Nonresection management of the pancreas for grade III and IV blunt pancreatic injuries in children: a single center’s experience

This study adheres to the ethical principles of the Declaration of Helsinki. After approval of informed consent for consent waiver from the Ethics Committee of Beijing Children’s Hospital (2020-k-12), the medical record system at Children’s National Medical Center, China, was retrospectively searched for children diagnosed with “pancreatic injury” between January 2010 and June 2018. Their medical information, including demographics, imaging data and treatments, was analyzed. Children who were diagnosed with high-grade pancreatic injuries (n = 23) based on guidelines from the Organ Injury Scale grading of the American Association for the Surgery of Trauma (AAST-OIS) were included (Table 1) [5].

All children with suspected pancreatic injury were first examined in our center by the abdominal US. Repeat US was performed to assess progress. Regardless of positive US findings, contrast-enhanced computed tomography (CE-CT) or magnetic resonance imaging (MRI) including magnetic resonance cholangiopancreatography (MRCP), was performed to assess the grade of blunt pancreatic injury for 24–48 h after the vital signs became stable. Typical CE-CT, MRI and MRCP images of grade III and IV blunt pancreatic injuries are shown in Fig. 1. Endoscopic retrograde cholangiopancreatography (ERCP) is not routinely carried out in our center.

The children with blunt pancreatic injury were admitted to the Department of General Surgery in our center. The treatments were performed in accordance with the center’s regulations (Fig. 2). Except for children with gastrointestinal injury or continued hemorrhagic shock after initial blood transfusion, all children received conservative treatments, nil-per-os (NPO), somatostatin injected continuously with a microinfusion pump until no symptoms after oral intake or treatment by external drainage and internal drainage (3.5–5 μg/kg/h in children until the same dosage as adults), intravenous omeprazole inhibition of acid secretion until oral intake (1–2 mg/kg in children until the same dosage as adults), anti-infection treatment, total parenteral nutrition (TPN) and enteral nutrition (EN). Severe abdominal infection was mainly based on abdominal pain, high fever, signs of peritonitis and elevated laboratory indicators including WBC, CRP and procalcitonin after conservative treatment. It was relieved by simple, not continuous peritoneal lavage and drainage, which was achieved through open or laparoscopic surgical techniques, and included opening the lesser omental sac, peritoneal irrigation, and placing the abdominal drainage tube [6‐8]. For children who developed immature pancreatic pseudocysts (PCs) during conservative treatment, external drainage was performed after US-guided percutaneous drainage failed (Fig. 3). The indication for US-guided percutaneous drainage was enlargement measured over 5 cm in diameter or threatening rupture of the acute pseudocysts and chronic pseudocysts causing abdominal symptoms or recurring fever [9, 10]. The indication for external drainage was US localization failure, no relief of symptoms, and no change or enlargement in PCs after US-guided percutaneous drainage. The criteria for removing external drainage included the following: 1) drainage fluid < 5 ml per day during oral intake; 2) no symptoms when the drainage tube was clipped for 48–72 h during oral intake; and 3) no change in PCs monitored by the US when the drainage tube was clipped for 48–72 h during oral intake. For children with mature PCs after conservative treatment, internal drainage, such as gastric-cyst and jejunum-cyst anastomosis, was performed.

Our primary outcome was treatment strategy, as defined above, including pancreatic tissue resection and operative interventions. Peritoneal lavage and drainage, external drainage and internal drainage were considered operative interventions. Secondary outcomes were mortality and major complications, including abdominal infection and PCs. Length of stay was also evaluated. The discharge criteria included the following: 1) no symptoms during oral intake; 2) US indicated a PC diameter less than 2 cm and no significant increase after oral intake; and 3) no evidence of infection based on laboratory findings. Normalization of serum amylase level was not included as an essential discharge criterion unless the value was over 200 U/L due to its limited value to reflect damage to the pancreas [11‐13]. All children were followed up regularly at 2, 4, 8 and 12 weeks after discharge in the outpatient department to monitor serum amylase, serum lipase and blood glucose, as well as to perform the abdominal US. The long-term prognosis was followed up by telephone in February 2020.

During the study period, 20 children with grade III (16/20, 80.0%) or IV (4/20, 20.0%) blunt pancreatic injuries met the inclusion criteria. Children with biliary tract injury (n = 1), who died without treatment (n = 1), or who refused treatment (n = 1) were excluded (Fig. 4). The average age at the time of injury was 6.9 (3.3–12.3) years and 70.0% were males. The most common cause of injury was traffic accidents (8/20, 40.0%). Of these patients, 13 (13/20, 65.0%) had isolated pancreatic injuries and 7 (7/20, 35.0%) had pancreatic injuries combined with organ injuries (Table 2). The median time for the diagnosis of pancreatic injury was 2.5 (1–30) days. Thirteen (13/20, 65.0%) patients had a delayed diagnosis (≥24 h after injury). All children received the initial US with an accuracy of 65.0% (13/20), and there were no significant differences in accuracy between children with isolated injury and those with combined pancreatic injury (9/13 vs. 4/7, P = 0.651). Sixteen children received CT with an accuracy of 68.8% (11/16), and there was a significant difference in accuracy between children with isolated injury and those with combined pancreatic injury (9/9 vs. 2/7, P = 0.005). Five patients with isolated pancreatic injury received MRI, with an accuracy of 100.0% (5/5) (Table 3).

Nine children who presented with intractable symptoms, such as advanced peritonitis and PCs, after treatments at external hospitals were transferred to our center. One patient received emergency surgery at an external hospital due to a diagnosis of appendicitis. The patient was found to have a pancreatic injury during the operation and received abdominal drainage, which led to the development of PCs. The patient ultimately recovered through percutaneous drainage performed in our center. Another patient received emergency surgery due to a preoperative diagnosis of intestinal perforation but was not found to have a pancreatic injury at another external hospital, which resulted in severe abdominal infection. The patient later recovered through peritoneal lavage and drainage performed in our center. The remaining 7 patients were transferred to our center due to the development of PCs after conservative treatment at external hospitals, of whom 1 recovered after conservative treatment, 1 recovered through percutaneous drainage, 3 recovered through external drainage, and 2 had longer postinjury periods (45 days and 80 days) recovered at our center after receiving a gastric-cyst and jejunum-cyst anastomosis. The average length of stay of the above patients in our center was 33.8 (8–63) days.

Eleven children were admitted to our hospital directly after pancreatic injury. One child who had continued hemorrhagic shock ultimately received an emergency laparotomy and underwent Roux-en-Y drainage of the distal pancreas. The remaining 10 children received conservative treatment, among whom 7 developed PCs, 2 developed an abdominal infection and 1 recovered uneventfully. Among the patients who developed PCs, 3 recovered after conservative treatment, 2 recovered through percutaneous drainage, and 2 recovered through external drainage. The patients who developed severe abdominal infection recovered after peritoneal lavage and drainage. The average length of stay of the above patients in our center was 47.5 (23–69) days.

None of the children underwent pancreatic tissue resection. Fifteen (15/20, 75.0%) developed PCs, and 7 (7/15, 46.7%) of them received an operative intervention, including 5 external drainage and 2 internal drainage. Three (3/20, 15.0%) developed an abdominal infection, and they all recovered after abdominal irrigation and drainage. No patient was admitted to the ICU, and there was no death and no incidence of impaired pancreatic function during hospitalization.

All children were regularly followed up for 12 weeks after discharge, and 4 (4/20, 20.0%) were diagnosed with local pancreatic atrophy by the abdominal US, without obvious abnormalities in serum amylase, serum lipase or blood glucose. With a mean follow-up time of 7.8 (1.7–10.1) years after discharge, no other long-term complications were observed.