Introduction

Surgery remains the treatment of choice for the majority of patients with benign and malignant biliary obstruction. Biliary obstruction may occur in a variety of situations ranging from benign biliary strictures to bile duct injuries following cholecystectomy and biliary malignancies. In most cases, a bilio-enteric anastomosis with or without bile duct resection is a sufficient method to relieve biliary obstruction. In general, biliary surgery requires fine dissection and advanced suturing.

Laparoscopic bile duct surgery has recently been shown to be a safe treatment option for selected patients with choledochal cysts [1, 2]. Reports on laparoscopic surgery for biliary obstruction are scarce and these procedures are generally performed by an open approach. The introduction of robotic surgery has, however, allowed for better three-dimensional visualization, tremor filtering, increased dexterity and ease of suturing. Due to these advantages, recent multicenter studies report favourable results of robotic surgery compared to conventional laparoscopic surgery for pancreas and liver resections. This includes decreased rates of conversion for pancreatic surgery and potential technical advantages in high-demanding liver resections [3,4,5,6].

Until now, only few case series have been published describing the role of robotic surgery in patients with biliary obstruction [7]. The aim of this paper is to report our initial experience and feasibility of robotic bile duct surgery in a group of patients with biliary obstruction due to a variety of causes.

Materials and methods

Robotic hepatobiliary surgery was introduced in our center in the beginning of 2020. Initial robotic experience was gained with inguinal hernia cases, distal pancreatectomies and minor liver resections. Following our progressive experience with advanced robotic pancreatic and liver surgery, patients requiring complex bile duct surgery were consecutively selected for minimally invasive robotic surgery by surgeon preference from September 2020. There were no exclusion criteria for a robotic approach. Patients who underwent a pancreatoduodenectomy were not selected for this analysis as they represent a different patient category.

The indication for surgery was discussed by a multidisciplinary team for all cases. All patients underwent preoperative evaluation with computed tomography imaging and magnetic resonance imaging with cholangiopancreatography to define biliary anatomy.

Information regarding patient demographics, indication for surgery, intra-operative details and postoperative course were prospectively collected in our database. Institutional review board approval and written consent was not needed for this analysis.

Surgical technique

All surgeries were undertaken using the da Vinci Xi Surgical System® (Intuitive Surgical, Sunnyvale, CA, USA). Patients were placed in supine position with legs parted and a laparoscopic assistant standing between the legs. The operating table was placed in slight reverse Trendelenburg (12 degrees). Three robotic 8 mm trocars were placed in a straight line two hands below the xiphoid with a fourth trocar on the left side slightly more cranially (Fig. 1A). One or two additional 12 mm assistant trocars were generally placed in the lower abdomen in between two robotic trocars. The robot was docked, coming from the patient’s right side.

Fig. 1
figure 1

A Port placement for robotic complex biliary reconstruction; B creation of the posterior layer of the hepaticojejunostomy; C locking both threads of the anterior and posterior layer with a Hem-o-lok clip

Full size image

Meticulous dissection of the liver hilum was generally done using the monopolar robotic hook, monopolar scissors and bipolar Maryland or bipolar forceps. Robotic ultrasound and indocyanine green were used selectively to assess individual biliary and vascular anatomy. In cases where a liver resection was performed, parenchymal dissection was done using the Kelly clamp crush technique with robotic Maryland and robotic vessel sealer. In general, total pedicular clamping was used during parenchymal dissection to reduce blood loss.

When a bilio-enteric anastomosis was needed, a Roux-en-Y loop was prepared 30 to 40 cm from Treitz ligament which was divided by an endoscopic stapler. The distal loop was brought to the liver hilum through a mesocolonic window to create a tension-free anastomosis. Since most patients had dilated bile ducts in this series, the hepaticojejunostomy was generally performed end-to-side using two separate resorbable v-loc™ 4/0 (Covidien, Mansfield, MA, USA) running sutures for the anterior and posterior wall (Fig. 1B). Both threads were progressively tightened and locked together with a Hem-o-lok clip (Weck Closure Systems, Research Triangle Park, NC, USA) after finishing the anastomosis (Fig. 1C). In patients with non-dilated bile ducts, the hepaticojejunostomy is usually performed using separate sutures with vicryl 4/0 for the anterior and posterior wall. In case of slight dilated bile ducts, we sometimes use a running v-loc™ 4/0 suture for the posterior wall and separate vicryl 4/0 sutures for the anterior wall. In patients with separated bile ducts that need a bilio-enteric diversion (for example right and left duct or right anterior and right posterior duct), we generally perform a double barrel anastomosis resulting in a single hepaticojejunostomy (Fig. 2A).

Fig. 2
figure 2

A Double barrel anastomosis of the right and left hepatic duct; B common limb tattoo following Roux-en-Y hepaticojejunostomy to guide possible future ERCP

Full size image

An intracorporeal robotic end-to-side entero-enterostomy was made 60 cm more distally. The anastomosis was a two layered anastomosis using two layers of a resorbable v-loc™ 3/0 running suture. In recent cases, we used a small syringe with a fine needle to inject blue dye in the common limb to guide possible future endoscopic retrograde cholangiopancreatography (Fig. 2B).

In all cases, a silicone drain was placed near the hepaticojejunostomy to detect postoperative bile leakage.

Postoperative outcome

The surgical drain was removed on postoperative day 2 when drain production was negative for bile. Postoperative complications were classified using the Clavien–Dindo classification [8]. After discharge, patients were followed in our outpatient clinic at 1 month after surgery and on indication thereafter.

Statistical analysis

Descriptive statistics including percentages, medians and ranges were calculated using Microsoft Excel (Microsoft, Redmond, WA, USA).

Results

Between September 2020 and May 2021, 14 patients underwent robotic biliary surgery for benign and malignant biliary obstruction (Table 1). In the same period, 113 other patients had robotic hepatobiliary surgery, including 35 pancreatic resections (pancreatoduodenectomy or distal pancreatectomy) and 78 liver resections. Since the start of our learning curve, before starting complex robotic biliary surgery, we had performed already 42 technically or anatomically major robotic hepatic resections and 22 pancreatic resections. Our experience with robotic pancreaticoduodenectomy was useful to gain experience in constructing bilio-enteric anastomosis before starting complex robotic biliary surgery.

Table 1 Indications for surgery and performed robotic procedures
Full size table

This series of patients that had robotic biliary surgery included 4 males and 10 females between 23 and 77 years old (median age: 66 years) (Table 2). Eight of fourteen patients (57%) were referred to us from other hospitals. Injury to the common bile duct following cholecystectomy was the main indication for surgery (4 patients). Other indications included complicated gallstone disease (3 patients), malignant disease (4 patients) and benign biliary strictures following open surgery (3 patients). Six of fourteen patients (43%) had previous open abdominal surgery.

Table 2 Patient and operative characteristics and postoperative outcome
Full size table

In addition to biliary reconstruction, additional liver resections were performed in 4 patients, including two major hepatectomies. One patient underwent additional closure of a cholecystocolic fistula with wedge resection of the colon and Heineke–Mikulicz plasty (patient nr. 5) (Fig. 3) and another patient had a choledochoplasty with the remaining wall of the gallbladder (patient nr 8) [9, 10].

Fig. 3
figure 3

Mirizzi type IV with cholecystocolic fistula

Full size image

Operative characteristics

Median operative time was 205 min (range 135–360 min) (Table 2). Most cases were completed within 4 h surgical time. Only one patient with a type IV perihilar cholangiocarcinoma that underwent a left hepatectomy with segment I and resection of the extrahepatic bile duct with lymphadenectomy with a hepaticojejunostomy on the separate right anterior and posterior branch had a significant longer operative time of 360 min (patient nr 12). The other patient that underwent a concomitant left hepatectomy had a history of recurrent abscesses in the left liver lobe due to a common bile duct injury following cholecystectomy (patient nr. 6). Operative time in this patient was 270 min.

Median blood loss was 25 mL (range 10–120 mL). There were no intra-operative complications or technical issues and no conversions.

Postoperative outcome

Postoperative length of stay ranged between 3 and 11 days (median: 6 days). Eleven patients were discharged within 7 days after surgery. One patient in whom a previous hepaticoduodenostomy was converted to a hepaticojejunostomy on the right hepatic duct had a postoperative bile leak that resolved on postoperative day 8 (Clavien–Dindo grade I) (patient nr. 10). She had concomitant delayed gastric emptying and was discharged on day 11. Another patient also had a postoperative bile leak that resolved on postoperative day 8 (Clavien–Dindo grade I) (patient nr 13). One other patient had delayed gastric emptying that needed a gastric tube for five days and total parenteral nutrition (Clavien–Dindo grade II). She was also discharged on postoperative day 11 (patient nr 12).

Discussion

Recently, robotics has been introduced in different fields of abdominal surgery as it is able to overcome most of the major limitations of conventional laparoscopy. Surgery of the biliary tree is technically demanding with fine dissection and advanced suturing limiting a standard laparoscopic approach. In most centers, complex biliary reconstruction is, therefore, most often performed by an open approach. The robotic platform allows for enhanced visualization, increased stability and dexterity creating the possibility of complex minimally invasive manoeuvres. The use of robotic surgery for complex biliary problems is, however, still scarce [11].

In this series, we describe the application of robotic surgery in a variety of patients with complex benign and malignant biliary obstruction. In most of these cases, a bilio-enteric anastomosis is necessary to provide a definitive treatment with sufficient biliary drainage. The results in our paper show that complex robotic biliary reconstruction is feasible with minor morbidity and limited length of stay. Also, median blood loss was less than 50 mL. Moreover, we experienced no conversions although almost half of patients had previous open upper abdominal surgery. Also, a number of patients underwent a concomitant robotic hepatectomy.

In the current literature, mostly anecdotal case reports regarding robotic biliary surgery have been published so far. This mainly concerns robotic-assisted hepaticojejunostomy following bile duct injury after cholecystectomy [12,13,14,15,16,17]. Cuendis-Velazquez et al. describe a series of 30 consecutive patients and report that robotic hepaticojejunostomy can be performed safely with acceptable short-term results [18]. Another series of 12 patients confirm these results with, however, higher blood loss [19]. One study from southeast Asia included 10 robotic hepaticojejunostomies (including two pancreatoduodenectomies) in a total series of 27 robotic biliary cases. They also showed a low anastomotic complication rate [20]. The largest series of robotic hepaticojejunostomies includes 152 consecutive patients of whom 96 had a pancreatoduodenectomy [21, 22]. They reported similar rates of anastomotic leak and strictures as compared to open surgery. Other published case reports describe successful choledochal cyst excision with hepaticojejunostomy [23,24,25].

We included a mix of patients including three patients with Mirizzi syndrome (type III and IV) [26]. In such complicated cases, a conventional laparoscopic approach is associated with high conversion rates to open surgery [27]. However, using the robotic platform, we successfully managed a case of cholecystocolic fistula with dense surrounding inflammation as well as a large choledochal defect that was closed using the remaining gallbladder wall that was discharged uneventfully on postoperative day 3 [9].

Regarding our technique for the construction of the hepaticojejunostomy we generally used two separate v-loc™ 4/0 (Covidien, Mansfield, MA, USA) running sutures for the anterior and posterior wall that are locked together with a Hem-o-lok clip since most patients had a dilated common bile duct. Some authors suggest that continuous suturing carries a risk of anastomotic strictures, especially in patients with non-dilated bile ducts [28]. However, this was not confirmed in a recently reported large robotic series [21]. Indeed, we believe that a running suture with a single thread may lead to a narrow anastomosis. However, in case of dilated bile ducts, the tension-free locking of both threads with a clip instead of tightening them together might help to overcome the risk of anastomotic strictures. The follow-up in our series is, however, too short to analyze this hypothesis.

For patients with biliary obstruction, a Roux-en-Y hepaticojejunostomy is the preferred technique of reconstruction in our center. However, in one patient who previously had a gastric bypass (patient nr. 13), we performed a hepaticoduodenostomy to avoid the construction of a second Roux-en-Y limb and changing the length of the common channel. Indeed, it was previously showed that a hepaticoduodenostomy is a safe technique in selected patients [29].

The additional costs of robotic surgery compared to open or standard laparoscopic surgery are frequent mentioned as limiting factor in its wide implementation. However, the possibility to perform complex biliary surgery in a minimally invasive way will certainly result in a decreased length of hospital stay. The enhanced visualization and dexterity of the robotic platform may possibly also lead to a lower anastomotic complication rate favoring the use of robotic surgery. The three-dimensional view also facilitates adhesiolysis after previous open surgery with a negligible risk of conversion to open surgery.

Biliary surgery and reconstruction for benign and malignant diseases is generally considered as complex surgery that needs referral to specialized centers to optimize patient outcome [30, 31]. Even more, when performed robotically, sufficient robotic case load and experience is of significant importance.

Conclusion

Robotic surgery for benign and malignant bile duct obstruction is efficient and safe in experienced hands. Our previous experience with robotic and laparoscopic hepatobiliary surgery has possibly favored these positive outcomes. Nonetheless, robotics overcome most of the major limitations of conventional laparoscopic surgery required for advanced biliary surgery. By this way, these complex patients can also have the advantages of minimally invasive surgery. Referral to a high-volume expert center is, however, advised and further matched comparison with open and laparoscopic procedures is warranted.