The importance of retinal vascular manipulation for the treatment of vasculopathies has received increasing attention during recent years. To date, however, only minimal experimental or clinical interventions such as puncturing, probing, and sheathotomy have been tried.
1 2 3 4 5 6 7 In the present study, we have demonstrated the feasibility of additional surgical maneuvers on retinal vessels in a manner analogous to vascular surgery elsewhere in the body. Using a step-by-step approach in a porcine eye model, we have shown that several basic maneuvers can be combined for the performance of more complex manipulations.
Intravascular access, probing, and catheterization were the first maneuvers that were investigated. We were able to duplicate the findings of previous researchers who have shown that puncturing, probing, and intravenous injections can be performed in retinal vessels.
5 6 7 8 For arterial puncture, we found that pretreatment with the Er:YAG laser, to thin the tougher arterial wall, allowed subsequent puncture with a sharp instrument to be easily and reliably achieved. Conversely, the handling of the relatively more compliant veins was facilitated by the intravascular injection of a supporting material. In addition to venous injections of air and aqueous solutions, we were also able to inject a viscous material that provided internal support for the thin-walled veins in our porcine eye model and facilitated their separation from the retinal tissue and mobilization, although this could be accomplished with more difficulty without injection.
Probing and catheterization of the vascular lumen was achieved with sutures, glass tubes, and polyimide tubes. The end point in this study was the advancement of the catheterization device into the vascular lumen for a sufficient length and not the mere insertion of its tip into the lumen. Beveling of the catheterization tube tip was critical for easing entry. The most important factor affecting the ease of vascular catheterization was the catheter diameter. We achieved the best results with polyimide tubes with an internal diameter of 50.8 μm and a wall thickness of 7.6 μm. With these, we were able to catheterize both veins and arteries. This size seems to represent the maximum catheter diameter that can be used for arterial catheterization. Because porcine and human retinal arterioles share similar sizes, these findings can be used as a reference for future experiments in humans.
9 Taking into account that the average luminal diameter of the large human retinal arteries is 120 μm and decreases to 8 to 15 μm in the periphery,
10 it is obvious that tubes of smaller diameter must be used to catheterize peripheral arteries. Although micropuncture of both arteries and veins with fine tubes has been reported in the past,
5 6 7 the advancement of a tube into a vessel has not been described in the literature, as far as we know. In their significant work concerning surgical approaches to retinal vascular occlusions, Tang and Han
8 used only suture material as an experimental probing device. They stressed, however, the importance of retinal vessel cannulation with a catheter for the delivery of thrombolytic agents. A number of other possible applications of these basic maneuvers have been described in the literature, including measurement of retinal circulation pressures, and analysis of local vascular parameters such as metabolites and volume flow rates.
11 Intravascular procedures using the Er:YAG laser after the development of appropriately sized laser probes might be added to this list.
The application of well-established macrovascular procedures, such as anastomosis and vessel bypass, to the microvascular level required by retinal vessels will necessitate the performance of more invasive maneuvers. Dissection and isolation of the vessels from the surrounding tissue represents the first prerequisite step.
12 In this study, we were able to demonstrate that separation of retinal vessels from the retinal tissue can be achieved. Separation of retinal vessels from the retinal surface, known as retinal vascular avulsion, is known to happen automatically as a result of traction forces applied on the vessels.
13 14 Similarly, surgical separation of retinal arteries from the retina is used as part of the newly developing sheathotomy surgery for branch retinal vein occlusions.
1 2 We were able to achieve a controlled separation of retinal arteries from the retina to a considerable extent, in many cases up to several millimeters in length. This length, after severance of small branch vessels, is sufficient to allow for unconstrained mobilization of major arterial branches. Arterial separation and mobilization were achieved with minimal damage to the underlying retinal tissue, whereas venous mobilization was much more difficult because of their friable nature and deeper location. The intravenous injection of viscous material facilitated vein mobilization, but full-thickness retinal damage was more common, especially early in the experiments.
The connection of vascular branches represents another significant step for the completion of a major microvascular procedure. With combined mobilization of two arterial branches, we were able to arrange the vessels in a way that would facilitate a side-to-side or an end-to-end anastomosis. In addition, we were able to connect different vascular branches with glass or polyimide tubes. The combination of puncture, catheterization, and mobilization permitted the establishment of a connection similar to an end-to-end anastomosis. The demonstration of feasibility of vascular branch connection by means of fine tubes advanced into the vascular lumen may offer the basis for utilization of synthetic grafts in retinal microvascular surgery, although several questions regarding anastomosis functionality remain. In this work, we did not check the patency of the anastomosis achieved with the tubes. Moreover, no measures were taken to obtain a seal connection between the vessels and the tubes. In future experiments, these questions should be resolved and the patency of the anastomosis should be verified with dye injection. However, even if immediate patency is demonstrated, the long-term patency of small-caliber synthetic grafts remains a problem.
15 Other methods of vascular connection may also have a role in retinal vascular anastomosis. Although suturing may not be feasible because of size restrictions, the use of glue materials, such cyanoacrylate or thrombin glue, may be useful alternatives. These glues have been used for microvascular anastomosis in other body locations,
16 17 in addition to some work in retinal tissues such as gluing peripheral or macular holes.
18 19 20
In the current work, we used a porcine eye model and an open-sky approach. In so doing, we eliminated two major sources of difficulty. The first was the active blood flow of a living, retinal circulation. Management of the blood flow during retinal vascular procedures in the living eye represents a serious challenge. Tang and Han
8 reported that hemostasis could be accomplished by simply increasing the intraocular pressure after penetration of the vascular wall in a living animal. They also proposed a bimanual technique and intravitreal injection of perfluorocarbon liquids to achieve hemostasis. Development of miniature vascular clips or “sandbags” could represent another option for hemostasis during retinal microvascular procedures. The second source of difficulty in future animal experiments is the added level of difficulty if vascular manipulations are performed through the small openings used in standard three-port pars plana vitrectomy. The development of specialized vitreoretinal instruments and the appropriate modification of some of these maneuvers are needed to overcome these hurdles.
In conclusion, we have demonstrated the feasibility of several microsurgical maneuvers on retinal vessels. Further experimentation and innovations in a living animal model are necessary to establish the feasibility of these maneuvers in the living eye. Classic microvascular techniques have been used with great success to reperfuse ischemic tissues in other fields, such as cardiothoracic and vascular surgery. Our results represent further steps toward the application of such classic microvascular techniques that could lead to vascular bypass and anastomosis at the level of retinal vasculature and the exciting promise of novel treatments for retinal vascular diseases.