Proliferative diabetic retinopathy

Published on 08/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2153 times

CHAPTER 61 Proliferative diabetic retinopathy

Management of proliferative diabetic retinopathy is guided by the findings and recommendations of the Diabetic Retinopathy Study (DRS)¹^–³, Early Treatment Diabetic Retinopathy Study (ETDRS)⁴, Diabetic Retinopathy Vitrectomy Study (DRVS)^5,⁶ and the Diabetes Control and Complication Trial (DCCT)⁷. The combination of successful results achieved with the DRS, ETDRS, DRVS, and DCCT have clearly established the indications, techniques, and benefits of photocoagulation and pars plana vitreous surgery, and intensive control of blood sugars for people with diabetes. Advances in anti-angiogenic therapies have the potential to influence the way we care for patients with proliferative diabetic retinopathy in the future.

Clinical features

The severity of diabetic retinopathy is strongly related to the duration of diabetes: the prevalence of proliferative diabetic retinopathy ranges from 0% in individuals with fewer than 5 years of diabetes to as high as 50% in individuals with 20 or more years of diabetes ⁸. In addition to duration of diabetes, other factors are influential in the development and progression of proliferative diabetic retinopathy. These include puberty, pregnancy, poor metabolic control, hypertension, hypercholesterolemia, renal disease, and intraocular surgery. Recent reports of lower rates of proliferative diabetic retinopathy may reflect improvements in primary care for diabetes and these other systemic associations^9,¹⁰.

Proliferative diabetic retinopathy is characterized by the presence of newly formed blood vessels arising from the retina or optic disc. The risk of proliferative diabetic retinopathy is related to the extent of ischemia within the retina ^11,¹². As the retina becomes progressively more ischemic, a cascade of events occurs that ultimately results in the proliferation of abnormal blood vessels; vascular endothelial growth factor (VEGF) plays a significant role in this process¹³. Clinically, the most important findings in predicting progression to proliferative diabetic retinopathy include intraretinal hemorrhages and microaneurysms, venous beading, and intraretinal microvascular abnormalities (IRMA)¹⁴. The ‘four-two-one rule’ is helpful in assessing the severity of diabetic retinopathy and predicting the progression to proliferative retinopathy. Severe non-proliferative diabetic retinopathy is defined as the presence of either four quadrants of dot/blot hemorrhages or microaneurysms, two quadrants of venous beading, or one quadrant of IRMA. In eyes with severe non-proliferative diabetic retinopathy, approximately 50% will progress to proliferative diabetic retinopathy within 1 to 3 years.

Most eyes with neovascularization will have neovascularization on the disc itself or within one disc diameter of the disc (NVD). Neovascularization of the disc appears as fine lacy vessels lying on the surface of the disc or bridging the physiologic cup (Fig. 61.1). As the NVD progresses, the vessels become more prominent and extend into the vitreous cavity or along the vascular arcades; the new vessels may be associated with significant fibrous proliferation (Fig. 61.2). Neovascularization elsewhere (NVE) is defined as new vessels located at retinal sites other than on or within one disc diameter of the optic disc. Neovascularization elsewhere is usually associated with the retinal veins and arteriovenous crossing sites between areas of perfused and non-perfused retina (Fig. 61.3).

Fig. 61.1 Neovascularization of the disc appears as fine lacy vessels lying on the surface of the disc or bridging the physiologic cup.

Fig. 61.2 As the neovascularization of the disc progresses, the vessels become more prominent and extend into the vitreous cavity or along the vascular arcades; the new vessels may be associated with significant fibrous proliferation.

Fig. 61.3 Neovascularization elsewhere is usually associated with the retinal veins and arteriovenous crossing sites between areas of perfused and non-perfused retina. This fluorescein angiogram demonstrates widespread retinal ischemia and areas of intense hyperfluorescence consistent with neovascularization elsewhere.

The vitreous plays a critical role in the development and progression of proliferative diabetic retinopathy ¹⁵. It appears that the posterior vitreous surface provides a ‘scaffold’ for the progression of neovascularization. As the posterior vitreous detachment progresses, there is increased traction on the fibrovascular tissue. This may result in recurrent preretinal/vitreous hemorrhage or traction retinal detachment, or both.

Iris neovascularization or rubeosis may develop during the course of proliferative diabetic retinopathy. Iris neovascularization is characterized by the proliferation of fine lacy vessels on the iris surface; progressive iris neovascularization may result in angle closure and neovascular glaucoma.

Major diabetic retinopathy studies

The DRS and ETDRS provide clear guidelines for the management of proliferative diabetic retinopathy. Definitions used by the DRS and ETDRS are listed in Table 61.1.

Table 61.1 Diabetic Retinopathy Study (DRS) and Early Treatment Diabetic Retinopathy Study (ETDRS) definitions

Moderate visual loss	Loss of 15 or more letters on the ETDRS reading chart (equivalent to a doubling of the initial visual angle, e.g., 20/20 to 20/40 or 20/50 to 20/100)
Severe visual loss	Visual acuity less than 5/200 on two consecutive follow-up visits scheduled at 4-month intervals
Severe non-proliferative retinopathy	Any one of the following:
	Hemorrhages and/or microaneurysms in four quadrants
	Venous beading in two quadrants
	Intraretinal microvascular abnormalities (IRMA) in one quadrant
Early proliferative retinopathy	Proliferative retinopathy without DRS high risk characteristics
Less severe retinopathy	Mild or moderate non-proliferative retinopathy
More severe retinopathy	Severe non-proliferative or early proliferative retinopathy
High risk proliferative diabetic retinopathy	At least three of the following risk factors:
	New vessels present
	New vessels located on or within 1 disc diameter of the disc (NVD)
	Vitreous and/or preretinal hemorrhage
	Moderate to severe new vessels
	NVD greater than standard photo 10a ( disc area) (Fig. 61.2)
	In absence of NVD, neovascularization elsewhere (NVE) greater than disc area

Diabetic retinopathy study

Does photocoagulation surgery reduce the risk of severe visual loss in diabetic retinopathy? The DRS evaluated individuals with proliferative diabetic retinopathy in at least one eye or severe non-proliferative diabetic retinopathy in both eyes ^1,². One eye was randomly selected for photocoagulation, while the other eye was observed. Extensive scatter panretinal photocoagulation (PRP) and focal laser of new vessels on the surface of the retina was performed using either argon laser or xenon arc. Typical argon laser parameters included: 800–1600 burns, 500 μm spot size, 0.1 second duration, and intensity to obtain definite whitening.

The DRS identified four retinopathy risk factors for severe visual loss in diabetic retinopathy: new vessels present, new vessels located on or within one disc diameter of the disc, vitreous or preretinal hemorrhage, and moderate to severe new vessels (NVD greater than one-third disc area or, in the absence of NVD, NVE greater than one half the disc area)³. The risk of severe visual loss increased progressively as more risk factors were added: the risk of severe visual loss remained relatively low with two or fewer risk factors (roughly 5% at 2 years) but increased considerably with three or more risk factors (roughly 25% at 2 years). Therefore, the presence of three or four of these retinopathy risk factors indicates ‘high risk’ for severe visual loss.

Extensive PRP reduced the risk of severe visual loss by greater than 50% regardless of the baseline severity of retinopathy (Fig. 61.4). After 2 years, severe visual loss occurred in 16.3% of all untreated eyes compared to only 5.3% of eyes that received xenon arc and 7.4% of eyes treated with argon laser.

Fig. 61.4 Severe visual loss. Cumulative rates of severe visual loss, including and excluding observations made after the 1976 Diabetic Retinopathy Study protocol change, argon and xenon groups combined. Extensive scatter photocoagulation reduces the risk of severe visual loss by 50% or more.

From The Diabetic Retinopathy Study Research Group: Photocoagulation treatment of proliferative diabetic retinopathy. Clinical application of Diabetic Retinopathy Study [DRS] findings. DRS Report Number 8. Ophthalmology 1981;88:583–600.

Although xenon arc was slightly more effective than argon laser in reducing the risk of severe visual loss, this advantage was outweighed by the greater adverse effects of xenon arc. Visual field constriction attributable to PRP occurred in 10% of argon-treated eyes compared with 35% of xenon arc-treated eyes. Persistent decreases of one or more lines of vision occurred in approximately 10% of argon-treated eyes compared with 20% of xenon arc-treated eyes.

The DRS findings strongly indicate that prompt panretinal laser PRP should be performed in eyes with high risk proliferative diabetic retinopathy. In these eyes, the risk of severe visual loss is high. Laser surgery effectively reduces the risk of severe visual loss by greater than 50%. In high risk proliferative retinopathy, the beneficial effects of PRP clearly outweigh the adverse effects. In eyes with severe non-proliferative or proliferative diabetic retinopathy without high risk characteristics, the DRS does not provide a clear choice between prompt treatment and careful follow-up with deferral of PRP until high risk characteristics develop. The risk of severe visual loss in these eyes is relatively low. Although laser surgery reduces the risk of severe visual loss, the beneficial effects of PRP must be weighed against the adverse effects.

Early treatment diabetic retinopathy study

When, during the course of diabetic retinopathy, is it most appropriate to initiate photocoagulation therapy? The ETDRS enrolled patients with mild to severe non-proliferative diabetic retinopathy or early proliferative retinopathy without high risk characteristics in both eyes ⁴. One eye of each patient was assigned randomly to early full or mild PRP and the other to deferral. For eyes assigned to deferral, full PRP was performed immediately if high risk proliferative diabetic retinopathy developed. Eyes with macular edema were also randomized to immediate or delayed focal laser surgery.

The ETDRS provided very important information on the natural history of diabetic retinopathy. For eyes with mild or moderate non-proliferative retinopathy, the risk of progression to high risk proliferative diabetic retinopathy was low: approximately 10% over 3 years. For eyes with more severe retinopathy, the risk of progression to high risk proliferative retinopathy increased to about 50% over 3 years. Early PRP effectively reduced the rate of developing high risk proliferative retinopathy by over 50% regardless of baseline severity of retinopathy.

The development of severe visual loss for all baseline categories of retinopathy was low for both the treatment and the deferral groups (2.6% vs. 3.7%, respectively).

Adverse effects of PRP on visual acuity and visual field were observed. Eyes assigned to immediate full PRP had significantly greater loss of visual field as measured by the Goldmann 1/4e test object than those eyes assigned to deferral. Treated eyes were more likely to have early moderate visual loss than untreated eyes, particularly when early PRP was applied in eyes with pre-existing macular edema. In eyes with macular edema, immediate focal photocoagulation with delayed PRP – added only if more severe retinopathy developed – was the most effective strategy for reducing the risk of moderate visual loss.

Although early PRP reduces the risk of developing high-risk proliferative retinopathy by more than 50%, the ETDRS demonstrated that the risk of severe visual loss for all eyes studied is low, whether treated with early PRP or followed closely and treated as soon as high risk proliferative retinopathy developed. Early full PRP is associated with higher rates of moderate visual loss and visual field loss than deferral. Provided careful follow-up can be maintained, PRP is not recommended for eyes with less severe retinopathy. As retinopathy approaches the high risk stage, the benefits and risks of early PRP may be roughly balanced. Initiating PRP early in at least one eye seems particularly appropriate when both of a patient’s eyes are approaching the high risk stage. Once high risk proliferative diabetic retinopathy develops, PRP should not be delayed. Focal treatment should be considered for eyes with clinically significant macular edema, preferably before PRP for high risk proliferative retinopathy becomes urgent.

Performing panretinal photocoagulation

Eyes with high risk proliferative diabetic retinopathy are at significant risk for severe visual loss and should undergo PRP without delay. Recall that high risk proliferative diabetic retinopathy is defined as the presence of three or four of the following risk factors: new vessels present, new vessels on the disc, preretinal or vitreous hemorrhage, and moderate to severe new vessels (new vessels on the disc greater than one third disc area or new vessels elsewhere greater than one half disc area)³. The presence of three or four high risk characteristics translates into essentially three clinical situations. These include moderate to severe neovascularization on the disc in the presence or absence of preretinal or vitreous hemorrhage, less severe neovascularization on the disc associated with preretinal or vitreous hemorrhage, or moderate to severe neovascularization elsewhere associated with preretinal or vitreous hemorrhage. Iris neovascularization is another indication for prompt PRP, whether or not high risk characteristics are present.

In eyes with less severe proliferative diabetic retinopathy or severe non-proliferative diabetic retinopathy, the treating ophthalmologist must weigh the benefits of early PRP versus the potential risks.

Panretinal photocoagulation is tolerated well by many individuals with topical anesthesia alone. However, retrobulbar, peribulbar, subTenon’s, or subconjunctival anesthesia may be necessary.

Panretinal photocoagulation may be performed in one or multiple sessions. Eyes undergoing treatment in one session are more likely to have transient adverse effects than eyes treated in multiple sessions ¹⁶. These adverse effects include exudative retinal detachment, choroidal detachment, and angle closure. However, the long-term beneficial effects of laser are equal whether laser is administered in one or multiple sessions. In eyes with macular edema, extensive fibrovascular proliferations, or traction retinal detachments, it is advisable to perform PRP in two or more sessions separated by 2 to 3 weeks.

The type or wavelength of laser may vary. Studies have demonstrated similar efficacies of argon, krypton, diode, and dye lasers ¹⁷^–²⁰. In cases of dense cataract or vitreous hemorrhage, the red wavelength may be more useful.

Several wide-field funduscopic lenses are available for PRP. These lenses allow visualization to the ora serrata and give an inverted, reversed view of the retina. Most of the available lenses magnify the laser spot size 1.5–2 times at the retina.

With the retina in focus, the laser power is increased until whitening of the retina is achieved. In general, surgeons may start with the following parameters: power: 200 mW; duration: 0.1 seconds; with a spot size of 500 μm at the retina. If it is not possible to obtain a satisfactory burn with a power setting of 1 watt or less, a smaller spot size or longer exposure time may be used.

Laser burns are spaced one-half to one spot-width apart (Fig. 61.5). Peripheral PRP is advocated to reduce the risk of visual loss, exacerbation or development of macular edema, and extensive visual field loss²¹. The posterior fundus is demarcated with two to three rows of laser to reduce the risk of extension into the macula or optic disc. The burns are placed one disc diameter from the optic disc and the major temporal vascular arcades. Special care must be used when demarcating the posterior extent of treatment of the temporal retina. Generally, laser should not extend posterior to an imaginary line drawn at two fovea-disc diameters from the optic disc. Panretinal photocoagulation is then extended peripherally to the equator and beyond.

Fig. 61.5 Panretinal photocoagulation. 500 μm spots using enough energy to obtain slight whitening of the RPE are placed one-half to one burn width apart to the retina.

Approximately 1600 burns of 500 μm or larger are common for eyes with high risk proliferative diabetic retinopathy. However, the extent of PRP – not necessarily the number of burns placed – is important in assessing the completeness of surgery.

A follow-up visit is scheduled in 2 or 3 weeks for additional laser, if divided sessions are used, or in 1 month following completion of PRP. In eyes that fail to regress despite standard PRP, supplemental laser may be helpful.

Regression of proliferative diabetic retinopathy

The efficacy of PRP is related to the induced regression of retinopathy risk factors (Fig. 61.6)^22,²³. Regression of high risk characteristics may take several forms: complete regression of new vessels, marked reduction in the severity of new vessels, or resolution of preretinal and vitreous hemorrhage.

Fig. 61.6 Regression of proliferative diabetic retinopathy. Panretinal photocoagulation induces a rapid and stable reduction of retinopathy risk factors in most eyes. (A) High-risk proliferative diabetic retinopathy; (B) one month following PRP. Notice the reduction in venous dilation and tortuosity following successful PRP.

A subset of eyes fail to show regression of retinopathy risk factors despite standard PRP; failure to demonstrate regression of high risk characteristics may be associated with a poorer prognosis ²²^–²⁵. Supplemental laser may induce regression of high risk retinopathy in eyes that fail to respond to initial therapy²⁴^–²⁹. The DRS and the ETDRS did not evaluate the efficacy of supplemental PRP in eyes that fail to respond to standard therapy. However, the DRS showed that the risk of severe visual loss decreased with an increasing amount of initial treatment as measured by treatment density in the post-treatment photographs³⁰.

In addition to regression of retinopathy risk factors, manifest changes in the retinal vascular system are associated with a favorable response to laser treatment. There is a significant increase in the venous caliber as the stage of diabetic retinopathy progresses ³¹. Panretinal photocoagulation has been shown to reduce retinal blood flow and retinal vessel diameter³²^–³⁵. The diameter of the retinal vessels is correlated to the amount of regression of proliferative diabetic retinopathy^36,³⁷.

How does panretinal photocoagulation work?

Photocoagulation destroys the photoreceptor–retinal pigment epithelial (RPE) complex and produces significant thinning of the outer retina ³⁸. By decreasing the oxygen consumption at the photoreceptor–RPE complex, more oxygen is available to diffuse into the inner retina and vitreous³⁸^–⁴¹. Enhanced oxygen diffusion into the inner retina and vitreous reduces inner retina ischemia and the stimulus for neovascularization.

VEGF is an angiogenic molecule whose expression is markedly induced by retinal hypoxia ⁴². Vitreous VEGF levels are significantly higher in eyes with proliferative diabetic retinopathy than in eyes without proliferative diabetic retinopathy⁴³. VEGF concentrations in vitreous fluid decline significantly after successful PRP, suggesting that PRP reduces retinal ischemia and the hypoxia-induced expression of VEGF⁴⁴.

In addition, photocoagulation may inhibit neovascularization by inducing changes in the RPE. Retinal pigment epithelial cells have the ability to inhibit vascular endothelial cell proliferation and induce regression of new blood vessels ⁴⁵. Photocoagulation of RPE cells may induce changes in the production of growth factors. In particular, it appears that photocoagulated RPE cells may release potential inhibitors of neovascularization⁴⁶.

Complications of panretinal photocoagulation

Laser photocoagulation effectively reduces the risk of severe visual loss in eyes with high risk proliferative diabetic retinopathy. However, adverse effects may occur (Box 61.1).

Subjective changes in vision after PRP include blurred vision, decreased accommodation, difficulty adjusting to changing levels of illumination, color vision abnormalities, constriction of peripheral visual fields, and poor night vision ⁴⁷. Adverse effects to the cornea include corneal burns, erosions, and superficial punctate keratitis^48,⁴⁹.

Panretinal photocoagulation may affect intraocular pressure. Blondeau and associates noted transient intraocular pressure elevations soon after treatment, which resolved within several hours ⁵⁰. Conversely, transient decreases in intraocular pressure may occur^51,⁵². Elevation of intraocular pressure may be associated with anterior chamber shallowing and frank angle closure glaucoma.

Iris and pupillary abnormalities include pupillary border hyperpigmentation or atrophy, posterior synechiae, reduced pupillary reactivity, mydriasis, miosis, accommodative palsy, light-near dissociation, and sector palsies of the iris sphincter ⁵³.

Stable, non-progressive lens opacities may occur after laser treatment ^54,⁵⁵.

Posterior segment complications of laser surgery include inadvertent macular/foveal burns, exacerbation or development of macular edema, exudative macular detachment, ciliochoroidal effusion, hemorrhage, rupture of RPE/Bruch’s membrane, cilioretinal and choriovitreal neovascularization, extension of tractional retinal detachment, and optic neuropathy ^4,⁵⁶^–⁶⁸.

Retrobulbar anesthesia may be employed safely to reduce the discomfort of PRP ^69,⁷⁰. Risks of retrobulbar anesthesia include penetration or perforation of the globe, retrobulbar hemorrhage, central retinal artery or central retinal vein occlusion, optic neuropathy, Purtscher’s retinopathy, central nervous system depression, or even respiratory arrest.

Intravitreal medications

As more information is uncovered about the underlying mechanisms of diabetic retinopathy, novel treatment strategies are being explored. A variety of anti-VEGF agents, including intravitreal bevacizumab, ranibizumab, and pegaptanib have been used as alternative or adjunctive therapy for proliferative diabetic retinopathy, vitreous hemorrhage, and neovascular glaucoma. Preliminary studies suggest these agents may be effective alternative treatment strategies alone or in combination with standard therapeutic methods for vision stabilization or improvement ^71,⁷². Further randomized clinical trials are needed.

Cryosurgery

Peripheral retina ablation with cryosurgery may be useful in proliferative diabetic retinopathy when media opacities such as severe cataract or vitreous hemorrhage prevent PRP, or in individuals who fail to demonstrate regression despite extensive laser surgery ⁷³^–⁷⁸. This method of surgery may be particularly helpful for individuals who are not candidates for pars plana vitrectomy with intraoperative PRP.

Pars plana vitrectomy

Pars plana vitrectomy is an important surgical option for complicated proliferative diabetic retinopathy (Fig. 61.7). Indications include non-clearing vitreous hemorrhage, retinal detachment, premacular fibrosis, macular heterotopia, proliferative diabetic retinopathy/iris neovascularization unresponsive to conventional therapy, hemolytic glaucoma, and dense cataract impeding photocoagulation. The technique of vitrectomy is described in other chapters on vitreoretinal surgery.