Minimally invasive pancreatectomy is a treatment method for pancreatic tumors. Robotic pancreatectomy, as a type of minimally invasive pancreatic surgery, is subdivided into robotic distal pancreatectomy (RDP) and robotic pancreatoduodenectomy (RPD). The location of the pancreatic tumor determines the choice between RDP and RPD. Usually, pancreatic body and tail lesions are treated with RDP and pancreatic head lesions are treated with RPD. RDP was first reported in 2003 by Melvin et al,¹ and Giulianotti et al² first reported RPD in 2003, in a study that found RPD to be feasible. Since then, several studies have been published on the efficacy and advantages of robotic pancreatectomy. An appropriate surgical view controlled by the operator, correction of the operator’s tremor, high-definition three-dimensional images, articulation of instruments, and ergonomic comfort of the operator are well-known advantages of robotic systems. Despite these advantages, robotic pancreatectomy has not yet become popular due to expensive equipment, the costs of surgery, and the unfamiliarity of operators with robotic surgery. However, as the advantages of robotic surgery are emphasized, the application of robotic pancreatectomy is expected to increase gradually. This study aims to describe the indications and contraindications, surgical methods, and perioperative outcomes of current robotic pancreatectomy.

Indications and contraindications

The Miami International Evidence-Based Guidelines recommend that minimally invasive distal pancreatectomy (MIDP) should be considered over open distal pancreatectomy (ODP) for left-sided benign and low-grade malignant tumors.³ Left-sided pancreatic cancer also could be an indication of MIDP. A pan-European propensity score-matched study on MIDP and ODP for pancreatic cancer showed comparable survival outcomes between the two groups.⁴ The Miami International Evidence-Based Guidelines also recommend that both laparoscopic distal pancreatectomy (LDP) and RDP are safe and feasible options for left-sided pancreatic tumors.³ Therefore, it can be suggested that all indications of left-sided benign, borderline-malignant and malignant pancreatic lesions that can be applied to current LDP are also applicable to RDP. However, there is still a lack of consensus regarding the advanced stages of left-sided pancreatic cancer. Celiac axis or portal vein resection using a robotic system requires high-quality surgical techniques. However, several reports have described performing this high-morbidity operation. Ocuin et al⁵ reported 11 cases of RDP with celiac axis resection. Rao et al⁶ also reported that RDP with celiac axis resection was safe and feasible, with the potential of offering long-term survival in selected patients. Radical antegrade modular pancreatosplenectomy, which provides a sufficient tangential margin and a sufficient number of harvested lymph nodes, can be performed with a robotic system.⁷ However, studies on these surgical procedures are still in the stage of case reports or comparative studies. Therefore, it is too early to conclude that locally advanced left-sided pancreatic cancer can be an indication for RDP, and it is reasonable to suggest that RDP be implemented according to the operator’s experience and skills. Prior abdominal surgery with adhesions is also a relative contraindication that can be overcome with the surgeon’s experience and appropriate patient selection.

Surgical technique

The RDP procedure at our institution has been presented previously in detail,⁸ but RDP methods may differ depending on the operator or institution. Usually, RDP is performed using five trocars, including four robotic trocars (one robotic camera and three robotic arms) and one 12-mm accessory port used by the assistant. However, a study has reported that robotic single-site plus one-port distal pancreatectomy was safe and feasible in terms of short-term outcomes.⁹ Therefore, the number and location of the ports can vary depending on the operator’s preference. In Fig. 1, three robotic instruments and the assistant’s suction are shown in front of the pancreas. To prepare for a proper surgical view, the stomach is retracted upward. At some institutions, laparoscopic instruments are used to secure a proper surgical field for performing RDP. The lesser sac is opened by dividing the omentum using a robotic energy device or bipolar instruments. The transverse colon and left colon are mobilized and taken down to identify the lower border of the pancreas. The superior mesenteric vein is exposed with preservation of the gastrocolic trunk. If there is a tumor at the far tail of the pancreas, the pancreas should be mobilized from the tail to the body direction without identifying the superior mesenteric vein or splenic vein. After identifying the superior mesenteric vein, the hepatic artery and gastroduodenal artery are exposed and preserved. After tunneling the lower border of the pancreas, pancreatic transection is performed using a robotic or laparoscopic stapler according to the surgeon’s preference (Fig. 2). If spleen preservation is performed using the Kimura method,¹⁰ the splenic vein and artery are isolated and preserved until the splenic hilum. If spleen preservation is performed using the Warshaw method,¹¹ the splenic artery and vein are encircled and divided between locking clips. The pancreas is dissected from the retroperitoneum to the splenic hilum. In patients with malignant tumors of the pancreas, splenectomy is concomitantly performed.

Figure 1. A robotic electrocautery hook, robotic Maryland bipolar forceps, robotic Prograsp forceps, and assistant suction are placed for robotic distal pancreatectomy.

Figure 2. Pancreas transection is performed by a laparoscopic stapler used by the assistant. During robotic distal pancreatectomy, the assistant helps with laparoscopic suction, irrigation, clipping, and stapling.

Perioperative outcomes

Several reports have found that RDP was associated with higher spleen and splenic vessel preservation rates^12,13 and a lower open conversion rate^14,15 than LDP and ODP. However, there is still insufficient evidence to support these claims because of the lack of randomized clinical trials and insufficient published data. Several recent comparative studies of LDP and RDP have been conducted (Table 1),^8,13,16–18 but the absence of randomized clinical trials is a limitation; hence, the extant literature cannot provide proper evidence of the advantages of RDP. Nevertheless, as the number of RDP cases increases, the benefits of this procedure are highlighted. Two meta-analyses on RDP and LDP reported that RDP required a longer operation time than LDP.^19,20 Many recent studies reported differences in the operation time, but the differences were not statistically significant.^21,22 As the number of RDP cases increases, the operating time of RDP decreases.²³ Since the number of RDP cases is gradually increasing worldwide, it is believed that the difference in the operation time between LDP and RDP will continue to decrease. Furthermore, three case-matched studies reported that the open conversion rate was lower in RDP than in LDP.^8,17,18 Additionally, a systematic review and a multicenter propensity score-matched analysis reported a high spleen preservation rate of RDP.^18,24 Another report stated that the splenic vessel preservation rate was higher in RDP.¹⁸ However, no reports have described LDP as resulting in a lower open conversion rate or a higher spleen and splenic vessel preservation rate than RDP. These results can be interpreted as showing that RDP has a high success rate if surgery is performed as planned. This can be explained by characteristics of the robotic system, such as its three-dimensional high-definition images, optimal viewing by the operator, tremor filtration, articulation of instruments, and hardware-software upgradability.⁸ Most retrospective studies reported that there were no differences in other perioperative outcomes, such as blood loss during the operation, postoperative pancreatic fistula, major complications, and postoperative hospital stays between RDP and LDP.^8,13,16–18 Therefore, RDP can be suggested as a suitable treatment option for benign and borderline-malignant pancreatic tumors, which are currently indications for LDP. However, hospital costs are a disadvantage of robotic surgery. According to a meta-analysis comparing the surgical costs of RDP and LDP, RDP was more expensive.²² Depending on the country and socioeconomic status of the individual patient, the cost to be borne by an individual may be different. This issue of the cost of surgery is not only an academic issue, but also a social topic to be discussed.

Table 1 . Perioperative Outcomes of Robotic Distal Pancreatectomy Compared to Laparoscopic Distal Pancreatectomy.

Author (yr)	Number	Operation time (min)	Open conversion, n (%)	Estimated blood loss (mL)	POPF, n (%)	Major morbidity*, n (%)	LOS (day)	Spleen preservation, n (%)	Vessel preservation†, n (%)
Lyman et al (2019)16	RDP (108) LDP (139)	252‡ 196	2 (1.9) 6 (4.3)	406 377	21 (19.4) 28 (20.1)	NA NA	5.0 5.0	34 (31.5)‡ 12 (8.6)	NA NA
Hong et al (2020)13	RDP (46) LDP (182)	166 141	0 (0.0) 0 (0.0)	NA NA	4 (8.7) 19 (10.4)	1 (2.2) 3 (1.6)	8.0 8.1	30 (65.2)‡ 47 (25.8)	NA NA
Kwon et al (2022)8	RDP (104) LDP (208)	180 164	0 (0.0)‡ 8 (3.8)	351 411	16 (15.4) 31 (14.9)	3 (2.9) 9 (4.3)	7.9 8.6	57 (54.6) 105 (50.4)	21 (20.2) 48 (23.1)
Liu et al (2017)17	RDP (102) LDP (102)	207 200	3 (2.9)‡ 10 (9.8)	100 100	7 (6.9) 13 (12.8)	7 (6.9) 13 (12.8)	7.7‡ 8.6	62 (60.8) 50 (49.0)	26 (25.5) 22 (21.6)
Lof et al (2021)18	RDP (402) LDP (402)	285‡ 240	27 (6.7)‡ 61 (15.2)	150 150	99 (24.6) 106 (26.5)	57 (14.2) 66 (16.5)	8.5‡ 7.0	153 (81.4)‡ 122 (64.2)	112 (27.9)‡ 82 (20.4)

POPF, postoperative pancreatic fistula; LOS, length of hospital stay after operation; RDP, robotic distal pancreatectomy; LDP, laparoscopic distal pancreatectomy; NA, not available..

*Major morbidity was defined as a Clavien-Dindo classification grade IIIa or above..

†Vessel preservation was defined as preservation of the splenic vein and splenic artery..

‡A statistically significant difference existed between two groups..

Indications and contraindications

The Miami International Evidence-Based Guidelines reported that there were insufficient data to recommend minimally invasive pancreatoduodenectomy (MIPD) over open pancreatoduodenectomy (OPD).³ The guidelines also reported no evidence of superiority between laparoscopic pancreatoduodenectomy (LPD) and RPD.³ Retrospective studies comparing RPD with OPD and LPD have been published recently (Table 2),^25–29 but it is still too early to conclude that the results of RPD are comparable to those of OPD or LPD. One reason for this is that the indications for MIPD are not yet equivalent to those for OPD. As LEOPARD II has announced that LPD is a high-level operation and has a high risk of morbidity and mortality,³⁰ there is a consensus that it is necessary to carefully consider the indications for MIPD and the proper institutions that can perform this operation. Another reason is that randomized clinical trials comparing RPD with other surgical procedures have not yet been published. However, it can be expected that periampullary diseases, such as periampullary adenocarcinoma, borderline malignancy of the pancreatic head, and resectable pancreatic head cancer, which can be treated with LPD, can also be treated with RPD. The rationale for this is that 1) there exists hybrid RPD wherein robot reconstruction is performed after laparoscopic resection, and 2) robotic resection uses the same surgical method as LPD. Whether it is possible to perform RPD in pancreatic head cancer after neoadjuvant therapy or in pancreatic head cancer requiring major vascular resection for R0 is also difficult to determine. No comparative data exist, and further experiences and studies are warranted. However, there are many reports on RPD with superior mesenteric vein or portal vein resection.^31,32 Therefore, even in cases of advanced pancreatic head cancer, RPD can be implemented according to the surgeon’s skill and the institution’s volume.

Table 2 . Perioperative Outcomes of Robotic Pancreatoduodenectomy Compared to Open or Laparoscopic Pancreatoduodenectomy.

Author (yr)	Number	Operation time (min)	Estimated blood loss (mL)	Intraoperative transfusion, n (%)	Major vessel resection*, n (%)	Open conversion, n (%)	POPF, n (%)	Major morbidity†, n (%)	LOS (day)
Kim et al (2022)25	RPD (328) OPD (656)	339‡ 290	417 470	40 (12.2) 81 (12.3)	NA NA	NA NA	32 (9.8) 73 (11.1)	56 (17.1) 120 (18.3)	10.8‡ 15.6
Shyr et al (2022)26	RPD (65) OPD (65)	498‡ 420	251‡ 562	NA NA	9 (13.8) 11 (16.9)	NA NA	6 (9.2) 2 (4.6)	4 (6.2) 4 (6.2)	NA NA
Weng et al (2021)27	RPD (105) OPD (210)	300 300	300 300	NA NA	9 (8.6) 20 (9.5)	2 (1.8) NA	6 (5.7) 14 (6.7)	NA NA	17 17
Kim et al (2022)28	RPD (74) LPD (74)	412‡ 453	NA NA	5 (6.8) 4 (5.4)	NA NA	0 (0.0)‡ 6 (8.1)	9 (12.2) 10 (13.5)	16 (21.6) 11 (14.9)	11.9‡ 14.6
van Oosten et al (2021)29	RPD (90) LPD (90)	471‡ 445	125‡ 300	1 (1.1)‡ 11 (12.2)	NA NA	NA NA	9 (10.0) 4 (4.4)	NA NA	8 8

POPF, postoperative pancreatic fistula; LOS, length of hospital stay after operation; RPD, robotic pancreatoduodenectomy; OPD, open pancreatoduodenectomy; NA, not available; LPD, laparoscopic pancreatoduodenectomy..

*Major vessel resection was defined as resection of the superior mesenteric vein, portal vein, hepatic artery, or superior mesenteric artery during operation..

†Major morbidity was defined as a Clavien-Dindo classification grade IIIa or above..

‡A statistically significant difference existed between two groups..

Surgical technique

Hybrid RPD (laparoscopic resection and robotic reconstruction)

The surgical technique for RPD has been well described elsewhere.^33,34 Several institutions use hybrid RPD to take advantage of RPD when the operator is accustomed to LPD. The hybrid RPD method that is mainly used at our institution and has been previously published is as follows.³⁵

The patient is placed in a supine position, and an anti-Trendelenburg position (10°–30°) is used to expose the operating field. The operator and laparoscopist stand to the right of the patient, with the assistant positioned on the left. Three 12-mm ports and two 5-mm ports are used for the operation. The scope is inserted through the 12-mm port at the right McBurney’s point. After abdominal access is established, the greater omentum is divided using an energy device, and the right colon is separated and fully mobilized from the duodenum. The retropancreatic superior mesenteric vein is then exposed, and the right gastroepiploic vessels are transected. After removing the soft tissue from around the superior mesenteric vein, the vessel is hung using a vessel loop. The stomach and duodenum are divided using an endoscopic linear stapler with pylorus preservation, if possible. Mobilization of the duodenum to the ligament of Treitz is performed with traction of the duodenum in the opposite direction by the surgical assistant. The gastrohepatic ligament is opened to visualize the superior border of the pancreas and to identify the common hepatic artery. The right gastric and gastroduodenal arteries are identified and transected using a Hem-o-lock clip.

The right and left hepatic arteries are identified and isolated, and lymph node dissection is performed. Dissection of the hepatoduodenal ligament and isolation of the common bile duct are performed after cholecystectomy. The pancreas is divided above the superior mesenteric vein. After retracting the resected pancreas to the right side of the patient’s abdomen, the portal vein is identified and hung using a vessel loop. The jejunum is divided approximately 10–15 cm distal to the ligament of Treitz with an endoscopic linear stapler. An energy device and endoscopic electrocautery are used to separate the remaining soft tissue and branches of the superior mesenteric artery between the uncinate process of the pancreas and the superior mesenteric artery, to complete the resection. After resection is completed, the robotic system is docked. The laparoscopic port is changed to a robotic port, and the first assistant uses the 12-mm camera port at McBurney’s point for needle insertion. A robotic scope is inserted through the umbilical port. The da Vinci SI or Xi robotic platform (Intuitive Surgical Inc., Sunnyvale, CA, USA) is used for anastomosis. Pancreatojejunostomy is performed using a double-layered, end-to-side, duct-to-mucosa method (Fig. 3). A polyethylene internal stent is inserted into the pancreatic duct. End-to-side hepaticojejunostomy is performed using laparoscopic continuous suturing at the posterior wall and interrupted or continuous suturing of the anterior wall (Fig. 4). Duodenojejunostomy or gastrojejunostomy with jejunojejunostomy is performed intracorporeally or extracorporeally via the specimen extraction site, namely, the umbilicus port placed after extension (5 cm). Two closed suction drains are placed at the superior and inferior borders of the pancreatojejunostomy site.

Figure 3. Robotic pancreatojejunostomy. The duct-to-mucosa method is usually performed. The free articulation angle of the needle driver makes it possible to comfortably suture even a small pancreas duct and jejunal hole.

Figure 4. Robotic hepaticojejunostomy. An absorbable monofilament is used for the posterior wall of the hepatic duct and jejunal wall. An interrupted suture is comfortable to perform using a robotic needle holder.

RPD (robotic resection and reconstruction)

The surgeon’s positions are the same as above. Six trocars are used for the operation: 8 mm, four trocars for the robotic platform and two trocars for the assistant (12 mm trocar, 5 mm trocar). Barring the use of the robot arm, the surgical procedure is the same as that for hybrid RPD. A vessel sealer is used as an energy device, and the assistant helps in the operation with suction, irrigation, insertion of needle and endoscopic linear stapler. After hepaticojejunostomy, the specimen is extracted through the extended robot camera port site. Duodenojejunostomy or gastrojejunostomy with jejunojejunostomy is performed either intracorporeally or extracorporeally, similar to the hybrid RPD.

Perioperative outcomes

RPD has no clear advantage over OPD or LPD because this operation is still in its infancy and a more extensive registry of RPD cases needs to be gathered. In a meta-analysis comparing RPD and OPD, Yan et al³⁶ reported that RPD had the advantages of less blood loss, a shorter hospital stay, and a lower wound infection rate, but longer operative time. Two of three recently published propensity score-matched studies reported that RPD had a longer operative time than OPD, but each study presented different results about blood loss, major morbidities, and hospital stay.^25–27 In a meta-analysis of RPD and LPD, six comparative studies showed that RPD was associated with lower open conversion rates, lower transfusion rates, and shorter hospital stays.³⁷ Two recently published propensity score-matched studies about RPD and LPD also reported that RPD had advantages, such as a lower open conversion rate, less blood loss, and a shorter hospital stay.^28,29 More studies are needed to confirm the advantages and disadvantages of RPD. However, the advantages of LPD compared to OPD, such as less blood loss, lower pain scores, and shorter hospital stays, are expected to exist for RPD as well.

Studies on the pathologic and survival outcomes of robotic pancreatectomy are insufficient for a meaningful discussion. The efficacy of robotic pancreatectomy in benign and borderline malignancies has already been mentioned, with advantages in perioperative outcomes, but the effect of robotic pancreatectomy on pancreatic cancer still needs further research. Whether robotic pancreatectomy is also possible for advanced pancreatic cancer remains to be established. A few studies have reported that the pathologic and survival outcomes of pancreatic cancer after robotic pancreatectomy were similar to those of conventional pancreatectomy through a case-matched analysis,^26,27,38 but these results were obtained from selective pancreatic cancer cases. Thus, whether robotic pancreatectomy demonstrates comparable pathologic and survival outcomes to conventional pancreatectomy for all pancreatic cancers remains debatable. However, with appropriate pancreatic cancer case selection for robotic pancreatectomy, the pathologic and survival outcomes can be similar to those of conventional pancreatectomy.

As a treatment for benign or borderline-malignant pancreatic tumors, robotic pancreatectomy is an appropriate method with similar advantages to the laparoscopic approach, such as less intraoperative blood loss, reduction of postoperative pain, and a shorter hospital stay after surgery. In addition, RDP can be expected to show a lower open conversion rate and a higher spleen preservation rate than LDP. For pancreatic cancer, proper case selection is required to perform robotic pancreatectomy, but there are no absolute contraindications. More studies on robotic pancreatectomy, including randomized clinical trials, should be conducted.

None.

No potential conflict of interest relevant to this article was reported.