Feasibility of a sutureless anastomotic technique in robot-assisted vascular surgery

Abstract

While other specialties embraced robotic technology, vascular surgery was revolutionized by endovascular techniques. The combination of open surgical techniques with endovascular solutions became standard practice; however, it is considered an unexplored frontier in the evolving field of robotic vascular procedures. One of the criticisms against robotics is the difficulty of vascular anastomosis. Hybrid grafts—although now discontinued—were once used for sutureless vascular anastomosis during open reconstructions. This article evaluates the feasibility of a sutureless anastomotic technique by reincarnating the idea of the hybrid vascular graft: a PTFE graft connected to a self-expanding covered stent.

The Da Vinci surgical robot (Intuitive Surgical) has revolutionized surgical practices, but it has yet to find its place in vascular surgery, despite the abundance of procedures involving core vascular surgical techniques performed by urologists, general surgeons, and gynecologists.¹ The robotic platform has proven to improve outcomes, shorten the length of stay, and decrease the need for postoperative pain management in various fields.2, 3, 4, 5

One of the criticisms from the vascular community is directed at the safety and difficulties of vascular anastomosis creation. Perhaps this is one of the factors that prevented the platform’s wider adoption. On the other hand, vascular surgeons embraced endovascular techniques that are now considered first-line approaches in a wide selection of indications ranging from aortic pathologies to peripheral arterial disease.^6,7 Therefore, most vascular surgeons are comfortable with the endovascular toolkit but generally have limited to no experience in laparoscopic techniques.

A hybrid vascular graft (HVG) consisting of a self-expanding stent incorporated into the end of an expanded polytetrafluoroethylene (ePTFE) graft, although now discontinued, was once used to facilitate anastomosis creation.^8,9 We aimed to assess the feasibility of utilizing the HVG concept to simplify robotic vascular anastomosis creation.

This research utilized only anonymized cadaver specimens obtained through informed consent (including their use for research purposes) from donors or their legal representatives prior to death, in compliance with all applicable legal and ethical standards for the use of human tissue in research.

Experiments

Sutureless iliac anastomosis with robotic assistance

A fresh frozen human cadaveric model was used at room temperature. The cadaver was positioned in a supine position on an interventional operating table tilted in a 15-degree Trendelenburg position, with the cadaver’s left side elevated in 30° to 45°. Pneumoperitoneum was established (and held between 12-15 mmHg) through a small 5-mm port inserted in the Palmer’s point (3 cm below the costal margin in the left midclavicular line). This port is then exchanged to an 8-mm robotic port. An additional three 8-mm trocars were inserted in the abdomen along the left midclavicular line approximately 8 cm apart, with two additional 12-mm assistant ports placed along the posterior axillary line.¹⁰ The robot was docked from the cadaver’s right side, and four robotic instruments, including a 30-degree robotic endoscope, were introduced (Fig 1).

Fig 1.

Patient positioning and port setup used for the procedure.

The infrarenal aorta, the right common iliac artery, and the first portion of the external iliac artery (EIA) were dissected free. We began by reflecting all the small bowels and the omentum to the patient’s right side and towards the diaphragm. We have used the technique of transperitoneal direct approach described by Dr Stadler.¹⁰ According to this, the retroperitoneum was opened on the left side of the aorta from the aortic bifurcation to the left renal vein alongside the medial border of the left gonadal vein. The posterior retroperitoneum, including the periaortic fat, was dissected and elevated from the aorta. This flap is then stitched up to the parietal peritoneum using multiple (usually 3-4) 3-0 nylon monofilament sutures with 60-mm long straight needles inserted through the abdominal wall, threaded through the flap, and then punched out and fixated with hemostats. This essentially creates a curtain that helps with bowel retraction and exposure. The tightness of this curtain can be adjusted externally by pulling or releasing the suture. After exposing the infrarenal aorta, dissection was carried down on top of both common iliac arteries inferiorly. Due to the use of the cadaveric model, clamping was not performed but could have been carried out distally with the use of a Rummel tourniquet, proximally a laparoscopic clamp. In this case, the guidewire should have been introduced first using the Seldinger technique. Although we did not perform this step in the current experiment, we have tried it in a later one that we plan to publish. A long, 18 G angiographic needle can be inserted through the abdominal wall into the target vessel; then a guidewire can be introduced, advanced to the desired location, and held in place by the robotic instruments. Next, the needle is removed, and the wire is pulled inside the abdomen. From there, the other end of the wire can be externalized through an assist port. The other option is to introduce a sheath through the abdomen to act as a “port” and perform the intervention through that.

Due to our focused approach on proving the feasibility of the creation of the anastomosis and the limiting factor of flow simulation, we proceeded to open the EIA via a transverse arteriotomy without clamping.

Next, we brought in a 6-mm ePTFE graft through an assist port. This was threaded onto an 8 × 40 mm Gore Viabahn stent (W. L. Gore & Associates, Inc) with the robotic instruments (a fenestrated grasper and a needle driver) that were introduced through the assist port on a 0.35 profile guidewire.

The Viabahn graft was then placed in the EIA, and the ePTFE graft was positioned near the entry site.

A preoperative computed tomography scan can be used to measure the length of the landing zone. In a previous series with HVG, generally 2 cm was used.¹¹ The stent can be measured and marked externally, or fluoroscopy can be used to plan out the landing zone.

After positioning, the Viabahn was opened halfway in the artery, halfway inside the ePFTE graft. While removing the delivery system, the robotic arms were used to hold the wire in place. Then, a 7 × 40 mm balloon was introduced over the wire, positioned, and inflated. Finally, we removed all endovascular devices. The time to completion was 14 minutes and 50 seconds for the endovascular part of the procedure (Fig 2; Supplementary Video 1, online only).

Fig 2.

Steps of the first experiment. (A) Threading the covered stent through the graft. (B) Inserting the stent inside the external iliac artery (EIA). (C) Positioning the graft to the entry point. (D) Deployment of the covered stent. (E) Balloon angioplasty to reach tight seal. (F) Opened polytetrafluoroethylene (PTFE) graft with the fully expanded covered stent inside.

Sutureless aortic anastomosis with robotic assistance

In the second phase of the experiment, we wanted to deploy an aortic stent graft with robotic assistance in the transected infrarenal aorta. This serves as the proximal anastomosis, while the limbs of the graft could be connected to the desired outflow either by sutured or sutureless anastomosis techniques.¹² In theory, the endograft could be used to fixate a Dacron or PTFE graft similarly, as it was reported previously by Segers et al and Donas et al.^12,13

The infrarenal aorta was dissected free and completely transected at the infrarenal portion.

A Gore Excluder bifurcated aortic stent graft was introduced through an assist port over a stiff guidewire. The graft was advanced in the aorta and subsequently opened.

We managed to acquire a tight seal and strong fixation due to the graft’s endo-fixation system. We were unable to dislodge the stent after it was opened (Fig 3; Supplementary Video 2, online only).

Fig 3.

(A) The aortic stent graft is introduced into the transected infrarenal aorta. (B) Deployed stent graft. (C) Inside view of the attachment point inside the aorta. (D) Fully deployed graft seen in the infrarenal aorta. (The polytetrafluoroethylene [PTFE] graft anastomosed to the aorta is from a previous experiment and does not play a role in the current one.)

Discussion

In these experiments, we combined robot-assisted and endovascular techniques to facilitate anastomosis creation.

The basis of both experiments is two similar sutureless telescoping anastomotic techniques, referenced as the Viabahn Open Rebranching Technique (VORTEC) or EndoVascular REtroperitoneoScopic Technique (EVREST).12, 13, 14 These involve a covered stent deployed partially in the native vessel and partially in a vascular graft to create a “side branch” that can be used for bypass creation. The VORTEC technique was described to facilitate and speed up the anastomosis of visceral branches during thoracoabdominal repair.¹¹ The discontinued Gore Hybrid Vascular Graft (W. L. Gore & Associates, Inc) was used in complex aortic reconstruction similarly with good results.^8,9

To the best of our knowledge, our experiment is the first to describe the use of this technique in a robot-assisted vascular procedure. In the first experiment, the iliac artery was used as a target. After robot-assisted dissection, we managed to implant a 6-mm PTFE graft end-to-side with the use of a 7-mm balloon expandable covered stent.

Although we did not gain vascular access using the Seldinger technique in this particular experiment, we did not encounter any issues with the entry angle of the guidewire or difficulties in maintaining wire access during our experiments. The guidewire entry angle and position can be kept shallow enough by utilizing the farther assist port from the target vessel as the wire exit/entry site.

Notably, the guidewire length has to be carefully planned depending on the anatomical location and the shaft length of the stent, with altogether a longer wire length preferred, although if the wire is too long, it can complicate manipulation.

Because robotic arms have seven degrees of freedom, needle drivers or fenestrated graspers were found to be useful in manipulating endovascular tools without issues. Device exchange was seamless as the wire could be held tight and kept in place by the robotic arm and with the countertraction of the assistant externally.

The procedure required 14 minutes and 50 seconds from the time of introducing the first endovascular device through a port to removing the guidewire. Of note, this was our first attempt with this technique. Therefore, the anastomotic time is expected to further improve along the learning curve. However, even this attempt was roughly twice as fast as the time required for a sutured anastomosis. Median anastomotic times in the largest series of vascular robotic procedures were reported to be around 30 minutes.¹⁵ Compared with this, the hybrid anastomotic technique promises a significant improvement.

Although this experiment demonstrates the concept’s feasibility, its potential application in patients remains a topic of further research. This is due to the following limitations. Due to the restrictions of a cadaver model and not having access to a pulsatile flow model at that time, we tested the created seal by longitudinally cutting the graft to visualize the anastomosis.¹⁶ The graft was completely expanded with a landing zone in the EIA. Then, we proceeded to remove the graft to test its fixation. Due to the lack of haptic feedback on the robotic arms, resistance could not be felt. However, visual clues suggested tight fixation. We considered the fixation of the stent with sutures, as it was proposed in the description of VORTEC and used with HVGs, but we did not perform it in this experiment.^11,12 Our focus was to create a truly sutureless anastomosis, reducing the need for suturing as it can be cumbersome or can even dislodge the implanted stent. In the second experiment, the hooks of the Gore Endurant graft proved to be appropriate to hold the graft in place.

Our future goals include the development of a fixation device with a similar mechanism as endo-anchors or repurposing vascular closure devices like the Celt ACD (Vasorum) to act like a sutureless fixation system.

We did not use clamping on the blood vessels; therefore, we could not test whether the technique would be hindered by the use of vascular clamps or if vascular control would be adequate. Clamping (if necessary) can be done proximally through an additional skin incision or with the insertion of an additional port with a laparoscopic clamp, endovascularly with balloons, or laparoscopically with modified Rummel torniquets. Hypothetically, not all patients will require clamping when using the variation of the technique with needle access. Four of 12 patients required clamping of the aorta due to anastomotic bleeding after opening the end-to side endograft using the EVREST technique, where they accessed the aorta with needle puncture. Additionally, the end of the graft was suture-ligated in advance to avoid the need for clamping.¹³ In our preliminary experiment, we focused on the feasibility of the anastomosis creation. Testing anastomotic leaks and scenarios where clamping is necessary is among our future ideas.

Furthermore, we did not use fluoroscopy to position our devices, and we did not perform an angiography to confirm their position after deployment. This could be exchanged with detailed preoperative imaging-based planning, the measuring and marking of the device of choice.

Finally, the host vessel’s patency can potentially be compromised during the end-to-side anastomosis creation by the inserted covered stent. In theory, this could be solved by the placement of another covered stent inside the host vessel in a kissing or double barrel configuration to maintain distal flow. Perforation of the inserted portion or even designing a special stent with an uncovered portion could be possible answers to this issue. Further questions, such as expected patency, remain unanswered.

Apart from the concerns mentioned above, the value of this experiment lies in proving the feasibility of merging robotic and endovascular techniques, highlighting the unique ability of vascular surgeons to perform minimally invasive operations with the combination of various techniques. The declining number of open reconstructions due to frequent short-term complications and the long-term reliability issues of endovascular aortic repair call for a new approach. Robotics could be a possible answer because it retains core, time-proven vascular surgical techniques while providing a minimally invasive approach.^17,18 Learning curves of certain robotic vascular surgical tasks like anastomosis creation and clamping times have shown to be short, although vascular procedures as a whole are complex and difficult to evaluate.¹⁹ Focus in training should be put on individuals with prior laparoscopic or robotic experience and notably younger generations, whose familiarity with modern technology has been proven to boost the uptake of robotic skills.²⁰ This makes incoming general surgical fellows prime candidates when building a vascular robotic program.¹

Conclusions

Sutureless vascular anastomosis during robotic vascular procedures using covered stents seems feasible. More studies are warranted to further evaluate the technique’s applicability before its use in patients.

Funding

None.

Disclosures

P.C. is a senior scientist at Occam Labs; and an interventional consultant at Siemens Medical Solutions USA Inc. A.B.L. receives research support from W. L. Gore & Associates; consults with Boston Scientific, W. L. Gore & Associates, and Siemens; and is a shareholder in Hatch Medical, Egg Medical, and Brijjit.

From the Southern Association for Vascular Surgery

Appendix (online only)

Supplementary Video 1 (online only)

First experiment. End-to-side sutureless anastomosis (external iliac artery [EIA]).

Supplementary Video 2 (online only)

Second experiment. End-to-end sutureless anastomosis (aorta).