Microvascular anastomosis of blood vessels is required during free tissue transfer. The quality of blood vessel anastomosis is important for the success of the procedure. End-to-side anastomosis, in particular, requires a more specialized technique than end-to-end anastomosis. The choice of the microsurgical anastomotic technique, end-to-end or end-to-side anastomosis, must be made according to recipient vessel quality and accessibility. There was no statistically significant difference in surgical outcomes, and end-toside anastomosis did not increase the risk of complications, while preserving distal perfusion.^1,2 If end-to-side anastomosis is facilitated, anastomosis between vessels of varying sizes and in other difficult cases may become more efficient.

For many years, hand-sewn techniques have been the mainstay for microvascular anastomosis. However, anastomotic thrombosis occurs in clinical cases where skilled surgeons perform microvas-cular anastomosis using hand-sewn techniques. Analyses of the surgical complications of 1000 microvascular free flaps, 4.5% of arterial thrombosis, and 6.8% venous thrombosis have been reported.³ Therefore, devices that facilitate microvascular anastomosis and simplified procedures that do not require specialized techniques are needed.⁴

Anastomotic coupling devices were made available in 1962 by Nakayama,⁵ and Östrup⁶ subsequently reported modified versions. Chernichenko et al⁷ reported a free-flap survival rate of 97.6% in 124 arterial anastomoses using unlink coupling devices. However, the use of unlink coupling devices for performing arterial anastomoses has not gained widespread acceptance. This is because these devices are better suited for venous anastomoses due to the reliance on the pliability of the vessel wall for eversion over the pins of coupling devices.⁸

Preliminary in vivo studies on microvascular end-to-end anastomosis with a bare-metal stent report numerous advantages over standard microvascular anastomosis.^4,9 Saegusa et al⁴ reported that sutureless anastomosis using a bare-metal stent yields good results that were not inferior to those obtained by hand-sewn techniques, as thrombus formation, patency rate, blood vessel strength, and operation time for anastomosis were all significantly improved.

End-to-side anastomosis using metal stents has not yet been reported, although many end-to-side anastomosis procedures have been described in clinical practice. Direct bypass in patients with moyamoya disease is classically treated by end-to-side anastomosis of the superficial temporal artery to the M₄ division of the middle cerebral artery.¹⁰ Three-vessel coronary artery disease is currently treated with coronary artery bypass grafting. The left internal mammary artery is used for bypass grafting and offers superior long-term outcomes compared to venous grafting. The Tector procedure uses arterial end-to-side grafting in coronary artery bypass and is safe and feasible.¹¹

Some studies have reported sutureless anastomosis using metal stents; however, most of these studies involved end-to-end anastomoses in rats.^4,9,12–16 Prabhu and Homer-Vanniasinkam¹⁷ and Liang et al¹⁸ conducted studies in rabbits. Mbaidjol et al¹⁹ reported laser-assisted vascular end-to-end anastomosis in piglets. As swine models are anatomically close to humans, we selected porcine models for our experiment.²⁰ However, to the best of our knowledge, no reports on end-to-side sutureless anastomosis with metal stents using a swine model have been published so far.

We invented T-shaped metal stents and an end-to-side sutureless anastomosis technique, which involves the subsequent use of adhe-sives around the circumference of the site of anastomosis. In this experiment, T-shaped metal stents and our anastomosis technique were used in piglets. We evaluated the patency of blood vessels and the presence of thrombosis at 4 weeks postoperatively to verify the safety, effectiveness, and operability of the newly developed stents.

MATERIALS AND METHODS

All experimental protocols were approved by the animal experiment committee at Jichi Medical University (approval no. 16012).

Animal Models

Four 7- to 11-month-old male Mexican hairless piglets were purchased from the Ibaraki branch of the National Livestock Breeding Center. The piglets weighed between 31.1 and 45.0 kg at the time of the study. All piglets were treated with ticlopidine hydrochloride 200 mg/day from day 3 preoperatively and acetylsalicylic acid 200 mg/day from day 2 preoperatively. The administration of both agents was continued until the end of the study. General anesthesia was induced in the piglets using medetomidine at 0.6 mg/kg body weight and midazolam at 0.3 mg/kg body weight intramuscularly. Anesthesia was maintained with 3.0% sevoflurane. Heparin sodium at 150 μg/kg was used perioperatively and immediately before autopsy for anticoagulation. The piglets were observed carefully to identify any decrease in activity, and their wounds were inspected once a week. If a potential problem was identified, a detailed physical examination was performed, and appropriate measures were taken. For instance, a bandage roll was applied for wound dehiscence.

Stent

We developed a T-shaped metal stent that was constructed with nickel titanium. The design was intended to avoid interference with blood flow when placed in the microvascular lumen (Fig. 1A-B).

The stents were made by laser material processing and heat treatment. We minimized the surface of the stent to prevent corrosion when the stents are retained within the body. We made the stents as thin as possible, and a thickness of 100 μm was used to minimize the size of any indentation and thus prevent thrombosis from turbulent blood flow. The stents were T-shaped, comprising a flat base with a cylindrical part that diverged from the base.

The stents were made of a memory alloy shape and could be compressed when inserted into the blood vessels. On the tip of the cylindrical part of the device, a triangular projection was created to minimize the risk of positional changes.

We prepared T-shaped metal stents of various sizes. Before the procedure, we used a gauge to unlink coupling devices with stumps of resected femoral artery for the selection of an appropriate stent size for the anastomosis.

Surgical Techniques

After inducing anesthesia, each piglet was fixed in a supine position. The operative field was sterilized, and the surrounding areas were covered with a sterile drape.

An incision was made on the skin over the medial side of the left thigh. The sartorius and gracilis muscles were divided, and the saphenous artery was exposed. The saphenous artery was dissected as far as the bifurcation of the femoral artery. The medial vastus and pectineus muscles were then separated, exposing the femoral artery. The exposed femoral artery was then resected by approximately 8 cm. In the same way, the femoral artery in the right thigh was exposed. The superior and posterior stumps of the resected left femoral artery were used for end-to-side anastomosis with the right femoral artery (Fig. 1C).

We then performed a sutureless end-to-side anastomosis technique using the T-shaped metal stents under surgical microscopy (M300; Leica, Ikegami, Japan). Cyanoacrylate glue (Aron alpha A; Sankyo, Tokyo, Japan) was then used to reinforce the anastomotic region (Fig. 1D). A surgical image for the anastomosis using T-shaped metal stents is shown in Figure 2.

The technique for end-to-side anastomosis of the T-shaped metal stents was as follows. Temporary vessel clips were placed at both anastomotic sites of the resected left femoral artery and right femoral artery. For the right femoral artery, 2 clips were placed at the proximal and distal parts of the blood vessels to interrupt blood flow. Two holes were made at the side of the right femoral artery between the 2 clips. At these 2 holes, end-to-side anastomoses were carried out with the resected left femoral artery. Insertion of the T-shaped metal stents was performed using microforceps. After the T-shaped metal stents were inserted, cyanoacrylate glue was used to reinforce the anastomotic region. Blood flow was interrupted by placing titanium ligating clips (LIGACLIP; Ethi-con-Johnson & Johnson Japan, Tokyo, Japan) between the 2 sites of the end-to-side anastomosis on the right femoral artery. The temporary clamps were removed, and blood flow was restored. When the operation was concluded, the incision of the skin over the thigh was closed by interrupted hand sutures with 3-0 nylon. The piglets were then extubated and carried to the breeding room.

Imaging Studies

Perfusion and patency of the vessels that were anastomosed using the T-shaped metal stents were assessed using ultrasonogra-phy (Aplio 300; Toshiba, Tochigi, Japan) or contrast-enhanced computed tomography (SOMATOM Definition AS; Siemens, Tokyo, Japan). In 3 cases, blood flow was evaluated using ultraso-nography at 4 weeks after the microvascular anastomosis. Contrast-enhanced computed tomography was performed in 1 case, and this subject was prepared for the examination using this modality.

Histology

An autopsy was performed 4 weeks after the end-to-side anastomosis, and the blood vessels were harvested. The blood vessels were fixed in 4% formalin and placed in acrylic resin blocks. Samples that were prepared in this manner were cut into 5-μm-thick sections. All samples were stained with hematoxylin and eosin stain (Sept Sapie, Tokyo, Japan). We observed and measured sections using a virtual slide system (NanoZoomer and NDP.View, Hamamatsu, Japan). We evaluated the probe patency of blood vessels and histological changes in the blood walls around the stent branches.

RESULTS

All the animals (100%) survived the procedure. Four piglets underwent the procedure, resulting in the anastomoses of 8 blood vessels. Supplementary Digital Content, Table 1, https://links.lww.com/SCS/D515 shows a summary of the individual surgical outcomes.

We did not record the exact duration of the blood-vessel anastomoses. However, the procedures, including preparations of the blood vessels and 2 sites of end-to-side anastomosis using metal stents, were performed under microscopy and required approximately 50 to70 minutes per subject. The time required for the insertion of metal stents and application of cyanoacrylate glue was 12 to 20 minutes per anastomosis site.

As shown in Supplementary Digital Content, Table 1, https://links.lww.com/SCS/D515, no thrombosis was seen in any of the 8 anastomotic sites under any imaging modalities at 4 weeks post-operatively. Nevertheless, the overall 4-week patency rate was 100%.

Doppler ultrasonography confirmed blood flow at the anasto-motic site and no thrombosis was apparent in 3 piglets at 4 weeks postoperatively (Fig. 3). In the remaining piglet that underwent contrast-enhanced computed tomography, clear contrast was apparent in the anastomosed vessels (Fig. 4). We could have performed a Doppler ultrasonography; however, we did not do so because the result from the contrast-enhanced computed tomography was sufficient.

The anastomosed vessels were harvested at 4 weeks postopera-tively to evaluate chronic histological changes after anastomosis with the T-shaped metal stents. A strong scar tissue was seen around the anastomosed vessels. We could confirm the patency of the blood vessels and the existence of neointimal cell proliferation around the stent branches on histological examination (Fig. 5).

DISCUSSION

We evaluated the metal stents reported by Saegusa et al⁴ in a murine model for safety, and the materials and design of our T-shaped metal stents included various improvements on that design. The diameter of the femoral artery in our model was suitable as a model for end-to-side anastomosis using T-shaped metal stents in humans. The exact diameter of the femoral arteries of the piglets was not measured at the time of anastomosis, but the longer axis of the internal lumen was 1.5 to 2.0 mm on histological examination.

Perfusion and patency of the anastomosed vessels with T-shaped metal stents were assessed using ultrasonography or contrast-enhanced computed tomography. We confirmed blood flow through the T-shaped metal stents. Ultrasonography was efficient as a method of evaluation; however, contrast-enhanced computed tomography provided clearer images than the ultrasonography. We evaluated only 1 case using contrast-enhanced computed tomography. Future studies should therefore be conducted using contrast-enhanced computed tomography for evaluation.

Saegusa et al⁴ found that the histopathological examination of metal stents in rats showed mild local intimal thickening that resembled similar phenomena at the placement sites of cardiac stents. However, the diameter of the lumen was unchanged. Similarly, our porcine study could confirm probe patency of the blood vessels and neointimal cell proliferation around the stent branches on histological examination.

After the T-shaped metal stents were inserted, cyanoacrylate glue was used to reinforce the anastomotic region because retention was weak with the T-shaped metal stents alone. A recent study used temporary intravascular stenting during sutureless microvascular anastomosis with a cyanoacrylate adhesive, and the short-term safety outcomes of this technique were established at the experimental stage.²¹ However, the end-to-side anastomosis may have exerted a different pattern of hydrodynamic stress on the glued areas; future experiments using ex-vivo stress tests may be necessary.

We considered improvements to optimize retention and opera-bility as future problems. Insertion of the T-shaped metal stents at each artery stump was relatively easy using the microforceps. However, fixing each blood vessel with cyanoacrylate glue was slightly difficult. A hematoma could form if a blood vessel is not fixed by the liberal application of the cyanoacrylate glue. Close attention should be paid during end-to-side anastomosis of blood vessels to ensure that the size of the insertion point is appropriate for the metal stents used. Additionally, this adhesive does not harden in wet environments. As a result, anastomosis with a T-shaped metal stent would entail a similar duration as the hand-sewn technique.

Anastomotic couplers offer a superior alternative to hand-sewn methods, particularly, in cases of the anastomosis with a vein. Stanix et al²² reported that the rates of complications and flap failure were similar between reconstructions performed using a venous coupler and those performed using hand-sewn venous anastomosis in lower extremity reconstructions, which have the highest reported rates of complications and flap failure among all anatomical regions. Conversely, in a systematic review of venous anastomotic couplers by Wu et al,²³ unique complications, such as coupler extrusion and palpability, were reported in cases with previous radiation exposure or thin, soft tissues (eg, head and neck, hands, and feet). Methods such as the use of metal stents completely buried in the target blood vessels may prevent such complications.

When using anastomotic couplers for arterial anastomosis, everting thicker vessel walls onto the pins and abutment of the vessels is difficult. Sando et al²⁴ designed an arterial everter device to simplify arterial anastomosis with the use of anastomotic couplers in a swine model. That technique was easier and significantly more efficient when compared to the standard hand-sewn method in cases of arterial anastomosis. However, eversion of the arterial wall is required until the coupler pin is passed, and thrombosis occurred in 1 case due to increased tension along with potential intimal injury. Additionally, performing end-to-side anastomosis may be technically difficult even in cases of anastomotic couplers involving veins. T-shaped metal stents can be used in difficult cases that require eversion of the vessel wall.

The end-to-side venous anastomosis may be possible using the approach we have described for performing the arterial anastomosis. Venous couplers are already widely accepted in clinical practice for end-to-end anastomosis, and end-to-side venous anastomosis appears to be feasible. We hope to evaluate end-to-side venous anastomosis with these T-shaped metal stents in future studies.

This study had some limitations. The small number of animals included in this study is a limitation. Relatively few parameters were examined, and the accumulation of further data is necessary. We did not record the exact duration of blood vessel anastomosis, and this was not compared with those of other anastomotic techniques. This study tested the operability, but there was no comparison with a standard technique (manual suturing) to conclude its safety or effectiveness. T-shaped metal stents have limitations of poor operability and anastomotic retention, and further improvements were needed to improve their practical application. We could confirm the patency of the blood vessels and the existence of neointimal cell proliferation around the branches of the stent on histological examination; however, there were no comparisons with the conventional techniques using Nylon sutures.

In this study, we performed the end-to-side anastomosis technique in piglets and observed no thrombosis in any of the 4 subjects. Improvements in retention and operability are required before practical use of our T-shaped metal stents can be considered. We hope to continue improving the stents in order to develop a type of stent that does not require adhesive glue.

Supplemental Digital Content

• SCS_2021_10_17_SUGIURA_SCS-21-01268_SDC1.doc; [Word] (34 KB)