Robotically Assisted Laparoscopic Roux-en-Y Hepaticojejunostomy

Abstract

Introduction:

This study evaluates the feasibility and safety of using robotically assisted laparoscopy to perform a Roux-en-Y hepaticojejunostomy. This new method was compared with the open and standard laparoscopic approaches.

Methods:

Eighteen pigs underwent a needlescopic common bile duct ligation to create a jaundice model. Three to 5 days later, transabdominal ultrasound was performed, and the common bile duct diameter was documented. For the Roux-en-Y hepaticojejunostomy, the pigs were randomly assigned to the open group (n=6), standard laparoscopy group (n=6), or robotically assisted laparoscopy group (Zeus) (n=6). One surgeon performed all 3 approaches with 1 assistant. Operative times, techniques, and complication rates were documented.

Results:

The open approach was faster in all instances. At the hepaticojejunostomy, no difference was noted between the groups with the total number of stitches used. The robot required fewer stitches and less time in the posterior wall of the hepaticojejunostomy (P=0.0083 and P=0.02049, respectively). The hepaticojejunostomy time was similar for the laparoscopy and robotically assisted groups.

Conclusion:

Robotically assisted laparoscopic Roux-en-Y hepaticojejunostomy is a feasible procedure. When compared with standard laparoscopy, operating time is similar.

INTRODUCTION

In the early days of laparoscopic surgery, many surgeons used clips and staplers to avoid suturing. Most minimally invasive procedures were ablative rather than reconstructive until laparoscopic suturing was mastered. However, laparoscopic suturing requires advanced skills, which are mastered during advanced laparoscopic training.

Laparoscopic surgery requires many small and precise movements to manipulate and suture tissue in a confined space; it has shown applicability in pancreatic tumor for staging the disease and at the time of biliary bypass.^1–7

In 1997, robotically assisted surgery was proposed for cardiovascular applications.^8–14 Robotic surgery can increase precision, accuracy, and filter out hand tremor in a micro-anastomosis. The initial success in this specialty opened the spectrum to general surgery. Applications in the field of minimally invasive surgery began with camera voice activation.^15–17 Now, sophisticated robot devices can reproduce the surgeon's movements with the advantage of increased range of motion. The robot can reproduce a surgeon's moves rapidly, precisely, and can filter out the tremor of the human hand caused by fatigue. Still, the current status of robotics in surgery remains investigational.

This study evaluates the feasibility and safety of robotically assisted laparoscopy to perform a Roux-en-Y hepaticojejunostomy. This new method was compared with the open and standard laparoscopic approaches.

METHODS

The Animal Research Committee of Duke University Medical Center approved the study. Pigs received humane care in compliance with the guide for Care and Use of Laboratory Animals published by the National Institutes of Health.

Eighteen female Yorkshire pigs, weighing 18 kg to 31 kg, underwent a common bile duct (CBD) ligation to create a jaundice model. With the pigs under general anesthesia, 3-ports were placed with mini-laparoscopic instruments (2 mm to 3 mm). Pneumoperitoneum was performed, a window behind the CBD wall was created, and a distal duct ligation with silk was made. Transabdominal ultrasound was performed 3 to 5 days later, and the CBD diameter was measured and documented for follow-up. Once the CBD distention was confirmed, the animals were randomly assigned to 3 groups for the Roux-en-Y hepaticojejunostomy (HJ): open (n=6), standard laparoscopy (lap) (n=6), and robotically assisted laparoscopy (Zeus) (n=6). One surgeon, who had clinical experience in the assistant position in laparoscopic hepaticojejunostomy, performed all 3 approaches with the help of 1 assistant. Neither the surgeon nor the assistant had experience with robotics in surgery.

In the open group, under general anesthesia, the abdomen was prepped and draped in sterile fashion for ventral midline laparotomy. Once the common bile duct dilation was confirmed, the Roux-en-Y was performed as a side-to-side enteroenterostomy at 40 cm from the angle of Treitz by using a 3.0 running suture. A 1-cm incision at the common hepatic duct and small bowel was made, and the side-to-side hepaticojejunostomy was performed with a double running 4-0 suture. Starting with the superior vertex, the posterior wall was initially made with a running suture and later tied with the separate suture from the anterior wall, tying both sutures at the inferior vertex. Closure of the abdominal wall was accomplished in 2 layers.

In the standard laparoscopic group, six 10-mm ports were used (Figure 1). Pneumoperitoneum with a Veress needle was made at the umbilicus and maintained with a 12 mm Hg pressure. The liver was retracted with cephalic gallbladder traction with an Endograsp through the right lower port to expose the common bile duct.

Figure 1.

Port placement in standard laparoscopy group and robotically assisted group (Zeus system, Computer Motion, Inc); Surgeon's instruments (A,B); Endocamera (AESOP) (C); Assistant (D,E,F).

The Roux-en-Y was performed by exposing the small bowel through the umbilical port, and side-to-side enteroenterostomy was performed outside the peritoneal cavity. After completion, it was reintroduced into the peritoneal cavity, pneumoperitoneum was reestablished, and the Roux-en-Y was retracted cephalic close to the CBD. A 1-cm hepatic duct incision and enterostomy were made with laparoscopic scissors. The HJ was performed in the same fashion as described for the open group, with a double running suture using intracorporeal knots with laparoscopic needle drivers.

The Zeus system (Computer Motion, Inc., Goleta, CA) incorporates 3 robotic arms. Voice control of the AESOP camera allows the surgeon to simultaneously manipulate the laparoscopic camera and 2 laparoscopic surgical instruments from a remote location. The surgeon operated seated in an armchair in front of a console including a monitor and control instruments. Movements of robotic handles are transmitted to a computer controller that filters the moves, reduces the tremor, and then translates the surgeon's movements to the robotic arms. The scale can be managed in a way that a robot reproduces in smaller scale the surgeon's movements for more precision (Figure 2).

Figure 2.

The Zeus system (Computer Motion, Inc.) incorporates 3 robotic arms (A). While voice control of the AESOP (B) manipulates remotely the laparoscopic camera, movements of robotic handles are transmitted to a computer controller translating the surgeon's movements to the robotic arms.

In the Zeus approach, all the surgical steps were reproduced as in the standard laparoscopy group until the time of HJ when the Zeus sterile devices were set up. The surgeon operated from the console while the assistant stood between the robotic arms for tissue traction. The HJ was performed as in the other groups (Figure 3).

Figure 3.

Laparoscopic view of robotically assisted hepaticojejunostomy.

The time required for the HJ for the 3 groups, number of stitches on the posterior and anterior wall, and the total time of surgery were analyzed. Complications were also documented. The animals were recovered from the anesthesia with proper care after surgery and were sacrificed 2 weeks later for anastomosis inspection.

To analyze the data regarding the differences between the groups, we used 3 types of tests: multiple comparison tests, 2-sample t tests, and paired t tests.

RESULTS

Three to 5 days after creation of the jaundice model with mini-laparoscopic CBD ligation, ultrasound was performed prior to the biliary bypass. The CBD was measured, confirming dilatation with a mean of 17, 15, and 16.8 mm for the open, lap, and Zeus groups, respectively. One of the 18 pigs in the lap group developed a small ischemic hole on the anterior wall of the CBD after ligation. No free collections were shown in the ultrasound; this was probably an ischemic consequence of the ligation. This case was diagnosed at the time of surgery, and the CBD wall was sutured (4.0 running suture) adding 20 minutes with a laparoscopic approach before performing the HJ. The pig had a good postoperative course, with no other complications.

The open approach was faster in all instances (Table 1). Analyzing Roux-en-Y time, a statistical difference was found between the open (20.2 min) and the lap and Zeus (39.2 and 40.5 min) procedures.

Table 1.

Summary of Surgical Time for Open, Standard Laparoscopy, and Robotically Assisted Laparoscopy Groups

Group	Set-up (min)	Roux-en-Y (min)	Hepaticojejunostomy (min)	Total Surgery (min)
Open	0	20.2	22	100
Standard lap	0	39.25	78.5	243.16
Robotically assisted (Zeus)	30.16	40.5	77.33	260.83

Before the hepaticojejunostomy, the Robot set up for sterilized procedures included covering 3 arms with a sterile plastic cover required a mean of 30.16 minutes (range, 18 to 45) with a fast learning curve.

No difference was noted in the total number of stitches used in HJ among the 3 groups. When comparing Zeus to lap (2-sample t test), the former required fewer stitches on the posterior wall of the HJ (P=0.0083).

The time needed for the posterior wall of the HJ between lap and Zeus showed Zeus was faster. The total time required for HJ and surgery compared with that in lap and Zeus were similar (Table 2).

Table 2.

Hepaticojejunostomy for the Open, Standard Laparoscopy, and Robotically Assisted Laparoscopy Groups

Group	Posterior Stitches	Anterior Stitches	Total Stitches	Posterior Time (min)	Anterior Time (min)	Total Anastomosis Time (min)	Minutes per Stitch
Open	10	10.2	20.2 NS	7.4	13.8	22	1.07
Standard lap	11.8	9.25	18.5 NS	40.2	35	78.5	3.66
Robotically assisted (Zeus)	7.4 (P=0.008)	9.4	16.8 NS	27.4 (P=0.0204)	35.4	77.33 NS	4.08

Postoperative complications related to Roux-en-Y and HJ included 2 small bowel obstructions (11.1%) and 2 instances of ischemia/necrosis in the small bowel (11.1%) in the lap group. Only 1 pig (5.5%) developed a bile leak at the anastomosis due to a broken suture (first case of Zeus group). One pig developed a subcapsular hepatic hematoma found at necropsy (5.5%) (Zeus group), probably due to instrumentation.

DISCUSSION

The Roux-en-Y hepaticojejunostomy by a laparoscopic approach is a technically demanding procedure that requires knowledge of advanced laparoscopic suturing and the intracorporeal knot technique. Significant suture practice is required, especially when smaller instruments are used to master the technique. Trainees usually spend many hours practicing before competency is achieved.

A strong limitation while suturing with laparoscopic instruments is the angle of work between the direction of the needle drivers and the direction of the anastomosis line. In standard and robotically assisted laparoscopy, the most time-consuming and critical factor in performing the hepaticojejunostomy was establishing the proper working angle in the anastomosis and the instruments.

Robotics appears to provide some advantages to the surgeon, such as improved range of motion and filtering out hand tremor to allow micro suturing. A potential disadvantage is the cost of the device. Still, no proven benefits to the patient have been shown with these models in general surgery.

The laparoscopic instruments are fixed to the port and the tip is rigid. The robotic device used at Duke University was an initial prototype with a fixed tip of the instrument similar to that in laparoscopic instruments. Today, new robots have a “micro-wrist” in the instrument tip that allows the surgeon to have a complete flexible tip like a little “hand” inside the abdominal cavity with wrist mobilization through a small access port. Robotics may have an application in laparoscopic suturing, the learning curve is mainly to achieve the suturing task, and minimizing work angle difficulty.

Several studies reported about robotically assisted micro-suturing.^18–23 Garcia-Ruiz et al¹⁸ compared manual with robotically assisted laparoscopic maneuvering for suturing and found that the laparoscopic manual approach was faster when the size of the suture increased. We did not find differences in the total time of anastomosis (HJ) in these 2 groups; we performed all anastomoses with the same suture size.

We found a difference on the posterior wall number of stitches and time required. The Zeus group required fewer stitches than did the lap group (P=0.008). The posterior wall of the HJ took less time for the Zeus group than for standard laparoscopy group (Table 2) even after excluding one laparoscopic case (L4) which presented with a difficult angle of work (2-sample t test, P=0.02049).

The biliary leak complication rate was low (5.5%). When using the Zeus at the first robotically assisted anastomosis, this animal developed a postoperative bile leak due to a broken suture at the superior vertex of the anastomosis. This complication was caused by a loss of tactile perception with the robot instruments and a mesh in the suture, feedback that is important when tying the suture knot. A surgeon must learn how to handle the suture while maintaining visual control through the monitor.

We found that once the surgeon has some training in laparoscopic intracorporeal suturing, the learning curve for the robotic device is short. The basic concept of intracorporeal suturing is the same in laparoscopy and robotics because the instruments are similar. We believe that when the technique is mastered, the laparoscopic or the robotic-assisted approach does not make a difference, especially in this non-microwrist device. The critical factor of the angle may be minimized with more advanced robotic prototypes, such as flexible tip instruments. This brings a new concept, hands inside the peritoneal cavity through mini-invasive access, and it may show a difference for intracorporeal suturing techniques. Further studies are needed to analyze anastomosis time with the new microwrist devices.

Today, practical applications of Robotics technology in general surgery is investigational. The future potential is outstanding as technological advances continue and cost becomes more affordable.

CONCLUSION

Robotically assisted laparoscopic Roux-en-Y hepaticojejunostomy is a feasible procedure that requires the same operative time as that of standard laparoscopy.