There were 309 patients with grade I or II injuries in whom the treatment strategy could clearly be extracted. Of these patients, 258 (83%) were initially managed nonoperatively, whereas 51 patients (17%) were initially managed operatively, with the surgical technique reported in 39 patients. In total, surgical drainage was performed in 19 patients, distal pancreatectomy in 12, cystogastrostomy in 6, and gastrostomy in combination with a feeding jejunostomy in 1 patient. Primary repair was performed in 1 patient. Nonoperative management was successful in 248 (96%) of the 258 patients. Arkovitz et al15 reported 6 patients with grade II injuries who failed nonoperative management. Of those 6 patients, 2 required a distal pancreatectomy and 3 required a cystogastrostomy. The procedure performed for 1 patient could not be extracted from the study. Snajdauf et al21 reported 4 patients with grade II injuries who developed symptomatic pseudocysts and who were treated with cystogastrostomy at an average of 3.8 months after admission.
FIGURE 2. Pseudocyst occurrence in grade III to V injuries.
1092
www.pancreasjournal.com
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
Management of Blunt Pancreatic Trauma in Children
TABLE 1. The AAST Pancreatic Injury Classification7 Grade* I II III IV V
Type of Injury
Description of Injury
Hematoma Laceration Hematoma Laceration Laceration Laceration Laceration
Minor contusion without duct injury Superficial laceration without duct injury Major contusion without duct injury or tissue loss Major laceration without duct injury or tissue loss Distal transection or parenchymal injury with duct injury Proximal† transection or parenchymal injury involving ampulla Massive disruption of pancreatic head
*Advance 1 grade for multiple injuries up to grade III. † Proximal pancreas is to the patients' right of the superior mesenteric vein.
One of these patients had previous computed tomography– guided drainage.
Grade III, IV, or V Pancreatic Injuries There were a total of 365 patients with grade III, IV, or V injuries in whom the treatment strategy could clearly be extracted from the article. Nonoperative management was initially chosen for 167 patients (46%), whereas 198 patients (55%) had initial operative management. The operative technique could be determined for 139 patients. Distal pancreatectomy was performed in 99 patients, operative drainage in 19 patients,
pancreaticojejunostomy in 17 patients, cystogastrostomy in 3 patients, and feeding jejunostomy in 1 patient. Nonoperative management was successful in 148 patients (89%). There were 5 patients who required surgery after failed percutaneous drainage.20 Distal pancreatectomy was performed in 4 patients, and pancreaticojejunostomy was performed in 1 patient. Endoscopic stenting was unsuccessful in 1 patient who had a distal pancreatectomy 2 days later for a persistent leak.34 Six patients underwent surgery for progressive symptoms, distal pancreatectomy was performed in 4 patients, and pancreaticojejunostomy was performed in 2 patients.16,21,29 Eight patients required cystogastrostomy for pseudocysts that occurred after nonoperative management, one of which was after failed endoscopic
TABLE 2. Study Characteristics Reference 15
Arkovitz et al Firstenberg et al16 Canty et al17 Wales et al18 Jacombs et al19 Stringer20 Snajdauf et al21 Houben et al22 Mattix et al6 Vane et al23 Juric et al24 Sutherland et al25 Wood et al26 Klin et al27 Paul et al28 Borkon et al29 Cigdem et al30 Cuenca et al31 Yamana et al32 Beres et al33 Abbo et al11 Iqbal et al10 Garvey et al34 Dai et al35
Year
Country
Study Design
Level of Evidence
Sample Size
Male, %
Age*
ISS
1997 1999 2001 2001 2004 2005 2007 2007 2007 2009 2009 2010 2010 2011 2011 2011 2011 2012 2012 2013 2013 2014 2015 2015
United States United States United States Canada Australia United Kingdom Czech Republic United Kingdom United States United States Croatia Australia United States Israel United States United States Turkey United States Japan Canada France United States United States China
Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series Retrospective case series
4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4
26 9 18 9 65 9 13 15 173 14 7 91 43 10 43 25 31 79 6 39 36 167 7 33
85 67 44 67 60 66 69 73 63 79 57 72 66 90 70 48 90 66 29 69 72 64 42 —
7.4 8.7 — 8.0 6.5 — 8.5 10.0 7.7 6.9 8.3 8.7 7.3 9.7 — 10.5 9.1 8.6 7.6 9.2 8.7 9.1 7.0 —
— — — 25 — — — — — — — — 10 — 12 20.7 — 9 7.1 19.2 16.5 15.1 — —
*Reported as mean or median. ISS indicates injury severity score.
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1093
Pancreas • Volume 46, Number 9, October 2017
Koh et al
TABLE 3. Methodological Quality Selection Reference
Outcome
Representativeness Selection Ascertainment
Arkovitz et al15 Firstenberg et al16 Canty et al17 Wales et al18 Jacombs et al19 Stringer20 Snajdauf et al21 Houben et al22 Mattix et al6 Vane et al23 Juric et al24 Sutherland et al25 Wood et al26 Klin et al27 Paul et al28 Borkon et al29 Cigdem et al30 Cuenca et al31 Yamana et al32 Beres et al33 Abbo et al11 Iqbal et al10 Garvey et al34 Dai et al35
Conflicted FU Adequacy Total Interest Comparability Assessment Length of FU Points
1 1
1 1
1 1
0 0
0 0
1 1
0 0
0 0
4 4
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1
0 0 0 0 0 0 0 0 0 0
0 N/A 0 0 1 1 N/A 0 N/A N/A
1 1 1 1 1 1 1 1 1 1
0 1 0 0 1 1 N/A 0 1 0
0 1 0 0 1 1 N/A 0 1 0
4 6 4 4 7 7 4 4 6 4
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
1 1 1 1 1 1 1 1 1 1 1 1
0 0 0 0 0 0 0 0 0 0 0 0
1 0 0 N/A 0 0 0 1 N/A 0 1 N/A
1 1 1 1 1 1 1 1 1 1 1 1
0 0 0 0 1 0 0 0 0 0 0 0
0 0 0 0 1 0 0 0 0 0 0 0
5 4 4 4 6 4 4 5 4 1 5 4
FU indicates follow-up; N/A, not applicable.
drainage.10,17,22 Overall, success of nonoperative management was 96% for grade I or II injuries versus 89% for grade III, IV, or V injuries (Fisher exact test, P < 0.01).
Indications for Operative Management In 16 studies, the indications for operative management were discussed, of which 8 studies provided individual numbers. A total of 19 patients with grade I or II injuries who underwent operative management were identified. Of these patients, 18 had operative exploration for pancreatic indications. In the remaining patients, the decision to pursue operative management was due to a positive diagnostic peritoneal lavage, after which during exploration a grade I pancreatic contusion was identified and a gastrostomy and feeding jejunostomy were placed. Of the 37 patients with grade III to V injuries, all had surgery for identified pancreatic injuries. None of the studies provided criteria for operative or nonoperative management.
Endoscopic Retrograde Cholangiopancreatography Endoscopic retrograde cholangiopancreatography (ERCP) was performed in 49 patients. In 3 separate studies, stenting was performed in 14 patients (7 grade IV injuries, 6 grade III injuries, 1 grade II injury).17,22,34 Sphincterotomy was performed in 2 patients (both grade III injuries).34 In 1 patient, stenting and sphincterotomy were performed simultaneously (grade III).
1094
www.pancreasjournal.com
Houben et al22 reported 2 technical failures because of difficult anatomy. Two patients required distal pancreatectomy after ERCP (1 after stent and 1 after sphincterotomy, 1 grade III and 1 grade IV injury). In total, 3 patients required repeat ERCP for stent replacement.
Complications Five patients (2%) had surgical complications that required repeat surgery (2 cases of bowel obstruction, 2 cases of ongoing leakage, and 1 case of wound dehiscence). Eleven patients (4%) had postoperative complications that did not require surgical intervention (3 cases of line sepsis, 2 cases of pancreatitis, 2 cases of pancreatic fistula, and 1 case each of small bowel obstruction, high drain output, intraabdominal abscess, and Clostridium infection).20,27–29 During nonoperative management, a total of 12 patients (3%) developed complications (6 cases of sepsis, 2 cases of abscesses, and 1 case each of infected fluid collection, pancreatitis, and a Mallory-Weiss tear).19,24,30,33,34 All patients recovered successfully without surgical intervention.
Pseudocysts In total, 140 patients (15%) developed pseudocysts with sufficient available data for further analysis in 85 patients (61%). There were 76 patients (18%) who developed pseudocysts after nonoperative management, compared with 9 patients (4%) who © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
developed pseudocysts after operative management Fisher’s exact test, P < 0.01; number needed to treat = 7.14). We were able to identify a total of 22 patients (9%) with minor pancreatic injury who developed pseudocysts after nonoperative management and 1 patient (2%) who developed a pseudocyst after operative management (Fisher’s exact test, P = 0.14; number needed to treat = 14.3). Pseudocysts developed in 54 patients (32%) with major injuries after nonoperative management versus 8 patients (4%) after operative treatment (Fisher’s exact test, P < 0.01; number needed to treat = 3.6) (Fig. 2). Percutaneous drainage of the pseudocyst was performed in 26 patients (31%), whereas endoscopic drainage was performed in 9 patients (11%). In 8 patients (23%), percutaneous or endoscopic drainage was unsuccessful. A total of 29 patients (35%) had surgical cystogastrostomy.
Total Parenteral Nutrition Eight studies reported on the use of TPN. On average, TPN was administered in 62% of patients for a period of 15.1 days. The TPN was administered in 65% of patients treated nonoperatively for an average of 19.4 days and 68% of patients treated surgically for an average of 10 days. The median TPN duration between nonoperative and operative management was similar (nonparametric t test, P = 0.10)
Length of Stay Thirteen studies reported on length of stay, which overall was mean 21.7 (standard deviation, 12.2) days. The median (interquartile range) hospitalization was 20.5 (11.3–37.6) days after nonoperative management in comparison with 15.1 (12–20) days after operative management (nonparametric t-test, P = 0.414).
Mortality Mortality occurred in 49 patients (5%), often as a result of associated injuries. There were 4 patients who died of traumatic brain injury.15,17,29 Another 10 patients died of either traumatic brain injury or multiorgan failure.31 Four patients died because of pancreatic injuries.18,23 One patient had a grade V injury, although the severity of injury in the other patients is unknown. Of the remaining 31 patients, the exact cause of death could not be extracted. However, both studies reported that these deaths were not attributable to pancreatic injury.18,25
Follow-Up In total, 3 studies with 37 patients reported follow-up results. One study reported excellent follow-up results of patients treated nonoperatively; after 47 months follow-up, none of the 9 patients had exocrine or endocrine insufficiency.19 Another study reported uneventful follow-up in 15 patients treated nonoperatively, 1 of which eventually required surgery, without any pancreatic insufficiency after a median follow-up of 2 years.22 Snajdauf et al21 also reported excellent follow-up results of operative management in a series in which distal pancreatectomy was performed in 6 patients, pancreaticoduodenectomy in 4 patients, and cystogastrostomy in 3 patients. In a follow-up ranging from 1 to 12 years, none of the 13 patients had either exocrine or endocrine insufficiency.21
DISCUSSION This systematic review comparing nonoperative and operative management demonstrates that the vast majority of children with grade I or II blunt pancreatic injuries and about half of patients with grade III – V injuries can be managed nonoperatively with excellent success rates. Although pseudocysts are more © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Management of Blunt Pancreatic Trauma in Children
common (18%) in patients managed nonoperatively, these can be treated nonoperatively in two thirds of patients. Nonoperative management does not lead to prolonged hospital stay. A previous Cochrane systematic review failed to include any studies because the authors were only interested in randomized trials, which are nonexistent in this topic.12 Therefore the present review brings together the best available evidence, but at the same time acknowledging that these are all retrospective studies with a significant risk of selection bias. None of the included studies reported on specific selection criteria for operative management. Little is known about the epidemiology of blunt pancreatic trauma in children. Our results suggest that pancreatic trauma is more common in males and that almost a quarter of these injuries are caused by bicycle accidents. No relationship between age and severity of injury could be identified.21,22,27,32,34 Low body weight has been associated with these injuries.22 Data from the included studies showed an almost equal distribution of minor and major pancreatic injuries, but it is important to consider that minor injuries may be underrepresented in the literature because these patients may not always seek treatment. According to current guidelines, identifying ductal injury is crucial to determine which type of intervention is indicated. Computed tomography and magnetic resonance imaging both have shown high accuracy in assessing pancreatic injury.36 The diagnostic accuracy of computed tomography is reported to be 75%.37 Endoscopic retrograde cholangiopancreatography is an accurate method of investigating the pancreatic duct but is less attractive because it is invasive with associated risks such as pancreatitis. Magnetic resonance cholangiopancreatography has been reported as an accurate, noninvasive method of determining whether ductal injury is present.38,39 Because of the low incidence of blunt pancreatic trauma in children, large, high-quality studies, such as randomized trials, are not available and will probably never be performed. This review suggests that many patients can be managed nonoperatively. However, this comes at a cost of a higher rate of pseudocysts after nonoperative management (18% vs 4% in this review). Nevertheless, there are many concerns about operative management as well. Patients may develop endo- or exocrine insufficiency after distal pancreatectomy. Surprisingly, this was not reported by studies included in this systematic review. Studies have shown that after elective pancreatic surgery in children, postoperative diabetes occurs in less than 10% of children, whereas exocrine dysfunction occurs in up to 60% of children.40–43 Overall, pseudocysts developed in 142 children (15%) with blunt pancreatic trauma. Currently, there are several options for managing these pseudocysts. Observation is an acceptable initial step, because pseudocysts may spontaneously resolve.44 In case of persistent symptoms or infection, interventional radiology, endoscopy, or surgery can be used to drain the pseudocyst internally. Both radiologic and endoscopic drainages show good results, and technical success is over 90%.44–47 When comparing both techniques, endoscopic drainage seems to have a greater procedural risk but has a lower risk of reinterventions, residual collections, and the need for subsequent surgery due to internal drainage of pseudocysts.48 Laparoscopic techniques also show favorable results, although selection bias may be present.49 This review has several limitations. First, all included studies were retrospective and hence selection bias has played a clear role in the decision. Second, the heterogeneous nature of the studies led to frequent missing data. Despite contacting corresponding authors, we were unable to attain all missing data. Therefore, we were unable to compare outcomes according to overall injury severity using the injury severity score and among different www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1095
Pancreas • Volume 46, Number 9, October 2017
Koh et al
age groups. Lastly, not all studies had sufficient follow-up to determine whether pseudocysts developed or not.
CONCLUSIONS In conclusion, blunt pancreatic trauma in children can be managed nonoperatively in most patients with high success rates. Higher injury severity is associated with an increased need for surgical intervention, although selection criteria are lacking and should be developed. Pseudocysts are more common after nonoperative management but do not require surgery in two thirds of patients. The exact role of endoscopy and interventional radiology in these patients remains unclear. Further studies are warranted to validate the risks and benefits of endoscopy, interventional radiology, or surgery for these patients to establish definitive guidelines.
14. Howick J, Chalmers I, Glasziou P, et al. OCEBM Levels of Evidence Working Group. The Oxford 2011 Levels of Evidence. 2011. Available at: http://www.cebm.net/ocebm-levels-of-evidence/. Accessed September 29, 2016. 15. Arkovitz MS, Johnson N, Garcia VF. Pancreatic trauma in children: mechanisms of injury. J Trauma. 1997;42:49–53. 16. Firstenberg MS, Volsko TA, Sivit C, et al. Selective management of pediatric pancreatic injuries. J Pediatr Surg. 1999;34:1142–1147. 17. Canty TG Sr, Weinman D. Management of major pancreatic duct injuries in children. J Trauma. 2001;50:1001–1007. 18. Jacombs AS, Wines M, Holland A, et al. Pancreatic trauma in children. J Pediatr Surg. 2004;39:96–99. 19. Wales PW, Shuckett B, Kim PC. Long-term outcome after nonoperative management of complete traumatic pancreatic transection in children. J Pediatr Surg. 2001;36:823–827. 20. Stringer MD. Pancreatic trauma in children. Br J Surg. 2005;92:467–470.
ACKNOWLEDGMENTS The authors thank Dr Mary Brindle (Department of Surgery, Alberta Children's Hospital, Calgary, Canada) for providing additional data from her original study. REFERENCES 1. Holmes JH 4th, Wiebe DJ, Tataria M, et al. The failure of nonoperative management in pediatric solid organ injury: a multi-institutional experience. J Trauma. 2005;59:1309–1313.
21. Snajdauf J, Rygl M, Kalousová J, et al. Surgical management of major pancreatic injury in children. Eur J Pediatr Surg. 2007;17:317–321. 22. Houben CH, Ade-Ajayi N, Patel S, et al. Traumatic pancreatic duct injury in children: minimally invasive approach to management. J Pediatr Surg. 2007;42:629–635. 23. Vane DW, Kiankhooy A, Sartorelli KH, et al. Initial resection of potentially viable tissue is not optimal treatment for grades II-IV pancreatic injuries. World J Surg. 2009;33:221–227. 24. Jurić I, Pogorelić Z, Biocić M, et al. Management of blunt pancreatic trauma in children. Surg Today. 2009;39:115–119.
2. Bradley EL 3rd, Young PR Jr, Chang MC, et al. Diagnosis and initial management of blunt pancreatic trauma: guidelines from a multiinstitutional review. Ann Surg. 1998;227:861–869.
25. Sutherland I, Ledder O, Crameri J, et al. Pancreatic trauma in children. Pediatr Surg Int. 2010;26:1201–1206.
3. Lin BC, Chen RJ, Fang JF, et al. Management of blunt major pancreatic injury. J Trauma. 2004;56:774–778.
26. Wood JH, Partrick DA, Bruny JL, et al. Operative vs nonoperative management of blunt pancreatic trauma in children. J Pediatr Surg. 2010; 45:401–406.
4. Dervan LA, King MA, Cuschieri J, et al. Pediatric solid organ injury operative interventions and outcomes at Harborview Medical Center, before and after introduction of a solid organ injury pathway for pediatrics. J Trauma Acute Care Surg. 2015;79:215–220. 5. Partrick DA, Moore EE, Bensard DD, et al. Operative management of injured children at an adult level I trauma center. J Trauma. 2000;48: 894–901. 6. Mattix KD, Tataria M, Holmes J, et al. Pediatric pancreatic trauma: predictors of nonoperative management failure and associated outcomes. J Pediatr Surg. 2007;42:340–344. 7. Moore EE, Cogbill TH, Malangoni MA, et al. Organ injury scaling, II: pancreas, duodenum, small bowel, colon, and rectum. J Trauma. 1990;30: 1427–1429. 8. Biffl WL, Moore EE, Croce M, et al. Western Trauma Association critical decisions in trauma: management of pancreatic injuries. J Trauma Acute Care Surg. 2013;75:941–946. 9. Ho VP, Patel NJ, Bokhari F, et al. Management of adult pancreatic injuries: a practice management guideline from the Eastern Association for the Surgery of Trauma. J Trauma Acute Care Surg. 2017;82:185–199. 10. Iqbal CW, St Peter SD, Tsao K, et al. Operative vs nonoperative management for blunt pancreatic transection in children: multi-institutional outcomes. J Am Coll Surg. 2014;218:157–162. 11. Abbo O, Lemandat A, Reina N, et al. Conservative management of blunt pancreatic trauma in children: a single center experience. Eur J Pediatr Surg. 2013;23:470–473.
27. Klin B, Abu-Kishk I, Jeroukhimov I, et al. Blunt pancreatic trauma in children. Surg Today. 2011;41:946–954. 28. Paul MD, Mooney DP. The management of pancreatic injuries in children: operate or observe. J Pediatr Surg. 2011;46:1140–1143. 29. Borkon MJ, Morrow SE, Koehler EA, et al. Operative intervention for complete pancreatic transection in children sustaining blunt abdominal trauma: revisiting an organ salvage technique. Am Surg. 2011;77:612–620. 30. Cigdem MK, Senturk S, Onen A, et al. Nonoperative management of pancreatic injuries in pediatric patients. Surg Today. 2011;41:655–659. 31. Cuenca AG, Islam S. Pediatric pancreatic trauma: trending toward nonoperative management? Am Surg. 2012;78:1204–1210. 32. Yamana I, Kawamoto S, Kamitani T, et al. Pancreatic injury in children: review of 7 cases and the pertinent literature. Hepatogastroenterology. 2012;59:574–577. 33. Beres AL, Wales PW, Christison-Lagay ER, et al. Non-operative management of high-grade pancreatic trauma: is it worth the wait? J Pediatr Surg. 2013;48:1060–1064. 34. Garvey EM, Haakinson DJ, McOmber M, et al. Role of ERCP in pediatric blunt abdominal trauma: a case series at a level one pediatric trauma center. J Pediatr Surg. 2015;50:335–338. 35. Dai LN, Chen CD, Lin XK, et al. Abdominal injuries involving bicycle handlebars in 219 children: results of 8-year follow-up. Eur J Trauma Emerg Surg. 2015;41:551–555.
12. Haugaard MV, Wettergren A, Hillingsø JG, et al. Non-operative versus operative treatment for blunt pancreatic trauma in children. Cochrane Database Syst Rev. 2014;2:CD009746.
36. Panda A, Kumar A, Gamanagatti S, et al. Evaluation of diagnostic utility of multidetector computed tomography and magnetic resonance imaging in blunt pancreatic trauma: a prospective study. Acta Radiol. 2015;56: 387–396.
13. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Ann Intern Med. 2009;151:264–269, W264.
37. Moschetta M, Telegrafo M, Malagnino V, et al. Pancreatic trauma: the role of computed tomography for guiding therapeutic approach. World J Radiol. 2015;7:415–420.
1096
www.pancreasjournal.com
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
Pancreas • Volume 46, Number 9, October 2017
38. Fulcher AS, Turner MA, Yelon JA, et al. Magnetic resonance cholangiopancreatography (MRCP) in the assessment of pancreatic duct trauma and its sequelae: preliminary findings. J Trauma. 2000;48: 1001–1007.
Management of Blunt Pancreatic Trauma in Children
44. Russell KW, Barnhart DC, Madden J, et al. Non-operative treatment versus percutaneous drainage of pancreatic pseudocysts in children. Pediatr Surg Int. 2013;29:305–310.
39. Nirula R, Velmahos GC, Demetriades D. Magnetic resonance cholangiopancreatography in pancreatic trauma: a new diagnostic modality? J Trauma. 1999;47:585–587.
45. Akshintala VS, Saxena P, Zaheer A, et al. A comparative evaluation of outcomes of endoscopic versus percutaneous drainage for symptomatic pancreatic pseudocysts. Gastrointest Endosc. 2014;79:921–928; quiz 983.e2, 983.e5.
40. Marchegiani G, Crippa S, Malleo G, et al. Surgical treatment of pancreatic tumors in childhood and adolescence: uncommon neoplasms with favorable outcome. Pancreatology. 2011;11:383–389.
46. Coelho DE, Ardengh JC, Carbalo MT, et al. Clinicopathologic characteristics and endoscopic treatment of post-traumatic pancreatic pseudocysts. Pancreas. 2011;40:469–473.
41. Muller CO, Guérin F, Goldzmidt D, et al. Pancreatic resections for solid or cystic pancreatic masses in children. J Pediatr Gastroenterol Nutr. 2012;54: 369–373.
47. Thomson DA, Krige JE, Thomson SR, et al. The role of endoscopic retrograde pancreatography in pancreatic trauma: a critical appraisal of 48 patients treated at a tertiary institution. J Trauma Acute Care Surg. 2014; 76:1362–1366.
42. Park HH, Kim HY, Jung SE, et al. Long-term functional outcomes of PPPD in children—nutritional status, pancreatic function, GI function and QOL. J Pediatr Surg. 2016;51:398–402. 43. Sugito K, Furuya T, Kaneda H, et al. Long-term follow-up of nutritional status, pancreatic function, and morphological changes of the pancreatic remnant after pancreatic tumor resection in children. Pancreas. 2012;41: 554–559.
© 2017 Wolters Kluwer Health, Inc. All rights reserved.
48. Keane MG, Sze SF, Cieplik N, et al. Endoscopic versus percutaneous drainage of symptomatic pancreatic fluid collections: a 14-year experience from a tertiary hepatobiliary centre. Surg Endosc. 2016;30:3730–3740. 49. Palanivelu C, Senthilkumar K, Madhankumar MV, et al. Management of pancreatic pseudocyst in the era of laparoscopic surgery—experience from a tertiary centre. Surg Endosc. 2007;21:2262–2267.
www.pancreasjournal.com
Copyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.
1097
