Since the first description of solid pseudopapillary tumor (SPT) by Frantz in 1959, SPT is considered a low-grade malignant neoplasm that predominantly affects adolescent girls and young women [12]. Although SPT is rare, its incidence has been steadily increasing for the last two decades [345]. SPT accounts for approximately 1%–3% of all primary exocrine pancreatic tumors [678] and classically presents as a solitary, large, well-demarcated lesion which can be purely cystic, solid, or mixed in ultrasonography or computed tomography (CT) [9]. These neoplasms rarely metastasize, with an incidence of 9%–15%. Common sites include the liver, peritoneum, greater omentum, and regional lymph nodes [310].

The mainstay treatment of SPT in pediatric patients is surgical resection with excellent prognosis [3611]. Surgical plans will vary according to the tumor location. Pancreaticoduodenectomy (PD) or pylorus-preserving pancreatoduodenectomy (PPPD) is performed for pancreatic head tumors and distal pancreatectomy (DP) for pancreatic body or tail tumors. Furthermore, enucleation may be a treatment option for biopsy-proven SPT without main pancreatic duct involvement [1213].

Open pancreatectomy (OP) is still primarily considered in large tumors since huge SPT can abut nearby structures obscuring dissection. However, with increasing expertise and feasibility in the era of minimally invasive surgery (MIS), laparoscopic pancreatectomy (LP) is widely and increasingly performed in pediatric patients [14151617]. However, pediatric patients with large tumor size should be carefully selected for MIS because of the high chance of open conversion due to bleeding, vessel invasion and comparatively small abdominal cavity. Currently, evidence and guidelines that support the feasibility and safety of MIS in pediatric patients are limited.

In this study, we aim to evaluate the perioperative and long-term outcomes of MIS and to establish indications of MIS in SPT in pediatric patients.

From October 1992 to April 2018, 66 patients (age, <18 years) diagnosed with SPT underwent either OP or MIS. We retrospectively reviewed patient's clinical characteristics, surgical and pathologic data, and surgical outcome from medical records. This study reviewed by Institutional Review Board of Asan Medical Center (approval number: 2018-1317). SPT had been diagnosed pathologically on the basis of gross and microscopic appearances of each tumor. In addition, all specimens had been assessed through immunohistochemical staining, including staining with antibodies against beta-catenin, E-cadherin, and synaptophysin, to confirm the diagnosis. Data variables included patient demographics such as age and sex, as well as their clinical presentation, tumor size, tumor grade, metastasis, tumor markers, surgery type, complications, hospital stay, time to oral feeding, duration of follow-up, and status at the end of follow-up.

To evaluate the difficulty of MIS in each patient, we investigated tumor size and mass size/abdominal diameter (MS/AD) ratio. The degree of vascular abutment in the portal vein (PV), superior mesenteric vein (SMV), and splenic vein (SV) was also calculated. These parameters were measured using the preexisting abdominal CT. Since the size of the abdominal cavity in children varied by age and physical condition, the tumor size alone does not directly reflect the degree of operative difficulty. We measured the width of the abdominal cavity (abdominal diameter [AD]) at the level of the largest point of each tumor and calculated relative figures for the standardization of tumor sizes by the size of the abdominal cavity. We also calibrated the degree of compression in each vessel around each tumor (Fig. 1).

SPSS software version 23.0 for Windows (SPSS Inc., Chicago, IL, USA) was used for statistical analysis. Significance level was set at a p-value of less than 0.05.

Sixty-six patients underwent pancreatectomy for SPT. Surgical procedures in these patients are described in Table 1 and Fig. 2. Thirty-five patients underwent open surgery, whereas 31 patients underwent MIS; whom 2 underwent robotic DP and 29 underwent LP. Patient demographics, as summarized in Table 2, showed no difference in proportion in sex, age, weight, and most of the preoperative laboratory findings. Amylase levels were significantly higher in the open surgery group (p=0.040). Lipase levels showed no statistical significance but tended to be higher in the open surgery group (p=0.059).

Duration of operation in MIS was significantly shorter in the OP group (p=0.005) (Table 3). Furthermore, tumor size in the MIS group was significantly smaller than the OP group (p<0.001), and hospital stay was longer in the OP group than in the MIS group (p=0.001). Time to initial oral intake and time to full feeding was significantly shorter in the MIS group (p=0.027 and 0.010, respectively). Although postoperative pancreatic fistula showed statistical value (p=0.006), most patients experienced International Study Group for Pancreatic Fistula (ISGPF) grade A, which had no effect on patient clinical course. All of our patients with ISGPF grade A had elevated Jackson-Pratt drain amylase levels but all received routine postoperative management. Moreover, ISGPF grade B showed no statistical value (p<0.638). Other postoperative complications including postoperative fluid collection, splenic infarct, hematoma, PV thrombus, and chylous drainage showed no difference. For solid pseudopapillary carcinoma (SPC), there were 6 cases of whom four had open surgery and 2 had laparoscopic surgery (p=0.662). Invasion in the capsule, soft tissue, perineural tissue, and lymphovascular tissue showed no statistical difference (Table 3).

The median tumor size and MS/AD ratio in patients undergoing enucleation was larger in the OP group (7.8±2.8 cm vs. 4.1±1.6 cm, p=0.007 and 0.40±0.12 cm vs. 0.20±0.08 cm, p=0.002, respectively; Table 4). Furthermore, the median tumor size and MS/AD ratio was significantly greater in the OP than in the MIS DP/spleen sparing distal pancreatectomy (SSDP) group (8.8±4.0 cm vs. 4.4±2.0 cm, p=0.002 and 0.33±0.16 cm vs. 0.20±0.08 cm, p=0.440, respectively).

After removing 11 cases from the early 1990s when open surgery was considered the standard treatment, tumor size and MS/AD ratio between 19 OP and 31 MIS cases were compared (Table 5). The OP group showed higher MS/AD ratio (0.36±0.10 vs. 0.20±0.08, p<0.001) and larger mass size (7.2±2.9 vs. 4.3±1.8, p<0.001). The total average of the MS/AD ratio was 0.3. Of 19 open surgery cases, 13 cases had MS/AD ratio higher than 0.3 and only 3 cases of 31 in the MIS group had higher MS/AD ratio of 0.3 (p<0.001).

A total of 28 patients had either PV or SMV abutment (Table 6). The number of PV or SMV abutment in the OP group was larger compared with the MIS group (p<0.001). Although the number of grades 1 and 2 PV and SMV compression between the two groups did not show statistical difference, grades 3 and 4 showed statistical difference (p<0.001). There were total of 34 SV compressions, none of which showed any difference.

Of 66 patients, 4 patients experienced recurrence (Table 7). Of the 4 patients with recurrence, 2 were pathologically confirmed as SPC and other 2 were confirmed as SPT. Recurrence sites included liver, aortocaval node, and resection margin from the previous operation. These recurred patients received liver segmentectomy, aortocaval node dissection and mass excision according to the recurred site. Until this day, these patients are followed in our outpatient clinic.

SPT causes abdominal discomfort and pain often after abdominal trauma or can be found incidentally upon physical examination. In adults, SPT is usually detected incidentally during health screening [316]. The majority of SPT is located in the pancreatic body or tail (59.3%) with some involving the pancreatic head or uncinate process (36%). The mean tumor size in adult patients was 9.8 cm before 2000 and has decreased to 8.1 cm in 2000–2012 (p<0.05) [3]. The mean tumor size is known to be significantly larger in children [18]. Despite its size, the prognosis of SPT is favorable with a survival rate of 95% even with metastasis in both adult and pediatric patients [1019]. Comparative analysis revealed that malignant SPT, also known as SPC, tends to be larger in size and have more solid components [20]. Although SPT shows excellent survival outcomes, recurrence has been observed without pathologic or clinical risk factors. Four patients in this study experienced recurrence. One patient had incomplete resection margin, and three had tumors exceeding 8 cm. These patients are visiting the outpatient clinic for routine check-ups.

LP has become one of the mainstay treatments in pancreatic disease with increased safety. Laparoscopic distal pancreatectomy (LDP) in pancreatic body or tail cancer in adult patients has been shown to be feasible. Kim et al. [15] showed that LDP had superior clinical outcomes over open distal pancreatectomy (ODP). In this study, nasogastric tubes were removed significantly earlier in laparoscopic surgery compared with its open counterpart (0.6±0.7 days vs. 1.7±1.0 days, p<0.001). Furthermore, patients were started on diets earlier in LDP (2.8±1.3 days vs. 4.5±1.6 days, p<0.001). Moreover, patients were discharged significantly earlier in laparoscopic surgery (median [range] of hospital stays: 10 [5–36] days vs. 16 [8–65] days, p<0.001). There was no statistical difference in postoperative complications. Similar results were observed in a systematic review published recently [17]. In this study, LDP showed significantly faster oral intake, shorter hospital stays, and lower postoperative complication rate. Stewart et al. [4] and Kang et al. [16] also reported similar results. In these studies, the authors state several key factors of utilizing MIS for pancreatic diseases. First, laparoscopic surgery provides excellent cosmetic outcomes in young women. Second, LDP can be performed because most SPT occurs frequently in the tail. Third, SPT smaller than 5 cm can be considered benign, disregarding malignant potentials.

Laparoscopic pylorus-preserving pancreatoduodenectomy (L-PPPD), first introduced by Gagner and Pomp [21] in 1994, is also becoming more prevalent. Zureikat et al. [22] showed that OPD and LPD had no statistical difference in postoperative outcomes. Although the number of total patients was small, R0 resection was achieved in all LPD patients. Tumor size was smaller in the LPD group (2.2 [0.8–4.7] cm vs. 3.6 [3–5] cm, p=0.02), but the mean number of retrieved lymph node showed no difference. Kim et al. [23] stated that, even though they have not directly compared L-PPPD from open PPPD, L-PPPD is technically safe and feasible with acceptable levels of morbidity in benign and malignant periampullary cancer. Recently, published literature comparing the long-term results of LPD and OPD in pancreatic cancer showed positive results towards the LPD group [24]. LPD was associated with significantly longer operative time (median: 375 minutes vs. 518 minutes, p<0.001), but decreased blood loss (median: 600 mL vs. 250 mL, p<0.001) and decreased use of packed red blood cell transfusions (46.6% vs. 25.9%, p=0.005).

Although there is increasing evidence proving superiority of MIS over OP, there is still a lack of evidence regarding MIS in pediatric patients. According to our previous study [6], LDP was superior than open DP in terms of shorter time to oral intake, time to full feeding, and hospitalization. There was no difference in perioperative morbidity. Similar results were observed in this study. However, in case of PPPD, open PPPD was mainly performed in pediatric patients for three reasons. First, L-PPPD is not widely accepted among surgeons; second, pancreatic main ducts are too small for laparoscopic anastomosis; and third, laparoscopy is not accessible for huge masses. Nonetheless, MIS is becoming more popular and being acknowledged by many surgeons. In our center, we are opening the possibility of MIS for all eligible SPT cases, and we have already performed two cases of L-PPPD. Thus, we would like to set some guidelines as to when MIS is feasible based on our experience.

Among many parameters, the most important factors are tumor size, MS/AD ratio, and the degree of abutment or compression on PV or SMV. In this study, tumor size, MS/AD ratio, and operative time were analyzed according to the surgery and approach, whether open or laparoscopic/robotic. The tumor size and MS/AD ratio was significantly different between open and minimally invasive enucleation and DP/SSDP. From these results, patients with larger tumor size or higher MS/AD ratio should undergo open surgery rather than MIS. In our experience, tumors with compression grade 3 or higher do not offer appropriate surgical field for the surgeon, making ligation of the perforation vessel extremely difficult. Since clipping was not applicable in most cases, open suture ligation was necessary. In this study, there were 4 cases of open conversion due to vessel encasement (n=2), bleeding (n=1), and pancreatic duct injury during enucleation (n=1).

There is a tendency in recent literature siding with MIS. Even in our center, we can notice that the MS/AD ratio in the MIS group is increasing (Fig. 3). With increasing evidence proving superiority of MIS over OP, surgeons are performing MIS despite bigger masses and harder cases. To reduce the risk of MIS, surgeons must carefully select patients feasible for MIS. We carefully suggest MS/AD ratio smaller than 0.3 and PV/SMV compression of grade 2 or lower, to be considered a feasible candidate for MIS. As mentioned before, SV compression is not an absolute indication. Since, SPT cases are rare, it is very difficult for a single pediatric surgeon to complete a learning curve. However, with the advancement in laparoscopic technique in other diseases, we believe that the range of surgical indications will expand.

Although this study has suggested some indications for MIS in pediatric patients, it has several limitations. This study was retrospective in nature, and the sample size in open vs. laparoscopic/robotic enucleation, DP/SSDP, and PD/PPPD was small. Furthermore, most of our patients received open PPPD compared to its laparoscopic counterpart (14 open PPPD vs. 2 L-PPPD). Since patients with pancreatic head or uncinate process received open PPPD when there was high degree of either PV or SMV compression, selection bias naturally occurred. However, this study is the first to suggest indications for MIS in pediatric patients. Further validation is needed.

In conclusion, MIS is widely utilized in pediatric patients including pancreatectomy. Careful selection of MIS is required in pediatric patients using parameters such as MS/AD ratio and vessel abutment. MS/AD ratio lower than 0.3 and vessel compression of grade 2 or below are considerable candidates for MIS.