Surgery of pancreas tumors in pediatric and adolescent patients: a single institution experience in South America

Pancreas tumors are extremely rare in pediatric and adolescent patients. Surgical resection is the mainstay of treatment; however, the data are limited with respect to morbidity and mortality. We aimed to evaluate short- and long-term outcomes of pediatric and adolescent patients who underwent surgical resection of pancreatic tumors. Patients ≤\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\le$$\end{document} 18-year-olds who underwent resection of pancreas tumor at the National Institute of Neoplastic Diseases INEN during 2000–2020 were included. Thirty-four patients were diagnosed; 28 patients were female and 6 were male. The median age was 13.4-years-old. Histological diagnosis was solid pseudopapillary neoplasm (SPN) (n = 29, 85.3%), pancreatoblastoma (n = 3), neuroendocrine carcinoma (n = 1), and insulinoma (n = 1). No patient experienced postoperative mortality and 15 (44.1%) patients developed postoperative complications including pancreatic fistula as the most frequent. Under a median follow-up period of 33.8 (0.5–138) months, four (11.8%) patients died. Of the 29 patients with SPN, the 3- and-5-year OS rates were 100% and 83.1%, respectively. SPN was the most frequent cause of surgical treatment for pediatric and adolescent patients in the high-volume cancer center in Peru and was associated with favorable survival. Pancreaticoduodenectomy was safely performed in this patient group with acceptable morbidity and zero mortality.


Introduction
Pancreas tumors are extremely rare in pediatric and adolescent patients. Solid pseudopapillary neoplasm (SPN) and pancreatoblastoma are relatively common in this patient group [1,2]. SPN was the most frequent pancreas tumor in patients with 10-20-years-old, and pancreatoblastoma was generally found in patients < 10 years-old [3,4]. Other histological types including neuroendocrine tumors (NET), acinar carcinoma, ductal carcinoma, neuroblastoma, and lymphoma were reported in pediatric and adolescent patients [2,[5][6][7]. Surgical resection is the mainstay of treatment and provides a better survival and lower recurrence rate [5]. Generally, radiotherapy and chemotherapy were used as adjuvant treatments. However, the data are limited with respect to the rates of morbidity and mortality in pediatric and adolescent patients who underwent pancreatectomy [8]. The aim of this study is to evaluate clinical and pathological characteristics, short-term (morbidity and mortality) and long-term outcomes (overall and disease-free survival) in pediatric and adolescent patients who underwent pancreas resection for benign and malignant pancreas tumors.

Study population
Patients who underwent resection of pancreas benign and malignant primary tumors with curative intent at the National Institute of Neoplastic Diseases INEN from January 2000 to January 2020 were identified from a prospectively compiled database. Patients who underwent surgery at the age ≤ 18-years-old (i.e., pediatric and adolescent patients) were included in the study. Demographic characteristics and short-and long-term outcomes were evaluated. Pancreas tumors were diagnosed based on the current WHO classification [9]. The study was approved by the institutional review board at the National Institute of Neoplastic Diseases INEN.

Preoperative evaluation
All patients underwent computed tomography of the thorax, abdomen, and pelvis. A preoperative tumor biopsy was not routinely performed. Blood test including a complete blood count, hepatic, renal function, and a coagulation profile was performed. A cardiological, psychological, and nutritional evaluation was performed routinely before surgery. No patients received neoadjuvant chemotherapy and radiotherapy.

Surgical technique
The surgical procedure was performed based upon the tumor location and involvement of adjacent organs. For tumors located at the pancreas head, pancreaticoduodenectomy (PD) was performed using the conventional or pylorus-preserving method. D2 lymph node dissection was generally performed. For patients with pancreatic tumors invaded to the portal vein and/or superior mesenteric vein, combined resection of the vessels and reconstruction were performed. The anastomosis of pancreas stump was performed depending on pancreas texture and main pancreatic duct diameter. In this series, we used duct to mucosa or modified telescopic anastomosis [10,11]. For tumors located at the pancreas tail, distal pancreatectomy was performed, with or without splenectomy. For spleen-preserving distal pancreatectomy, Warshaw or Kimura technique was used depending on the surgeon's preferences [12,13]. Central pancreatectomy was performed for tumors located at the neck and proximal body of the pancreas and the distal pancreatic stump was anastomosed using a Roux-en-Y jejunal loop or a pancreatic-gastric anastomosis [14]. For small tumors which are located away from the main pancreatic duct, enucleation was selected [15].

Postoperative management
An enteral feeding tube was routinely used in patients who underwent PD and central pancreatectomy. In patients who underwent distal pancreatectomy, an enteral feeding tube was only used in case of malnutrition and albumin < 35 g/ dL. At the postoperative day (POD) 1 or 2 enteral nutrition was started through an enteral feeding tube, oral intake was started at POD 3 in patients who underwent an enteral anastomosis. Amylase levels in blood and fluid from abdominal drains were measured at POD 1, 3, 5, and 7. The discharge amount from abdominal drains was measured. When patients had pancreatic fistula, somatostatin analog was used [16]. Abdominal drains were removed when amylase level was < 5000 U/L; after POD 5 or on the day of patient discharge. Postoperative morbidity and mortality were defined as complications and mortality which occurred within 30 days after surgery. All complications were categorized according to the Clavien-Dindo classification [17].

Variable studied
Demographic characteristics (age, symptoms, tumor location, tumor size, and histology), surgical and pathological outcomes (surgical procedure, type of anastomosis, and nodal status), postoperative outcomes (hospital stay and postoperative complications), and long-term outcomes were evaluated. Tumors were classified according to the current WHO classification for pancreatic tumors [9].

Statistical analysis
Categorical variables are expressed as number (%) and continuous variables are expressed as median (ranges). Overall survival (OS) and recurrence-free survival (RFS) curves were constructed using the Kaplan-Meier method. Deaths without recurrence were censored for the RFS analysis. A p value < 0.05 was considered statistically significant. Statistical analysis was conducted using SPSS 22.

Discussion
Primary pancreatic tumors are rare in pediatric and adolescent patients. Our study showed outcomes of the largest series in South America and the second largest single institution series worldwide. A total of 34 patients underwent resection without mortality (Table 4).
In line with previous studies [1,2,5,8,18], the proportion of female sex was higher (82.4%) and the most common histology type was SPN in this population. Most of the patients were ≥ 10-years-old (88.2%). The most frequent symptom was abdominal pain. These were also in line with previous studies, showing that the demographic characteristics of pediatric and adolescent pancreatic tumors may be common regardless of the regions worldwide [1,2,19,20].
SPN is a low grade malignant pancreas tumors which contain solid and cystic components and mainly found in female patients in the 2nd to 4th decades of life [4]. A maleto-female ratio in incidence of SPN was 1:9.5. The size of SPN is generally smaller in male patients than in female patients, and SPN diagnosed in male patients mainly contains solid components [19]. In pediatric and adolescent patients, SPN is generally found at the age from 10-to 20-years-old [4,21,22]. Our study showed that the 3-and 5-year OS rates were 100% and 83.1% in pediatric and adolescent patients with SPN, respectively. This is in line with previous studies which showed that the OS in pediatric and adolescent patients with SPN was more than 95% and was better than OS in this patient group with other histology types of pancreas tumors [1,5,8,23].
Pancreatoblastoma is the second most common histology type in our study. Pancreatoblastoma was originally termed as infantile pancreatic carcinoma by Becker in 1957 [24] and is an extremely rare pancreas neoplasm which contains typical squamous corpuscles and tumor cells with acinar, glandular, or undifferentiated appearance [25]. It is predominant in male patients who are from 0 to 20-years-old and has an association with Beckwith-Wiedemann syndrome [3,26]. Serum alpha-fetoprotein level is elevated in most patients. Mylonas, et al. reported 21 pediatric and adolescent patients with pancreatoblastoma. The median age of the patients 5.5-years-old at the time of diagnosis [22]. In our study, of the three patients, two are alive with no evidence of recurrence after surgery and one died because of hepatic recurrence at 46 months after surgery. Bien, et al. reported 20 pediatric and adolescent patients with pancreatoblastoma in a multicenter study [27]. The 5-year OS was 79.4% and the 5-year event-free survival was 58.8%. Complete surgical resection was a prognostic factor for a better event-free survival in univariate analysis.
The largest single institution retrospective study about this topic was reported in Asia [2]. The study included 104 pediatric and adolescent patients between 2007 and 2018. Other studies included approximately 30 patients mainly from North America. Our study is the largest series in the population of South America and found that female sex is dominant and SPN was the most common histology types in line with previous studies in other regions. Mylonas et al. [22], reported a retrospective multicenter study using Surveillance, Epidemiology and, End Results (SEER) database, and included 114 pediatric and adolescent patients from 1973 to 2013. In this study, the most common histology type was NET with 40 (35%) patients followed by epithelial tumors with 29 (25%). A total of 73 (64%) patients had surgical treatment. Another study using the SEER database from 1973 to 2004, showing that the incidence of pancreas tumors in pediatric and adolescent patients was 0.018 per 100,000 people [5].
Pancreaticoduodenectomy was the most frequent surgical procedure in our study which accounts for 19 patients (55.9%). Pancreaticoduodenectomy in pediatric and adolescent patients had been reported but the number of cases was low [28]. In children, pylorus-preserving pancreaticoduodenectomy may be considered to preserve gastric function [29,30]. Mansfield et al. reported that morbidity rate after pancreaticoduodenectomy was lower in patients ≤ 30-years-old than in adult patients [31]. In our study, no patient experienced postoperative mortality and 15 patients (44.1%) developed postoperative complications including 5 patients (14.7%) who had major complications. Pancreatic fistula was found in six (31.5%) patients undergoing pancreaticoduodenectomy versus one (10%) patient undergoing distal pancreatectomy.
Our study is a retrospective study with a small number of patients. Nonetheless, given the rarity of pancreas tumors in pediatric and adolescent patients, our study is the second largest series worldwide and showed outcomes of pancreas tumors for this patient group in South America. Another limitation is that the functional change after pancreatectomy was not followed because most patients lived in remote areas far from our institution.
In conclusion, SPN was the most frequent cause of surgical treatment for pediatric and adolescent patients in the high-volume cancer center in Peru and was associated with favorable survival. Pancreaticoduodenectomy was safely performed in this patient group with acceptable morbidity and zero mortality.