Laparoscopic distal pancreatectomy (DP) is regarded as a safe and feasible minimally invasive surgical approach in benign and borderline malignant pancreatic tumors [1]. With advances in laparoscopic techniques and instruments and surgeons’ experience, much effort to reduce the number of the trocar site has been made. Minimally invasive surgery appears to be developing in multiple directions, including single-port laparoscopic (SPL) [2,3,4], single-port robotic [5], reduced port laparoscopic [6], or single-site plus one-port robotic (RSS + 1) approach [7].

The first laparoscopic DP was performed in 1994 by Cuschieri [8], and the first robotic DP was reported by Melvin et al. [9]. In the same year, Giulianotti et al. [10] published their robotic experience on the first series of five DP cases. Daouadi et al. [11] reported that robot-assisted DP was as safe and effective as laparoscopic DP but it significantly reduced the conversion to open resection, despite a statistically greater probability of malignancy in the robotic cohort. Short-term results of robotic DP seem to be similar to the laparoscopic approach, but the high cost of robotic approach makes this approach less cost-effective [12].

The single-port surgical approach, which has begun to gain attention with the introduction of natural orifice transluminal endoscopic surgery, has gradually expanded its field from simple to complex surgery. However, the application of this approach has not increased rapidly due to apprehension about the technical difficulty and mental burden of single-port approach and/or the increased operative risk in terms of surgeon’s experiences; there have been great concerns regarding operative safety and feasibility compared with the established outcomes of the single-port approach complex surgery, such as hepatectomy or pancreatectomy. The first SPL-DP was reported by Barbaros et al. [13]. In 2014, Han et al. [14] reported their initial experience on SPL-DP, which was similar to conventional laparoscopic DP, except in terms of duration of operation and hospital stay.

The advantages of the robotic platform over the laparoscopic approach include motion stabilization, absence of the fulcrum effect, reduction of operator fatigue, three-dimensional and high-definition vision, seven degrees of freedom, and improved ergonomics for the surgeon. All these features enable the surgeon to assess challenging complex surgical procedures while maintaining all the benefits of a minimally invasive approach and with greater precision and improved functional and cosmetic outcomes [15,16,17].

To the authors’ best knowledge, Kim and colleagues from Yonsei first reported on RSS + 1-DP [18, 19], which is currently available and has the potential to make SPL-DP much easier and ergonomic, providing some room to expand on more minimally invasive surgery. In addition, Peng et al. [20] recently reported that single-port robotic DP using commercial single-port device was safe and efficient in their initial experience. Until now, no clinical investigations compared the perioperative outcomes between SPL-DP and RSS + 1-DP.

The aim of study was to compare the short-term perioperative outcomes in patients undergoing minimally invasive DP via the SPL or the RSS + 1 approach.

Materials and methods

Patient and data

This study is a bi-institutional collaborative study between Korea University and Yonsei University in the Republic of Korea. The medical records of patients who consecutively underwent SPL-DP at the Department of Surgery, Korea University Ansan Hospital and RSS + 1-DP at the Department of Surgery, Yonsei University Severance Hospital between January 2012 and January 2018 for benign and low-grade malignant pancreatic tumor were retrospectively reviewed. Clinicopathologic data including demographic characteristics, clinical presentation, perioperative results, complications, and pathologic findings were collected. The postoperative pancreatic fistula (POPF) was defined by the revised 2016 International Study Group of Pancreatic Fistula guidelines [21]. Additionally, the Clavien–Dindo classification (CDC) [22] was surveyed for overall complications and the comprehensive complication index (CCI) [23] was obtained using the CCI calculator available online (http://www.assessurgery.com). Tumor location was reviewed based on preoperative abdominal computed tomography results. Postoperative pain intensity was registered at postoperative days (POD) 1 and 3. All patients were asked to rate their present intensity of pain by (1) placing a vertical mark on a 10-cm visual analogue scale (VAS), which ranged from “no pain” on the left to “worst pain” on the right end, (2) choosing a number between 0 and 10, where 0 was “no pain” and 10 was “worst pain,” and (3) describing their present pain, from no pain, mild, moderate, severe, and excruciating [24].

Indications for splenectomy in the SPL and RSS + 1-DP groups were pancreatic neoplasm located in the splenic hilum, intraoperative intractable bleeding from the spleen during spleen preservation, and highly suspicious malignancy in the preoperative imaging study.

Surgical techniques of SPL-DP and RSS + 1-DP

The SPL technique was performed using a pure SPL technique, which was described in detail in 2014 [14]. Additionally, we created new channels on the body of the glove port with purse-string sutures and sterile Fixomull stretch (BSN Medical, Hamburg, Germany) for laparoscopic stapling device using an SPL technique. The remnant pancreas was reinforced with nonabsorbable 3–0 continuous barbed sutures using an SPL technique. The RSS + 1-DP technique was described in our recent studies [18, 19].

Statistical analysis

Each variable’s distributional characteristics were assessed for normality. Continuous data were reported in terms of mean ± standard deviation and/or median with interquartile range based on variance. The chi-squared test or Fisher’s exact test was used to compare categorical variables, and independent t test or Mann–Whitney test was used to compare continuous variables. The p value was adjusted under Bonferroni correction after Mann–Whitney test. The POPF risk factor analysis was performed in the univariate logistic regression analysis and variables with p value < 0.2 were analyzed in the multivariate logistic regression method. Statistical analyses were performed using SPSS (version 21.0 for Mac, SPSS Inc., Chicago, IL, USA). The institutional review board of the ethics committees (IRB) of Korea University Ansan Hospital and Yonsei University Severance Hospital approved the study protocol. The written informed consent was waived by the IRB owing to the study’s retrospective nature.

Results

Clinicopathologic characteristics

Twenty-two patients underwent SPL-DP for pancreatic neoplasm in Korea University Ansan Hospital (SPL group) and 13 patients underwent RSS + 1-DP for pancreatic neoplasm in Yonsei University Severance Hospital (RSS + 1 group). Table 1 shows that the mean age in the SPL group was significantly higher than that in the RSS + 1 group (58.3 vs 46.1 years, p = .023). The mean body mass index (BMI) in the SPL group was higher than that in the RSS + 1 group (23.9 vs 20.9 kg/m2, p = .034). There was no statistically significant difference in male -to-female ratio, American Association of Anesthesiologist (ASA) score, and the presence of previous abdominal operation history between the two groups. The pathologic examination demonstrated some differences between the two groups. The proportions of intraductal papillary mucinous neoplasm (18.2 vs 0%), and mucinous cystic neoplasm (27.3 vs 0%) were higher in the SPL group. However, the proportion of neuroendocrine tumor (18.2 vs 30.8%) and solid pseudopapillary tumor (4.5 vs 30.8%) were higher in the RSS + 1 group.

Table 1 Clinicopathologic characteristics of patients undergoing SPL or RSS + 1
Full size table

Perioperative characteristics

The perioperative characteristics were compared between the groups in Table 2. The mean operation time in the SPL group was significantly higher than that in the RSS + 1 group (281 vs 192, p = .001). The mean blood loss in the SPL group was significantly larger than that in the RSS + 1 group (163 vs 12 mL, p = .002), but there was no transfusion in either group. The spleen was significantly preserved in the SPL group (54.5 vs 7.7%, p = .001), which was explained by the observation that pancreatic neoplasms in the RSS + 1 group were mainly located in the pancreas tail abutting splenic hilum (38.1 vs 100.0%, p < .001). The mean number of trocars in the SPL group was significantly lower than that in the RSS + 1 group (1.1 vs 2, p < .001). The mean tumor size in the SPL group was similar to that in the RSS + 1 group (3.0 vs 2.7 cm). The mean length of free resection margin in the SPL group was significantly longer than that in the RSS + 1 group (2.1 vs 0.4 cm, p = .001). There were no significant differences in the mean pain VAS score on POD 1 (2.6 vs 2.3, p = .395) and POD 3 (2.3 vs 2.5, p = .985), mean CCI (8.3 vs 3.3, p = .566), and complication rate (36.4 vs 38.5%, p > .99) between the two groups. However, SPL had significantly higher CDC grade complications (p = .014). There was no significant difference in POPF between the two groups (p = .274); however, the SPL group had slightly higher POPF grade than the RSS + 1 group (p = .120). Moreover, the mean hospital stay in the SPL group was significantly longer than that of RSS + 1 group (14.4 vs 7.4 days, p = .004). The mean follow-up period in SPL group was slightly longer than that in the RSS + 1 group (23.4 vs 7.8 months, p = .005).

Table 2 Perioperative characteristics of patients undergoing SPL or RSS + 1
Full size table

Multivariate logistic regression analysis

The results of the analysis of the overall complications and POPF risk factors are shown in Table 3. Overall complications were related with previous abdominal operation history (p = .043) but were not significant in the multivariate analysis (p = .053). POPF development was related with longer operation time (p = .043) and smaller tumor size (p = .037), but based on the multivariate analysis, there was no statistically different risk factor of POPF in patients between the two groups.

Table 3 Complications and postoperative pancreatic fistula risk factor analysis after SPL or RSS + 1
Full size table

Table 4 shows the factors that increased the duration of operation and hospital stay. In terms of operation time, higher BMI (p = .008), SPL approach (p = .014), and larger amount of blood loss (p = .027) were significantly correlated with longer operation time, but in the multivariate analysis, only higher BMI and higher amount of blood loss were significantly important factors (p = .034). Prolonged hospital stay was related to spleen preservation (p = .008), higher complication rates (p = .013), grade IIIa complications (p = .003), POPF (p = .032), grade B POPF (p = .002), and higher CCI (p = .001). In the multivariate analysis, there are statistically significant differences in spleen preservation (p = .014) and complications (p = .020).

Table 4 Analysis of factors that increased operation time or hospital stay after SPL or RSS + 1
Full size table

Discussion

The RSS + 1 approach was associated with a significantly shorter operation time, blood loss, postoperative hospital stays, and lower CDC grade of complications, but the SPL approach was associated with significantly fewer trocars, better spleen preservation, larger free resection margin, and wider operation field for more centrally located pancreatic neoplasm. There were no significant differences in the incidence of overall complications, CCI, and POPF, but more grave complications were significantly more common in the SPL group. The advantages of robotic approach to distal pancreatic neoplasms are closely related with the merits of minimally invasive surgery in our study. Reduction of postoperative hospital stay is one of the main objectives of minimally invasive pancreatic surgery, as it can balance the increased costs of laparoscopic or robotic equipment, as demonstrated recently for minimally invasive liver surgery [25]. For hepatobiliary surgery in particular, progress made in the last 5 years in robotically assisted minimally invasive surgery has been astounding [26]. Robotic DP was known to be associated with significantly better spleen preservation rate and reduced operative time, blood loss, transfusion requirement, and postoperative hospital stay for patients undergoing spleen preservation compared with conventional laparoscopic DP, but it offered fewer benefits for patients undergoing splenectomy [27, 28]. The advantages of the laparoscopic approach using a single port to distal pancreatic neoplasm are significantly more focused on the benefits of SPL itself, such as widening of the operation field and extension of the surgical scope that robotic surgery do not have.

The duration of operation and hospital stay in the SPL group was much longer than that in the RSS + 1 group, which was similar to the findings of other studies on robotic DP [27, 28]. However, in our study, there were significant differences in the mean age, mean BMI, and pathologic reports between the two groups. The operation time was influenced and prolonged by the SPL approach, higher BMI, and larger blood loss in our study. Resection of pancreatic tumors near the splenic hilum is thought to be easily performed by laparoscopy or robotic DP with planned splenectomy; however, tumors near the neck or the body of the pancreas sometimes require division of the pancreatic neck portion, which is a more complex and relatively difficult procedure. Simultaneous splenectomy is performed with DP mainly for technical reasons, such as to make resection easier, to shorten operation time, and to minimize bleeding from dissection of the splenic vessels. The use of laparoscopy was reported to improve the spleen preservation rate compared with laparotomy [29], although spleen-preserving laparoscopic DP relatively needs more time and effort. A major limitation regarding the laparoscopic technique is its less effective control of bleeding from the splenic vessels, because splenic vessel bleeding is a major risk factor for failed spleen preservation. The distance from the umbilicus to the spleen in the SPL approach is different from that of the conventional laparoscopic approach. Although bleeding around the spleen is also a major risk factor for failed spleen preservation in the SPL approach, spleen-sacrificing SPL-DP is still more time consuming and labor consuming. From that point of view, we can explain why SPL-DP has a much longer operation time even if the docking time of the robotic approach is not considered. The operative time in robotic approach is influenced by the docking method and the surgeon’s experience in robotic pancreatic surgery. However, a surgeon with enough experience and the latest version of the da Vinci Xi, which has narrower arms and a more straightforward docking method, contributes to a decrease in operative time. Expert surgical experience is important in achieving superior robotic DP surgical outcomes. It would be important to know the number of minimally invasive distal pancreatectomies performed in the institution and the average experience of each of the surgeons. A recent study has reported improvements in complications and length of hospital stay after 30 cases of learning curves in laparoscopic DP [30].

However, we do not believe that SPL-DP should increase more clinical experiences to obtain the learning curve and improve satisfactory perioperative outcomes. The robotic approach has been introduced to overcome the limitations of conventional laparoscopic surgery, and it may be useful for pancreatic surgery, which requires delicate laparoscopic surgical techniques. However, many laparoscopic surgeons have overcome the limitations of conventional laparoscopic techniques as a result of their surgical experiences [31]. It is necessary for the SPL approach to develop a brand-new single-port system to untangle the complexity of the arrangement of the laparoscopic instruments and expand the number of working laparoscopic instruments in the single-port. The RSS + 1 pancreas approach is a modified surgical platform of a complete minimally invasive surgical technique, but SPL pancreatic surgery is a developing, time-consuming, and technical-demanding technique in minimally invasive surgery.

Prolonged duration of hospital stay was related with complications in our study. The index of complications including CDC grade, CCI, and POPF grade indicated that more severe complications are associated with longer duration of hospital stay. Additionally, significant relationship was noted between spleen preservation and longer duration of hospital stays. There is no evidence that a higher spleen preservation rate is associated with longer duration of hospital stays until now. However, spleen preservation was frequently observed in the SPL approach, which was directly related with longer operation time.

The gateway in comparison between laparoscopic and robotic surgery is the problem of cost. The meta-analyses between laparoscopic and robotic DP by Gavriilidis et al. [32] could not demonstrate that the robotic approach was more expensive because of a small number of cost-analysis studies and differences in medical costs between countries. In our study, the mean operation cost in the SPL group was significantly lower than that in the RSS + 1 group ($2262 ± 587 vs. $7409 ± 0, p < .001). The mean total cost in the SPL group was $10,661 ± 3070, of which patients were only charged $4402 ± 1121, as a benefit from the National Health Insurance Service. Zhou et al. [33] in a meta-analysis of two studies that reported the costs, reported no statistical difference between the laparoscopic and robotic groups, which was explained by the shorter hospital stay in the robotic group.

Our study has several limitations, including the small sample size, different laparoscopic pancreatic surgical experiences of the two surgeons, different volumes of laparoscopic pancreatic surgery in each hospital, heterogeneous pancreatic disease, and limitation of the study design. However, despite the limitations, our results can be used as groundwork for the future direction of reduced port minimally invasive surgery. It may be that the baseline itself is different when comparing SLP and RSS + 1 surgeries. However, this study is not an comparison study of conventional laparoscopic and robotic systems currently being performed. Our study is a comparison between reduced port robotic surgery and SPL surgery, which will be conducted predominantly in the recent future. As reduced port robotic surgery had a number of advantages over SPL surgery, we should provide opportunities for single-port surgery, including laparoscopic and robotic, to be developed as a new low-entry barrier surgical platform by technical advances, rather than not having SPL surgery.

Conclusions

Both SPL and RSS + 1 are technically feasible and safe. RSS + 1-DP is superior to SPL-DP in terms of operation time, blood loss, severe complications, and hospital stay. The SPL approach shows advantages in terms of single wound site, fewer trocar usage, higher rate of spleen preservation, and wider range of operative field.