New treatments for old retinal problems

In the last several years, a number of innovative treatments have been proposed for some common retinal diseases. While these therapies remain controversial and largely unproven to date, we felt it would be worthwhile to review what is known about three particularly exciting new procedures that may have utility for select patients: intravitreal tPA and gas injection for the treatment of submacular hemorrhage, YAG laser "hyaloidotomy" for the treatment of subhyaloid hemorrhage, and laser-induced chorioretinal venous anastomosis for central retinal vein occlusion. In this edition of Focus on Retina, we’ll try to distill the relevant literature on these procedures and summarize our own personal experience and approach.

Intravitreal gas and tPA for submacular hemorrhages

Submacular hemorrhages usually occur as a complication of choroidal neovascularization (CNV) and may result in severe loss of central vision. Exudative age-related macular degeneration (AMD) is the most common underlying etiology although CNV from any cause may lead to submacular hemorrhage (ocular histoplasmosis, myopia, angioid streaks, idiopathic CNV, etc). The visual prognosis for patients with submacular hemorrhages is relatively poor and is related to the size and thickness of the subfoveal hemorrhage. Until recently, the only treatment available was pars plana vitrectomy surgery to surgically evacuate these thick hemorrhages. Unfortunately, even with surgery, the visual results were usually disappointing.

A few years ago, an Australian retina specialist, Wilson Heriot, M.D., reported a pilot series of submacular hemorrhage patients whom he treated with a novel approach: intravitreal tissue plasminogen activator (tPA) injection, intravitreal gas injection, and face-down positioning. Without going to the operating room, he successfully displaced the majority of the submacular blood into the more peripheral subretinal space in every patient resulting in improved visual acuity.

The idea is simple: the tPA dissolves the blood clot making it relatively liquid, then the gas bubble "steamrolls" the blood out of the macula. The procedure can be performed in an office setting, requires only a few minutes, is less expensive than vitrectomy surgery, and appears likely to have fewer complications and better visual results than vitrectomy.

While most retina specialists believe that this procedure will prove valuable for select patients with submacular hemorrhages, some controversies remain. One crucial question is whether tPA is essential to the procedure.

Naturally occurring tPA is produced by vascular endothelial cells. It converts plasminogen to plasmin which then breaks down fibrin clots. tPA in clinical use is made through recombinant DNA techniques, must be stored at –70 degrees, and is expensive. It has also been found to have retinal toxicity in animals in doses as low as 50 micrograms. If the use of tPA were proven unnecessary, the procedure would become safer, less expensive, and could be performed more easily in an office setting.

There is significant controversy surrounding the question of whether tPA is even capable of crossing the retina when injected into the vitreous cavity. Furthermore, Ohji and coworkers recently reported a series of five submacular hemorrhage patients treated with gas alone (Arch Ophth Oct 1998). Blood was significantly displaced in 3 of the 5 eyes.

The age of the hemorrhage is also crucial; older hemorrhages appear less amenable to displacement with gas. Ohji was unable to displace an 18 day old clot, but was successful with a clot 11 days old.

Other unresolved questions include the following: what size hemorrhages respond best?, how long is face down position required?, for older hemorrhages that fail to respond to gas alone, would subsequent tPA be beneficial?

Finally, the procedure is not free of complications. Bleeding, infection, and elevated intraocular pressure are all potential problems but are rarely serious. Vitreous hemorrhage may occur but can usually be safely removed with vitrectomy surgery if it fails to spontaneously resolve. Retinal detachment has also been reported and may be complicated by proliferative vitreoretinopathy. Detachment presumably occurs secondary to the gas causing vitreous traction.

We at The Retina Center consider intravitreal gas injection (without tPA) for subfoveal hemorrhages which are less than three weeks old in patients who are able to maintain face down position. We follow these patients daily until no further displacement of blood is observed. We then discontinue prone positioning and perform angiography to look for a treatable source of the bleeding. If sufficient blood displacement is not achieved, vitrectomy surgery is considered.

Laser-induced chorio-retinal anastomosis for central retinal vein occlusion

Central retinal vein occlusion is a relatively common retinal vascular disorder which may result in severe visual loss. The underlying cause is thrombus formation in the central retinal vein which obstructs the outflow of venous blood from the eye. This leads to a pressure build-up in the retinal vascular system which causes small capillaries to break and leak fluid. As a result, widespread retinal hemorrhages and macular edema (swelling of the central retina) develop. If the CRVO is severe (I.e. significant retinal ischemia is present), neovascularization of the iris may develop which may ultimately cause neovascular glaucoma, and blindness. Currently, there is no proven treatment for the visual loss associated with central retinal vein occlusion.

Recent investigation has focused on attempting to bypass the occluded central retina vein by creating a chorioretinal venous anastomosis with a laser. A high power laser spot is applied on or next to a retinal vein in an attempt to create a hole both in the vein and in the underlying RPE/Bruch’s membrane complex. It is hoped that during healing, vascular ingrowth will occur from both the retinal and choroidal circulation creating a vascular anastomosis. This would then allow an alternate route for venous blood to enter the choroidal circulation, thus decompressing the retinal venous system. The optimal laser power, wavelength, and exact technique remain undetermined.

While creating such an anastomosis is theoretically desirable, it is difficult to achieve. McAllister et al. successfully created anastomoses in 8 of 24 eyes (33%) in their initial series (Arch Ophth Apr 1995), while Fekrat and coworkers succeeded in 38% of eyes (Arch Ophth Jan 1998). By making multiple anastomotic attempts in a given eye (up to 8 sites), anastomosis formation may be successful in as many as 54% of eyes.

When an anastomosis is successful, resolution of the funduscopic appearance of vein occlusion may occur, often accompanied by dramatic visual improvement. In their most recent series, McAllister observed visual improvement in 84% of eyes with successful anastomoses (AJO Aug 1998). The average visual improvement was 4.3 lines (range 2 to 20 lines). The remaining 16% of eyes either worsened or remained unchanged.

Neovascular complications are the most common concern when considering laser-induced chorioretinal venous anastomosis. Approximately 20% of treated eyes will develop intravitreal, intraretinal, or subretinal neovascular membranes. The greater the degree of retinal ischemia, the greater the risk of this complication. This has made most clinicians very reluctant to recommend this procedure to patients with ischemic central retinal vein occlusions. Scatter laser treatment to the area of retinal ischemia may reduce this complication.

At The Retina Center, we consider laser-induced chorioretinal venous anastomosis for select patients with central retinal vein occlusion. We do not usually recommend the procedure for patients with significant ischemia for fear of neovascular complications. We also do not offer the procedure to patients whose visual potential is poor such as individuals with a long-standing vein occlusion (greater than three years), or significant preexisting macular pathology. Finally, we observe patients with recent CRVOs for several months to be certain that spontaneous improvement does not occur. An ideal candidate would thus have a non-ischemic CRVO with acuity of 20/100 or less due to perfused macular edema who worsens or fails to improve after 4 months of follow-up.

YAG laser hyaloidotomy for premacular hemorrhage

Subhyaloid hemorrhage is an uncommon cause of visual loss in patients with retinal neovascularization such as diabetic retinopathy and branch retinal vein occlusion. Subhyaloid hemorrhages may also occur in the absence of neovascularization in Terson’s syndrome, Valsalva retinopathy, and retinal arterial macroaneurysms. The hemorrhage may be loculated in the space beneath the internal limiting membrane or in the space between the ILM and hyaloid, although it is usually impossible to make this anatomic distinction clinically. Subhyaloid hemorrhages may clear spontaneously without permanent visual loss over weeks to months. Alternatively, macular pigment alterations or epiretinal membranes may develop with resultant loss of acuity. Until recently, the only available options to manage subhyaloid hemorrhage were observation or vitrectomy surgery. YAG laser hyaloidotomy now offers a third therapeutic option for these patients.

Several case series have demonstrated the feasibility of using a Nd:YAG laser to puncture the posterior hyaloid face and/or internal limiting membrane and allow drainage of subhyaloid hemorrhage into the vitreous cavity. This may allow more rapid restoration of visual acuity and reduce the visual loss associated with long-standing subhyaloid hemorrhage without the need for vitrectomy surgery.

The largest case series investigating this procedure was published in the November 1998 issue of Archives of Ophthalmology. Ulbig and coworkers retrospectively reviewed 21 eyes treated with the YAG laser for subhyaloid hemorrhages of varying etiologies. In 16 treated eyes, visual acuity improved within one month. One clotted hemorrhage could not be successfully drained. Four eyes developed dense, non-clearing vitreous hemorrhages that ultimately required vitrectomy. Complications included a macular hole in one patient and a retinal detachment in another. Not surprisingly, eyes without pre-existing macular pathology had the best visual recovery (Valsalva retinopathy, Terson’s syndrome, blood dyscrasia).

As always, the procedure does have drawbacks. First, old hemorrhages may not drain. Ulbig et al successfully drained a 21 day old clot, but could not drain a 35 day old hemorrhage. Second, patients with small subhyaloid hemorrhages may not be good candidates. The authors speculated that in small hemorrhages, the photodisruptive effect of the laser may be too close to the macula and may have caused the macular hole observed. Third, one third of eyes in this study still required vitrectomy surgery usually because of non-clearing vitreous hemorrhage. Finally, in eyes with neovascularization, it is important to apply panretinal photocoagulation prior to performing YAG laser hyaloidotomy since a vitreous hemorrhage may preclude PRP application afterwards.

At The Retina Center, we consider YAG laser hyaloidotomy for patients with premacular subhyaloid hemorrhages larger than 3 disc diameters in size and less than one month old.

The Retina Center is the vitreoretinal subspecialty practice of Dr. Edward J. Goldman, a board-certified ophthalmologist with fellowship training in the management of vitreoretinal disease. The Retina Center provides state of the art retinal care in either of two convenient locations. Both offices are staffed with experienced photographers, and are equipped with laboratory facilities to develop fluorescein angiograms for immediate evaluation when needed. Other services available in-office include laser treatment, ocular echography, cryotherapy, and pneumatic retinopexy.

Retinal emergencies are common and we are happy to see patients on very short notice when indicated. Also, please feel free to call anytime for consultations over the phone.