Vitreoretinal surgical anatomy

The vitreous can be considered as a three-dimensional matrix of collagen fibers and hyaluronic acid gel (Fig. 101.1). In the normal state, the outer surface of the vitreous is in contact with the retina, pars plana, and ciliary body in a roughly spherical shape with an anterior facet for the lens. Disease-induced RPE and glial cell migration, attachment of cells to the extracellular matrix and hypocellular contraction of the collagen matrix take place with the majority of the relevant changes occurring at the cortex surface. The anterior vitreous cortex (AVC) is contiguous with the posterior vitreous cortex (PVC) and, for the most part, is nonfenestrated.

Fig. 101.1 The vitreous is a three-dimensional matrix of collagen fibers and hyaluronan gel.

Activated retinal glial cells, retinal pigment epithelial (RPE) cells, and cells of hematogenous origin migrate along the front and back surfaces of the retina and vitreous. These cells have coated pits lined with fibronectin, allowing them to attach to and contract the collagen matrix.

A detailed understanding of the abnormal vitreoretinal interface and its derivative geometry is requisite to undertaking vitreoretinal surgery. The task involves visualization of vitreous structures and a systematic search for membranes based on observed retinal topology. In general, membranes are white and matte-finish, whereas the retina has a reflective surface luster and appears pale yellow. If a complete posterior vitreous separation has not occurred, there is usually continuity between areas of epiretinal membrane (ERM) and adjacent detached PVC. Because the retina itself does not contract or develop intraretinal proliferation, changes in contour occur because of perpendicular or oblique vitreoretinal traction (funnel, plateau, or ridge-like elevations) or tangential epiretinal membrane traction (star folds, epimacular membranes).

Retinal breaks result in loss of the normal transretinal pressure gradient. Trans-hole flow is related to intraocular pressure, capability of the RPE pump, viscosity of the fluid, and the area and dimensions of the retinal break.

Mechanics of vitreoretinal surgery

An understanding of the physical principles of surgical tools enhances the capabilities of the vitreoretinal surgeon. Discussion of the forces available for cutting and thermal effects follows. Cutting may simply be defined as the separation of a tissue into two parts.

Peeling

Force along the axis of a collagen fiber bundle causes non-elastic collagen fibers to slightly stretch and ultimately to fail. Damage to attached structures is a function of the number of fibers and the strength of the attachment and the substrate. Membrane peeling requires force perpendicular and tangential to the retina, which causes failure of the attachment at the vitreoretinal interface by elongation. Because dense, highly adherent membranes such as those associated with diabetic traction retinal detachments are approximately 100 times stronger than the retina, retina surface defects or retinal breaks usually occur before removal of the membrane occurs (Fig. 101.2). For this reason, membrane peeling is inappropriate in diabetic traction retinal detachment cases.

Fig. 101.2 Forceps designed to place one blade under the epiretinal membrane (ERM) damage the retinal surface. Similarly, pics and membrane scrapers damage the retinal surface.

Shear

Shear cutting occurs when force is applied along two opposing parallel edges moving past each other. Vitreous cutters and scissors use shearing to cut tissue. Inclusive shears such as a vitreous cutter prevent the push-out force that occurs as exclusive shears (scissors) close (Fig. 101.3), pushing the tissue away from the fulcrum, but require elastic deformation of tissue into the port caused by transport pressure gradient (Fig. 101.4).

Fig. 101.3 Scissors create a push-out force; if they are inserted open and then closed they tear the retina at the epiretinal membrane attachment points.

Fig. 101.4 Inclusive shears (vitreous cutters) require elastic deformation of tissue into the port.

Infusion system management

Vitrectomy surgeons have occasionally experienced excessively low intraocular pressure (IOP) during vitreoretinal surgery throughout the history of vitreoretinal surgery. Gravity-fed infusion systems were simplistic and could only cause low intraocular pressure if the bottle was too low or the infusion fluid was depleted. Sutured 20-gauge (G) vitrectomy resulted in low intraoperative IOP if the cannula was initially positioned into the suprachoroidal space without surgeon recognition or displaced intraoperatively thereby causing suprachoroidal infusion. Digital display of infusion driven by an air pressure source in the machine created the false impression that IOP was controlled when in fact as much as 25 mmHg of difference existed between infusion pressure and IOP with typical flow rates during core vitrectomy and fragmenter use. The fragmenter lumen, unlike the vitreous cutter is not obstructed by an inner needle or port opening and closing. Low intraoperative IOP is prevented by using a sufficient infusion pressure to compensate for infusion line and cannula resistance-based losses. Ohm’s law of fluidic resistance; pressure = flow × resistance, explains pressure drop in the infusion system during flow. There are many causes of excessively low intraocular pressure during vitrectomy; each of these will be discussed. Inadvertent suprachoroidal infusion is a relatively common cause of low pressure as well as other more serious complications during vitrectomy with 23G and 25G surgery as well as the 20G sutured systems used for over three decades. Sutureless 25G vitrectomy initially utilized straight-in trocar cannula trajectories to produce sclerotomies perpendicular to the sclera. When 23G, sutureless surgery was introduced subsequently, oblique trocar-cannula entry was utilized in order to construct a scleral tunnel to reduce wound leakage. Initially, surgeons used a two-plane approach; the initial trocar-cannula insertion segment was approximately 30° relative to the sclera and the second segment trajectory perpendicular to the sclera. Some surgeons even believed that a biplanar incision was constructed, although this is not true because the scleral tunnel was created before changing the trajectory. More recently, surgeons using both 23G and 25G systems have switched to oblique entry in order to create a long scleral tunnel^1–3; unfortunately some surgeons use excessively steep angles (5–10°). Although near tangential entry creates a long scleral tunnel; it increases the chances of infusing into the suprachoroidal or subretinal space (Fig. 101.5). If the cannula is in the suprachoroidal space, the peripheral choroid, which is not observed by the surgeon, expands early in the case allowing infusion without hypotony, later the choroid can no longer expand and infusion becomes limited, alerting the surgeon to the problem (Fig. 101.6). Thinking that hypotony caused a choroidal effusion leads to incorrect management. A single plane, 20–30° trajectory is better compromise between the benefits of a long scleral tunnel and the catastrophe of suprachoroidal infusion. Inspecting the infusion cannula with the operating microscope after insertion and before initiating infusion was standard practice with sutured 20G vitrectomy. Many surgeons have discontinued this practice since 23/25G sutureless vitrectomy began; clearly the crucial step of observing the tip of the infusion cannula must not be omitted. It is best practice to insert the infusion port in the cannula with the infusion running to prevent bubbles followed by immediate inspection of the tip of the infusion cannula. The naked eye and endoilluminator provide insufficient magnification to make the determination that the cannula has penetrated the choroid and non-pigmented pars plana epithelium; microscope visualization is essential. Adhesively fastening the infusion cannula tubing and associated stopcock(s) and connectors to the drape is imperative to prevent traction on the infusion cannula and the eye. Unrecognized pulling on the tubing by the assistant or surgeon can easily cause the cannula to partially pull out causing a suprachoroidal infusion. Adhesively fastening the infusion cannula tubing to the drape with eye in the primary position with a short tubing loop can result in a suprachoroidal infusion when the eye is rotated to view the periphery creating tension on the cannula. Scleral depression is another cause of inadvertent suprachoroidal infusion by causing torque on the cannula as the eye is rotated by the depressor. In addition, scleral depression can force blood clots, dense scar tissue, peripheral vitreous, or silicone oil into the infusion cannula and tubing, effectively plugging it, giving the false impression of infusion system failure. Placing the infusion cannula too close to the lower lid rather than just inferior to the horizontal meridian is a common cause of suprachoroidal infusion created when the eye is rotated down to visualize the inferior periphery and the cannula is rotated into the suprachoroidal space. Kinking of the more flexible silicone tubing terminal segment of the infusion cannula can be caused by the surgeon or assistant accidentally pulling on the tubing. This problem is exacerbated by using excessively low infusion pressure settings (10–25 mmHg) insufficient to straighten out the tubing kink. The author has always used 45 mmHg except when operating on children or patients with very low systemic blood pressure, typically under general anesthesia. Some surgeons have recently advocated using infusion settings of 10–20 mmHg because of a completely unfounded belief that occult ischemia is common during vitrectomy. Using infusion settings of 10–20 mmHg causes miosis, bleeding, and corneal astigmatism from contact lens pressure on the cornea and instrument forces on the sclerotomies as well as scleral infolding often mistakenly thought to be choroidals. Kinking as well as multiple bubbles in the infusion line increase resistance to flow and cause pressure drop ultimately resulting in excessively low IOP. Surgeon remediation of intraoperative low IOP should be systematic; the first step is to inspect the cannula with the microscope to make sure it extends all the way through the choroid and non-pigmented ciliary epithelium. If not, a 25G MVR blade can be used to incise the tissue covering the cannula while pressing the cannula into the eye with smooth forceps (Fig. 101.7). Another option is to move the infusion system port to the supranasal cannula. The infusion tubing should be examined for kinking or inadvertent disconnection. Kinking is most common when excessively low infusion pressure settings are used and the tubing bends at the fluid–air stopcock/valve. If a suprachoroidal infusion occurs, infusion through a cannula with a 25G needle (if 25G surgery) into the middle of the eye will compress the choroid against the sclera and cause the suprachoroidal fluid to disappear often via egress around the cannulas (Fig. 101.8). Cut down drainage is never necessary. If choroidals are present at the inception of surgery, a 6 mm instead of 4 mm cannula can be used and/or infusion initiated with a 25G needle as described above.

Fig. 101.5 An excessively steep entry angle creates a long scleral tunnel, but increases the risk of infusing into the suprachoroidal or subretinal space.

Fig. 101.6 Infusion into the suprachoroidal space causes expansion of the peripheral choroid.

Fig. 101.7 If the infusion cannula does not extend into the vitreous cavity, a 25G MVR blade can be used to incise the tissue covering the tip.

Fig. 101.8 Infusion through a 25G needle will compress the choroid and cause egress of the suprachoroidal fluid around the cannula.

Vitreous cutter considerations

All current vitreous cutters utilize suction and inclusive shearing. Ideal tissue cutting is defined as that producing zero displacement of the tissue to be removed and no vitreoretinal traction. It is safest to use the lowest suction force sufficient to imbricate tissue or vitreous into the cutter port. High cutting rates (≥5000 cuts/minute) increase port-based flow limiting and thereby decrease pulsatile fluid flow and pulsatile vitreoretinal traction (Fig. 101.9). Port-based flow limiting via high cutting rates also limits fluid surge after sudden elastic deformation of dense ERM, scar tissue or lens material through the port. Dual actuation eliminates the spring, which facilitates higher cutting rates and enables duty cycle control (Fig. 101.10). In summary, using the highest available cutting rate is the best approach for all tasks and all cases unless all the vitreous has been removed first.

Fig. 101.9 High cutting rates reduce pulsatile vitreoretinal traction.

Fig. 101.10 Dual actuated cutters eliminate the return spring and enable higher cutting rates.

Control systems

Analog parameters using surgical force such as suction levels, power scissors and forceps, and power injector rate are best controlled by proportional depression of the surgeon’s foot (linear control). Switching of valves, pumps, electronic devices, and lasers are ideally controlled by a single integrated system, with functions controlled by the surgeon rather than the circulating nurse or even scrub tech.

Microscope requirements

A stereo operating microscope with magnification up to ×30 with coaxial illumination is required. The microscope should have a beam splitter to enable stereo-optical viewing by the assistant and television viewing for the operating room team. Illumination on–off should be controllable by the surgeon’s foot.

Microscope XY positioning must be controlled via surgeon foot switch. Power zoom and focus positioning are required. Physical stability of the microscope and patient’s head is required to preserve the dimensional stability of the surgical view for microsurgery. Ceiling-mounted microscopes are less mechanically stable than floor-mounted microscopes because of longer moment arms and inherent lack of ceiling rigidity. Wrist rests and surgeon’s chair with armrests should not be used; the surgeon’s hands should be in contact with the patient’s head so hand motion can be coordinated with head motion. The goal is to provide a stable mechanical relationship between the microscope, the patient’s head, the surgeon’s hands, and the OR table resting on the floor.

All power and control sources for surgical tools should be integrated into a single system for greater efficiency. Advanced vitreoretinal surgery systems have a unified human–machine interface combining all surgical functions into an integrated system (Fig. 101.11). Illumination, diathermy, and infusion are referred to as global functions and are always available. Infusion is best controlled by digital, sensor-based, pressurized infusion systems.

Fig. 101.11 Advanced vitreoretinal surgical systems utilize an intuitive human–machine interface and combine all surgical functions into an integrated system.

Tool ergonomics

All surgical tools should be as light as possible and held in the surgeon’s fingertips. They should be contoured rather than cylindrical to reduce the force required to prevent dropping and constrain grip at a consistent position. They should be no longer than the distance from the fingertips to the point of contact with the hand. Shorter handles reduce the torque produced by the weight and reduce friction as the cables, fibers, and tubing used to connect surgical tools slide on the drape. Minimizing forces required to hold tools increase the surgeon’s proprioceptive sense (Weber–Fechner law) and decrease fatigue and tremor. Standardization of tip-to-grip across all tools facilitates so-called muscle memory; the surgeon’s cerebellum, motor strip and frontal lobe trajectory generator knows where the tip is located if view is suddenly lost.

Surgical steps