PEER REVIEWED

Managing Giant Retinal Tears

A video-enhanced guide to surgical repair.

JOHN D. PITCHER, III, MD

Rhegmatogenous retinal detachments (RRDs) related to giant retinal tears (GRTs) present unique challenges to vitreoretinal surgeons. Greater than three clock hours in size, GRTs have a tendency to roll back posteriorly and can extend radially.

Historically, outcomes in these cases were poor. In 1975, Kanski reported a 58% retinal reattachment rate using intentional retinal incarceration.¹ Other strategies included prone air-fluid exchange and retinal tacks or sutures. Since then, several innovations in surgical technique have coincided with improved outcomes (Figure 1).^2-8

Figure 1. Primary anatomic success rates in the surgical repair of GRT-related RRD have improved over time with new techniques and technologies.

A recent case series from The Retina Service of the Wills Eye Hospital reported an 88% single surgery success rate.⁹ To achieve this level, each step of the procedure must be optimized.

The next time a patient walks into your office with a RRD due to GRT, consider this step-by-step guide with some clinical pearls to help increase the chance of a favorable outcome.

TO BUCKLE OR NOT TO BUCKLE?

Conflicting reports exist regarding the benefit of adding an encircling scleral buckle (SB) to vitrectomy for GRT-RRDs. Authors who favor buckling argue that GRTs occur in patients with an abnormal vitreous base, and supporting this region decreases the tractional forces that could result in reopening of the break or extension into uninvolved retina.^10,11 This mechanism may be especially important in phakic patients, in whom the crystalline lens hinders thorough vitrectomy.

In contrast, some argue that GRTs, especially if extending greater than 180º, already have significant release of traction, similar to a retinectomy, so a SB may not be helpful.

In cases with pre-existing proliferative vitreoretinopathy, a buckle should be strongly considered. Most will agree that, when initially in doubt, placement of a SB is seldom regretted.

If adjuvant scleral buckling is performed, a low lying buckle is preferred (41 or 42 band). A high buckle could increase the chance of retinal slippage or radial folds.

This is also true if larger elements are utilized. In all cases, the goal should be to position the posterior edge of the GRT to fall on the plateau of the buckle.

VITREOUS GEL REMOVAL

Small-gauge vitrectomy is now the preferred method for GRT-RRD repair. Both 23- and 25-gauge platforms have sufficient stiffness of instrumentation to complete the necessary tasks. Valved cannula systems decrease turbulent flow and lessen vitreous incarceration, which are especially important in complex cases, in which insertion and removal of instruments are more frequent.

High speed cutting with low vacuum levels allows for safer shaving of peripheral vitreous over detached retina. Confirm a posterior vitreous detachment intraoperatively.

In cases associated with trauma or genetic syndromes (Stickler, Marfan, etc.), a PVD is less commonly present. A thorough dissection of the vitreous base is advised, with the aid of scleral depression if available.

Lensectomy is not necessary for the vast majority of cases, unless lenticular opacity is an impediment.⁸ As in all cases, proper lighting is critical for gel visualization. Give extra attention to the horns of the GRT, the most common site where the break may reopen.

The anterior flap of the tear is devitalized (ischemic) retina that can serve as a scaffold for anterior-loop PVR — it should be preferentially trimmed, when possible (Video 1).

MARKING THE BREAKS

Endodiathermy along the posterior edge of the GRT can ensure hemostasis in areas of amputated retinal blood vessels, which the surgeon may encounter in subsequent steps. The lateral extent of the tear is helpful to mark, because the surgeon’s view may become more limited later in the case.

Cautery should be gentle enough to avoid retinal contraction but generous enough to aid in visualization of the edges of the tear (Video 2). Of course, other retinal defects outside of the GRT should be marked for similar purposes.

UNFOLDING THE RETINA

The widespread integration of perfluorocarbon liquid (PFCL) in the 1990s revolutionized the repair of GRT-associated RRDs. Gradually inject PFCL over the optic disc to ensure that a single enlarging bubble expands past the peripheral edges of the GRT.

Tilting the eye away from the break can help displace all subretinal fluid. A soft (silicone) tipped cannula can facilitate unrolling any residual areas of retinal folding (Video 3).

LASER

Adequate retinopexy is critical in GRT-RRD repair. Endolaser should be applied in at least two to three rows along the posterior edge of the tear and continued anteriorly to the ora serrata.

Moderate intensity burns are preferred, with a near-confluent pattern (approximately one-half burn-width apart). More generous barricade should be placed at the horns or radial extensions of the GRT, where primary failure most often occurs (Video 4).

Other retinal breaks should be treated appropriately, but no evidence exists to support the addition of 360º endolaser in the absence of lattice degeneration or other pathology.⁸

PREVENTING POSTERIOR RETINAL SLIPPAGE

The fluid-air exchange is arguably the most critical step in the surgical repair of GRT-associated RRDs. Improper technique can result in retinal slippage or folds. A meticulous, slow aqueous removal with an extrusion cannula at the air-PFCL interface and at the edge of the tear is essential.

Rotating the eye to place the GRT in the most dependent position, maintaining suction, and draining for a longer duration than in typical RRDs will ensure that residual subretinal fluid is nonexistent (Video 5).

GAS OR OIL?

Many factors contribute to the decision of choice of tamponade agent. The superotemporal quadrant is the most common location for spontaneous GRT formation. Longer acting gas (C₃F₈) is appropriate in these cases, if uncomplicated.

The presence of PVR, primarily inferior GRT, or suspected poor compliance with positioning generally necessitate the use of silicone oil.

Avoid the use of short-acting gas (SF₆) because evidence has suggested higher rates of redetachment.⁸ Seven days of postoperative head positioning is recommended for most cases.

WHAT ABOUT THE OTHER EYE?

Patients who have had a spontaneous GRT are at risk of developing a GRT in their fellow eye. Because of the substantial potential visual morbidity associated with a GRT-RRD, many have debated the role of laser and even surgical prophlyaxis.^12,13

Most advocate for retinopexy of any peripheral retinal breaks or lattice degeneration in the fellow eye. Less evidence exists for preventative treatment in the absence of pathology. Patients with a syndromic predisposition are at especially high risk, and these cases deserve close monitoring.

CASE STUDIES

A Previously Lasered GRT

A 45-year-old phakic myopic man presented to the Retina Service of the Wills Eye Hospital with a macula-on RRD associated with a five clock hour superotemporal GRT. One week earlier he underwent laser retinopexy by an outside retina specialist to treat the GRT.

Subretinal fluid progressed through the barricade, and surgical intervention was necessary. The pathology was nearly completely contained by the previous treatment, so I did not include a scleral buckle in the surgical strategy.

I performed 23-gauge pars plana vitrectomy (Figure 2A, page 19), using scleral depression to remove inferiorly settled vitreous hemorrhage and ensure the vitreous base was thoroughly dissected (Figure 2B). I used the vitrector to relieve all vitreous traction from the break and amputated the anterior edge of the GRT (Figure 2C).

Figure 2. While performing pars plana vitrectomy (A), scleral depression (B) helped to ensure the vitreous base was thoroughly dissected. After relief of vitreous traction from the break and amputation of the anterior edge of the GRT (C), endodiathermy was used to cauterize bleeding vessels at the edge of the GRT (D). During fluid-air exchange, the cannula was used to pull the retina peripherally (E). A laser probe was used to complete retinopexy (F) and C₃F₈ gas was perfused (G).

I used endodiathermy to cauterize bleeding retinal blood vessels at the edge of the tear (Figure 2D). This also aided in visualization and identification of the extent of the GRT later in the case.

During the fluid-air exchange, I used a silicone soft-tipped extrusion cannula to gently engage the retina and pull it peripherally (Figure 2E). This maneuver was possible due to the localized nature of the GRT, and it facilitated retinal flattening under the air tamponade, obviating the need for PFCL.

I used a 23-gauge illuminated laser probe to complete three rows of barricade retinopexy surrounding the previous treatment zone (Figure 2F). At the end of the case, I infused C₃F₈ gas to a physiologic intraocular pressure (Figure 2G).

I instructed the patient to avoid supine positioning for five days. Three months later, his retina remained attached, and his Snellen visual acuity was 20/20.

A Nearly 360º GRT

A 50-year-old pseudophakic man presented to the Retina Service of the Wills Eye Hospital with a macula-off RRD due to an 11+ clock hour GRT (Figure 3A).

Figure 3. A patient with an 11+ clock hour GRT (A) underwent vitrectomy. The vitreous base was trimmed and the anterior flap of the GRT amputated (B). Care was taken to avoid cutting the intact retina (C). PFCL was injected to facilitate unfolding of the retina (D). Endolaser was performed along the posterior edge of the GRT (E). After ensuring no subretinal fluid was present (F), C₃F₈ tamponade was performed (G).

I performed 23–gauge PPV without a scleral buckle. I completed a meticulous trim of the vitreous base and amputated the anterior flap of the GRT (Figure 3B).

Only a small amount of peripheral retina in the superonasal quadrant was unaffected by the GRT. Although several small breaks were present in this region, I was careful to avoid cutting the intact retina (Figure 3C). Connecting the horns would have resulted in a 360º GRT, increasing the risk of torsional translocation.

I slowly injected PFCL with the globe rotated toward the superonasal quadrant. This facilitated unfolding of the retina (Figure 3D). I then placed three rows of barricade endolaser 360º along the posterior edge of the GRT (Figure 3E). I also used an extrusion cannula to perform a fluid-air exchange.

I removed the fluid layer first, leaving only air and PFCL. I spent additional time at the edge of the GRT in each quadrant to ensure that no subretinal fluid was present (Figure 3F) and used C₃F₈ gas for long-acting tamponade (Figure 3G). I asked the patient to maintain prone positioning for five days after surgery.

Three months later, the patient’s retina was flat, and his Snellen visual acuity was 20/60.

CONCLUSION

When RRDs are associated with GRTs, they pose a difficult challenge. We now have the tools to achieve improved outcomes, even in such difficult cases.

Focusing on the optimization of each step of the procedure will ensure that your results are as favorable as possible. RP

REFERENCES