Part III: Imaging Techniques for the Clinical Evaluation of Age-Related Macular Degeneration

INTRODUCTION

Fluorescein angiography (FA) revolutionized the diagnosis of retinal disorders (1,2). However, there are certain limitations to this technique. Overlying hemorrhage, pigment, or serosanguineous fluid can block underlying pathologic changes and prevent adequate visualization by FA. Indocyanine green (ICG) is a Food and Drug Administration-approved tricarbocyanine dye that has several advantageous properties over sodium fluorescein as a dye for ophthalmic angiography. The clinical usefulness of indocyanine green angiography (ICGA) in the past has been limited by our inability to produce high-resolution images. However, enhanced high-resolution ICG angiograms can now be obtained owing to the technological advance of coupling digital imaging systems to ICG cameras (3,4). Thus, digital ICGA finally allows the theoretical advantages of ICG as an ophthalmic dye to be realized.

SPECIAL PROPERTIES OF ICG

The ICG absorbs and fluoresces in the near-infrared range. Owing to the special characteristics of the dye, there is less blockage by the normal eye pigments, which allows enhanced imaging of the choroid and choroidal abnormalities. For example, Geeraets and Berry (5) have reported that the retinal pigment epithelium (RPE) and choroid absorbs 59% to 75% of blue–green (500 nm) light, but only 21% to 38% of near-infrared (800 nm) light. The activity of ICG in the near-infrared range also allows visualization of

pathologic conditions through overlying hemorrhage, serous fluid, lipid, and pigment that may block structures by FA. This property allows enhanced imaging of occult choroidal neovascularization (CNV) and pigment epithelial detachment (PED) in age-related macular degeneration (AMD) (4,6).

A second special property of ICG is that it is highly protein-bound (98%). Therefore, less dye escapes from the choroidal vasculature, which allows enhanced imaging of choroidal abnormalities.

HISTORICAL PERSPECTIVES

ICG dye was first used in medicine in the mid-1950s at the Mayo Clinic to obtain blood flow measurements (7). In 1956, ICG was used for determining cardiac output and characterizing cardiac valvular and septal defects. In 1964, studies of systemic arteriovenous fistulas and renal blood flow were reported. The finding that exclusively the liver excreted the dye soon led to development of its application for measuring hepatic function. Recently, the use of real-time intraoperative ICGA provided information about vessel patency during neurosurgical aneurysm repair (8,9).

ICG first became attractive to ophthalmologists interested in better ways to image the choroidal circulation because of its safety and its particular optical and biophysical properties. Kogure and coworkers (10) in 1970 first performed choroidal absorption angiography in monkeys, using intraarterial ICG injection. The first ICG angiogram in a human was performed by David (11) during carotid angiography.

160 OLIVER ET AL.

In 1971, Hochheimer (12) described choroidal absorption angiography in cats using intravenous ICG injections and black-and-white infrared film instead of color film. One year later, Flower and Hochheimer performed intravenous absorption ICGA for the first time in a human (13,14). These same investigators then described the superior technique of ICG fluorescence angiography (15,16). Further technological improvements followed (17), and, in 1985, Bischoff and Flower (18) reported on their 10-year experience with ICGA, which included 500 angiograms of various disorders.

However, the sensitivity of infrared film was too low to adequately capture the low-intensity ICG fluorescence, as the fluorescence efficacy of ICG is only 4% of that of sodium fluorescein. The resolution of ICGA was improved in the mid-1980s by Hayashi and coworkers, who developed improved filter combinations and described ICG videoangiography (19–21). However, their video system lacked freezeframe image recording and possessed cumulative light toxicity potential due to its 300-W continuous halogen lamp illumination. In 1985, Destro and Puliafito (22) described a similar video system. In 1989, Scheider and Schroedel (23) reported the use of the scanning laser ophthalmoscope for ICG videoangiography; refinements of their technique allowed for improved imaging of choroidal neovascular membranes (24,25).

In 1992, Guyer and coworkers (3) and Yannuzzi and associates (4) introduced the use of a 1024-line digital imaging system to produce high-resolution enhanced ICG images. These systems have improved the resolution of ICGA such that this technique is now of practical clinical value.

PHARMACOLOGY

ICG is a sterile, water-soluble tricarbocyanine dye, which is anhdyro-3,3,30 ,30 -tetramethyl-1-10 -di- (4-sulfobutyl)-4,5,40 ,5-dibenzoindotricarbocyanine hydroxide sodium salt. Its empirical formula is C43H47N2NaO6S2 and its molecular weight is 775 Daltons (26). It is highly protein-bound (98%). Although it has been thought that ICG is primarily bound to albumin in the serum (27), 80% of ICG in the blood is actually bound to globulins, such as A1lipoproteins (28).

ICGs spectral absorption is between 790 and 805 nm (28–30). The dye is excreted by the liver via bile. ICG is not reabsorbed from the liver, is not detected in cerebrospinal fluid (31,32) and does not cross the placenta (33).

TOXICITY

ICG is a relatively safe dye, with only a few side effects reported in clinical use (7,27,34–36). In our experience,

it is safer than sodium fluorescein. In contrast to FA, nausea and vomiting are extremely uncommon during ICG angiography. We have observed two serious vasovagal-type reactions during ICGA.

No complications were reported in one study using intravenous ICG doses of 150 to 200 mg. No side effects were noted in another series of 700 procedures (18). In a study 1226 consecutive patients undergoing ICGA, there were three (0.15%) mild adverse reactions, four (0.2%) moderate reactions, one (0.05%) severe reaction, and no deaths (36).

ICGA should not be performed on patients allergic to iodide, since it contains approximately 5% iodide by weight. In addition, it should not be performed on patients who are uremic (18) or who have impaired hepatic clearance. Appropriate emergency equipment should be readily available, as with FA.

TECHNIQUE OF INJECTION

ICGA can be performed immediately before or after FA. We inject intravenously 25 to 50 mg of ICG (Cardio-Green: Hynson, Westcott & Dunning Products, Cockeysville, Maryland, U.S.A.) which has been diluted in the aqueous solvent supplied by the manufacturer. Rapid injection is essential and should be followed by a 5-mL normal saline flush. For wide angle angiography, the dosage is increased to 75 mg.

Bindewald and associates (37) recently tested the lower limits fluorescein and ICG dye doses for angiography. Using a confocal scanning laser ophthalmoscope (cSLO) (Heidelberg retina angiograph 2, Heidelberg Engineering, Dossenheim, Germany), they found that a fluorescein dose as low as 166 mg, and an ICG dose as low as 5 mg, allowed adequate resolution for diagnosis and management of neovascular AMD. Resolution was impaired, however, in late phase images, compared to standard doses.

DIGITAL IMAGING SYSTEMS

The coupling of a digital imaging system with an ICG camera allows production of enhanced high-res- olution (1024-line) images, which are necessary for ICGA. The instantaneous images from these systems produce images which decrease patient waiting time and expedite treatment. Digital imaging systems also allow image archiving, hard-copy generation, and direct qualitative comparison between fluorescein and ICGA findings. These systems are useful for planning preoperative treatment strategies and for monitoring the adequacy of treatment postoperatively.

Imaging systems contain film, video, or digital cameras with special antireflective coatings and appropriate excitatory and barrier filters. Flash

synchronization allows high-resolution image capture. The digitally charged coupling device camera captures the digitized images and transmits them to a digital imaging workstation. These images are captured at one frame per second, stored in buffer memory, and displayed on a high-contrast, high-resolution video monitor. The images can be printed to photographs or slides, transferred via a variety of storage media, or networked to other stations in treatment areas and in other offices.

INTERPRETATION OF ICGA FINDINGS IN AMD

Definitions

The terminology used to describe the angiographic manifestations of AMD corresponds, with certain exceptions described below, to definitions previously reported by the Macular Photocoagulation Study Group (38). Most relevant to the interpretation of ICGA in AMD are the definitions of serous pigment epithelium detachment (SPED), vascularized pigment epithelial detachment (VPED), classic CNV, and occult CNV (4,19,22,39).

Serous Pigment Epithelial Detachment

The SPED is an ovoid or circular detachment of the RPE. On FA study there is rapid filling with dye of the fluid in the subRPE space. This corresponds to early hyperfluorescence beneath the PED, which increases in intensity in the late phase of the study resulting in a bright and homogeneous well-demarcated pattern. ICGA reveals a variable, minimal blockage of normal choroidal vessels, more evident in the mid-phase of the angiogram. Thus, a SPED is bright (hyperfluorescent) on FA and dark (hypofluorescent) on ICG. This difference is caused by the fact that ICG molecules are larger and almost completely bound to plasma proteins, which prevents free passage of ICG dye throughout the fenestrated choriocapillaris in the subRPE space. Also, it is important to remember the difference of appearance on ICGA between a SPED in AMD and a SPED in central serous chorioretinopathy (CSC). In fact, in CSC there is increased permeability of the choriocapillaris that causes leakage of ICG molecules under the PED. As a result, a SPED in CSC appears bright (hyperfluorescent) with ICGA. Approximately 1.5% of newly diagnosed patients with exudative AMD present with a pure SPED.

Choroidal Neovascularization

CNV is defined as a choroidal capillary proliferation through a break in the outer aspect of Bruch’s membrane under the RPE and/or the neurosensory retina. CNV is divided into classic and occult based on the FA angiography appearance.

Classic CNV

Classic CNV is an area of bright, fairly uniform hyperfluorescence identified in the early phase of the FA. The fluorescence increases through the transit phase with leakage of dye obscuring the boundaries of this area by the late phase of the angiogram. With ICGA, a classic CNV has a similar appearance to that seen with FA angiography, but is usually less well delineated (Fig. 1) and exhibits little or no leakage in the late phases of the ICG study. Only 12% of newly diagnosed patients with exudative AMD present with classic CNV.

Occult CNV

Occult CNV is characterized as either fibrovascular pigment epithelial detachment (FVPED) or late leakage of undetermined source (LLUS). FVPED consists of irregular elevation of the RPE consisting of stippled hyperfluorescence not as bright or discrete as classic CNV within one to two minutes after fluorescein injection, with persistence of fluorescence 10 minutes after injection. LLUS consists of areas of leakage at the level of the RPE in the late phase of the angiogram not corresponding to an area of classic CNV or FVPED discernible in the early or middle phase of the angiogram to account for the leakage. Also any area of blocked fluorescence contiguous to the CNV is considered occult CNV. More than 85% of newly diagnosed patients with exudative AMD present with occult CNV (Fig. 2). Two main types of occult CNV are recognized on ICGA.

Figure 1 Classic choroidal neovascularization. Early phase indocyanine green angiogram shows a well-defined hyperfluorescent vascular network consistent with a classic choroidal neovascular membrane.

Without SPED. The first type of occult CNV is caused by subRPE CNV that is not associated with a PED. The early stages of FA study reveal minimal subretinal hyperfluorescence of undetermined source that slowly increases over a period of several minutes to produce an irregular staining of the subRPE tissue. The ICG angiogram reveals early vascular hyperfluorescence and late staining of the abnormal vessels. If the ICG angiographic image has distinct margins, it is considered to be a well-defined CNV on ICGA. Twothird of newly diagnosed patients with an occult CNV present without an associated SPED.

With SPED. The second type of occult CNV is associated with a SPED of at least 1-disc diameter in size. Combined CNV and SPED are called a VPED. This lesion is the result of subRPE neovascularization associated with a serous detachment of the RPE. Onethird of newly diagnosed patients with AMD have an associated SPED. The determination of whether a SPED is present is best made on the basis of the FA study. FA may also demonstrate occult vessels as late, indistinct, subretinal hyperfluorescence beneath, or at the margin of the SPED. ICGA reveals early vascular hyperfluorescence and late staining of the CNV. The SPED, as noted previously, is comparatively hypofluorescent, because only minimal ICG leakage occurs beneath the serous detachment. ICG is more helpful than FA in differentiating between a SPED and a VPED. It also permits better identification of the vascularized and serous component of VPEDs. These differentiations between the vascularized and serous components are often not possible with FA alone

because the serous and vascularized portions of a PED demonstrate late hyperfluorescence and leakage respectively. Although fluorescein staining is more intense in the serous portion of the detachment than in the vascularized component, differences in intensity are often too minimal for accurate interpretation. However, the ICG angiographic findings are infinitely more reliable for this differentiation; the serous component of a PED is hypo-32#fluorescent and the vascularized component is hyperfluorescent.

Occult CNV is also sub grouped in two types, one with a solitary area of well-defined focal neovascularization (hot spot) and the other with a larger and delineated area of neovascularization (plaque).

Hot Spot (Focal CNV)

Focal CNV or a “hot spot” is an area of occult CNV that is both well-delineated and no more than 1-disc diameter in size on ICGA. Also a hot spot represents an area of actively proliferating and more highly permeable areas of neovascularization (active occult CNV). Chorioretnal anastomosis and polypoidaltype CNV may represent two subgroups of hot spots (see below).

Plaque

A plaque is an area of occult CNV larger than 1-disc diameter in size. A plaque often is formed by latestaining vessels, which are more likely to be quiescent areas of neovascularization that are not associated with appreciable leakage (inactive occult CNV).

Plaques of occult CNV seems to slowly grow in dimension with time. Well-defined and ill-defined plaques are recognized on ICG study. A well-defined plaque has distinct borders throughout the study and the full extent of the lesion can be assessed. An illdefined plaque has indistinct margins or may be one in which any part of the neovascularization is blocked by blood.

In a review of our first 1000 patients with occult CNV by FA, which were imaged by ICGA, we categorized occult CNV into three morphologic categories: focal CNV or hot spots, plaques (well-defined and ill-defined), and combination lesions in which both hot spots and plaques were noted (39). The results of that study are discussed later in this chapter under clinical applications.

Two other forms of occult CNV are identified by ICGA: polypoidal choroidal vasculopathy (PCV) and retinal angiomatous proliferation (RAP).

POLYPOIDAL CHOROIDAL VASCULOPATHY

PCV is a primary abnormality of the choroidal circulation characterized by an inner choroidal vascular network of vessels ending in an aneurysmal bulge or outward projection, visible clinically as a reddish orange, spheroid, polyp-like structure. The disorder is associated with multiple, recurrent, serosanguineous detachments of the RPE and neurosensory retina, secondary to leakage and bleeding from the peculiar choroidal vascular abnormality (40,41).

ICGA has been used to detect and characterize the PCV abnormality with enhanced sensitivity and specificity (Fig. 3) (42–55). In the initial phases of the ICG study, a distinct network of vessels within the choroid becomes visible. Optical coherence tomography (OCT) (Fig. 3C) delineates the polypoidal extensions of the choroidal vasculature. In patients with juxtapapillary involvement, the vascular channels extend in a radial, arching pattern and are interconnected with smaller spanning branches that become more evident and numerous at the edge of the PCV lesion (Fig. 4).

Early in the course of the ICG study, the larger vessels of PCV network start to fill before the retinal vessels, but the area within and surrounding the network is relatively hypofluorescent compared with the uninvolved choroid. The vessels of the network appear to fill more slowly than the retinal vessels. Shortly after the network can be identified on the ICG angiogram, small hyperfluorescent “polyps” become visible within the choroid.

These polypoidal structures correspond to the reddish, orange choroidal excrescence seen on clinical examination. They appear to leak slowly as the surrounding hypofluorescent area becomes increasingly

(A)

(B)

(C)

Figure 3 Polypoidal choroidal vasculopathy. Color photograph (A) demonstrates hemorrhagic detachment of the macula. Latephase indocyanine green study (B) reveals a peripapillary polyplike vascular network. Note central hypofluorescence indicative of a pigment epithelial detachment. Optical coherence tomography (C) delineates the polypoidal extensions of the choroidal vasculature.

hyperfluorescent. In the later phase of the angiogram there is a uniform disappearance of the dye (“washout”) from the bulging polypoidal lesions. The late ICG staining characteristic of occult CNV is not seen in the PCV vascular abnormality.

While the first reports of PCV were in middleaged black females, it is now recognized that PCV may be a variant of CNV seen in white patients with AMD, it may be localized in the macular area without any

peripapillary component (Figs. 5 and 6), and it may be formed by a network of small branching vessels ending in polypoidal dilation difficult to image without ICGA (Fig. 6).

Ahuja and colleagues sought to determine the prevalence of PCV among British patients in their

Figure 5 Macular polypoidal choroidal vasculopathy. Midphase indocyanine green angiogram demonstrates a prominent lesion of polypoidal channels in the macula.

practice (50). Of 40 consecutive patients with hemorrhagic or exudative PEDs, 34 (85%) were attributed to PCV. Of those with PCV, 65% were female, the mean age was 65 years (range 44–88), 74% were white, 20% black, and 6% Asian. Eight had a history of hypertension. Sixty-eight percent of lesions were located in the macula.

RETINAL ANGIOMATOUS PROLIFERATION

RAP is a distinct subgroup of neovascular AMD, manifested by intraretinal neovascularization (IRN) that extends into the deep retinal, subretinal, and subRPE spaces.

Clinical evidence of pre-, intra-, or subretinal hemorrhage, sometimes with associated exudates or cystoid macular edema, in the setting of a PED suggests a RAP lesion. Often dilated compensatory vessels perfuse and drain the neovascularization, forming a retinal–retinal anastomosis. Extension of the neovascular complex to the subretinal space may result in a retino-choroidal anastomosis (RCA).

On FA, indistinct RPE staining, often with associated PED, resembles occult CNV. Presence of active IRN extending into a PED is difficult to distinguish from a standard VPED. ICG allows better characterization of a VPED, revealing the neovascular hotspot contained within the hypofluorescent PED (Fig. 7A,B). The OCT (Fig. 7C) shows

intraretinal cystic changes overlying a PED along with a hyperreflective area suggestive of a RCA. Late intraretinal leakage may arise from the IRN. ICG permits visualization of the direct communication between the retinal and the choroidal components of the neovascularization as they form an RCA (Fig. 8) Lafaut and coworkers (56) documented the histopathology of an RCA, in which neovascularization grows out from the neuroretina into the subretinal space.

Kuhn et al. (57), in 1995, first identified RCA as a potential manifestation of this form of neovascular AMD in the setting of a VPED. With ICGA for enhanced choroidal imaging, this group found RCA in 50 of 186 (28%) patients with AMD and an associated VPED. Slakter et al. (58) detected RCA in 34 of 150 eyes (21%) with occult AMD and a focal hot spot in ICG. Fernandez and coworkers (59) reported a series of 190 patients with neovascular AMD in which ICGA revealed 34 eyes (16%) with RAP lesions.

(A)

Figure 7 Retinal angiomatous proliferation. Fluorescein angiogram (A) indicates a pigment epithelial detachment, while the indocyanine green (B) reveals a focal area of hyperfluorescence adjacent to two retinal arterioles. The optical coherence tomography (C) shows intraretinal cystic changes overlying a pigment epithelial detachment along with a hyperreflective area suggestive of a retino-choroidal anastomosis.

(B)

(C)

Yannuzzi and colleagues (60) classified RAP into three stages: stage I involves IRN, stage II results from extension to subretinal neovascularization, and stage III occurs once CNV is documented.

Clinical knowledge and recognition of RAP is important because this form of neovascular AMD may have a natural course, visual prognosis, and response to treatment distinct from other forms of neovascular AMD. Different forms of treatment may be preferable

for each stage of the disorder. For example, we have found that an uncomplicated focal area of IRN may be amenable to conventional thermal laser treatment; whereas, a more advanced stage involving a VPED and an RCA is less likely to respond to any form of currently available treatment.

Bottoni and colleagues (61) retrospectively reported results of 99 eyes of 81 patients with RAP treated with direct laser photocoagulation of the

vascular lesion, laser photocoagulation of the feeder retinal arteriole, scatter gridlike laser photocoagulation, photodynamic therapy (PDT), or transpupillary thermotherapy. Complete obliteration was achieved in 24 (57%) cases of stage I lesions (73% closure from direct laser and 45% closure from PDT, 11 (26%) of stage II lesions (38% closure from scatter gridlike photocoagulation and 17% closure from direct photocoagulation of the vascular lesion), and only 3 (15%) of stage III lesions. The uncontrolled use of therapeutic interventions in this study makes it difficult to draw definitive conclusions about a superior treatment modality, but the study makes clear the difficulty in effectively treating more advanced stages of RAP.

More recent series using PDT alone or PDT with triamcinolone confirm the challenge in effectively treating RAP lesions. Boscia and colleagues (62) treated 21 eyes with stage II or III RAP using PDT alone and reported an overall decline in vision from 20 out of 80 to 20 out of 174, stabilization of vision in six eyes (29%), occlusion of RAP and PED flattening in three eyes (14%), and an RPE tear in four eyes (19%). Nicolo and colleagues (63) reported 10 eyes with stage II RAP treated with 20 mg of intravitreal triamcinolone acetonide (IVTA) followed one month later by PDT. All patients experienced flattening of the PED prior to PDT, six patients (60%) showed improved vision of at least three early treatment of diabetic retinopathy study (ETDRS) lines at three, six, and nine months, and four patients (40%) maintained visual improvement at 12 months.

PDT with a photosensitizing dye such as verteporphin may have a different effect on RAP than on classic-CNV or occult-CNV lesions (64,65). Given the tendency for ICG dye to stain the retina in eyes with RAP, there is a possibility that similar staining may occur with the verteporphin molecule, theoretically predisposing the retina to photochemical damage when exposed to the excitatory light used in PDT. This possibility is speculative, since verteporphin has not yet been imaged successfully with good spatial and temporal definition in the human.

Because eyes with RAP are generally classified as pure occult CNV based on FA, it is possible that patients with RAP were actually treated in the treatment of age-related macular degeneration with photodynamic therapy (TAP) trial (65). ICGA was not used in the TAP trial, so the frequency of RAP in the subset of patients classified as occult-CNV is unknown. Future studies of AMD that use ICG will be able to delineate between these two distinct forms of macular degeneration.

CLINICAL APPLICATION OF ICGA

TO THE STUDY OF AMD

Patz and associates (26) were the first to study CNV by ICG videoangiography. They could resolve only 2 of 25 CNVs with their early model. Bischoff and Flower (18) studied 100 ICG angiograms of patients with AMD. They found “delayed and/or irregular choroidal filling” in some patients. The significance of this finding is unclear, however, because these authors did not include an age-matched control group. Tortuous vessels and marked dilation of macular choroidal arteries, often with loop formation, were also observed.

Hayashi and associates (19,21) found that ICG videoangiography was useful in the detection of CNV. ICG videoangiography was able to confirm the fluorescein angiographic appearance of CNV in patients with well-defined CNV. It revealed a more welldefined neovascularization in 27 eyes with occult CNV by FA. In a subgroup of patients with poorly defined occult CNV, the ICG angiogram, but not the FA, imaged a well-defined CNV in 9 of 12 (75%) cases. ICG videoangiography of the other three eyes revealed suspicious areas of neovascularization. Hayashi and coworkers (19,21) were also the first to show that leakage from CNV with ICG was slow compared to the rapid leakage of sodium fluorescein. While the results of these investigators concerning ICG videoangiographic imaging of occult CNV were promising, the spatial resolution that they could obtain was limited by the 512-line video monitor and analog tape of their ICG system.

Destro and Puliafito (22) reported that ICG videoangiography was particularly useful in studying occult CNV with overlying hemorrhage and recurrent CNV. Guyer and coworkers (3) used a 1024-line digital imaging system to study patients with occult CNV. These authors reported that ICG videoangiography was useful in imaging occult CNV and that this technique could allow photocoagulation of otherwise untreatable lesions. Scheider and coinvestigators (25) have reported enhanced imaging of CNV in a study of 80 patients using the scanning laser ophthalmoscope with ICG videoangiography.

Yannuzzi and associates (4) have shown that ICGA is extremely useful in reclassifying occult CNV into “well-defined CNV.” In their study, 39% of 129 patients with occult CNV were reclassified as welldefined CNV based on information added by ICGA. Five of seven (72%) cases of occult CNV with SPED were reclassified as “well demarcated” CNV by ICG. In 17 of 38 (45%) VPED cases and in 11 of 19 (58%) combined VPED and SPED cases, ICGA allowed occult CNV to be reclassified as well defined CNV. These authors concluded that ICGA was especially

168 OLIVER ET AL.

useful in identifying occult CNV in patients with SPED or with recurrent CNV.

Lim et al. found that ICG demonstrated welldemarcated hyperfluorescence in 50% of eyes thought to have occult CNV by FA and in 82% of eyes with PED (66). Baumal et al. found that ICG demonstrated underlying CNV in 19 of 23 eyes (83%) with an isolated PED and in 21 of 21 eyes (100%) with PED and occult CNV (67).

Yannuzzi and coworkers (68) studied with ICGA 235 consecutive AMD patients with occult CNV and associated VPED. These eyes were divided into two groups, depending on the size and delineation of the CNV. Of the 235 eyes 89 (38%) had a solitary area of neovascularization that was well delineated, no more than 1-disc diameter in size, and defined as focal CNV. The other 146 (62%) eyes had a larger area of neovascularization, with variable delineation defined as a plaque CNV.

In a further report, 657 consecutive eyes with occult CNV determined by FA were studied with ICGA. Of 413 eyes with occult CNV without pigment epithelium detachments, focal areas of neovascularization were noted in 89 (22%). Overall, 142 (34%) eyes had lesions that were potentially treatable by thermal laser photocoagulation based on additional information provided by ICGA. Of the 235 eyes with occult CNV and VPEDs, 98 (42%) were eligible for photocoagulation therapy based on ICGA findings. The authors calculated that ICGA enhances the treatment eligibility by approximately one-third (69).

In a expanded series (39) the same authors reported their results on ICGA study of 1000 consecutive eyes with occult CNV by FA. They recognized three morphologic types of CNV, which included focal spots, plaques (well defined and poorly defined), and combination lesions (in which both focal spots and plaques are noted). Combination lesions were further subdivided into marginal spots (focal spots at the edge of plaque of neovascularization), overlying spots (hot spots overlying plaques of neovascularization), or remote spots (a focal spot remote from a plaque of neovascularization).

The relative frequency of these lesions was as follows: focal spots 29%, plaques 61% consisting 27% of well-defined plaques and 34% of poorly defined plaques, and combination lesions 8%, consisting of 3% of marginal spots, 4% of overlying spots and 1% of remote spots (39). A follow-up study from the same authors of patients with newly diagnosed unilateral occult CNV secondary to AMD showed that the patients tended to develop the same morphologic type of CNV in the fellow eye (70).

Chang et al. (71) reported on the clinicopathologic correlation of AMD with CNV detected by ICGA. Histopathologic examination of the lesion revealed a

thick subRPE CNV corresponding to the plaque-like lesion seen with ICGA.

Watzke and colleagues analyzed 104 consecutive AMD patients to determine the sensitivity of ICG in detecting lesions originally identified by FA (72). ICG hyperfluorescence was present in 87% of eyes with classic CNVand in 93% of eyes with fibrovascular pigment epithelium detachments (FVPEDs). Of eyes diagnosed with LLUS by FA, 50% were hyperfluorescent and 50% were isofluorescent by ICG. Additionally, three fellow eyes with dry AMD had hyperfluorescent lesions by ICG, but it is unknown whether these eyes progressed to neovascular AMD.

Finally, Lee et al. (73) reported on 15 eyes with surgically excised subfoveal CNV that underwent preoperative and postoperative ICGA. All excised membranes were examined by light microscopy, and all surgically excised ICG-imaged membranes corresponded to subRPE and subneurosensory CNV.

The above studies demonstrate that ICGA is an important adjunctive study to FA in the detection of CNV. FA is more sensitive than ICGA in imaging fine capillaries that connect larger vessels and capillaries at the proliferating edge of well-defined CNV. While FA images well-defined CNV better than ICGA in some cases, ICGA allows reclassification of FA-defined occult CNV into well-demarcated CNV eligible for ICG-guided thermal laser treatment in about 30% of cases (74).

The best imaging strategy to thoroughly classify CNV is the combination of FA and ICGA. Helbig et al. studied 502 patients using simultaneous FA and ICG to characterize AMD, and found that 3% of eyes had a hot spot within an occult lesion, 4% had plaques within an occult lesion, 9% had RAP, and 6% had PCV (75). Yanagi and colleagues (76) compared simultaneous fluorescein and ICG injection with FA-guided ICGA, in which FA was used to detect an area of leakage, allowing a lower dose ICG injection and focusing the ICG detector only on the lesion in question. Overall detection of feeder vessels (FVs) was similar between the simultaneous and FA-guided ICG groups, but the latter group required lower quantities of ICG and had shorter examination times. The benefits of simultaneous procedures, such as convenience and accurate diagnosis of treatable cases, must be weighed against the disadvantages of increased cost and adverse effects.

RECURRENT CNV IN AMD

Recurrent CNV following photocoagulation treatment is a major cause of treatment failure. Although most recurrences can be detected and imaged with clinical biomicroscopic examination and FA, a significant number of patients demonstrate new exudative

manifestations and visual symptoms without a clearly defined area of recurrent neovascularization identified by FA. These patients may exhibit diffuse staining and leakage at the site of previous treatment or may demonstrate no FA evidence of recurrence despite the new exudative manifestations identified clinically. ICGA has proven to be often useful in detecting the recurrence.

Sorenson et al. (77) reported the use of ICGguided laser treatment in 66 cases of recurrent occult CNV secondary to AMD. Only 29 (44%) were eligible for laser retreatment, and of these 29 eyes 18 (62%) had anatomic success with an average follow-up of six months (54). Similar results were reported by Reichel et al., who reported 58 eyes with recurrent CNV from AMD (78). In 14 eyes (24%), a well-defined recurrent CNV could be identified by evaluating the fluorescein angiogram. In 6 (14%) of the remaining 44 eyes, a welldefined recurrent CNV was identifiable by ICGA.

However, clinical evidence of recurrence must accompany a hot spot detected by ICG. Chen and colleagues (79) performed ICGA two weeks after krypton laser treatment on 230 consecutive eyes with exudative AMD. Forty patients (18%) developed ICG hot spots after treatment, and these hot spot spontaneously resolved without development of CNV in 31 patients. Recurrent CNV was present at the hot spot in four patients and away from the hot spot in five patients.

ICG-ASSOCIATED TREATMENT STRATEGIES FOR CNV IN AMD

In the past, patients were considered potentially eligible for laser photocoagulation therapy by ICG guidance if they had clinical and FA evidence of occult CNV. Of the two types of occult CNV identifiable by ICG study, hot spots and plaques, direct laser photocoagulation was recommended only to hot spots. In fact as mentioned above, hot spots represent areas of actively leaking neovascularization that can be obliterated by laser photocoagulation in attempt to eliminate the associated serosanguineous complications, and stabilize or improve the vision. On the contrary, plaques seem to represent a thin layer of neovascularization, which is not actively leaking, and which may benefit from PDT (80) or intravitreal antiangiogenic agents (81–86).

In the case of a lesion comprised of a hot spot and a plaque, and in which the hot spot is at the margin of the plaque (that may extend under the fovea), laser photocoagulation to the extrafoveal hot spot spares the fovea. This treatment approach was successful in obliterating the CNV and stabilizing the vision in 56% of a consecutive series of AMD patients (74). On the contrary we have had poor success with

direct laser treatment of hot spots overlying plaques, or confluent treatment of the entire plaque.

Slakter and associates (87) performed ICGguided laser photocoagulation in 79 eyes with occult CNV. The occult CNV was successfully eliminated with stabilized or improved visual acuity in 29 (66%) of 44 eyes with occult CNV associated with neurosensory retinal elevations, and in 15 (43%) of 35 eyes with occult CNV associated with PED. This study demonstrated that in some cases ICGA imaging can successfully guide laser photocoagulation of occult CNV.

Another pilot study of ICG-guided laser treatment of occult CNV had similar results (88).

Guyer and coworkers (74) reported a pilot study with ICG-guided laser photocoagulation of 23 eyes with occult CNV secondary to AMD with focal spots at the edge of a neovascular plaque of the ICG study. ICG-guided laser photocoagulation was applied solely to the focal spot at the edge of the plaque. At 24 months of follow-up anatomic success with resolution of the exudative findings was obtained in 6 (37.5%) of 16 eyes. Importantly, these studies set the foundation for future prospective studies of ICG-guided laser treatment. In addition, they proved that the presence of a PED is a poor prognostic factor in the treatment of exudative AMD.

Lim et al. reported the visual acuity outcome after ICGA-guided laser photocoagulation of CNV associated with PED in 20 eyes with AMD (89). At three months after laser photocoagulation, visual acuity had improved two or more Snellen lines in two eyes (10%), worsened by two or more lines in 10 (50%), and remained unchanged in eight of 20 (40%). At nine months after laser photocoagulation, visual acuity had improved by two or more lines in one eye (9%), worsened by two or more lines in nine (82%), and remained unchanged in one of 11 (9%). They concluded that ICG-guided laser photocoagulation may temporarily stabilize visual acuity in some eyes with CNV associated with PED, but final visual acuity decreases with time.

More recently, Da Pozzo and associates evaluated the efficacy of ICG-guided photocoagulation in 86 eyes with occult CNV and a hot spot on ICG (90). Of the 53 eyes without PED, 32 (60%) had stable or improved vision at one year, but 27 (51%) had recurrence of the CNV. Of the 33 eyes with PED, only five (15%) had stable or improved vision at one year, and 23 (70%) had CNV recurrence.

Another potential therapeutic application using ICG is ICG dye-enhanced diode laser photocoagulation. The peak absorption of ICG (795 to 810 nm) is at a similar wavelength as the peak emission of the diode laser (805 nm). Thus, dye-enhanced laser photocoagulation may allow selective ablation of the ICG-containing CNV with relative sparing of the

170 OLIVER ET AL.

normal neighboring retina. However, leakage of ICG into the intraretinal space, which occurred in 11% of 149 eyes in a series reported by Ho and colleagues, may be a contraindication to ICG dye-enhanced diode photocoagulation (91).

A pilot study by Reichel and associates of 10 patients with poorly defined CNV resulted in closure of the CNV in all cases, but a severe immediate vision loss occurred in one patient (92). A larger series by Obana et al. studied 38 eyes with classic or occult CNV, and found that CNV occlusion was achieved in 92% of eyes, with and 18% recurrence rate over an average follow-up of 26 months (93). Ten eyes (26%) showed improved visual acuity, 16 (42%) showed no change, and 12 eyes (32%) worsened.

A pilot study by Arevalo et al. compared ICG dye-enhanced diode laser photocoagulation alone with dye-enhanced laser combined with IVTA (94). In their initial series of 19 eyes selected irrespective of lesion subtype, none of the 9 eyes receiving combination therapy required retreatment at seven months, while 4 of 10 eyes receiving ICG dyeenhanced laser alone needed retreatment. A follow up paper by the same group reported 31 eyes treated with dye-enhanced laser and 4 mg of IVTA followed for a mean of nine months (95). Nineteen eyes (61%) showed stable vision, seven eyes (23%) improved, and five eyes (16%) worsened. In the occult subgroup, however, the proportion of patients who worsened was greater (33%). No severe acute vision loss was reported. Topically treatable glaucoma occurred in five eyes.

NEW TECHNIQUES IN ICG ANGIOGRAPHY

Recent advances in ICGA are real-time angiography, contrast enhancement ICGA, wide-angle angiography, digital subtraction-indocyanine green (DS-ICG) angiography, dynamic ICG-guided FV laser treatment of CNV, ICGA for dry AMD, and use of cSLO-ICG.

Contrast Enhanced ICG Angiography

Contrast enhancement of ICG angiographic images using digital imaging software may enhance the diagnostic sensitivity and specificity of the study. Maberley and Cruess compared nonenhanced and contrastenhanced ICG angiographic images of 50 consecutive patients with occult CNV from AMD (96). Only 36% of the nonenhanced images demonstrated well-defined membranes, whereas 58% were well defined with the contrast-enhanced images.

Real-Time ICGA

Real-time ICGA (97) uses a modified Topcon 50IA camera with a diode laser illumination system that

has an output at 805 nm (Topcon 50IAL camera), can produce images at 30 frames per second, and allows continuous recording. The images can be acquired either as a videotape, or as single image at a frequency of 30 images per second. To make printed copies of these images single frames are digitized, but the resolution is limited to 640 by 480 pixels.

Wide-Angle ICGA

Wide-angle images of the fundus can be obtained by performing ICG videoangiography with the aid of wide angle contact lenses. The contact lenses used are the Volk SuperQuad 160, the Volk Quadraspheric, or the Volk Transequator (Volk, Mentor OH). Because the image formed by these lenses lies about 1 cm in front of the lens, the fundus camera is set on A or C so that the camera is focused on the image plane of the contact lens.

This technique allows instantaneous imaging of a large area of the fundus. The combined use of the contact lens and of the laser illumination system in a high-speed digital fundus camera allows realtime imaging of a 1608 of field of view. Staurenghi and colleagues recently developed a combined contact and noncontact system to achieve wide-field images up to 1508 with a cSLO ICG (98).

Digital Subtraction-Indocyanine Green Angiography

DS-ICGA uses DS of sequentially acquired ICG angiographic frames to image the progression of the dye front in the choroidal circulation (99,100). A method of pseudocolor imaging of the choroid allows differentiation and identification of choroidal arteries and veins. DS-ICGA allows imaging of occult CNV with greater detail and in a shorter period of time than with conventional ICGA.

Matsumoto et al. performed DS-ICGA on 20 patients with CNV accompanied by subretinal hemorrhage (101). In six of the 20 eyes, DS-ICGA distinguished hyperfluorescence due to a slowly expanding, poorly defined, large lesion from simple leakage with a well-defined lesion. The DS-ICGA technique made clear the expanding wave of hyperfluorescence from a more slowly filling, illdefined lesion.

FV Therapy

Staurenghi et al. (102) considered a series of 15 patients with subfoveal CNVM in whom FVs could be clearly detected by means of dynamic ICGA but not necessarily with FA. Based on the pilot study, the authors simultaneously reported a second series of 16 patients with FVs smaller than 85 mm. FV were treated with argon green laser. The ICGA was repeated

immediately after treatment, and at two, seven, and thirty, and then every three months, to assess FV closure. A FV that remained patent was immediately retreated, and ICGA follow-up started again. In the pilot study, 40% of FVs were successfully occluded; this result was affected by the width and number of FVs. The occlusion success rate in the second series, with FVs under 85 mm, was 75%. The authors concluded that dynamic ICGA may detect small FVs that are more successfully occluded by argon photocoagulation.

ICG FOR DRY AMD

Hanutsaha et al. studied 432 patients by ICG with exudative AMD in one eye and drusen without exudation in the fellow eye (103). Eighty-nine percent of eyes with drusen had normal fluorescence on ICG, while 11% of eyes with drusen had focal hot spots or hyperintense plaques. Over an average follow-up of 22 months, 27% of eyes with drusen and an abnormal ICG developed CNV, while only 10% of drusen eyes and a normal ICG developed exudative AMD. The authors suggested that ICG may be a predictive indicator of future exudative changes in eyes with drusen.

Patchy and slow choroidal filling on FA, in association with reduced choroidal fluorescence on ICGA, was associated by Pauleikoff and associates with early changes in AMD (104). One hundred eyes with early AMD were studied for the above characteristics, termed a prolonged choroidal filling phase (PCFP), which was associated with confluent drusen in the study eye, focal RPE atrophy in the study eye, and geographic atrophy in the fellow eye. The group postulated that PCFP was a clinical indicator of Bruch membrane deposits and a predictor for geographic atrophy from AMD.

Ultra-late phase ICGA, performed 24 hours after dye injection, demonstrates hypofluorescent geographic lesions in patients with both exudative and dry AMD, as shown by Mori and associates. They demonstrated that 95% of AMD eyes with CNV had geographic hypofluorescent lesions, and that all CNV detectable by FA or ICG was contained within these lesions. In 73% of eyes without CNV, the same geographic areas were present, while agematched normal subjects did not have the lesions. Mean fluorescence intensity was higher in a normal group older than 62 years, compared to normal subjects less than 36 years. The authors postulated that these geographic hypofluorescent areas may represent areas predisposed to CNV development.

CONFOCAL SCANNING LASER

OPHTHALMOSCOPE ICG

With the relatively recent availability of cSLO to perform ICG, many retinal physicians are choosing this modality over high-speed digital angiography because of the ability to use the cSLO for other functions, such as autofluorescence and FA. Gelisken and colleagues simultaneously compared cSLO ICG with high-resolution fundus camera ICG in 100 eyes with occult CNV (105). Confocal SLO was superior in delineating vessel architecture of the neovascular lesion; however fundus photography was much more sensitive than cSLO in detecting focal lesions (52% vs. 37%, respectively) and plaques (35% vs. 13%, respectively).

CONCLUSION

The role ICGA in the treatment of AMD is in evolution. As photocoagulation of extrafoveal CNV gave way to treatment of all types of CNV with PDT, ICG angiography has proven very useful in adding information to FA about lesion subtype. The ability of ICG to identify subtypes of occult CNV, such as VPED, hot spots, plaques, and RCA, allows targeted and sometimes effective therapy for these refractory types of CNV. Given that approximately 87% of new CNV from AMD is minimally classic or occult (106), many patients have derived some benefit from the additional information obtainable by ICGA.

The approval of pegaptanib sodium (Macugen) heralded a new era in AMD treatment (81). Shown to be equally efficacious for all lesion subtypes, pegaptanib was a departure from traditional laser-based, destructive procedures. Ranibizumab (Lucentis) has been shown to be even more efficacious and beavcizumab (Avastin) appears to show similar results to ranibizumab (82–85). Further research is necessary to determine whether lesion subtype remains an important predictor of treatment response with these new modalities.

A systematic evidence based review of the PubMed indexed literature in English or with an English abstract yielded a strong recommendation for the use of ICGA for the following conditions: PCV, occult CNV, neovascularization associated with PED, and recurrent choroidal neovascular membranes (107). The same review reported only modest evidence supporting the use of ICGA for routine choroidal neovascular membranes and for identifying FVs in AMD. Future advances in ICGA, such as wide angle, real-time, and DS techniques may improve our diagnostic ability in AMD.

172 OLIVER ET AL.

SUMMARY POINTS

&ICGA is a useful adjunctive technique to FA for the diagnosis of AMD. This is especially true in the presence of occult CNV. ICG allows better recognition of subtypes of occult CNV such as VPED, hot spots, plaques and retinal-choroidal anastomosis.

&ICGA is useful in the diagnosis of PCV, RAP, and recurrent choroidal neovascular membranes.

&Preliminary studies suggested that ICG-guided laser photocoagulation was beneficial in the treatment of CNV prior to the era of antivascular endothelial growth factor therapy.

&Further research is necessary to improve our understanding of all the information obtained by ICGA and its potential role in new therapeutic regimens.

&Real-time ICGA, wide-angle ICGA, and DS-ICGA may improve our diagnostic ability in AMD.

REFERENCES

1.Schatz H, Burton TC, Tannuzzi LA, Rabb MF. Interpretation of Fundus Fluorescein Angiograph. St. Louis: Mosby-Year Book, 1978.

2.Yannuzzi LA. Laser Photocoagulation of the Macula. Philadelphia, PA: JB Lippincott, 1989.

3.Guyer DR, Puliafito CA, Mones JM, Friedman E, Chang W, Verdooner SR. Digital indocyanine-green angiography in chorioretinal disorders. Ophthalmology 1992; 99(2):287–91.

4.Yannuzzi LA, Slakter JS, Sorenson JA, Guyer DR, Orlock DA. Digital indocyanine green videoangiography and choroidal neovascularization. Retina 1992; 12(3):191–223.

5.Geeraets WJ, Berry ER. Ocular spectral characteristics as related to hazards from lasers and other light sources. Am J Ophthalmol 1968; 66(1):15–20.

6.Fox IJ, Wood EH. Applications of dilution curves recorded from the right side of the heart or venous circulation with the aid of a new indicator dye. Mayo Clin Proc 1957; 32(19):541–50.

7.Fox IJ, Wood EH. Indocyanine green: physical and physiologic properties. Mayo Clin Proc 1960; 35:732–44.

8.Kuroiwa T, Kajimoto Y, Ohta T. Development and clinical application of near-infrared surgical microscope: preliminary report. Minim Invasive Neurosurg 2001; 44(4):240–2.

9.Raabe A, Nakaji P, Beck J, et al. Prospective evaluation of surgical microscope-integrated intraoperative nearinfrared indocyanine green videoangiography during aneurysm surgery. J Neurosurg 2005; 103(6):982–9.

10.Kogure K, David NJ, Yamanouchi U, Choromokos E. Infrared absorption angiography of the fundus circulation. Arch Ophthalmol 1970; 83(2):209–14.

11.David NJ. Infrared absorption fundus angiography. In: proceedings of the International Symposium on Fluorescein Angiography. Albi, France, 1969.

12.Hochheimer BF. Angiography of the retina with indocyanine green. Arch Ophthalmol 1971; 86(5):564–5.

13.Flower RW, Hochheimer BF. Clinical infrared absorption angiography of the choroid. Am J Ophthalmol 1972; 73(3):458–9.

14.Flower RW. Infrared absorption angiography of the choroid and some observations on the effects of high intraocular pressures. Am J Ophthalmol 1972; 74(4): 600–14.

15.Flower RW, Hochheimer BF. A clinical technique and apparatus for simultaneous angiography of the separate retinal and choroidal circulations. Invest Ophthalmol Vis Sci 1973; 12(4):248–61.

16.Flower RW, Hochheimer BF. Indocyanine green dye fluorescence and infrared absorption choroidal angiography performed simultaneously with fluorescein angiography. Johns Hopkins Med J 1976; 138(2):33–42.

17.Hyvarinen L, Flower RW. Indocyanine green fluorescence angiography. Acta Ophthalmol (Copenh) 1980; 58(4):528–38.

18.Bischoff PM, Flower RW. Ten years experience with choroidal angiography using indocyanine green dye: a new routine examination or an epilogue? Doc Ophthalmol 1985; 60(3):235–91.

19.Hayashi K, Hasegawa Y, Tazawa Y, de Laey JJ. Clinical application of indocyanine green angiography to choroidal neovascularization. Jpn J Ophthalmol 1989; 33(1):57–65.

20.Hayashi K, Hasegawa Y, Tokoro T. Indocyanine green angiography of central serous chorioretinopathy. Int Ophthalmol 1986; 9(1):37–41.

21.Hayashi K, de Laey JJ. Indocyanine green angiography of submacular choroidal vessels in the human eye. Ophthalmologica 1985; 190(1):20–9.

22.Destro M, Puliafito CA. Indocyanine green videoangiography of choroidal neovascularization. Ophthalmology 1989; 96(6):846–53.

23.Scheider A, Schroedel C. High resolution indocyanine green angiography with a scanning laser ophthalmoscope. Am J Ophthalmol 1989; 108(4):458–9.

24.Scheider A. Indocyanine green angiography with an infrared scanning laser ophthalmoscope. Initial clinical experiences. Ophthalmologe 1992; 89(1):27–33.

25.Scheider A, Kaboth A, Neuhauser L. Detection of subretinal neovascular membranes with indocyanine green

and an infrared scanning laser ophthalmoscope. Am J Ophthalmol 1992; 113(1):45–51.

26.Patz A, Flower RW, Klein ML, Orth DH, Fleishman JA, MacLeod D. Clinical applications of indocyanine green angiography. Doc Ophthalmol Proc Ser 1976; 9:245–51.

27.Cherrick GR, Stein SW, Leevy CM, Davidson CS. Indocyanine green: observations on its physical properties, plasma decay, and hepatic extraction. J Clin Invest 1960; 39:592–600.

28.Baker KJ. Binding of sulfobromophthalein (BSP) sodium and indocyanine green (ICG) by plasma alpha-1 lipoproteins. Proc Soc Exp Biol Med 1966; 122(4):957–63.

29.Leevy CM, Bender J. Physiology of dye extraction by the liver: comparative studies of sulfobromophthalein and indocyanine green. Ann N Y Acad Sci 1963; 111:161–76.

30.Goresky CA. Initial distribution and rate of uptake of sulfobromophthalein in the liver. Am J Physiol 1964; 207:13–26.

31.Ketterer SG, Weigand BD. The excretion of indocyanine green and its use in the estimation of hepatic blood flow. Clin Res 1959; 7:71.

32.Ketterer SG, Weigand BD. Hepatic clearance of indocyanine green. Clin Res 1959; 7:289.

33.Probst P, Paumgartner G, Caucig H, Frauolich H, Grabner G. Studies on clearance and placental transfer of indocyanine green during labor. Clin Chim Acta 1970; 29:157.

34.Leevy CM, Smith F, Kiernan T. Liver function test. In: Bockus HL, ed. Gastroenterology. 3rd ed., Vol. 3. Philadelphia, PA: Saunders, 1976:68.

35.Shabetai R, Adolph RJ. Principles of cardiac catheterization. In: Fowler NO, ed. Cardiac Diagnosis and Treatment. 3rd ed. Hagerstown, MD: Harper & Row, 1980:117.

36.Hope-Ross M, Yannuzzi LA, Gragoudas ES, et al. Adverse reactions due to indocyanine green. Ophthalmology 1994; 101(3):529–33.

37.Bindewald A, Stuhrmann O, Roth F, et al. Lower limits of fluorescein and indocyanine green dye for digital cSLO fluorescence angiography. Br J Ophthalmol 2005; 89(12):1609–15.

38.Macular Photocoagulation Study Group. Occult choroidal neovascularization. Influence on visual outcome in patients with age-related macular degeneration. Arch Ophthalmol 1996; 114(4):400–12.

39.Guyer DR, Yannuzzi LA, Slakter JS, et al. Classification of choroidal neovascularization by digital indocyanine green videoangiography. Ophthalmology 1996; 103(12):2054–60.

40.Kleiner RC, Brucker AJ, Johnston RL. The posterior uveal bleeding syndrome. Retina 1990; 10(1):9–17.

41.Yannuzzi LA, Sorenson J, Spaide RF, Lipson B. Idiopathic polypoidal choroidal vasculopathy (IPCV). Retina 1990; 10(1):1–8.

42.Spaide RF, Yannuzzi LA, Slakter JS, Sorenson J, Orlach DA. Indocyanine green videoangiography of idiopathic polypoidal choroidal vasculopathy. Retina 1995; 15(2):100–10.

43.Phillips WB, II, Regillo CD, Maguire JI. Indocyanine green angiography of idiopathic polypoidal choroidal vasculopathy. Ophthalmic Surg Lasers 1996; 27(6):467–70.

44.Yannuzzi LA, Ciardella A, Spaide RF, Rabb M, Freund KB, Orlock DA. The expanding clinical spectrum of idiopathic polypoidal choroidal vasculopathy. Arch Ophthalmol 1997; 115(4):478–85.

45.Yannuzzi LA, Nogueira FB, Spaide RF, et al. Idiopathic polypoidal choroidal vasculopathy: a peripheral lesion. Arch Ophthalmol 1998; 116(3):382–3.

46.Moorthy RS, Lyon AT, Rabb MF, Spaide RF, Yannuzzi LA, Jampol LM. Idiopathic polypoidal choroidal vasculopathy of the macula. Ophthalmology 1998; 105(8):1380–5.

47.Schneider U, Gelisken F, Kreissig I. Indocyanine green angiography and idiopathic polypoidal choroidal vasculopathy. Br J Ophthalmol 1998; 82(1):98–9.

48.Yannuzzi LA, Wong DW, Sforzolini BS, et al. Polypoidal choroidal vasculopathy and neovascularized agerelated macular degeneration. Arch Ophthalmol 1999; 117(11):1503–10.

49.Yannuzzi LA, Freund KB, Goldbaum M, et al. Polypoidal choroidal vasculopathy masquerading as central serous chorioretinopathy. Ophthalmology 2000; 107(4):767–77.

50.Ahuja RM, Stanga PE, Vingerling JR, Reck AC, Bird AC. Polypoidal choroidal vasculopathy in exudative and haemorrhagic pigment epithelial detachments. Br J Ophthalmol 2000; 84(5):479–84.

51.Escano MF, Fujii S, Ishibashi K, Matsuo H, Yamamoto M. Indocyanine green videoangiography in macular variant of idiopathic polypoidal choroidal vasculopathy. Jpn J Ophthalmol 2000; 44(3):313–6.

52.Uyama M, Wada M, Nagai Y, et al. Polypoidal choroidal vasculopathy: natural history. Am J Ophthalmol 2002; 133(5):639–48.

53.Sho K, Takahashi K, Yamada H, et al. Polypoidal choroidal vasculopathy: incidence, demographic features, and clinical characteristics. Arch Ophthalmol 2003; 121(10): 1392–6.

54.Nishijima K, Takahashi M, Akita J, et al. Laser photocoagulation of indocyanine green angiographically identified feeder vessels to idiopathic polypoidal choroidal vasculopathy. Am J Ophthalmol 2004; 137(4):770–3.

55.Nakajima M, Yuzawa M, Shimada H, Mori R. Correlation between indocyanine green angiographic findings and histopathology of polypoidal choroidal vasculopathy. Jpn J Ophthalmol 2004; 48(3):249–55.

56.Lafaut BA, Aisenbrey S, Vanden Broecke C, BartzSchmidt KU. Clinicopathological correlation of deep retinal vascular anomalous complex in age related macular degeneration. Br J Ophthalmol 2000; 84(11):1269–74.

57.Kuhn D, Meunier I, Soubrane G, Coscas G. Imaging of chorioretinal anastomoses in vascularized retinal pigment epithelium detachments. Arch Ophthalmol 1995; 113(11):1392–8.

58.Slakter JS, Yannuzzi LA, Schneider U, et al. Retinal choroidal anastomoses and occult choroidal neovascularization in age-related macular degeneration. Ophthalmology 2000; 107(4):742–53 (discussion 753-44).

59.Fernandes LH, Freund KB, Yannuzzi LA, et al. The nature of focal areas of hyperfluorescence or hot spots imaged with indocyanine green angiography. Retina 2002; 22(5):557–68.

60.Yannuzzi LA, Negrao S, Iida T, et al. Retinal angiomatous proliferation in age-related macular degeneration. Retina 2001; 21(5):416–34.

61.Bottoni F, Massacesi A, Cigada M, Viola F, Musicco I, Staurenghi G. Treatment of retinal angiomatous proliferation in age-related macular degeneration: a series of 104 cases of retinal angiomatous proliferation. Arch Ophthalmol 2005; 123(12):1644–50.

62.Boscia F, Parodi MB, Furino C, Reibaldi M, Sborgia C. Photodynamic therapy with verteporfin for retinal angiomatous proliferation. Graefes Arch Clin Exp Ophthalmol 2006; 244(10):1224–32.

63.Nicolo M, Ghiglione D, Lai S, Calabria G. Retinal angiomatous proliferation treated by intravitreal triamcinolone and photodynamic therapy with verteporfin. Graefes Arch Clin Exp Ophthalmol 2006; 244(10):1336–8.

64.Schmidt-Erfurth U, Hasan T. Mechanisms of action of photodynamic therapy with verteporfin for the treatment of age-related macular degeneration. Surv Ophthalmol 2000; 45(3):195–214.

65.Treatment of age-related macular degeneration with photodynamic therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in age-related macular degeneration with verteporfin: one-year results of 2 randomized clinical trials–TAP report. Arch Ophthalmol 1999; 117(10):1329–45.

66.Lim JI, Sternberg P, Jr., Capone A, Jr., Aaberg TM, Sr., Gilman JP. Selective use of indocyanine green angio-

graphy for occult choroidal neovascularization. Am J Ophthalmol 1995; 120(1):75–82.

67.Baumal CR, Reichel E, Duker JS, Wong J, Puliafito CA. Indocyanine green hyperfluorescence associated with serous retinal pigment epithelial detachment in agerelated macular degeneration. Ophthalmology 1997; 104(5):761–9.

68.Yannuzzi LA, Hope-Ross M, Slakter JS, et al. Analysis of vascularized pigment epithelial detachments using indocyanine green videoangiography. Retina 1994; 14(2): 99–113.

69.Guyer DR, Yannuzzi LA, Slakter JS, Sorenson JA, HopeRoss M, Orlock DR. Digital indocyanine-green videoangiography of occult choroidal neovascularization. Ophthalmology 1994; 101(10):1727–35 (discussion 1735-7).

174OLIVER ET AL.

70.Chang B, Yannuzzi LA, Ladas ID, Guyer DR, Slakter JS, Sorenson JA. Choroidal neovascularization in second eyes of patients with unilateral exudative age-related macular degeneration. Ophthalmology 1995; 102(9):1380–6.

71.Chang TS, Freund KB, de la Cruz Z, Yannuzzi LA, Green WR. Clinicopathologic correlation of choroidal neovascularization demonstrated by indocyanine green angiography in a patient with retention of good vision for almost four years. Retina 1994; 14(2):114–24.

72.Watzke RC, Klein ML, Hiner CJ, Chan BK, Kraemer DF. A comparison of stereoscopic fluorescein angiography with indocyanine green videoangiography in age-related macular degeneration. Ophthalmology 2000; 107(8):1601–6.

73.Lee BL, Lim JI, Grossniklaus HE. Clinicopathologic features of indocyanine green angiography-imaged, surgically excised choroidal neovascular membranes. Retina 1996; 16(1):64–9.

74.Guyer DR, Yannuzzi LA, Ladas I, Slakter JS, Sorenson JA, Orlock D. Indocyanine green-guided laser photocoagulation of focal spots at the edge of plaques of choroidal neovascularization. Arch Ophthalmol 1996; 114(6):693–7.

75.Helbig H, Niederberger H, Valmaggia C, Bischoff P. Simultaneous fluorescein and indocyanine green angiography for exudative macular degeneration. Klin Monatsbl Augenheilkd 2005; 222(3):202–5.

76.Yanagi Y, Tamaki Y, Sekine H. Fluorescein angiographyguided indocyanine green angiography for the detection of feeder vessels in subfoveal choroidal neovascularization. Eye 2004; 18(5):474–7.

77.Sorenson JA, Yannuzzi LA, Slakter JS, Guyer DR, Ho AC, Orlock DA. A pilot study of digital indocyanine green videoangiography for recurrent occult choroidal neovascularization in age-related macular degeneration. Arch Ophthalmol 1994; 112(4):473–9.

78.Reichel E, Pollock DA, Duker JS, Puliafito CA. Indocyanine green angiography for recurrent choroidal neovascularization in age-related macular degeneration. Ophthalmic Surg Lasers 1995; 26(6):513–8.

79.Chen CJ, Chen LJ, Miller KR. Clinical significance of postlaser indocyanine green angiographic hot spots in age-related macular degeneration. Ophthalmology 1999; 106(5):925–9 (discussion 929–31).

80.Photodynamic Therapy Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization-verteporfin in photodynamic therapy report 2. Am J Ophthalmol 2001; 131(5): 541–60.

81.Gragoudas ES, Adamis AP, Cunningham ET, Jr., Feinsod M, Guyer DR. Pegaptanib for neovascular agerelated macular degeneration. N Engl J Med 2004; 351(27):2805–16.

82.Avery RL, Pieramici DJ, Rabena MD, Castellarin AA, Nasir MA, Giust MJ. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006; 113(3):363–72. e5.

83.Michels S, Rosenfeld PJ, Puliafito CA, Marcus EN, Venkatraman AS. Systemic bevacizumab (Avastin) therapy for neovascular age-related macular degeneration twelve-week results of an uncontrolled open-label clinical study. Ophthalmology 2005; 112(6):1035–47.

84.Spaide RF, Laud K, Fine HF, et al. Intravitreal bevacizumab treatment of choroidal neovascularization secondary to age-related macular degeneration. Retina 2006; 26(4):383–90.

85.Heier JS, Antoszyk AN, Pavan PR, et al. Ranibizumab for treatment of neovascular age-related macular degeneration: a phase I/II multicenter, controlled, multidose study. Ophthalmology 2006; 113(4):642. e1–4.

86.Rosenfeld PJ, Heier JS, Hantsbarger G, Shams N. Tolerability and efficacy of multiple escalating doses of ranibizumab (Lucentis) for neovascular age-related macular degeneration. Ophthalmology 2006; 113(4):632.e1.

87.Slakter JS, Yannuzzi LA, Sorenson JA, Guyer DR, Ho AC, Orlock DA. A pilot study of indocyanine green videoan- giography-guided laser photocoagulation of occult choroidal neovascularization in age-related macular degeneration. Arch Ophthalmol 1994; 112(4):465–72.

88.Regillo CD, Benson WE, Maguire JI, Annesley WH, Jr. Indocyanine green angiography and occult choroidal neovascularization. Ophthalmology 1994; 101(2):280–8.

89.Lim JI, Aaberg TM, Capone A, Jr., Sternberg P, Jr. Indocyanine green angiography-guided photocoagulation of choroidal neovascularization associated with retinal pigment epithelial detachment. Am J Ophthalmol 1997; 123(4):524–32.

90.Da Pozzo S, Parodi MB, Ravalico G. A pilot study of ICG-guided laser photocoagulation for occult choroidal neovascularization presenting as a focal spot in agerelated macular degeneration. Int Ophthalmol 2001; 24(4):187–94.

91.Ho AC, Yannuzzi LA, Guyer DR, Slakter JS, Sorenson JA, Orlock DA. Intraretinal leakage of indocyanine green dye. Ophthalmology 1994; 101(3):534–41.

92.Reichel E, Puliafito CA, Duker JS, Guyer DR. Indocyanine green dye-enhanced diode laser photocoagulation of poorly defined subfoveal choroidal neovascularization. Ophthalmic Surg 1994; 25(3):195–201.

93.Obana A, Gohto Y, Nishiguchi K, Miki T, Nishi S, Asada A. A retrospective pilot study of indocyanine green enhanced diode laser photocoagulation for subfoveal choroidal neovascularization associated with age-related macular degeneration. Jpn J Ophthalmol 2000; 44(6):668–76.

94.Arevalo JF, Mendoza AJ, Fernandez CF. Indocyanine green-mediated photothrombosis with and without intravitreal triamcinolone acetonide for subfoveal choroidal neovascularization in age-related macular degeneration: a pilot study. Retina 2005; 25(6):719–26.

95.Arevalo JF, Garcia RA, Mendoza AJ. Indocyanine greenmediated photothrombosis with intravitreal triamcinolone acetonide for subfoveal choroidal neovascularization in age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol 2005; 243(11):1180–5.

96.Maberley DA, Cruess AF. Indocyanine green angiography: an evaluation of image enhancement for the identification of occult choroidal neovascular membranes. Retina 1999; 19(1):37–44.

97.Spaide RF, Orlock DA, Herrmann-Delemazure B, et al. Wide-angle indocyanine green angiography. Retina 1998; 18(1):44–9.

98.Staurenghi G, Viola F, Mainster MA, Graham RD, Harrington PG. Scanning laser ophthalmoscopy and angiography with a wide-field contact lens system. Arch Ophthalmol 2005; 123(2):244–52.

99.Miki T, Shiraki K, Kohno T, Moriwaki M, Obana A. Computer assisted image analysis using the subtraction

method in indocyanine green angiography. Eur J Ophthalmol 1996; 6(1):30–8.

100.Spaide RF, Orlock D, Yannuzzi L, et al. Digital subtraction indocyanine green angiography of occult choroidal neovascularization. Ophthalmology 1998; 105(4):680–8.

101.Matsumoto M, Shiraki K, Obana A. Detection of choroidal neovascularization by subtraction indocyanine green angiography. Osaka City Med J 2003; 49(2):85–91.

102.Staurenghi G, Orzalesi N, La Capria A, Aschero M. Laser treatment of feeder vessels in subfoveal choroidal neovascular membranes: a revisitation using dynamic indocyanine green angiography. Ophthalmology 1998; 105(12):2297–305.

103.Hanutsaha P, Guyer DR, Yannuzzi LA, et al. Indocyaninegreen videoangiography of drusen as a possible predictive indicator of exudative maculopathy. Ophthalmology 1998; 105(9):1632–6.

104.Pauleikhoff D, Spital G, Radermacher M, Brumm GA, Lommatzsch A, Bird AC. A fluorescein and indocyanine

green angiographic study of choriocapillaris in age-related macular disease. Arch Ophthalmol 1999; 117(10):1353–8.

105.Gelisken F, Inhoffen W, Schneider U, Stroman GA, Kreissig I. Indocyanine green videoangiography of occult choroidal neovascularization: a comparison of scanning laser ophthalmoscope with high-resolution digital fundus camera. Retina 1998; 18(1):37–43.

106.Freund KB, Yannuzzi LA, Sorenson JA. Age-related macular degeneration and choroidal neovascularization. Am J Ophthalmol 1993; 115(6):786–91.

107.Stanga PE, Lim JI, Hamilton P. Indocyanine green angiography in chorioretinal diseases: indications and interpretation: an evidence-based update. Ophthalmology 2003; 110(1):15–21 (quiz 22–3).

The widespread use of optical coherence tomography (OCT) has changed the way ophthalmologists evaluate and treat age-related macular degeneration (AMD). OCT has been added to the armamentarium of macular imaging that now includes color fundus photography, fluorescein angiography (FA), and indocyanine green angiography. Although each of these modalities is important in the management of macular degeneration, OCT provides useful information regarding retinal structure. Cross-sectional imaging with commercially available units gives an axial resolution of 10 to 15 mm, and ultrahigh resolution OCT (UHR-OCT) provides 2- to 3-mm resolution. This data can be used in the diagnosis and management of AMD; reliance on OCT as part of the decision making process in treatment with different antivascular endothelial growth factor (VEGF) agents is rapidly becoming standard of care.

OCT IMAGING PRINCIPLES AND

THE NORMAL OCT IMAGE

The different layers of the retina have a characteristic appearance on the OCT scan. The principle used in creating OCT images is Michelson interferometry, which uses the property of light passing through the eye and producing different reflections from different cell layers. A split beam of infrared light in commercially available OCT units reflects light from the layers of the retina and interacts with light reflected from a reference mirror (1). The interference pattern produced is digitally processed and the false-color map shown on the scan is a representation of the reflection characteristics of the cells in each layer, with hyperreflective areas being bright and hyporeflective areas being dark. When compared to in vivo specimens, layers of relative high reflectivity correspond to horizontally aligned retinal components (2).

The innermost layers of the retina, the nerve fiber layer, and ganglion cells, are bright on the false-color map. As the scan progresses to the outer retina, the

more densely packed nuclear layers are hyporeflective and dark, and the horizontally oriented plexiform layers are hyperreflective and bright. In the outermost retina, the photoreceptors are hyporeflective, although the junction between the inner segment (IS) and outer segment (OS), as best visualized with UHR-OCT, is bright. The outermost layers, those further distal than the OS of the photoreceptors, are hyperreflective and bright, and pathologic studies have shown the tissues responsible for this signal are the basement membrane of the retinal pigment epithelium (RPE) and the inner choroid (3).

The OCT image also reflects the topographical shape of the normal retina, and the OCT software can combine the topographical radial scans to create a volumetric measurement of retinal tissue. The retinal thickness is displayed on a separate false color map, with cool colors (greens and blues) representing areas of less retinal volume, and warm colors (yellows and reds) representing areas of greater retinal volume. The normal foveal depression of the umbo is visible on topographic and volumetric OCT, and the quantitative thickness of a normal fovea can be compared with known approximate normal central field (central 1-mm diameter) thickness, which is about 175 mm (4,5).

OCT in Assessment of Choroidal Neovascularization in AMD

OCT can help characterize retinal pathology, even when this information is difficult to discern on clinical examination or angiography. It is possible to define the location of choroidal neovascular membranes above or below the RPE. Solid fibrous tissue can be differentiated from subretinal fluid when these findings may be angiographically equivocal. Other features of AMD, including cystoid macular edema (CME), drusenoid RPE detachments and RPE tears can be imaged by OCT (6).

Optical coherence is useful for quantitative assessment of retinal thickness and subretinal fluid when associated with choroidal neovascularization (CNV). A characteristic appearance of CNV has also

Figure 1 Color picture of choroidal neovascularization in a patient’s left eye.

been described, consisting of thickening and fragmentation of the reflective layer corresponding to the RPE and choriocapillaris (Figs. 1–3) (7). The extent and location of subretinal fluid associated with CNV can be used to assess whether the pathology is subfoveal, as long as there is preservation of some foveal architecture (8).

As noted on clinical examination and FA, CME is frequently associated with CNV in wet AMD (9). However, the presence of CME may be difficult to definitively diagnose through those modalities alone. In addition to imaging subretinal fluid, OCT is effective in identifying intraretinal edema, compared to

Figure 3 Cross-sectional optical coherence tomography appearance of eye in Figures 1 and 2 revealing hyporeflective subretinal fluid and thickening of adjacent layer just anterior to the intensely hyperreflective retinal pigment epithelium. There is essentially preservation of the normal foveal anatomy and only a minimal increase in retinal thickness.

both clinical stereoscopic images (10) and FA (3,11). The appearance of CME on OCT images is seen as hyporeflective, dark spaces within retinal tissue. Its presence is important clinically, since CME as seen on OCT scan in wet AMD correlates with decreased visual acuity (Figs. 4–7) (9).

RPE detachments and sub-RPE neovascularization has been associated with occult CNV in AMD as defined histopathologically. This finding has been corroborated in studies using OCT imaging. In one series, new neovascular AMD lesions were characterized as occult or classic according to their angiographic findings (12). OCT scans of those same lesions revealed subretinal opacities separate from the RPE present in over 87% of lesions characterized as classic and only 13% of lesions characterized as occult. OCT findings consistent with RPE detachment were present in none of the lesions characterized as classic and in one-third of those characterized as

Figure 2 Fluorescein angiography of the eye in Figure 1 exhibiting classic choroidal neovascularization.

Figure 4 Transit phase fluorescein angiography of a right eye exhibiting a large neovascular membrane.

Figure 5 Recirculation phase fluorescein angiography of the same eye as Figure 4 revealing late leakage and cystoid macular edema from the membrane.

occult (13). The relationship of CNV and RPE detachments has been further elucidated in another OCT series, and a pattern of double RPE detachments separated by a notch, as well as highly reflective tissue beneath the dome of the RPE detachment have been correlated with CNV that was observed on angiography (Figs. 8–11) (14).

OCT in Assessment of Non-exudative Macular Degeneration

Although the utility of commercially available OCT scanning in the clinical diagnostic setting has been examined mostly for neovascular AMD, there are characteristic OCT findings in nonneovascular forms of AMD. “Drusenoid” RPE detachments, which appear angiographically as staining of drusen,

258

360

210 295 407 382 274

316

225

Microns

Figure 7 Topographical optical coherence tomography map of the lesion shown in Figures 4–6, showing marked thickening of retinal tissue.

are documented by OCT as elevations of the pigment epithelium itself (Figs. 12–14) (15).

The UHR-OCT is an experimental imaging modality which builds upon the interferometry principles of commercially available OCT imaging devices. The standard OCT axial resolution of 10 to 15 mm is improved to 3 mm using a femtosecond titanium-sapphire laser with a broader bandwidth. This improvement in resolution allows for better delineation and characterization of changes within the intraretinal layers, especially the IS and OS photoreceptors and the IS–OS junction (also known as the external limiting membrane) (16,17), which are the site of many early changes in AMD. In UHR-OCT, the outermost retina is seen as the hyperreflective RPE underlying hyporeflective OS photoreceptors, which

Figure 6 Cross-sectional optical coherence tomography (OCT) of the lesion in Figures 4 and 5, revealing large hyporeflective spaces typical of cystoid edema. It is present with thickening and fragmentation of the hyperreflective layer corresponding to the retinal pigment epithelium, which is the characteristic OCT appearance of choroidal neovascularization.

Figure 8 Color picture of a left eye showing a complex neovascular membrane. Lesion components include a large retinal pigment epithelium detachment, subretinal blood, and subretinal exudate.

Figure 9 Transit phase of the eye shown in Figure 12. There is some early filling of a superotemporal retinal pigment epithelium detachment with relative hypofluorescence of the central, neovascular component of the lesion.

are in turn distinguished from hyporeflective IS photoreceptors by the intensely hyperreflective IS–OS junction.

UHR-OCT has been used to evaluate eyes with dry AMD and subtle patterns associated with drusen have been observed. One pattern shows changes similar to those seen with drusen in commercially available OCT, with the RPE excrescences overlying reflective material consistent with drusen. A second patter seen on UHR-OCT in dry AMD is a saw-toothed

Figure 11 Cross-sectional optical coherence tomography of the lesion in Figures 12–14. There is diffuse retinal thickening and numerous hyporeflective spaces consistent with severe cystoid macular edema. There is a small amount of subretinal fluid. The retinal pigment epithelium (RPE) detachment is very apparent, underlying the retina and the hyperreflective band corresponding to the RPE.

configuration or bunching of the RPE. This pattern is seen associated with atrophy of the inner and outer photoreceptors and the outer nuclear layer, although there is no change in retinal tissue further inward (Fig. 15A–C). The third UHR-OCT pattern in dry AMD shows discrete nodular drusen which actually disrupt as opposed to distort the RPE and are associated with collections of reflective material. This third pattern corresponds to large hard drusen that are observed on clinical examination. Patients with dry AMD can have all three patterns present, or a variety of combinations. A very small percentage of patients in the series of dry AMD eyes studied with UHR-OCT were noted to have findings consistent with early CNV, despite no clinical or angiographic evidence

Figure 10 Late recirculation phase of the eye shown in Figures 12 and 13. There has been filling of the retinal pigment epithelium detachment, as well as filling of cystoid spaces in the center of the macula. There is leakage from the central and inferior occult choroidal neovascularization and blockage from the subretinal blood.

Figure 12 Color picture of a right eye showing moderate pigment atrophy centrally and confluent soft drusen.

Figure 13 Recirculation phase fluorescein angiography of the eye shown in Figure 12. There is staining of the soft drusen and transmission of dye fluorescence through the retinal pigment epithelium atrophy.

of CNV. This finding is particularly interesting, as very early diagnosis of neovascular AMD may be facilitated by UHR imaging (18).

Geographic atrophy has also been examined, both with conventional OCT and UHR-OCT. A typical pattern of geographic atrophy consisting of enhanced reflectivity of the choroid and significant thinning of overlying retinal tissue has been described using conventional OCT. The bright, well demarcated signal of the choroid is felt to be due to loss of the RPE, a common feature of geographic atrophy that has been well described in histopathologic studies (19). These findings are similar to those seen on UHR-OCT, with associated generalized retinal thinning and increased reflectivity from tissue below the absent RPE, which is likely choriocapillaris (19).

Figure 14 Cross-sectional optical coherence tomography of the eye shown in Figures 13 and 14. There are sub-RPE hyporeflective spaces typical of “drusenoid” RPE detachments. There is preservation of the retinal architecture. Of note, there is a mild epiretinal membrane.

OCT in the Treatment of AMD

The ability of OCT to accurately map and quantitate retinal findings has added an entirely new dimension for monitoring the response of neovascular AMD to treatment. Prior to OCT, qualitative clinical examination and FA were the only means by which one could assess disease progression, stability, or regression. In addition to the quantitative accuracy of OCT, its noninvasive nature, and absent risk of allergic response, OCT appears to be surpassing the use of angiographic techniques as the test of choice for visit- to-visit monitoring of neovascular AMD. (See also Chapter 11 on quantitative imaging).

Visudyne photodynamic therapy (PDT), which has been a Food and Drug Administration (FDA) approved treatment for neovascular AMD since 2000, was the first new AMD treatment introduced in the era of OCT, and the first pharmacologic treatment for CNV secondary to AMD. OCT findings 6- and 12-month following the initiation of CNV treatment with PDT show macular thickness declining after treatment, but did not show a decrease in the thickness of the CNV following treatment. However, that same study also found that OCT has an excellent sensitivity but only a fair specificity for monitoring CNV activity, though it served as a useful adjunct for verifying intraretinal or subretinal fluid, especially when angiography was inconclusive (20). Many of PDT effects have been shown by OCT, including a transient increase in intraretinal and subretinal fluid in the first week after treatment (21), which has also been described as an early response stage in an OCT grading system developed to monitor treatment by PDT and help clarify angiographic changes seen after treatment (22).

Pegaptanib (Macugenw) is a new class of drug for neovascular AMD, a molecular agent known as an aptamer, which targets VEGF isoform 165. OCT has been used to evaluate patients’ response to this therapy. Since the introduction of pegaptanib treatment, OCT has been used to document complications of therapy, and RPE rips have been documented with OCT after a single pegaptanib treatment of CNV with turbid pigment epithelial detachments (23,24). (It is important to note that RPE rips have also been documented with thermal laser treatment (25) and PDT treatment (26) of CNV, as well as other anti-VEGF agents including bevacizumab and ranibizumab.) Evaluation of eyes treated with a single intravitreal injection of pegaptanib showed that there was no difference in OCT anatomy compared to baseline (27).

OCT has played a pivotal role in the introduction of VEGF inhibitors in the treatment of neovascular AMD. Following the initial treatment of neovascular AMD with systemic bevacizumab (28), a full-length antibody approved by the FDA for the treatment of

182 EICHENBAUM AND REICHEL

200

μm

Figure 15 (A) Color photo of a left eye with retinal pigment epithelium (RPE) changes and pigment clumping. (B) Ultrahigh resolution optical coherence tomography (UHR-OCT) showing thickening and bunching of the RPE. There is a loss of both the outer segment (OS) and inner segment (IS) of the photoreceptors, with a fragmenting of the IS–OS junction. Intraretinal pigment migration is noted by the arrow. (C) Stratus OCT showing less distinction of intraretinal pigment migration. The photoreceptors cannot be distinguished, although the thickening of the RPE remains apparent. Abbreviations: IS, inner segment; OS, outer segment Source: From Ref. 18.

colon cancer, studies of intravitreal bevacizumab emerged in the ophthalmic literature supporting this off-label treatment. The results of intravitreal bevacizumab were also striking, with initially a single case report (29) followed by small prospective studies showing OCT evidence of improvement or resolution of intraretinal, subretinal, and sub-RPE fluid in a large percentage of patients as early as four weeks

after treatment that was also associated with improvement in visual acuity. Treatment effects persisted up to 12 weeks after treatment (30–32). In all of these studies, OCT results supported anatomical efficacy of treatment.

The PrONTO study, a small, single center, prospective, nonrandomized efficacy study, showed that patients given a “loading dose” of ranibizumab

Figure 16 Optical coherence tomography of the baseline lesion, showing hyperreflective retinal pigment epithelium (RPE) detachments, thickening of the tissue adjacent to the RPE, and intraretinal hyporeflective fluid with thickening of retinal tissue. The first treatment with intravitreal ranibizumab is administered.

Figure 17 Optical coherence tomography one month after the fourth ranibizumab treatment. There is complete resolution of the intraretinal and sub-retinal pigment epithelium (RPE) fluid, no subretinal scarring, and a normal contour to the retinal tissue. Note that the RPE band has a normal thickness to it.

monthly for three months and then given a customized dosing scheme based upon visual acuity and OCT findings had equivalent efficacy compared to monthly dosing with ranibizumab after one year of follow-up (33). This study is remarkable because it allowed a variable dosing regimen for all types of lesions at the discretion of the treating physician that was primarily guided by OCT. The following sequence (Figs. 16 and 17) is an example of a ranibizumab treatment of CNV with OCT used as guidance to retreat.

CONCLUSIONS

OCT has the ability to verify the presence of neovascular disease and quantitatively evaluate treatment response. The utility of OCT as a guide for treatment is suggested by early clinical experience from many centers and results from the PrONTO study. Further studies using UHR-OCT may determine the role of screening high risk eyes for the early detection of neovascular AMD. OCT has been utilized as supportive data in some trials for the treatment of macular degeneration, though its incorporation as a key endpoint in the evaluation of treatments for neovascular AMD has yet to be conclusively established. OCT will continue to be a key component in the evaluation, treatment and development of new therapeutic modalities in patients with AMD.

SUMMARY POINTS

&Cross-sectional imaging with commercially available units give an axial resolution of 10 to 15 mm, and UHR-OCT provides 2- to 3-mm resolution.

&OCT is useful to verify the presence of neovascular disease and quantitatively evaluate treatment response as suggested by clinical trials experience.

&Further studies using UHR-OCT may determine the role of screening high risk eyes for the early detection of neovascular AMD.

&OCT will continue to be a key component in the evaluation, treatment and development of new therapeutic modalities in patients with AMD.

REFERENCES

1.Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science 1991; 254:1178–81.

2.Toth CA, Narayan DG, Boppart SA, et al. A comparison of retinal morphology viewed by optical coherence tomography and by light microscopy. Arch Ophthalmol 1997; 115:1425–8.

3.Ghazi N, Dibernardo C, Ying HS, Mori K, Gehlbach PL. Optical coherence tomography of enucleated human eye specimens with histological correlation: origin of the outer “red line”. Am J Ophthalmol 2006; 141:719–26.

4.Hee MR, Puliafito CA, Duker JS, et al. Topography of diabetic macular edema with optical coherence tomography. Ophthalmology 1998; 105:360–70.

5.Massin P, Erginay A, Haouchine B, Mehidi AB, Paques M, Gaudric A. Retinal thickness in healthy and diabetic subjects measured using optical coherence tomography mapping software. Eur J Ophthalmol 2002; 12:102–8.

6.Voo I, Mavrofrides EC, Puliafito CA. Clinical applications of optical coherence tomography for the diagnosis and management of macular diseases. Ophthalmol Clin North Am 2004; 17:21–31.

7.Hee MR, Baumal CR, Puliafito CA, et al. Optical coherence tomography of age-related macular degeneration and choroidal neovascularization. Ophthalmology 1996; 103:1260–70.

8.Jaffe GJ, Caprioli J. Optical coherence tomography to detect and manage retinal disease and glaucoma. Am J Opthalmol 2004; 137:156–69.

9.Ting TD, Oh M, Cox TA, Meyer CH, Toth CA. Decreased visual acuity associated with cystoid macular edema in neovascular age-related macular degeneration. Arch Ophthalmol 2002; 120:731–7.

10.Strom C, Sander B, Larsen N, Larsen M, Lund-Andersen H. Diabetic macular edema assessed with optical coherence tomography and stereo fundus photography. Invest Ophthalmol Vis Sci 2002; 43:241–5.

11.Antcliff RJ, Stanford MR, Chauhan DS, et al. Comparison between optical coherence tomography and fundus fluorescein angiography for the detection of cystoid macular edema in patients with uveitis. Ophthalmology 2000; 107:593–9.

12.Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfinone-year results of 2 randomized clinical trials-TAP report. Arch Ophthalmol 1999; 117:1329–45.

13.Hughes EH, Khan J, Patel N, Kashani S, Chong NV. In vivo demonstration of the differences between classic and occult neovascularization using optical coherence tomography. Am J Ophtalmol 2005; 139:344–6.

14.Sato T, Iida T, Hagimura N, Kishi S. Correlation of optical coherence tomography with angiography in retinal pigment epithelial detachments associated with age-related macular degeneration. Retina 2004; 24:910–4.

15.Emfietzoglou I, Grigoropoulos V, Kipioti A, Alimisi S, Theodossiadis PG, Theodossiadis GP. Optical coherence tomography appearance of “drusenoid” pigment epithelial detachments. Ophthalmic Surg Lasers Imaging 2005; 36:147–50.

16.Drexler W, Morgner U, Ghanta RK, Ka¨rtner FX, Schuman JS, Fujimoto JG. Ultrahigh-resolution optical coherence tomography. Nat Med 2001; 7:502–7.

17.Drexler W, Sattmann H, Hermann B, et al. Enhanced visualization of macular pathology with the use of ultra- high-resolution optical coherence tomography. Arch Ophthalmol 2003; 121:695–706.

18.Pieroni CG, Witkin AJ, Ko TH, et al. Ultrahigh resolution optical coherence tomography in non-exudative age related macular degeneration. Br J Ophthalmol 2006; 90:191–7.

19.Hassenstein A, Ru¨hl R, Richard G. Optical coherence tomography in geographic atrophy—a clinicopathological study. Klin Monatsbl Augenheilkd 2001; 218:503–9.

184 EICHENBAUM AND REICHEL

20.Salinas-Alama´n A, Garcı´a-Layana A, Maldonado MJ, SainzGo´mez C, Alva´rez-Vidal A. Using optical coherence tomography to monitor photodynamic therapy in age related macular degeneration. Am J Ophthalmol 2005; 140:23–8.

21.Ozdemir H, Karacorlu SA, Karacorlu M. Early optical coherence tomography changes after photodynamic therapy in patients with age related macular degeneration. Am J Ophthalmol 2006; 141:574–6.

22.Rogers AH, Martidis A, Greenberg PB, Puliafito CA. Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization. Am J Ophthalmol 2002; 134:566–76.

23.Dhalla MS, Blinder KJ, Tewari A, Hariprasad SM, Apte RS. Retinal pigment epithelial tears following intravitreal pegaptanib sodium. Am J Ophthalmol 2006; 141:752–4.

24.Singh RP, Sears JE. Retinal pigment epithelial tears after pegaptanib injection for exudative age-related macular degeneration. Am J Ophthalmol 2006; 142:160–2.

25.Gass JDM. Retinal pigment epithelial rip during krypton red laser photocoagulation. Am J Ophthalmol 1984; 98:700–6.

26.Gelisken F, Inhoffen W, Partsch M, Schneider U, Kreissig I. Retinal pigment epithelial tear after photodynamic therapy for choroidal neovascularization. Am J Ophthalmol 2001; 131:518–20.

27.Schuman S, Rogers AH, Duker JS, Reichel E, Baumal CR. Six-week outcomes after pegaptanib. Ophthalmology 2006; 113:501.

28.Michels S, Rosenfeld PJ, Puliafito CA, Marcus EN, Venkatraman AS. Systemic bevacizumab (Avastin) therapy for neovascular age-related macular degeneration: twelveweek results of an uncontrolled open-label clinical study. Ophthalmology 2005; 112:1035–47.

29.Rosenfeld PJ, Moshfeghi AA, Puliafito CA. Optical coherence tomography findings after an intravitreal injection of bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmic Surg Lasers Imaging 2005; 36:331–5.

30.Avery RL, Pieramici DJ, Rabena MD, Castellarin AA, Nasir MA, Giust MJ. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006; 113:363–72.

31.Spaide RF, Laud K, Fine HF, et al. Intravitreal bevacizumab treatment of choroidal neovascularization secondary to age-related macular degeneration. Retina 2006; 26:383–90.

32.Rich RM, Rosenfeld PJ, Puliafito CA, et al. Short-term safety and efficacy of intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Retina 2006; 26:495–511.

33.Rosenfeld PJ. An OCT-guided variable-dosing regimen with Lucentis(TM) (ranibizumab) in neovascular AMD: one year results from the PrONTO study. In: Program and abstracts of the 24th Annual American Society of Retina Specialists and 6th Annual European Vitreoretinal Society Meeting. Cannes, France, September 9–13, 2006.

INTRODUCTION

Historically, clinical ophthalmology developed rapidly as the transparency of the ocular media permitted direct examination of the intraocular structures. The development of methods to examine the fundus of the eye considerably improved the understanding of retinal anatomy, structure, and circulation. Consequently, great advantages were gained in the diagnosis of retinal disease and comprehension of systemic disorders.

The development of innovative imaging techniques such as fluorescein and indocyanine green angiography followed by tomographic image capture has further improved the examination of the retina and its vasculature. These advances have allowed ophthalmologists to better describe retinal diseases and to develop classification systems in areas such as diabetic retinal disease and age-related macular degeneration (AMD). Investigators, for example, have developed systematic grading systems based on stereoscopic examination of color fundus and angiographic images to characterize disease progression and response to therapy (1,2). However, these classification systems are descriptive and categorical. By contrast, other medical specialties have made enormous strides in the development of quantitative methods to both describe and categorize diseases and to assess the efficacy of intervention. For example, an internist can use quantifiable markers such as blood pressure, cholesterol, and cardiac ejection fraction to predict patients at highest risk for a cardiovascular event and then offer preventive

treatments. Likewise, neurologists can image the brain with precision and accuracy to identify volumetric changes in tumor size and contour.

Similar advances in retinal imaging are clearly possible and a number of quantitative methods are currently being developed. To understand the complexities of the tissue layers and vasculature that can be imaged, it is necessary to review the various imaging modalities and techniques for interpretation that are currently available.

FUNDUS PHOTOGRAPHY

Color fundus photography has been a significant tool in the documentation of macular disease for many years (1,2). Clinical trials have used photographic documentation to ensure quality assurance and adherence to standards. Advances in image capture, allowing the transition from film-based photography to high-resolution digital acquisition, have resulted in many advantages. First, retrieval of information is easier. Second, computer-assisted image analysis can be easily applied. Third, parametric descriptors can be developed. All these contribute toward providing greater accuracy, objectivity, and reproducibility.

Application of quantitative techniques to evaluate eyes with AMD is becoming a reality. Quantitative techniques are useful for evaluating the type and number of drusen and retinal pigment epithelium pigmentary changes, both of which are key features of early AMD (3,4). The careful analysis of drusen number, size, area, and morphology allows clinicians to assess

186 ESMAILI ET AL.

disease severity and may aid in predicting conversion from non-neovascular to neovascular forms (5).

The goals of quantifying macular pathology such as drusen are multifactorial. Quantified data provide objective information that can be compared over time, giving an index of progression. Such data can be used to evaluate the response to treatment as well as to grade disease progression. Quantification of drusen would also allow for better designed epidemiological and clinical studies of the natural history of AMD.

Manual grading is the oldest form of photographic quantitation and is accurate, although labor and resource intensive (1,2). Currently, several semiautomated and automated modalities have been developed to address the shortcomings of manual grading (6–9). Shin and coworkers describe a supervised method of automated drusen grading with good correlation to manual grading, with interclass correlation coefficients of 0.92 and 0.93 (10). Semiquantitative methods have also been shown to have good interobserver reproducibility by graders at different institutions (11).

Several limitations have prevented the widespread adoption of such technology. To date, no automated system free of observer supervision exists that can accurately and reproducibly assess drusen burden. Obstacles that have limited progress in this field include difficulties in distinguishing features such as soft drusen with indistinct borders from other pale retinal lesions. The presence of branching and crossing points in retinal blood vessels also detract from detection of drusen as the former produce dark haloes around the lighter retinal background. This phenomenon causes lesion thresholds to be calculated below retinal backgrounds, thus creating a propensity for the false detection of drusen (8).

A major technical obstacle has been accounting for differences in macular reflectivity (6). As one moves toward the central macula, normal background reflectivity decreases in intensity. Thus, if there existed identical soft drusen, with one located in the central macula and the other parafoveally, the difference in normal background reflectance would affect the threshold at which the drusen could be detected. Semi-automated methods based on the geometry of fundus reflectance as well as newer, more automated technology utilizing mathematical modeling to reconstruct and then level the background reflectance are being explored to overcome such limitations (6,12).

The lack of true automation may limit the utility of drusen quantification in clinical practice. However, an automated method that is precise and reproducible would allow large-scale population-based studies to be performed with fewer resource implications. Furthermore, the technology could also be applied in trials assessing therapeutic interventions.

OPTICAL COHERENCE TOMOGRAPHY

Optical coherence tomography (OCT) was first described by Huang and colleagues as a method of utilizing near infrared light to provide a noninvasive means of evaluating ocular structures including the retina (13). In many ways, OCT is an ideal modality for obtaining quantifiable data for the evaluation of retinal pathology (14). The current generation of OCT scanners is able to provide topographic numerical determinations of retinal thickness, which in recent years has become an increasingly popular means of evaluating disease and, particularly, the localization of fluid within the different tissue layers of the fundus (15,16).

The role of OCT in quantifying changes due to AMD has yet to be established. This may partly be due to the fact that OCT is a relatively new modality, and that until recent years, the drive to create a numerical index of disease was perceived as irrelevant since treatment options were less than satisfactory. Currently, new therapeutic modalities leading to impressive improvements in visual acuity have been introduced in the management of neovascular AMD. OCT is now recognized as having an important role in the measurement of retinal thickness and subretinal pathology in monitoring choroidal neovascularization (CNV) activity after appropriate therapy (17). With an increasing number of treatment options available including combination therapies, a quantifiable understanding of the therapeutic response will be needed to allow meaningful comparisons of the morphological and functional outcomes.

The potential role of quantitative OCT in neovascular AMD is substantial. Automated image analysis algorithms are being developed to better define retinal layers (18). Software like the OCTOR system developed at the Doheny Eye Institute allows for delineation of retinal layers as well as pathologic entities such as subretinal fluid, area occupied by choroidal neovascular membranes, and the size of serous retinal pigment epithelial detachments. This software is currently employed as a research tool to evaluate the efficacy of various treatment modalities in neovascular AMD.

FLUORESCEIN ANGIOGRAPHY

Fluorescein angiography (FA) is a well-established modality for assessment of neovascular AMD. Quantitative FA can also provide similar benefits as OCT in the management and treatment of AMD. Current software allows for the measurement of neovascular membrane size, which is a parameter that can influence the therapeutic outcome. One such example is the use of photodynamic therapy (PDT) in the treatment

of occult CNV. Optimal outcomes were detected in eyes with lesions smaller than four disc diameters, while larger lesions were associated with a worse outcome than natural history (19).

Semi-automated detection and quantification of hyperfluorescent leakage has been described by Phillips et al., who manipulated digitized fluorescein angiograms via gradient threshold and regiongrowing techniques to detect leakage (20). This method superimposes paired images (early and late angiograms) to produce a composite with the area of leakage mapped onto a frame. A numerical value can then be determined to describe a total edema value.

Chakravarthy et al. (21) have described quantitative FA to assess the multiple pathological components of CNV using a unique algorithm that subtracts the background and adds a term for the positive change in fluorescence corrected to background [positive fluorescence quotient (PFQ)]. This work has shown that the PFQ for CNV and leakage are important determinants of visual function, and are better correlated with functional measures such as visual acuity and reading ability.

Shah and coworkers used quantitative methods to demonstrate the utility of quantitating the amount of hyperfluorescence intensity and area seen on FA images of AMD patients before and after treatment for CNV (22). The investigators used image processing to measure the area of hyperfluorescence and fluorescence intensity above background fluorescence. Values for each image were plotted against time after dye injection to generate curves. Each area under the curve (AUC) was calculated. The investigators found an 11% decrease in AUC for fluorescence area and a 32% decrease in AUC for fluorescence intensity in the patients who clinically improved with treatment, but increases of 131% and 292% in the patients who worsened after PDT. Similarly, a 38% decrease in AUC for fluorescence intensity and a 19% decrease in AUC for fluorescence area were observed in patients who received vascular endothelial growth factor trap compared with increases of 66% (pZ 0.004, Mann–Whitney U-test) and 21% (pZ0.07) for patients who received placebo.

FA quantification is still being developed and will undoubtedly serve as a means to furnish important clinical information regarding the progression of AMD. Like quantified fundus photography and OCT, developments in computer technology hold the promise of creating truly automated methods to better understand retinal pathology.

MICROPERIMETRY

Microperimetry, also known as fundus perimetry, is a noninvasive method by which focal areas of retinal

sensitivity loss can be measured in those with macular disease. Several systems are currently in practice, including the scanning laser ophthalmoscope (Rodenstock, Germany) and the Micro Perimeter 1 (Nidek Technologies, Italy) (23). By integrating real-time fundus imaging and computerized threshold perimetry, these systems can provide fixation control by accounting for eye movement disturbances that can be common in patients with central visual loss. As a result, this technology can provide point-to-point correlation of the area and magnitude of retinal sensitivity loss at a precise location in the macula. In other words, this technology serves to delineate absolute and relative scotomas while allowing for elucidation of preferential fixation location and fixation stability.

Traditionally, distance visual acuity has been the gold standard for assessment of macular function in those with AMD. Although visual acuity is a useful and easily assessed parameter, it does not provide a complete description of visual function, and correlations with self-reported visual functioning are generally poor. Performance in daily activities such as reading are better correlated to the integrity of the central visual field (24). Microperimetry has shown itself to be a useful tool in assessing the functional deficits due to AMD beyond that of visual acuity as it generates information on the location and depth of relative and absolute scotomas. For example, this technology has revealed changes in retinal sensitivity over drusen (25). Sunness and coworkers have also described changes in fixation patterns and quantified the area of scotomas in those with geographic atrophy (26,27).

Microperimetry has proved useful in the evaluation of neovascular AMD. Absolute scotomas have been measured over CNV, subretinal hemorrhage, and chorioretinal scars (28). Fujii and colleagues evaluated the characteristics of visual loss in subfoveal CNV and suggest that functional deterioration may begin with a mild decrease in retinal sensitivity that later evolves into fixation instability and eventual absolute central scotoma with subsequent eccentric fixation (29). This technology has also been used to assess macular function following treatment. Schmidt-Erfurth et al. quantified the size of absolute and relative scotomas in patients with CNV at baseline and following treatment with PDT (30). Microperimetry has also been applied to assess functional changes with other treatment modalities for CNV including laser photocoagulation, submacular surgery, and macular translocation (31–34).

In contrast to the other imaging modalities that assess structural integrity, microperimetry provides a functional measure of macular function. Much like the Humphrey visual field is used to assess visual loss

188 ESMAILI ET AL.

in glaucoma, microperimetry can functionally assess macular deterioration in AMD. As newer treatments emerge for AMD, this modality may supplement the anatomic evaluation of drug efficacy by providing a functional measure of the area and degree of retinal recovery or deterioration. Furthermore, microperimetric analysis to evaluate progression of macular deterioration may aid in clinical decision making.

CONCLUSION

The value of quantitative retinal imaging will undoubtedly increase as the image processing software and the quantification methods improve over time. These advances are already assisting ophthalmologist through improved education of the patient on the nature and progression of AMD disease with important implications for better physician–patient relationships and compliance with treatment. It now seems likely that quantitative retinal imaging will also provide robust endpoints for assessing the effectiveness of interventions in clinical trials and epidemiological studies.

SUMMARY POINTS

&Quantitative techniques are useful for objectively evaluating the type and number of drusen, retinal pigment epithelium pigmentary changes, and components of CNV lesions.

&OCT based software, such as the OCTOR system developed at the Doheny Eye Institute, allows for delineation of retinal layers as well as pathologic entities such as subretinal fluid, area occupied by choroidal neovascular membranes, and the size of serous retinal pigment epithelial detachments.

&Quantification of fluorescein angiographic information is being developed and will provide important clinical information regarding the progression of AMD and response to therapy.

&Microperimetry is a useful tool in assessing functional deficits due to AMD beyond that of visual acuity since it provides information on the location and depth of relative and absolute scotomas.

REFERENCES

1.Klein R, Davis MD, Magli YL, et al. The Wisconsin agerelated maculopathy grading system. Ophthalmology 1991; 98:1128–34.

2.Bird AC, Bressler NM, Bressler SB, et al. An international classification and grading system for age-related maculopathy and age-related macular degeneration. Surv Ophthalmol 1995; 39:367–74.

3.Bressler NM, Bressler SB, Seddon LM, et al. Drusen characteristics in patients with exudative versus non-exudative age-related macular degeneration. Retina 1988; 8:109–14.

4.Bressler NM, Maguire MG, Bressler SB, et al. Relationship of drusen and abnormalities of the retinal pigment epithelium to the prognosis of neovascular macular degeneration. The Macular Photocoagulation Study Group. Arch Ophthalmol 1990; 108:1442–7.

5.Ferris FL, Davis MD, Clemons TE, et al. A simplified severity scale for AMD: AREDS Report No. 18. Arch Ophthalmol 2005; 123:1570–4.

6.Smith RT, Nagasaki T, Sparrow JR, et al. A method of drusen measurement based on the geometry of fundus reflectance. Biomed Eng Online 2003; 2:10.

7.Peli E, Lahav M. Drusen measurement from fundus photographs using computer image analysis. Ophthalmology 1986; 93:1575–80.

8.Morgan WH, Cooper RL, Constable IJ, et al. Automated extraction and quantification of macular drusen from fundal photographs. Aust NZ J Ophthalmol 1994; 22:7–12.

9.Kirkpatrick JN, Spencer T, Manivannan A, et al. Quantitative image analysis of macular drusen from fundus photographs and scanning laser ophthalmoscope images. Eye 1995; 9:48–55.

10.Shin DS, Javornik NB, Berger JW. Computer-assisted, interactive fundus image processing for macular drusen quantification. Ophthalmology 1999; 106:1119–25.

11.Sivagnanavel V, Smith RT, Lau GB, et al. An interinstitutional comparative study and validation of computer aided drusen quantification. Br J Ophthalmol 2005; 89:554–7.

12.Smith RT, Chan JK, Nagasaki T, et al. A method of drusen measurement based on reconstruction of fundus background reflectance. Br J Ophthalmol 2005; 89:87–91.

13.Huang D, Swanson EA, Lin CP, et al. Optical coherence tomography. Science 1991; 254:1178–81.

14.Puliafito CA, Hee MR, Lin CP, et al. Imaging of macular diseases with optical coherence tomography. Ophthalmology 1995; 102:217–29.

15.Hee MR, Puliafito CA, Wong C, et al. Quantitative assessment of macular edema with optical coherence tomography. Arch Ophthalmol 1995; 113:1019–29.

16.Hee MR, Puliafito CA, Duker JS, et al. Topography of diabetic macular edema with optical coherence tomography. Ophthalmology 1998; 105:360–70.

17.Salinas-Alamon A, Garcia-Layana A, Maldonado MJ, et al. Using optical coherence tomography to monitor photodynamic therapy in age related macular degeneration. Am J Ophthalmol 2005; 140:23.e1–23.7.

18.Shahidi M, Wang Z, Zelkha R. Quantitative thickness measurement of retinal layers imaged by optical coherence tomography. Am J Ophthalmol 2005; 139:1056–61.

19.Verteporfin Photodynamic Therapy (VIP) Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in age related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization—Verteporfin in Photodynamic Therapy Report 2. Am J Ophthalmol 2001; 131:541–60.

20.Phillips RP, Ross PG, Tyska M, Sharp PF, Forrester JV. Detection and quantification of hyperfluorescent leakage by computer analysis of fundus fluorescein angiograms. Graefes Arch Clin Exp Ophthalmol 1991; 229:329–35.

21.Chakravarthy U, Walsh AC, Muldrew A, Updike PG, Barbour T, Sadda SR. Quantitative fluorescein angiographic analysis of choroidal neovascular membranes: validation and correlation with visual function. Invest Ophthalmol Vis Sci 2007; 48:349–54.

22.Shah SM, Tatlipinar S, Quinlan E, et al. Dynamic and quantitative analysis of choroidal neovascularization by fluorescein angiography. Invest Ophthalmol Vis Sci 2006; 47:5460–8.

23.Rohrschneider K, Springer C, Bultmann S, et al. Micro- perimetry—comparison between the micro perimeter 1 and scanning laser ophthalmoscope—fundus perimetry. Am J Ophthalmol 2005; 139:125–34.

24.Sunness JS, Applegate CA, Haselwood D, et al. Fixation patterns and reading rates in eyes with central scotomas from advanced atrophic age-related macular degeneration and Stargardt disease. Ophthalmology 1996; 103:1458–66.

25.Takamine Y, Shiraki K, Moriwaki M, et al. Retinal sensitivity measurement over drusen using scanning laser ophthalmoscope microperimetry. Graefes Arch Clin Exp Ophthalmol 1998; 236:285–90.

26.Sunness JS, Bressler NM, Maguire MG. Scanning laser

J Ophthalmol 1995; 119:143–51.

27.Sunness JS, Bressler NM, Tian Y, et al. Measuring geographic atrophy in advanced age-related macular degeneration. Invest Ophthalmol Vis Sci 1999; 40:1761–9.

28.Tezel TH, Del Priore LV, Flowers BE, et al. Correlation between scanning laser ophthalmoscope microperimetry

and anatomic abnormalities in patients with subfoveal neovascularization. Ophthalmology 1996; 103:1829–36.

29.Fujii GY, De Juan E, Jr., Humayun MS, et al. Characteristics of visual loss by scanning laser ophthalmoscope microperi-

metry in eyes with subfoveal choroidal neovascularization secondary to age-related macular degeneration. Am J Ophthalmol 2004; 136:1067–78.

30.Schmidt-Erfurth UM, Elsner H, Terai N, et al. Effects of verteporfin therapy on central visual field function. Ophthalmology 2004; 111:931–9.

31.Rohrschneider K, Gluck R, Becker M, et al. Scanning laser fundus perimetry before laser photocoagulation of well defined choroidal neovascularization. Br J Ophthalmol 1997; 81:568–73.

32.Fujii GY, de Juan E, Jr., Sunness J, et al. Patient selection for macular translocation surgery using the scanning laser ophthalmoscope. Ophthalmology 2002; 109:1737–44.

33.Hudson HL, Frambach DA, Lopez PF. Relation of the functional and structural fundus changes after submacular surgery for neovascular age-related macular degeneration. Br J Ophthalmol 1995; 79:417–23.

34.Loewenstein A, Sunness JS, Bressler NM, et al. Scanning laser ophthalmoscope fundus perimetry after surgery for choroidal neovascularization. Am J Ophthalmol 1998; 125:657–65.

INTRODUCTION

Autofluorescence (AF) is the intrinsic fluorescence emitted by a substance after being stimulated by excitation energy. Ocular structures that autofluoresce include the corneal epithelium and endothelium, lens, macular and retinal pigment epithelium (RPE) pigments, optic nerve drusen and RPE deposits in Best’s disease. The AF emitted by macular pigments is in the 520 to 800 nm range with peak emission at 590 to 630 nm. Clinically, AF of macular pigments can be produced in vivo by an exciting light source with a wavelength between 400 and 590 nm with peak excitation occurring between 490 and 510 nm (1). This can be achieved with a modified fundus camera or scanning laser ophthalmoscope (SLO).

The SLO uses blue laser light at 488 nm for illumination and a 500 nm barrier filter to isolate light from other ocular autofluorescent structures (2,3). The use of confocal scanning laser ophthalmoscope (cSLO) is considered superior to modified fundus camera images because cSLO helps eliminate the competing AF of the lens. As the plane of the cSLO detection system is conjugate to the plane of fundus, competing light signals from other planes are reduced (4). In addition, cSLO images have a higher reliability in comparing images in one patient from one visit to the next than modified fundus camera imaging (5). However, a cSLO requires the purchase of a new imaging device whereas most offices already have a fundus camera. Spaide has described an easy and inexpensive modification to preexisting fundus camera by adding 580 nm excitation and 695 nm barrier filters. With the proper barrier filter signals, wavelengths shorter than 695 nm including fluorescence emitted by the lens from 510 to 670 nm, would be blocked (4).

The predominant source of AF in the macula is lipofuscin, a complex mixture of fluorophores. When the RPE phagocytize photoreceptor outer segments, consisting of retinoids, fatty acids, and proteins,

lipofuscin accumulates as an oxidative byproduct within the RPE cells (1,6,7). Lipofuscin has an affinity for acidic organelles and thus accumulates in RPE lysosomes and it can account for as much as 20% of the free cytoplasmic space of a RPE cell (8,9). A loss of RPE cells has been shown to be accompanied by substantial loss of AF content (10).

The pigment within lipofuscin that causes this fluorescence was isolated and characterized by Eldred to be A2E, named for its derivation from two molecules of vitamin A aldehyde and one molecule of ethanolamine (Fig. 1) (12,13). A2E has been shown to inhibit human RPE cell growth and induce apoptosis in vitro. It exhibits detergent-like activity, disrupting membrane bound ATPase that maintain lysosomal pH (13–15). In mitochondria, A2E inhibits oxygen consumption synergistically with light by inhibiting cytochrome c oxidase (16). By mobilizing cytochrome c and apoptosis-inducing factor from mitochondria into the cytoplasm and nucleus, apoptosis is induced in RPE cells (17). A2E has also been shown to confer a dose related sensitivity to blue light damage in RPE cells via oxidative mechanisms (18).

AF is a rapid noncontact, noninvasive way to evaluate RPE function. AF can evaluate the amount of lipofuscin that is accumulated in RPE. By evaluating fundus autofluorescence images and thus lipofuscin accumulation, disturbances within the RPE can be readily detected. In a normal retina, lipofuscin is most concentrated in the macula with the exception of the fovea and decreases towards the periphery (1). The highest lipofuscin AF level in the eye is found 78 to 138 away from the fovea, correlating with the area with the highest distribution of rod photoreceptors (19). Even though high inter-subject variability is seen in the distribution of macular AF, Delori found similarity of AF level within the same retina, as well as that of the fellow eye (20,21). Lipofuscin fluorescence

levels increases linearly with age. In humans, intracellular lipofuscin levels occupy 1% of cell volume during the first decade, increasing to 12% to 13% in the 50to 80-year-olds and reaching 19% of the cell volume in the 81to 90-year-olds (9).

AF AND AMD

Several studies have found that the accumulation of lipofuscin over time may promote the development of AMD (Figs. 2 and 3). The age, spatial, and racial distribution of lipofuscin correlates well with the AF patterns seen with AMD. Dorey and colleagues found significant correlation between photoreceptor loss and

elevated lipofuscin levels in RPE within donor eyes of Caucasians over 50 years old. They hypothesized that lipofuscin accumulation may be indicative of increased phagocytic and metabolic stress on the RPE cells leading to photoreceptor death (22,23). Excessive lipofuscin accumulation may precede the development of GA and the enlargement of preexisting GA (24). Thus, AF imaging may be useful in evaluating the risk of AMD progression by mapping retinal AF and lipofuscin accumulation over extended periods. Previous work has shown a significant correlation in the amount of large, foveal, soft drusen and patterns of increased AF (5). Spaide reported greater levels of AF in fellow eyes of patients with exudative

AMD than in patients without a history of AMD (4). Delori and colleagues identified that RPE overlying drusen have a central area of decreased AF with surrounding ring of increased AF, suggesting damage to RPE health (7).

An International Fundus Autofluorescence Classification Group (IFAG) was organized by the Fundus autofluorescence in Age-Related Macular Degeneration (FAM) group to establish an international classification system to describe the abnormal fundus

AF patterns seen in AMD (25). The multinational group of clinicians established eight distinct patterns of AF in AMD: normal AF, minimal change pattern AF, focal increased pattern, patchy pattern AF, linear pattern AF, lacelike pattern AF, reticular pattern AF, and speckled pattern AF. Standardized photos of these AF patterns were established (Fig. 4).

1.Normal pattern. The normal AF pattern is characterized by a homogeneous background AF with a gradual decrease from the inner macula toward the

fovea due to the masking effect of yellow luteal macular pigment.

2.Minimal change pattern. The minimal change pattern is characterized by very small irregular increases or decreases of background AF without an obvious topographic pattern.

3.Focal increased pattern. Focal increased AF is described as the presence of at least one spot (less than 200 mm diameter) of markedly increased AF brighter than the surrounding fluorescence. The borders are well defined and some areas of focal

increased AF may be surrounded by a darkerappearing halo. Visible alterations (focal hyperpigmentation or drusen) seen on color fundus photos may or may not correspond to areas of AF.

4.Patchy pattern. Patchy AF is defined as at least one larger area (greater than 200 mm diameter) of markedly increased AF where the borders of the areas are typically less well-defined than the previous pattern. There is a gradual increase in AF from the back-

ground to the patchy area. This pattern may also correspond to large drusen, soft drusen and areas of hyperpigmentation seen on color photographs.

5.Linear pattern. The linear pattern describes the presence of at least one linear area of markedly increased AF with well-demarcated borders and no gradual decrease in AF. These AF areas usually correspond to hyperpigmented lines on the color fundus photograph.

6.Lacelike pattern. The lacelike pattern typically exhibits numerous branching linear structures of increased AF that form a lace pattern. The borders are poorly defined and a decline in AF is observed from the center of the AF areas to the surrounding areas. These lacelike areas can correspond to hyperpigmentation on the color image, but it may correspond to normal fundus areas as well.

7.Reticular pattern. The reticular pattern is exemplified by the presence of multiple small areas (less than

200 mm diameter) of decreased AF with poorly defined borders. Fundoscopically, there are usually visible small soft drusen, hard drusen, or areas with pigmentary changes overlying these areas, but the fundus can be normal as well.

8.Speckled pattern. The speckled AF pattern has the simultaneous presence of a variety of AF abnormalities that extend beyond the macular area. There can be multiple, small areas of irregularly increased and decreased AF which appear punctuate or

resemble linear structures. Color fundus photographs may include corresponding hyperand hypopigmentation and multiple subconfluent and confluent drusen.

Of 149 eyes studied within the IFAG group’s initial study of 107 patients (44 male, 63 female patients) with unilateral or bilateral GA, the diffuse pattern was the most common (57%), followed by the banded, focal and normal patterns at about 12% per group. The other patterns were less commonly seen (25).

AF with Drusen

There have been numerous reports examining AF in drusen associated with dry AMD. Delori and associates identified a specific pattern of AF spatially associated with hard and soft drusen ranging between 60 and 175 mm in size. The pattern is characterized by a central area of decreased AF surrounded, in most cases, by an annulus of increased AF around the drusen (7). It was hypothesized by Delori that this AF pattern is due to RPE impairment with secondary accumulation of lipofuscin around drusen, with RPE

198 SINGH ET AL.

atrophy overlying drusen. Spade interprets the appearance of this pattern as secondary to thinner RPE cells on top of the drusen, and thicker RPE cells around the base (26). These areas of increased AF around drusen showed normal or near-normal photopic sensitivity, but moderately reduced scotopic sensitivity (27). Soft drusen larger than 175 mm and confluent soft drusen show either a heterogenous distribution of AF or multifocal areas of decreased AF (7). But, both normal, hyperand hypofluorescence of the RPE cells overlying soft drusen has been reported with no proven biochemical explanation and thus future larger studies are needed to refine this classification (3,5,28,29).

AF in Geographic Atrophy

With loss of RPE containing lipofuscin, areas of GA appear dark under AF. Holz and colleagues found 83% of the geographic atrophic from AMD has increased AF pattern at the border (30). This ring of elevated AF from lipofuscin bordering the GA supports the concept that excessive lipofuscin may be associated with RPE damage (21). In fact, the increased AF in the junctional zone around GA is thought to be characteristic for AMD since only 9% of geography atrophy from other causes exhibit similar findings (3,30). When tested with fundus perimetry, a significant degree of retinal sensitivity loss is found in the junctional area between the inner dark zone and ring of increased AF (31). Photopic and scotopic fine matrix mapping of these areas has shown a scotopic sensitivity loss demonstrating a preferential loss of rods (27). This correlation of AF abnormality to a loss of function may further suggest a relationship between GA and its increased AF border. Holz found that 90% of the AMD patients with bilateral GA exhibited the same AF pattern in both eyes (30).

Longitudinal studies have demonstrated that AF is useful for the precise mapping and measuring of GA areas (Fig. 5) (24). Moreover, progression of increased AF from GA has been described in several studies (3,30). It has been noted that the rate of GA spread accelerates with expansion of GA area, then levels off at five disc areas (32,33). However, the natural progression of GA is poorly understood. Manual measurement of GA is time consuming and results in significant inter-observer variability (31). Previous studies have found that automated quantification and delineation of AF images is superior to fundus photography or fluorescein angiography in the delineation of GA (31). When combined with cSLO, the measurement of GA area improves significantly (34). The quantification of these lesions adds to the understanding of the natural history of GA formation and allows for the monitoring of future therapeutics to slowdown its progression.

In a three year prospective study of three patients, Holz and colleagues found the development and enlargement of GA within areas of increased AF (24). In another study by Holz, eyes with only diffuse patterns of AF were examined and found to have higher levels of AF on the transitional area between GA and healthy RPE. The study determined that the greater the area of increased AF adjacent to the GA, the faster the expansion of GA. They concluded a positive correlation between increased AF and GA expansion in this population (35).

With evidence linking lipofuscin to cell death in a dose-dependent manner, it is plausible to propose the zone of increased AF around GA may be the advancing border of GA. As GA advances, RPE cells absorb lipofuscin materials from adjacent dead RPE cells, thus increasing its susceptibility for destruction. This effect spread like dominos, and accelerates as the area of GA containing lipofuscin grows exponentially. Expansion of GA slows only when the spread reaches less vulnerable and healthier RPE cells.

However, in a retrospective study of AF photographs with differing baseline patterns of AF, Hwang and colleagues found that only 34% to 50% of the new areas of GA fell into an area of increased AF. The positive predictive value for increased AF to form new GA was no better than chance (36). Given these two conflicting reports, it difficult to form a consensus on whether increased AF at the border of GA truly predicts future progression of GA to this area.

AF in Choroidal Neovascularization

While the majority of studies with AF in AMD have concentrated on the dry form, there has been limited work on determining a correlation between AF and choroidal neovascularization (CNV). Eyes with early CNV lesions show various patterns of increased AF (Fig. 6). Einbock and colleagues studied eyes with exudative changes and found that 17% of eyes exhibited “patchy” AF pattern, as well as some with “focal increased” plaque and “reticular” patterns. Other less intense patterns were not associated with progression to late AMD during 18-month follow-up (37). Dandekar and colleagues studied AF in 65 consecutive eyes with CNV secondary to AMD (38). Patients were stratified by age of lesion. Eyes with recent onset lesions (one to six months) showed no AF abnormalities indicating relatively healthy RPE initially. Older CNV lesions (greater than six months) exhibited decreased AF levels indicating RPE damage and photoreceptor loss. Similar to pattern seen in GA, increased AF in the junctional zone is also seen in some eyes with CNV. Eyes with better visual acuity are those with intact AF in early onset lesions and others

with intact AF in the foveal area. This study highlighted the possible use of AF to determine visual prognosis for AMD lesions.

In an observational case series examining pigment epithelial detachments (PEDs) associated with AMD, increased AF was seen in all serous PEDs regardless of whether there was an underlying CNV in the area of detachment (Fig. 7). The authors concluded that the increased AF seen with serous PEDs may be due to AF of sub-retinal pigment epithelial fluid. In the case of a drusenoid PED, AF levels were dependent on pigment clumping with increased pigment correlating with lower AF levels. Larger numbers of patients are needed to verify this morphological features (39).

A few studies have attempted to classify CNV lesion type based on the AF pattern seen.

In a study examining AF of 68 eyes undergoing photodynamic therapy (PDT) treatment, Framme found 79% of the untreated classic lesions were associated with decreased AF and a junctional zone of increased or normal AF. In untreated occult membranes, a normal or mottled AF pattern with foci of hyperand hypofluorescence were seen. After PDT treatment, 90% of the classic CNV lesions showed decreased AF signal. There appeared to be no AF change in occult parts of CNV lesions after PDT (40). These baseline AF patterns were also described by McBain and collegues in patients with exudative AMD. Low AF signal at the site of classic CNV was detected in 90% of exudative AMD lesions. While multiple foci of low AF was seen in half of occult CNV lesions, focally increased AF was rarely seen with CNV lesions (41).

AF after Laser Treatment

AF has shown to be useful tools in localizing laser lesions after treatment. In a pilot study, Framme attempted Nd:YAG RPE selective laser treatment for diabetic maculopathy and central serous chorioretinopathy. Changes to AF levels around laser lesions 10 minutes after treatment were seen in 22 out of 26 patients. These laser lesions later exhibit a lower AF level in comparison to surrounding tissues at three to six months (5,42). In a study of subthreshold infrared diode laser for the reduction of drusen in dry AMD, visualization of laser lesions using AF was shown to be more sensitive than fluorescein angiogram in 75% of the eyes immediately post laser treatment, and 55% of the eyes three months after treatment (43).

LIMITATIONS

There are several limitations in the use of AF in the evaluation of patients for AMD. AF detection is limited in patients with significant media opacities such as cataract and vitreous hemorrhage. Furthermore, the comparative quantification of AF images cannot occur amongst patients. Rather, only images from the same patient can be compared to determine changes in intensities of AF seen over time. Thus, the use of AF is still in its infancy and further studies need to be performed to evaluate its role in the diagnosis and management of AMD.

Reports on the use of AF in the diagnosis and management of AMD are preliminary at best and sometimes conflicting. There remains a need for further data to clarify the relationship of AF patterns in formation and expansion of GA. More information is needed to substantiate the relationship between AF patterns and risk of CNV. There is a lack of studies identifying AF criteria useful in the classification of CNV subtype. The current literature thus far consists mostly of non-randomized small case series limiting its applicability to the general population. To address these limitations, the FAM study group, a multi-center study is currently under way, with some results already in press. The goal of the group is to investigate the correlation between fundus AF and the natural history of AMD. The group also intends to identify high-risk AF characteristics that can predict patients who will progress to late AMD (37).

SUMMARY POINTS

&The predominant source of AF in the macula is lipofuscin, a complex mixture of fluorophores.

&The pigment within lipofuscin that causes this fluorescence is A2E (named for its derivation from two molecules of vitamin A aldehyde and one molecule of ethanolamine).

&Fundus autofluorescence is a useful modality to image lipofuscin in RPE cells and is a unique way to assess RPE function in AMD.

&In GA, automated imaging analysis by AF has been shown superior to fundus photo or FA in assessing the extent of the atrophy. Increased AF, especially at the edge of GA area, may predict GA formation and expansion.

&AF has been helpful in assessing RPE health in exudative AMD, and can consistently visualize serous PED.

&AF can also be helpful in the early localization of previous retinal laser treatments where it may be up to three times more sensitive than fluorescein angiogram.

&AF is shown to be helpful in indicating graft visualization in RPE-choroidal grafts for AMD patients.

&Larger randomized controlled studies using AF are needed to further assess its potential in the detection and management of AMD.

REFERENCES

1.Delori FC, Dorey CK, Staurenghi G, Arend O, Goger DG, Weiter JJ. In vivo fluorescence of the ocular fundus exhibits retinal pigment epithelium lipofuscin characteristics. Invest Ophthalmol Vis Sci 1995; 36:718–29.

2.von Ruckmann A, Fitzke FW, Bird AC. Distribution of fundus autofluorescence with a scanning laser ophthalmoscope. Br J Ophthalmol 1995; 79:407–12.

3.von Ruckmann A, Fitzke FW, Bird AC. Fundus autofluorescence in age-related macular disease imaged with a laser scanning ophthalmoscope. Invest Ophthalmol Vis Sci 1997; 38:478–86.

4.Spaide RF. Fundus autofluorescence and age-related macular degeneration. Ophthalmology 2003; 110:392–9.

5.Lois N, Owens SL, Coco R, Hopkins J, Fitzke FW, Bird AC. Fundus autofluorescence in patients with age-related macular degeneration and high risk of visual loss. Am J Ophthalmol 2002; 133:341–9.

6.Gaillard ER, Atherton SJ, Eldred G, Dillon J. Photophysical studies on human retinal lipofuscin. Photochem Photobiol 1995; 61:448–53.

7.Delori FC, Fleckner MR, Goger DG, Weiter JJ, Dorey CK. Autofluorescence distribution associated with drusen in age-related macular degeneration. Invest Ophthalmol Vis Sci 2000; 41:496–504.

8.Kennedy CJ, Rakoczy PE, Constable IJ. Lipofuscin of the retinal pigment epithelium: a review. 1995; 9(Pt 6): 763–71.

9.Feeney-Burns L, Hilderbrand ES, Eldridge S. Aging human RPE: morphometric analysis of macular, equatorial, and peripheral cells. Invest Ophthalmol Vis Sci 1984; 25:195–200.

10.Katz ML, Eldred GE. Retinal light damage reduces autofluorescent pigment deposition in the retinal pigment epithelium. Invest Ophthalmol Vis Sci 1989; 30:37–43.

11.Parish CA, Hashimoto M, Nakanishi K, Dillon J, Sparrow J. Isolation and one-step preparation of A2E and iso-A2E, fluorophores from human retinal pigment epithelium. Proc Natl Acad Sci USA 1998; 95:14609–13.

202 SINGH ET AL.

12.Sakai N, Decatur J, Nakanishi K, Eldred GE. Ocular age pigment “A2E”: an unprecedented pyridinium bisretinoid. J Am Chem Soc 1996; 118:1559–60.

13.Eldred GE. Age pigment structure. Nature 1993; 364:396.

14.Mellman I, Fuchs R, Helenius A. Acidification of the endocytic and exocytic pathways. Annu Rev Biochem 1986; 55:663–700.

15.Sparrow JR, Parish CA, Hashimoto M, Nakanishi K. A2E, a lipofuscin fluorophore, in human retinal pigmented epithelial cells in culture. Invest Ophthalmol Vis Sci 1999; 40:2988–95.

16.Shaban H, Gazzotti P, Richter C. Cytochrome c oxidase inhibition by N-retinyl-N-retinylidene ethanolamine, a compound suspected to cause age-related macula degeneration. Arch Biochem Biophys 2001; 394:111–6.

17.Suter M, Reme C, Grimm C, et al. Age-related macular degeneration. The lipofusion component N-retinyl-N- retinylidene ethanolamine detaches proapoptotic proteins from mitochondria and induces apoptosis in mammalian retinal pigment epithelial cells. J Biol Chem 2000; 275:39625–30.

18.Nilsson SE, Sundelin SP, Wihlmark U, Brunk UT. Aging of cultured retinal pigment epithelial cells: oxidative reactions, lipofuscin formation and blue light damage. Doct Ophthalmol 2003; 106:13–6.

19.Delori FC, Goger DG, Dorey CK. Age-related accumulation and spatial distribution of lipofuscin in RPE of normal subjects. Invest Ophthalmol Vis Sci 2001; 42:1855–66.

20.Bellmann C, Jorzik J, Spital G, Unnebrink K, Pauleikhoff D, Holz FG. Symmetry of bilateral lesions in geographic atrophy in patients with age-related macular degeneration. Arch Ophthalmol 2002; 120:579–84.

21.Robson AG, Moreland JD, Pauleikhoff D, et al. Macular pigment density and distribution: comparison of fundus autofluorescence with minimum motion photometry. Vision Res 2003; 43:1765–75.

22.Dorey CK, Wu G, Ebenstein D, Garsd A, Weiter JJ. Cell loss in the aging retina relationship to lipofuscin accumulation and macular degeneration. Invest Ophthalmol Vis Sci 1989; 30:1691–9.

23.Dorey K, Staurenghi G, Delori FC. Lipofuscin in age and ARMD eyes. In: Hollyfield JG, ed. Retinal Degeneration. New York: Plenum Pub Corp., 1993:3–14.

24.Holz FG, Bellman C, Staudt S, Schutt F, Volcker HE. Fundus autofluorescence and development of geographic atrophy in age-related macular degeneration. Invest Ophthalmol Vis Sci 2001; 42:1051–6.

25.Bindewald A, Schmitz-Valckenberg S, Jorzik JJ, et al. Classification of abnormal fundus autofluorescence patterns in the junctional zone of geographic atrophy in patients with age related macular degeneration. Br J Ophthalmol 2005; 89:874–8.

26.Spaide RF. Macular autofluorescence. AAO Retina Subspecialty Day 2005; 1:192–7.

27.Scholl HP, Bellmann C, Dandekar SS, Bird AC, Fitzke FW. Photopic and scotopic fine matrix mapping of retinal areas of increased fundus autofluorescence in patients with agerelated maculopathy. Invest Ophthalmol Vis Sci 2004; 45:574–83.

28.Solbach U, Keilhauer C, Knabben H, Wolf S. Imaging of retinal autofluorescence in patients with age-related macular degeneration. Retina 1997; 17:385–9.

29.Sunness JS, Ziegler MD, Applegate CA. Issues in quantifying atrophic macular disease using retinal autofluorescence. Retina 2006; 26:666–72.

30.Holz FG, Bellmann C, Margaritidis M, Schutt F, Otto TP, Volcker HE. Patterns of increased in vivo fundus autofluorescence in the junctional zone of geographic atrophy of the retinal pigment epithelium associated with age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol 1999; 237:145–52.

31.Schmitz-Valckenberg S, Jorzik J, Unnebrink K, Holz FG, FAM Study Group. Analysis of digital scanning laser ophthalmoscopy fundus autofluorescence images of geographic atrophy in advanced age-related macular degene 132#ration. Graefes Arch Clin Exp Ophthalmol 2002; 240:73–8.

32.Schatz H, McDonald HR. Atrophic macular degeneration rate of spread of geographic atrophy and visual loss. Ophthalmology 1989; 96:1541–51.

33.Sarks SH. Drusen patterns predisposing to geographic atrophy of the retinal pigment epithelium. Aust J Ophthalmol 1982; 10:91–7.

34.Deckert A, Schmitz-Valckenberg S, Jorzik J, Bindewald A, Holz FG, Mansmann U. Automated analysis of digital fundus autofluorescence images of geographic atrophy in advanced age-related macular degeneration using confocal scanning laser ophthalmoscopy (cSLO). BMC Ophthalmol 2005; 5:8.

35.Schmitz-Valckenberg S, Bindewald-Wittich A, DolarSzczasny J, et al. Correlation between the area of increased autofluorescence surrounding geographic atrophy and disease progression in patients with AMD. Invest Ophthalmol Vis Sci 2006; 47:2648–54.

36.Hwang JC, Chan JW, Chang S, Smith RT. Predictive value of fundus autofluorescence for development of geographic atrophy in age-related macular degeneration. Invest Ophthalmol Vis Sci 2006; 47:2655–61.

37.Einbock W, Moessner A, Schnurrbusch UE, Holz FG, Wolf S, FAM Study Group. Changes in fundus autofluorescence in patients with age-related maculopathy. Correlation to visual function: a prospective study. Graefe’s Arch Clin Exp Ophthalmol 2005; 243:300–5.

38.Dandekar SS, Jenkins SA, Peto T, et al. Autofluorescence imaging of choroidal neovascularization due to age-related macular degeneration. Arch Ophthalmol 2005; 123:1507–13.

39.Karadimas P, Bouzas EA. Fundus autofluorescence imaging

in serous and drusenoid pigment epithelial detachments associated with age-related macular degeneration. Am J Ophthalmol 2005; 140:1163–5.

40.Framme C, Bunse A, Sofroni R, et al. Fundus autofluorescence before and after photodynamic therapy for choroidal neovascularization secondary to age-related macular degeneration. Ophthalmic Surg Lasers Imaging 2006; 37:406–14.

41.McBain VA, Townend J, Lois N. Fundus autofluorescence in exudative age-related macular degeneration. Br J Ophthalmol 2006; 91(4):491–6.

42.Framme C, Brinkmann R, Birngruber R, Roider J. Autofluorescence imaging after selective RPE laser treatment in macular diseases and clinical outcome: a pilot study. Br J Ophthalmol 2002; 86:1099–106.

43.Bessho K, Rodanant N, Bartsch DU, Cheng L, Koh HJ, Freeman WR. Effect of subthreshold infrared laser treatment for drusen regression on macular autofluorescence in patients with age-related macular degeneration. Retina 2005; 25:981–8.

INTRODUCTION

Until the initial Macular Photocoagulation Study (MPS) outcome data were published in June 1982, there were no reported treatments of proven benefit for patients with choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD). The MPS trials conducted from 1979 to 1994 showed that laser photocoagulation was preferable to observation for several categories of well-defined CNV based on the fluorescein angiographic location of the CNV with respect to the geometric center of the fovea, i.e., extrafoveal, juxtafoveal, and subfoveal (1–5). The MPS publications also described the factors which limited the utility of laser photocoagulation treatment.

1.Only a small proportion of symptomatic AMD eyes met MPS eligibility criteria as being appropriate for laser treatment (6,7);

2.There was a high rate of persistent and recurrent leakage even after initially successful closure of the CNV (8,9);

3.Laser photocoagulation caused immediate and permanent damage to the retina in the area treated and this damage typically resulted in an immediate decrease in visual acuity (VA) (2);

4.Treated as well as untreated eyes continued to lose central vision over time, despite initial closure of the CNV in laser-treated eyes.

In addition, because laser photocoagulation is a focal treatment, there is no expected beneficial effect beyond the area of laser application. Specifically, laser photocoagulation does not inhibit the development of new areas of CNV. With the advent of safe and effective antiangiogenic therapies which treat not only the existing neovascularization but also reduce the risk of developing CNV, it would appear that laser photocoagulation will have an extremely limited role in the

management of CNV secondary to AMD. Thus, the following narrative is presented primarily for an historical perspective on how treatment for CNV secondary to AMD developed over the last quarter century.

EPIDEMIOLOGY AND NATURAL HISTORY

AMD is a leading cause of severe and irreversible central vision loss in the developed world among people over the age of 55 (10–13). Up to 90% of the severe vision loss in AMD is caused by CNV (14–16).

Before the MPS was initiated in 1979, there were several natural history studies which documented the unfavorable visual prognosis of eyes with untreated CNV secondary to both AMD and ocular histoplasmosis (17,18). These natural history data were substantiated by the visual outcomes of untreated eyes among participants in the MPS. In the MPS trial of juxtafoveal CNV, 65% of untreated eyes lost six or more lines of acuity after five years follow up, and 93% progressed from juxtafoveal to subfoveal CNV (4,19).

The initial component of the MPS evaluated argon laser photocoagulation in patients with extrafoveal CNV secondary to AMD. At the time, this trial was known as the Senile Macular Degeneration Study (SMDS). Eyes with extrafoveal CNV were assigned randomly to immediate argon laser treatment or to observation. By 18 months after enrollment, 60% of untreated eyes had lost six or more lines of VA. By one year after enrollment, fluorescein angiography showed that 73% of untreated eyes had progressed from extrafoveal to subfoveal CNV (20).

In 1985, Guyer et al. reported that among 92 AMD patients with subfoveal neovascular lesions, 64% lost six or more lines of vision within two years (18). In the MPS trial of subfoveal lesions, 30% of untreated eyes lost six or more lines of VA at 12 months follow up, and 39% lost six or more lines of vision by two years (2).

The MPS Trials

The MPS documented that the visual outcome of laser treatment for eyes with extrafoveal CNV was better than the natural history (21–23). In fact, recruitment into the argon laser trial of extrafoveal CNV (SMDS) was halted early because 18 months after enrollment, only 25% of laser treated eyes compared to 60% of observed eyes had lost six or more lines of VA (21). Although laser treatment did not reverse or stop progression of vision loss, laser treated eyes continued to have better vision than untreated eyes even after five years of follow up (5). Trials of similar design conducted at Moorfields Eye Hospital in London, England and by Coscas and Soubrane in Creteil, France also demonstrated a benefit of laser treatment versus observation in AMD patients with selected CNV lesions (24,25). Several MPS trials reported that the difference in vision loss between laser-treated and untreated eyes was maintained over a four to five years course of follow up. The patient eligibility criteria defining the study population as well as the results from the key trials are summarized in Tables 1 and 2 (1–5,27).

Decreased Vision after Laser Treatment

The studies which showed a benefit of laser treatment compared to observation foreyeswith studyeligible CNV lesions also documented that laser treatment did not prevent the progressive vision loss associated with CNV. Significant vision loss occurred over time in most treated eyes. Follow up also showed that persistent and recurrent CNV were responsible for the progressive loss of vision. For example, 24 months after laser treatment of extrafoveal CNV lesions, 52% of eyes showed evidence of recurrence (28). Even for subfoveal lesions, after three years of follow up, nearly half the treated eyes had persistent or recurrent CNV (9). One MPS trial reported that eyes with recurrent CNV had less vision loss with laser treatment than with observation (Table 2) (3).

It must be noted that in eyes with subfoveal lesions and relatively good VA, there is greater loss of

vision in laser-treated versus untreated eyes within the first three months after laser treatment (3). This observation documents the immediate harmful effects of laser treatment to the fovea. However, when patients with subfoveal CNV were followed for longer periods, it became evident that laser treated eyes had less vision loss than observed eyes, indicating some long-term benefit of laser treatment even when applied to subfoveal CNV (3). This benefit was maintained over the three years course of follow up.

As indicated in the opening paragraphs, this review and the accompanying tables are provided for historical perspective. At present, anti vascular endothelial growth factor (AntiVEGF) therapy appears to be the preferred management strategy for all forms of CNV secondary to AMD irrespective of the geographic location of the CNV with respect to the foveal center. The reasons are listed below (20,29–33).

1.AntiVEGF therapy is not associated with immediate loss of vision due to destruction of visual elements in the retina.

2.AntiVEGF therapy is more effective than laser photocoagulation or photodynamic therapy.

3.AntiVEGF therapy discourages the formation of new vessels as well as treating the new vessels.

SUMMARY POINTS

&The MPS trials were conducted from 1979 to 1994 and showed that laser photocoagulation was preferable to observation for several categories of well-defined CNV based on the fluorescein angiographic location of the CNV with respect to the geometric center of the fovea, i.e., extrafoveal, juxtafoveal, and subfoveal.

&The MPS studies also documented that laser treatment did not prevent the progressive vision loss associated with CNV.

&Significant vision loss occurred over time in most treated eyes.

206 CUKRAS AND FINE

REFERENCES

1.Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions of age-related macular degeneration. Updated findings from two clinical trials. Arch Ophthalmol 1993; 111(9):1200–9.

2.Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions in age-related macular degeneration. Results of a randomized clinical trial. Arch Ophthalmol 1991; 109(9):1220–31.

3.Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal recurrent neovascular lesions in agerelated macular degeneration. Results of a randomized clinical trial. Arch Ophthalmol 1991; 109(9):1232–41.

4.Macular Photocoagulation Study Group. Laser photocoagulation for juxtafoveal choroidal neovascularization. Five-year results from randomized clinical trials. Arch Ophthalmol 1994; 112(4):500–9.

5.Macular Photocoagulation Study Group. Argon laser photocoagulation for neovascular maculopathy. Five-year results from randomized clinical trials. Arch Ophthalmol 1991; 109(8):1109–14.

6.Ciulla TA, Danis RP, Harris A. Age-related macular degeneration: a review of experimental treatments. Surv Ophthalmol 1998; 43(2):134–46.

7.Freund KB, Yannuzzi LA, Sorenson JA. Age-related macular

degeneration and choroidal neovascularization. Am J Ophthalmol 1993; 115(6):786–91.

8.Macular Photocoagulation Study Group. Persistent and recurrent neovascularization after krypton laser photocoagulation for neovascular lesions of age-related macular degeneration. Arch Ophthalmol 1990; 108(6):825–31.

9.Macular Photocoagulation Study Group. Persistent and

recurrent neovascularization after laser photocoagulation for subfoveal choroidal neovascularization of agerelated macular degeneration. Arch Ophthalmol 1994; 112(4): 489–99.

10.Fine SL, Berger JW, Maguire MG, Ho AC. Age-related macular degeneration. N Engl J Med 2000; 342(7):483–92.

11.Evans J, Wormald R. Is the incidence of registrable agerelated macular degeneration increasing? Br J Ophthalmol 1996; 80(1):9–14.

12.Vingerling JR, Dielemans I, Hofman A, et al. The prevalence of age-related maculopathy in the Rotterdam Study. Ophthalmology 1995; 102(2):205–10.

13.Klein R, Klein BE, Jensen SC, Meuer SM. The five-year incidence and progression of age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology 1997; 104(1):7–21.

14.Leibowitz HM, Krueger DE, Maunder LR, et al. The Framingham Eye Study monograph: an ophthalmological and epidemiological study of cataract, glaucoma, diabetic retinopathy, macular degeneration, and visual acuity in a general population of 2631 adults, 1973–1975. Surv Ophthalmol 1980; 24:335–610.

15.Ferris FL, III, Fine SL, Hyman L. Age-related macular degeneration and blindness due to neovascular maculopathy. Arch Ophthalmol 1984; 102(11):1640–2.

16.Hyman LG, Lilienfeld AM, Ferris FL, III, Fine SL. Senile macular degeneration: a case-control study. Am J Epidemiol 1983; 118(2):213–27.

17.Bressler SB, Bressler NM, Fine SL, et al. Natural course of choroidal neovascular membranes within the foveal

avascular zone in senile macular degeneration. Am J Ophthalmol 1982; 93(2):157–63.

18.Guyer DR, Fine SL, Maguire MG, et al. Subfoveal choroidal neovascular membranes in age-related macular degeneration. Visual prognosis in eyes with relatively good initial visual acuity. Arch Ophthalmol 1986; 104(5):702–5.

19.Macular Photocoagulation Study Group. Krypton laser photocoagulation for idiopathic neovascular lesions. Results of a randomized clinical trial. Arch Ophthalmol 1990; 108(6):832–7.

20.Barbazetto I, Burdan A, Bressler NM, et al. Photodynamic therapy of subfoveal choroidal neovascularization with verteporfin: fluorescein angiographic guidelines for evaluation and treatment—TAP and VIP Report No. 2. Arch Ophthalmol 2003; 121(9):1253–68.

21.Macular Photocoagulation Study Group. Argon laser photocoagulation for senile macular degeneration. Results of a randomized clinical trial. Arch Ophthalmol 1982; 100(6):912–8.

22.Macular Photocoagulation Study Group. Argon laser photocoagulation for ocular histoplasmosis. Results of a randomized clinical trial. Arch Ophthalmol 1983; 101(9): 1347–57.

23.Macular Photocoagulation Study Group. Argon laser photocoagulation for idiopathic neovascularization. Results of a randomized clinical trial. Arch Ophthalmol 1983; 101(9): 1358–61.

24.The Moorfields Macular Study Group. Treatment of senile disciform macular degeneration: a single-blind randomised trial by argon laser photocoagulation. Br J Ophthalmol 1982; 66(12):745–53.

25.Coscas G, Soubrane G. Argon laser photocoagulation of subretinal neovascularization in senile macular degeneration. Results of a randomized study of 60 cases. Bull Mem Soc Fr Ophtalmol 1982; 94:149–54.

26.Cukras C, Fine SL. Thermal laser treatment in AMD: therapeutic and prophylactic. Int Ophthalmol Clin 2007 Winter; 47(1):75–93. Review.

27.Yoken J, Duncan JL, Berger JW, et al. Laser photocoagulation for choroidal neovascularization in age-related macular degeneration. In: Lim JI, ed. Age-Related Macular Degeneration. New York: Marcel Dekker, 2002:181–201.

28.Macular Photocoagulation Study Group. Recurrent choroidal neovascularization after argon laser photocoagulation for neovascular maculopathy. Arch Ophthalmol 1986; 104(4):503–12.

29.Brown DM, Kaiser PK, Michels M, et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 2006; 355(14):1432–44.

30.Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006; 355(14):1419–31.

31.Rosenfeld PJ, Rich RM, Lalwani GA. Ranibizumab: phase III clinical trial results. Ophthalmol Clin North Am 2006; 19(3):361–72.

32.Ferrara N, Damico L, Shams N, et al. Development of ranibizumab, an anti-vascular endothelial growth factor antigen binding fragment, as therapy for neovascular agerelated macular degeneration. Retina 2006; 26(8): 859–70.

33.Ferrara N. Role of vascular endothelial growth factor in physiologic and pathologic angiogenesis: therapeutic implications. Semin Oncol 2002; 29(6 Suppl. 16):10–4.

INTRODUCTION

Feeder vessel treatment (FVT) of age-related macular degeneration (AMD)-related choroidal neovascularization (CNV); that is, occlusion of just the vessels delivering blood to CNV membranes has long been viewed as an attractive clinical approach, particularly when the neovascular membrane is very near or underlies the fovea. Available clinical evidence clearly indicates that this treatment approach is a successful one that, beyond lesion stabilization, often results in visual improvement. While in recent years the focus of an aggressive quest for an efficacious treatment for AMD-related CNV has been dominated by drug-based approaches, refinement of FVT has continued and reached the point that results from its application appear to rival and in some cases exceed those of other currently available methodologies. Moreover, as reaction to the unmet expectations of single-drug- based approaches has led to investigation of combining extant therapies, FVT—its attributes as a stand-alone treatment notwithstanding—is an attractive combination candidate, because it is not drug based, and it acts directly on the source of blood flow that must be present in every viable CNV membrane.

Although elegantly simple as a concept, successfully implementing a routine FVT method has been a protracted process. The history of its development spans a period of nearly 30 years, and the case can be made that its development has been coupled to the evolution of fundus angiography technology, especially choroidal angiography. Today FVT has been refined to take advantage of improvements not only in the devices used for angiogram acquisition and application of laser photocoagulation energy, but also in diagnostic angiogram analysis. In one method of FVT described here, even the method of applying laser energy to feeder vessels (FVs) has been optimized by introduction of dye-enhanced photocoagulation (DEP), wherein indocyanine green (ICG) dye transiting targeted FVs at the instant of photocoagulation acts

to selectively enhance absorption of the laser energy, thereby focusing the thermal tissue damage onto the targeted FV and sparing the surrounding tissues.

ORIGINS OF THE CONCEPT

Perhaps the earliest description of FVT in ophthalmology was in 1972 by Behrendt, who discussed argon laser photocoagulation of intraretinal and vitreous FVs of neovascular membranes associated with diabetic retinopathy (1). The then recent availability of visible light wavelength lasers led to numerous such novel approaches aimed at controlling ocular neovascularization. Understandably, all of those were related to retinal and anterior segment neovascularization, since they could be directly visualized by means of readily available optical devices. The choroidal vasculature, on the other hand, was not a popular target of interest, since direct visualization of it was obscured by retinal and choroidal pigments, and in sodium fluorescein angiography images it appeared mostly only as a diffuse “choriocapillaris (CC) flush.” The deeper-lying vascular layers remained obscured so far as routine clinical observations were concerned.

At about that same time, in the early 1970s, the concept of routine clinical angiography of the choroidal circulation using ICG dye was being developed. ICG fluorescence angiography initially had been explored as an investigative tool for studying choroidal blood flow in animal experiments. However, since ICG dye already had a long documented history of biocompatibility, exploring its use in human subjects as well was compelling. Since up to that time, relatively little attention had been paid to the choroidal circulation compared to the retina, there was no well-defined clinical goal at first in visualizing human choroidal blood flow beyond academic curiosity, so a rudimentary survey of both normal and diseased eyes was undertaken (2). One of the first groups of patients considered in the survey was those with macular degeneration.

Figure 1 shows simultaneously acquired fluorescein and ICG angiogram images of the first patient successfully studied by that methodology. The greatly improved ability to visualize the angioarchitecture of AMD-associated CNV lesions afforded by ICG angiography, coupled with the concept of FV photocoagulation, led to the first attempts at ICG-guided photocoagulation of CNV-FVs. Unfortunately, the results of those first attempts were not encouraging: clear differentiation between CNV afferent and efferent vessels was not easy—or in most cases not possible— since both spatial and temporal resolution of the early ICG fluorescence angiogram images was limited, the spot size and aiming precision of the first visible light laser photocoagulation delivery systems also was limited, and perhaps most importantly, the laser light wavelengths available were not ideally suited to the task.

For some time thereafter, the concept of FV photocoagulation was not seriously pursued as a clinical tool. Instead, the dominant treatment approach for AMD-associated CNV came to be based on Macular Photocoagulation Study (MPS) recommendations (3). These included destruction of the entire CNV membrane—as delineated by fluorescein angiography—along with an additional margin around the CNV, even when the procedure resulted in an immediate, non-recoverable additional loss of visual acuity (VA). The results of the MPS suggested

that despite an immediate vision loss, three years later a patient so treated statistically would have better VA than if untreated. Those results notwithstanding, few ophthalmologists remained comfortable with the notion of having to destroy the retina in order to save it, preferring for the most part to avoid photocoagulation near the fovea.

REVISITING THE CONCEPT

The first notable clinical application of ICG fluorescence angiography was its use in guiding laser photocoagulation of CNV. This method was applied to patients whose clinical and fluorescein angiographic features did not meet the eligibility criteria for laser therapy defined by the MPS recommendations; generally it was applied to cases of poorly defined, or occult, CNV. In this application, use of ICG angiography resulted in improved localization of abnormal choroidal vessels, thereby making treatment by photocoagulation possible (4,5). Whereas this clinical use of ICG undoubtedly contributed to sustaining interest in ICG angiography, arguably it was the commercial availability of the scanning laser ophthalmoscope (SLO) that contributed to increasing interest in ICG angiography. Compared to the predominantly available commercial ICG angiography systems based on fundus camera optics capable of acquiring images at a rate of about one per second,

the SLO afforded the ability to perform high-speed imaging. Ready access to high-speed ICG image acquisition systems was an important component of renewed interest in FV photocoagulation treatment.

The concept of FV photocoagulation was revisited as a treatment for AMD-associated CNV in February, 1998 by Shiraga and coworkers; they reported the results of a pilot trial to assess the feasibility of extrafoveal photocoagulation of subfoveal CNV secondary to AMD (6). Their use of SLO ICG angiography resulted in the identification of FVs in 37 of 170 consecutive patients (22%). In 70% of those 37 cases (26 cases) extrafoveal photocoagulation of the FVs, using 575to 630-nm wavelength light, resulted in resolution of the exudative manifestations and improved or stabilized VA. The following December, Straurenghi and coworkers (7), also using SLO ICG angiography, reported finding treatable FVs in 15 of 22 patients having subfoveal CNV not amenable to the treatment suggested by the MPS (3). They successfully obliterated the FVs in 40% of the cases, resulting in improved or stabilized VA after more than two years. In a second group of 16 patients they reported a much higher success rate of 75%, attributed to the smaller FV diameters (less than 85 mm) found in this group. In December 1997, yet another series of FV treatments was begun using a high-speed, pulsed-laser (HSPL) fundus camera system for FV identification. (Flower RW, Glaser BM, Murphy RP, Macula Soc. Presentation, 1999.) The HSPL used in this study consisted of a Zeiss fundus camera modified to include a pulsed 805-nm-wavelength diode laser for excitation of ICG dye fluorescence in the choroidal circulation; images were acquired at a rate of 30 per second (8). In this latter study, a higher incidence (about 66%) of FV identification was achieved, apparently due to use of the HSPL system and different angiogram analysis techniques. Nevertheless, treatment success of the latter study appears to be equivalent to that of the other groups, even though the follow-up period was shorter and it focused on occult CNV, whereas the other studies appear to have focused on classic CNV.

The common experience of all these studies was that FV photocoagulation appeared to be a viable treatment approach and worthy of continued pursuit, even though the exact nature of the vessels being treated and the most efficacious application of laser energy remain to be determined. Clearly, there is a catch-22 associated with this methodology. There are no histological data on treated CNV-FVs, per se, and the only proof currently available of the accuracy of angiographic CNV-FV identification is improvement or stabilization of the patient’s VA following treatment. But this standard of proof is biased toward failure, since conventional laser photocoagulation of CNV-FVs already has proven to be difficult or

incomplete in some cases. Therefore, if the full potential of FV treatment is to be accurately assessed and eventually realized, a more consistently successful approach to laser photocoagulation must be devised. And at the same time, a much better understanding of the hemodynamic consequences of FV photocoagulation must be developed in order to facilitate rational analysis of treatment successes and failures.

WHAT IS A FV?

Properly characterizing CNV-FVs in terms of their locations within the choroid, their vessel wall structure and the blood flow, is a necessary step in developing the most efficacious photocoagulation method. In that regard, however, histological data about CNV angioarchitecture appear to be at odds with the angiographic appearance of the so-called FVs being treated.

Histological Appearance of CNV-FVs

The vessels passing through breaks in Bruch’s membrane and connecting a CNV to the choroidal blood supply can be capillaries, arteries, or veins, as determined by the vessel wall structure. In general, CNV complexes up to 300 mm diameter have only one break containing a capillary-like vessel (9,10). Complexes on the order of 500 mm have two to four breaks, and at least one or two contain a capillary-like vessel; the others transmit only cells. CNV complexes of these dimensions consist of a single layer of capillary vessels on the inner surface of Bruch’s membrane, and they arise from a layer of vessels which lies just beneath, instead of between, the intercapillary tissue pillars; so it is assumed these are new vessels replacing the choroidal capillaries. Because many tissue sections must be cut to find and track these vessels, there are only a few examples in which the vessels can actually be tracked in the choroid, and even then it is not always clear whether they lead to an artery or a vein. (Sarks JP and Sarks SH, written communication, March 14, 1999). Complexes on the order of 2000 mm have more than four breaks, and the vessels passing through are of medium size. These complexes usually are two layers thick, but still lie beneath the retinal pigment epithelium (RPE), and they can be served by wellformed arterial and venous vessels. Complexes from patients with disciform scars have breaks containing larger arteries and veins; these disrupt the RPE and invaded the retina. (Sarks SP and Sarks SH, written communication, March 14, 1999).

It has been suggested that on average, there are 2.3 vessels passing through Bruch’s membrane and connecting each CNV to the underlying choroidal vasculature (11). The frequency with which these vessels are capillaries, arteries, or veins has not yet

been reported, but it is clear that the majority of penetrating vessels encountered are relatively short capillary-like vessels. It is clear also that such small vessels are not likely to be recognized in ICG angiogram images.

Angiographic Appearance of CNV-FVs

The most frequently identified and treated FVs reported in studies to date appear to be on the order of one to several millimeters long, a dimension quite large with respect to the penetrating vessels most frequently found in histological preparations. Figure 2 shows examples of FVs, identified using the HSPL fundus camera system, which have been successfully photocoagulated, resulting in improved or stabilized vision. In using that system, identification of FVs is made by first carefully examining the area surrounding the location of a known or suspected CNV complex in high-speed ICG angiogram images,

since the most obvious characteristic of a FV is proximity to CNV. Some FVs are easily identified, as in Figure 2 (top left and right), when they are prominent and easily distinguishable from adjacent choroidal vessels. Often, however, FVs are less prominent, as in Figure 2 (bottom left and right), and identification requires use of an analytical technique such as phi-motiona angiography, which helps

aPhi-motion is a phenomenon first identified by Wertheimer in 1912 (13); it refers to visual perception of motion where none exists. In a situation where there is a gap in visual information, the brain fills in what is missing. An example of the case in point is the appearance of two spatially separated points of light wherein first one is illuminated and, a finite time later, the second one is illuminated. The perception is that of a single point moving from the location of the first point to that of the second. By repeatedly viewing an appropriate segment of a high-speed angiogram image sequence in continuous loop fashion and at an appropriate speed, the phi-motion phenomenon accentuates perception of the movement of dye through vessels.

Figure 3 A scanning electron micrograph of a corrosion cast of the posterior (Sattler’s) layer of small diameter choroidal arteries and veins that feed and drain the choriocapillaris, which can be seen in the background. For the most part, the veins are oriented from the upper left-hand corner of the image toward the lower right-hand corner; they overlie the arteries. Source: From Ref. 12. Courtesy of Dr. Andrzej W. Fryczkowski.

differentiate a FV from its surroundings by enhancing visualization of blood flow through it, toward the CNV. Determining direction of flow is essential to correctly identifying CNV-FVs—as opposed to their draining vessels—even when their angioarchitecture seems obvious.

Reconciling Histological and Angiographic Data

Clearly, the vessels identified in histological specimens as the conduits of blood from the CC to CNVs appear to be different from the so-called FVs identified in

angiograms. Typically “FV” refers to an afferent vessel supplying blood to a particular vascular complex, one directly connected to the complex. To be precise, in the case of CNV that definition should apply to the short capillary-like vessels that penetrate Bruch’s membrane and form a CNV/CC connection. The vessels in ICG angiograms dubbed FVs in the recently reported studies of CNV-FV photocoagulation—especially in the case of occult CNV—meet the criterion of being afferent, but they appear to be much larger than the capillary-like vessels seen in the histological specimens. Strictly speaking, therefore, the term “CNVFV,” as applied in angiographic descriptions, appears to be a misnomer for some other choroidal vessel; most likely Sattler’s layer arterioles.

The so-called FVs seen in angiograms very much resemble vessels of the choroidal middle layer, or Sattler’s layer, which lies just beneath the CC. Comparison of the ICG angiogram images of the FVs in Figure 2 to the scanning electron micrographs of corrosion casts of the anterior aspect of the CC in Figure 3 demonstrates this similarity. Therefore, it seems a reasonable assumption that the FVs identified in ICG angiograms and reported to have been successfully treated by photocoagulation are Sattler’s layer arteriolar vessels.

There is additional evidence to support the notion that the angiographically-defined CNV-FVs are Sattler’s layer vessels: a commonly observed characteristic of successfully treated FVs is their “beaded” appearance in ICG angiograms (RP Murphy, symposium presentation, Chicago, June

212 FLOWER

3, 2000); an example of that appearance is shown in Figure 4A. The most likely explanation for the beaded appearance is that the dye-filled FV is crossed throughout its length by smaller non-dye-filled choroidal vessels. This same phenomenon is more pronounced in high-speed ICG angiograms of rhesus monkey eyes following carotid arterial dye injection, as demonstrated in Figure 4B, wherein carotid dye injection improves dye wave front definition, enhancing observation of the temporal filling differences between various layers of choroidal vessels. When crossed by small non-dye-filled vessels, the crossings result in dark segments along the FV; when crossed by small dye-filled vessels, the crossings result in hyperfluorescence, due to additivity of fluorescence from the overlapping vessels. The presence of small vessels between the FV and the CC fixes the FV location well below the CC, consistent with the notion that CNV-FVs are Sattler’s layer vessels.

Additionally, Arnold and coworkers (15) have shown the choroids of AMD eyes to be as little as half the thickness of those in age-matched normal eyes (e.g., 90 mm compared to 180 g), primarily due to a significant decrease in the number of vessels that normally occupy the middle choroidal layers (Sattler’s layer). So it is possible that the relatively high contrast of some FVs (Fig. 2) is a result of there being fewer than normal adjacent vessels in the same layer, and in the absence of the normal number of adjacent vessels, the FVs may have become preferential channels for blood flow through a diminished Sattler’s layer. Therefore, the assumption that many of the FVs investigators have identified and photocoagulated are Sattler’s layer arteriolar vessels is at least consistent with the evidence at hand.

THE RELATIONSHIP BETWEEN SATTLER’S LAYER

VESSELS (FVs) and CNVs

The explanation for apparently successful photocoagulation treatment of so-called CNV-FVs (i.e., Sattler’s layer vessels) lies in the hemodynamic relationship between the Sattler’s layer vessels and the capillarylike vessels that form the CC/CNV communication.

An Anthropomorphic Model of the CC/CNV Connection

The relationship proposed to exist between these two types of vessels is modeled in Figure 5, wherein there is no anatomical continuity between them, although functionally they behave as if there were. The figure also demonstrates how blood could move in a functionally contiguous manner from a Sattler’s layer FV, into the CC, and then through a nearby capillary vessel

leading to the CNV during the systolic phase of the intraocular pressure pulse.

By comparison to the sinusoid-like structure of the CC vascular plexus, it is likely that resistance to blood flow would be higher through a parallel CNV complex, connected to the CC by the capillary-like vessels that penetrate Bruch’s membrane. In this model, blood flow through the CNV would occur, but it would not be as great as through the underlying CC. In keeping with the pulsatile nature of CC blood flow shown to exist as the result of the perpendicular interface of arterioles and the wide, flat choriocapillaries (8,16), a high hydrostatic pressure head must exist at the interface early during systole, relative to the surrounding CC (as indicated by the collapsed state of the choriocapillaries and the CNV vessels in Fig. 5A,B). In addition to pushing blood into the choriocapillaries, the pressure head would be partially dissipated in forcing some blood into the adjacent penetrating vessel. Thus, a small, pulsatile pressure gradient would be established through the CNV, even though the majority of flow would be through the CC. In this model, closure of the FV or even significant partial closure would have the effect of reducing the pressure head available at the penetrating vessel to a level so low that resistance to flow through the CNV could not be overcome, thereby effectively closing the CNV as well.

Thus, there is considerable evidence to support the hypothesis that ultimately the source of blood supplying a CNV is a Sattler’s layer arteriole whose entry into the CC is situated near one of the capillarylike vessels that penetrate Bruch’s membrane, forming a CC/CNV communication. That is, the FVs identified for focal photocoagulation treatment of CNV appear to be Sattler’s layer arterioles that are functionally— but not directly physically—connected to the CNV. Throughout the rest of this discussion, the term CNVFV is intended to imply a Sattler’s layer vessel that is functionally contiguous with a CNV. This leads to the possibility that in some case a direct, anatomically contiguous connection between a Sattler’s layer vessel and a CNV eventually could evolve, obviating any CC involvement at all; indeed, such an evolution might be the path leading from occult to classic CNV.

A Model of the FV/CC/CNV Hemodynamic Relationship

The simple anthropomorphic model of FV/CNV blood flow described above was conceived to account for the clinically observed resolution of retinal edema following FV photocoagulation, even when only partial FV vessel closure is achieved (12). However, since the submacular CC is a true vascular plexus—fed and drained by multiple interspersed arteries and veins—a much more sophisticated

model is needed to describe the changes in CC blood flow beneath the CNV following FV photocoagulation. Therefore, a theoretical model for the human CC, based on available histologic and hemodynamic data, was developed to simulate the CC blood flow field before and after FV photocoagulation.b

Known angioarchitectural and hemodynamic parameters for the CC and CNV from the literature were used to construct the theoretical model of a section of submacular CC and a small overlying CNV membrane shown in Figure 6. The CC plexus consists of two parallel sheets separated by 7.5 mm, between which 10 in. diameter columns are placed at regular intervals, leaving 15 mm wide channels in between to simulate the CC plexus. Isolated, but well

bThis model was developed in collaboration with C. von Kerczek. L. Zhu, A. Ernest, C. Eggleton, and L.D.T. Topoleski from the Department of Mechanical Engineering University of Maryland, Baltimore, Maryland, U.S.A.

separated, precapillary arterioles and venules communicate with the CC plexus and perfuse it with blood. The cross-sectional dimensions of the arterioles and venules are of the same order as the CC thickness, h. The center-to-center spacing between adjacent arterioles and venules is much larger than h. Therefore, the CC was modeled as a planar porous medium containing a widely dispersed set of fluid inflows and outflows, simulating the feeding and draining vessels of Sattler’s layer. Feeding arteriolar and draining venous vessels consist, respectively, of 7.5 and 15 mm diameter tubes entering the CC from beneath.

An overlying CNV membrane was modeled as a parallel miniature version of the CC, but with smaller dimensions that will result in a significantly higher resistance to fluid flow. The communication between the CNV and the CC is by way of two capillary-dimensioned vessels that penetrate Bruch’s membrane. In the model, the position of the CNV

A 1

A 2

CNV 1

V 1

A 3 CNV 2

V 2

500 μm

A 4

Figure 7 The anterior aspect of the computer simulated segment of a human submacular choriocapillaris, marked with the actual locations of arteriolar and venous vessels Sattler’s layer vessels connected to its posterior aspect; the figure also shows the simulated choroidal neovascularization in two different locations. Abbreviation: CNV, choroidal neovascularization. Source: From Ref. 14.

could be changed in order to achieve various spatial relationships between the penetrating vessels and the Sattler’s layer vessels that feed and drain the CC.

This theoretical model became the basis for computer simulation of blood flow distribution in a segment of human sub-foveal CC approximately 1300!1000 mm in area. The actual placement of the multiple Sattler’s layer vessels to feed and drain blood from the simulated CC plexus segment was made according to the histologically determined locations of those vessels in one normal human eye (17). Figure 7 shows the anterior aspect of the computer simulated segment of that human submacular CC, marked with the actual locations of arteriolar and venous vessels, Sattler’s layer vessels connected to its posterior aspect; the figure also shows the simulated CNV in two different locations. Blood flow rates in the feeding arterioles and venules were then estimated by matching the predicted precapillary arteriole and venule pressure difference to experimentally measured data; the experimentally measured maximum pressure difference between a feeding arteriole and venule was found to be 4.5 mmHg (18).

216 FLOWER

Experimentally measured pressures and pressure differences were applied across the feeding and draining vessels in order to generate maps of blood flow through the computer-simulated model CC segment. Figure 8 shows the normal isobar and iso- blood-speed distributions in the computer simulated segment of CC from Figure 7; it also shows how those distributions are altered when one of the Sattler’s layer feeding arterioles is completely occluded.

A significant reduction in the local CC pressure probably results in significant changes in the blood flow through an overlying CNV network, since the driving force for CNV blood flow is the pressure difference between the capillary-like vessels that penetrate Bruch’s membrane, forming the CC/CNV communication. Clinical observations indicate that partial—as well as complete—photocoagulation of the (presumed Sattler’s layer) FV adjacent to a CNV’s penetrating vessel(s) is an effective means of decreasing the blood flow in the CNV (BM Glaser, RP Murphy, G Staurenghi, personal communications, 1999). Therefore, the model also was used to simulate blood flow through a CNV before and after FV laser photocoagulation; the simulation was performed for the CNV membrane situated in two different locations, as indicated in Figure 7. The first location, CNV #1, was between arteriole #2 and venule #1, while the second, CNV #2, was between arteriole #3 and a point equidistant from venules #1 and #2. Photocoagulation of arteriole #2 and of venule #1 resulted in significant reduction of CNV #1 blood flow (71% and 79%, respectively), with similar results in CNV #2 when arteriole #3 was photocoagulated (84% reduction). On the other hand, even the complete closure of venules #1 or #2 produced less than 30% decrease in blood velocity through CNV #2.

Implications of the FV/CC/CNV Hemodynamic Relationship

This model predicts that even 50% closure of a blood vessel entering the posterior aspect of the CC in the vicinity of a capillary-like vessel leading to a CNV can be effective in reducing or possibly stopping CNV blood flow, regardless of whether that vessel is a feeding arteriole or a draining venule. In other words, the important hemodynamic event with respect to reducing or stopping CNV blood flow is significant reduction of the blood pressure—hence, blood flow as well—in the local underlying CC. Thus, the predictions of the present computer simulated model support the novel approach to CNV management made previously, namely that (i) rather than total obliteration of a CNV (which frequently results in recurrence), the end point of laser photocoagulation treatment can be reduction of CNV blood flow to the extent that undesirable

manifestations of the CNV—most notably retinal edema—are halted or reversed and (ii) that CNV blood flow reduction can be mediated by reduction of blood flow through the underlying CC (12).

There are two important implications to that novel approach, one related to FV treatment and the other related to the mechanics of successful CNV treatments in general. Regarding FV photocoagulation treatment of CNV, the selection criterion for targeted FVs might be extended to include venous as well as arteriolar vessels entering the posterior CC in the vicinity of a CNV membrane. If, indeed, reduction of the underlying CC blood flow is the important treatment goal, then depending upon the orientation of the CNV’s penetrating vessels with respect to the field of vessels feeding and draining the CC, targeting veins or veins in conjunction with arteries may yield the best results. After all, the ramifications of occluding a venous drainage channel to a true vascular plexus, like the posterior pole CC, is not the same as occlusion of the drainage vein of a true end-arteriolar vascular complex. In the former case, blood is diverted to adjacent venous channels, without excessive increase in capillary transmural pressure; whereas in the latter case, venous occlusion likely results in blood flow stasis and elevation of capillary transmural pressure to a level near that across the feeding arterial vessel wall.

Since the predicted relationship between CC and CNV blood flows actually is independent of the specific means by which CC blood flow is reduced, the second implication of the results is that reduction of CC blood flow underlying a CNV membrane may be a component mechanism common to successful CNV photocoagulation treatments, including photodynamic therapy (PDT), transpupillaty thermal therapy (TTT), and drusen photocoagulation. It is well established that post-PDT angiograms routinely evidence reduced CC fluorescence (19), and that appears also to be the case following TTT (20). In the case of TTT, reduced CC blood flow may be due to increased resistance to plexus blood flow resulting from heat-induced interstitial tissue swelling and concomitant reduction of CC lulninal space. Angiographic data specifically related to submacular blood flow following photocoagulation destruction of macular drusen have not been presented anywhere; however, it has been demonstrated that CC obliteration occurs with application of moderate to heavy laser burns and that loss of choriocapillaries can add significant resistance to blood flow through the CC plexus (8).

If reduced CC blood flow is a component mechanism of successful CNV treatment, regardless of the photocoagulation modality used, then FV photocoagulation arguably might be viewed as the

most effective method. The difference between FV photocoagulation and the other methods is analogous to removing a weed from a lawn by pulling out its roots (FV) versus just cutting off the weed’s leaves. It can be argued that FV photocoagulation is the most precise of the various methods in terms of manipulating CC blood flow, and it minimizes the area of tissue–laser interaction. Moreover, since blood flow through a particular CC area apparently can be manipulated by modulation of adjacent venous or as arteriolar vessels connected to the plexus’ anterior side, it may be that the most precise manipulation of CC blood flow—and hence, treatment of CNV—will be by controlled, partial photocoagulation of carefully selected combinations of arterioles and venules in Sattler’s layer vessels.

DEVELOPMENT OF A MORE EFFICACIOUS

METHOD OF FV TREATMENT

The models of CNV-FVs are consistent with the clinical observation that often, even incomplete closure of a FV produces reduction of CNV dye filling, resolution of associated edema, and improved VA. Of course, partial closure of targeted FVs at present is an unintended end-point of Argon and Krypton laser photocoagulation application. In such cases, failure to completely close the relatively deep-lying targeted vessels may be attributable to generation of an insufficiently high temperature gradient, emanating from the RPE where laser light-to-heat transduction occurs. The temperature gradient that is produced does extend into the sensory retina and can produce significant damage there, so the location for FV photocoagulation must be chosen so as not to involve the fovea. It would be desirable, therefore, to avoid the concomitant retinal damage and to make FV photocoagulation more efficient and predictable. This would have the additional potential benefit of allowing such treatment to be applied much closer to the fovea than is presently possible, thereby increasing the number of patients who might benefit from CNV-FV treatment.

The Concept of ICG-DEP

An example of a successfully treated FV is shown in Figure 9, and it also shows an undesirable side effect as well: damage to the nerve fiber layer overlying the site of FV photocoagulation. Since CNV-FVs apparently lie below the plane of the CC, a method of photocoagulation that moves the epicenter of the lasergenerated heat closer to those vessels and away from the sensory retina would be an improvement over the presently available method. The concept of ICG-DEP has that potential and, therefore, should be revisited for this application, bearing in mind that its use must be optimized to accommodate characteristics of the

targeted choroidal vasculature. The main premise of DEP is that application of laser light energy with a wavelength matched to the primary wavelength absorbed by a bolus of dye passing through the target blood vessel produces the most efficient

(A)

(B)

(C)

Figure 9 Post-treatment indocyanine green angiogram images of a successfully treated feeder vessel. (A) Pre-treatment: the FV is indicated by asterisk. (B) Post-treatment: note lack of CNV filling. (C) Image shows an undesirable side effect as well: damage to the nerve fiber layer overlying the site of FV photocoagulation. Source: From Ref. 12.

218 FLOWER

photocoagulation burn in terms of vessel closure with minimum damage to surrounding tissue. Figure 10 demonstrates the main aspects of ICG-DEP and compares it to FV photocoagulation by conventional laser light photocoagulation. The concept of improving the efficiency of the photocoagulation process by ICG-dye enhancement is not new to

Conventional Visible Light Photocoagulation

Retina

RPE

(A)

ICG Dye Enhanced Photocoagulation

Retina

RPE

(B)

(C)

Figure 10 Schematic comparison of choroidal vessel photocoagulation by (A) conventional laser and (B) ICG dye-enhanced laser and (C) ICG angiogram image made immediately posttreatment with ICG-DEP demonstrating incarceration of ICG dye in the treated feeder vessel (arrow) and choroidal neovascularization membrane (circle). Abbreviations: ICG, indocyanine green; RPE, retinal pigment epithelium. Source: From Ref. 12. Courtesy of Dr. B. Eric Jones, Baltimore, Maryland, U.S.A..

treatment of AMD-related CNV, as Reichel and coworkers utilized it for treating poorly defined subfoveal CNV. Eventually they reported their initial clinical investigation in 10 patients (21), but in terms of visual outcome, their results were equivocal, and the technique did not achieve widespread use. The particular dye-enhancement technique they used relied on absorption of infrared laser light energy by dye-stained choroidal blood vessel walls minutes following dye injection. That apparently is a very inefficient process, compared to the one in which the same laser energy is absorbed by dye molecules within the target vessels during transit of a highconcentration dye bolus (12).

A Combined ICG Angiography/DEP System

Performance of ICG-DEP requires use of a laser delivery system that permits visualization of inter- venously-injected ICG dye as it traverses the vasculature. Such a system was constructed from a Zeiss fundus camera (Carl Zeiss, Oberkochen, Germany) modified to include a pulsed diode laser light source and a synchronized, gated CCD camera for performing high-speed ICG angiography, as previously described (8,22). The fundus camera was further modified so that the output tip of the fiber optic of an 810 nm diode laser photocoagulator (Oculight SLx, Iris Medical Instruments, Mountain View, California, U.S.A.) can be positioned in the plane of the fundus illumination optics pathway normally occupied by the internal fixation pointer; that plane is conjugate to the fundus of the subject’s eye. The He–Ne aiming beam emitted by the photocoagulator appears as a sharply focused spot when viewed through the fundus camera’s video system, and the position of the fiber optic with respect to the subject’s fundus can be controlled by the micromanipulator’s X- and Y-adjustments. With this configuration, it is possible to deliver 810 nm photocoagulation light pulses to precisely located areas of the fundus while observing the fundus with visible light through the fundus camera eyepiece, making it possible to synchronize photocoagulation laser pulse delivery with arrival of a dye bolus at a targeted vessel site. The fundus camera/laser photocoagulation system is shown in Figure 11.

Clinical Application of ICG-DEP

Use of the ICG dye-enhanced camera system is demonstrated in the three frames of Figure 12, which show ICG angiogram images from a patient treated with ICG-DEP. Incarceration of ICG dye immediately following laser photocoagulation (center frame) not only provides immediate feedback as to the success of the procedure, but the

incarcerated dye constitutes as a strongly absorbing target for further laser application without the need to inject additional dye boluses. The reduction in retinal tissue damage concomitant to FV laser photocoagulation using ICG dye-enhancement is demonstrated in Figure 13, which compares the extent of RPE damage resulting from application of identical laser burns to identical choroidal arteries of a rabbit eye, one with and one without presence of a transiting high-concen- tration dye bolus.

Recently, a single center, prospective, randomized study of FVT using ICG-DEP was conducted by Dr. G. Staurenghi (University of Brescia, Italy) under the auspices of Novadaq Technologies, Inc. (Toronto, Canada). The objective of the study was to evaluate the safety and effectiveness of choroidal FV closure in the presence of ICG using the above described fundus camera/laser photocoagulation system. In the study, forty patients were evaluated for presence of visible FVs associated with CNV. Upon identification of the FVs, the patients were randomized into one of two treatment arms: one group of 20 patients was treated by choroidal FV photocoagulation during ICG dye transit (ICG-DEP arm), the other group of 20 patients (Control arm) was identically using the same device system, but FV photocoagulation was done without ICG-DEP, using the laser energy alone. All patients were followed and/or treated at 2, 4, 8, 12 weeks, and 6 months; with 1 additional follow-up at 12 months post-first treatment.

The study demonstrated that the fundus camera/laser photocoagulation system was easy to use, and that treatment session times decreased with experience with the system. The entire diagnostic, treatment and post-treatment confirmation ICG angiography took 21 to 23 minutes; this was similar for both treatment arms. On average, four to five treatment sessions were required for complete treatment in both arms over the course of the study. And on average, the ICG-DEP arm used approximately seven times less energy/treatment session than the Control arm (5.7 J per treatment session versus 38.9 J per treatment session) to close targeted choroidal FVs.

Importantly, treatment was more effective and more durable in the ICG-DEP arm, as 90% of the patients were able to have their choroidal FVs closed or partially closed, with 70% of those vessels remaining closed at the last treatment assessment, compared to 77% and 44%, respectively, in the Control arm. During the course of the study, 45% fewer patients in the ICG-DEP arm went on to require alternative treatments for their wet AMD than patients in the Control arm. VA at the end of the treatment phase of this trial, as measured by the Early Treatment Diabetic Retinopathy Scale, showed that, for the whole treated population, on average the VA was stable, and 29% of all patients seen at this study milestone had an improvement in VA. Of those patients who completed the study as per the study prescribed treatment regimen, at the last scheduled treatment visit, 67% had stable or improved

VA, with 42% having one to four line improvement in VA, while 33% had a decrease of more than three lines of VA: none experienced severe vision loss (more than six lines of VA). Of the nine patients who followed the study prescribed treatment regimen and had a VA equal or better than 20/100 at entry, seven (78%) had stable or improved VA at the last treatment visit, with four (44%) having a one to four line improvement in VA and two (22%) had more than three line decrease in VA.

Overall, the study added to the body of data demonstrating the efficacy of the concept of FVT of wet AMD. In addition, it demonstrated that the fundus camera/laser photocoagulation system simplifies FV treatment by allowing for real-time visualization of choroidal FVs during treatment. Moreover, FV photocoagulation with ICG-DEP produced a more effective and more durable treatment outcome than FV photocoagulation using laser only.

THE FUTURE OF CNV-FV TREATMENT

The current anti-vascular endothelial growth factor (anti-VEGF) drugs, Avastin or Lucentis, have experienced significant clinical success to date. It appears that these anti-VEGF drugs have such strong anti-per- meability effects on CNV membrane vessels that fluid outflow into surrounding tissues is reduced or stopped, resulting in stabilized or even improved VA. This can occur early, before the CNV angioarchitecture is substantially changed in the process, leaving some CNV blood flow intact; but repeated injections are needed. Interestingly, this is analogous to what happens in FVT, where only partial FV closure occurs or where reperfusion recurs following complete closure. Apparently, even partial FV closure results in reduced transmural pressure across CNV membrane vessels, which in turn reduces fluid outflow. In these cases, CNV angioarchitecture also appears substantially unchanged, and there is no concomitant recurrent edema, resulting in stabilized or improved VA. It has been postulated that during the period of reduced transmural pressure, neovascular membrane maturation progresses to a level of vessel structural integrity such that fluid outflow no longer occurs once the higher pre-FVT transmural pressures are reestablished.

If the foregoing understanding of the methods of action of the anti-VEGF and FVTs continues to hold true, then their use in combination might prove to be symbiotic in a way that leads to a very effective treatment approach, since both act to reduce edema resulting from CNV membrane fluid outflow, but by different pathways. However, as a stand-alone treatment, FVT ultimately still may prove to be the most desirable approach, since even when repeated treatments are applied, those treatments are totally non-

invasive with respect to the peripheral retina and wall of the eye itself, and they are inexpensive. Moreover, consideration should be given to the long-term ramifications of successfully achieving the currently sought clinical treatment endpoint, namely CNV obliteration.

To the extent that CNV (especially “occult” CNV) serves to augment or replace functionally compromised choriocapillaries, successful destruction of the CNV ultimately would leave the adjacent sensory retina and RPE without adequate metabolic support from the choroidal circulation. In that situation, the RPE and retina likely would atrophy. It is for this reason that CNV blood flow obliteration as the treatment endpoint should be reconsidered in favor of modulating CNV blood flow just to the point that retinal edema is ameliorated, since that leaves a level of choroidal metabolic support for the RPE and retina in place. Owing to FVT’s highly localized application and the ability it affords for immediate CNV blood flow assessment, DEP-FVT allows for a level of individual patient treatment titration that drug-based treatment cannot provide.

Aggressive CNV behavior—rapid membrane growth, edema formation, etc.—has been viewed as a destructive event, and conventional treatment aims to remedy such behavior by complete CNV obliteration. But the frequent recurrence of CNV following such treatment could be nature’s continuing effort to compensate for the original—and perhaps now exacerbated—defect. Instead, such aggressive CNV behavior could be viewed as an over compensation for some metabolic or other blood flow related defect. And if laser treatment were to be applied in such a way as to just reduce the blood flow to aggressive CNV by an appropriate amount—perhaps until the CNV vasculature matures—then further aggressive behavior might be avoided; those cases of inadvertent incomplete FV closure resulting in improved vision would be examples.

Photocoagulating the FVs supplying CNV associated with AMD not only can be a successful treatment method (6,7) especially for occult CNV. Indeed, there may be an important difference between the response of CNV evoked by direct application of laser energy, as in conventional treatment, and that evoked by reducing blood flow through the otherwise undisturbed membrane. If ultimately FV photocoagulation treatment were to be refined along these lines, the laser would become more a precision instrument to modulate blood flow than a weapon for destruction of the very retinal tissue whose function we are trying to conserve. Additionally, because of the preand posttreatment high-speed ICG angiograms the method requires, information about choroidal hemodynamics is being accrued that otherwise probably would never be available.

222 FLOWER

SUMMARY POINTS

&Identification of FVs is made by first carefully examining the area surrounding the location of a known or suspected CNV complex in high-speed ICG angiogram images.

&It is a reasonable assumption that the FVs identified in ICG angiograms and reported to have been successfully treated by photocoagulation are Sattler’s layer arteriolar vessels.

&Clinical observations indicate that partial—as well as complete—photocoagulation of the (presumed Sattler’s layer) FV adjacent to a CNV’s penetrating vessel(s) is an effective means of decreasing the blood flow in the CNV.

&FV photocoagulation may be the most precise method of manipulating CC blood flow and minimizes the area of tissue/laser interaction.

&CNV blood flow eradication as the treatment endpoint should be reconsidered and replaced with modulation of the CNV blood flow just to the point that retinal edema is ameliorated, since that leaves a level of choroidal metabolic support for the RPE and retina in place.

REFERENCES

1.Behrendt T. Therapeutic vascular occlusions in diabetic retinopathy. Arch Ophthalmol 1972; 87:629–33.

2.Patz A, Flower RW, Klein ML, et al. Clinical application of indocyanine green angiography. In: DeLaey JJ, ed. International Symposium on Fluorescein Angiography. Documenta Ophthalmologica Proceedings Series. Vol. 9. The Hague: Dr. W. Junk b.v., 1976:245.

3.Macular Photocoagulation Study Group. Subfoveal neovascular lesions in age-related macular degeneration: guidelines for evaluation and treatment in the macular photocoagulation study. Arch Ophthalmol 1991; 109:1242–57.

4.Slakter JS, Yannuzzi LA, Sorensen JS, et al. A pilot study of indocyanine green videoangiography-guided laser treatment of primary occult choroidal neovascularizaton. Arch Opthtalmol 1994; 112:465–72.

5.Schwartz S, Guyer DR, Yannuzzi LA, et al. Indocyanine green videoangiography guided laser photocoagulation of primary occult choroidal neovascularizaton in age-related macular degeneration. Invest Ophthalmol Vis Sci 1995; 36:S244.

6.Shiraga F, Ojima Y, Matsuo T, et al. Feeder vessel photocoagulation of subfoveal choroidal neovascularization secondary to age-related macular degeneration. Ophthalmology 1998; 105:662–9.

7.Staurenghi G, Orzalesi N, La Capria A, et al. Laser treatment of feeder vessels in subfoveal choroidal neovascular membranes: a revisitation using dynamic indocyanine green angiography. Ophthalmology 1998; 105:2297–305.

8.Flower RW. Extraction of choriocapillaris hemodynamic data from ICG fluorescence angiograms. Invest Ophthalmol Vis Sci 1993; 34:2720–9.

9.Sarks SH. Aging and degeneration in the macular region: a clinicopathological study. Br J Ophthalmol 1976; 60:324–41.

10.Schneider S, Greven CM, Green WR. Photocoagulation of well-defined choroidal neovascularization in age-related macular degeneration. Retina 1998; 18:242–50.

11.Green WR, Enger C. Age-related macular degeneration: histopathologic studies. The 1992 Lorenz E Zimmerman lecture. Ophthalmology 1993; 100:1519–35.

12.Flower RW. Experimental studies of indocyanine green dye-enhanced photocoagulation of choroidal neovascularization feeder vessels. Am J Ophthalmol 2000; 129:501–12.

13.Wertheimer M. Experimentelle Studien ueber das Sehen von Bewegung. Z Psychol 1912; 61:161–265.

14.Flower RW, von Kerczek C, Zhu L, Ernest A, Eggleton C, Topoleski LDT. A theoretical investigation of the role of choriocapillaris blood flow in treatment of sub-foveal choroidal neovascularization associated with age-related macular degeneration, copyright 2001 (in press).

15.Arnold JJ, Sarks SH, Killingsworth MC, et al. Reticular pseudodrusen: a risk factor in age-related maculopathy. Retina 1995; 15:183–91.

16.Flower RW. High-speed ICG angiography. In: Yannuzzi LA, Flower RW, Slakter JS, eds. Indocyanine Green Angiography. Mosby: St. Louis, 1997:86–94.

17.Fryczkowski AW, Sherman MD. Scanning electron microscopy of human ocular vascular casts: the submacular choriocapillaris. Acta Anat 1988; 132:265–9.

18.Maepea O. Pressures in the anterior ciliary arteries, choroidal veins and choriocapillaris. Exp Eye Res 1992; 54:731–6.

19.Flower RW, Snyder WA. Expanded hypothesis on the mechanism of photodynamic therapy action on choroidal neovascularization. Retina 1999; 19:365–9.

20.Reichel E, Berrocal AM, Ip M, et al. Transpupillary thermotherapy (TTT) of occult subfoveal choroidal neovascularization in patients with age-related macular degeneration. Ophthalmology 1999; 106:1908–14.

21.Reichel E, Puliafito CA, Duker JS, et al. Indocyanine green dye-enhanced diode laser photocoagulation of poorly defined subfoveal choroidal neovascularization. Ophthalmic Surg 1994; 25:195–201.

22.Flower RW. Variability in choriocapillaris blood flow distribution. Invest Ophthalmol Vis Sci 1995; 36:1247–58.

INTRODUCTION

Photodynamic therapy (PDT) is a therapeutic modality that entails the administration of a photosensitizer with its subsequent accumulation in the target tissue and then its activation by non-thermal monochromatic light corresponding to the sensitizer’s absorption profile (1). Powerful oxidizing agents such as cytotoxic singlet oxygen and free radicals are produced causing irreversible cellular damage. PDT has traditionally focused on the treatment of cancer (2), but the potential for selective destruction of diseased vessels, while sparing normal overlying tissues, coupled with promising clinical efficacy, resulted in its use for the treatment of age-related macular degeneration (AMD), particularly subfoveal choroidal neovascularization (CNV). PDT selectivity for the CNV is achieved both through photosensitizer retention in CNV new vessels and through targeted light application. Illumination is restricted to the diseased area and the limited depth of light penetration restricts damage to underlying tissues.

VASCULAR TARGETING

PDT has been used successfully in the treatment of certain cancers due to the remarkable selectivity of many photosensitizers for tumor tissue. PDT causes direct cellular injury in addition to microvascular damage or “shutdown” within the illuminated tumor. Uptake is considered to be due to the increased expression of low-density lipoprotein receptors on tumor cells and neovascular endothelial cells. Porphyrin photosensitization in mammals was studied as early as 1910 when Hausmann investigated the effects of hematoporphyrin and light on mice (3). The results established the phototoxic propensity of

porphyrins, and Hausmann concluded that the peripheral vasculature was one of the primary PDT targets. In 1963, Castellani and coworkers demonstrated the microvasculature to be a crucial target (4). PDT-mediated neovascular damage became a mainstay in the treatment of wet AMD and has only recently began to be replaced by newer anti-vascular endothelial growth factor (VEGF) therapies.

Endothelial cells accumulate certain photosensitizers and are susceptible to PDT-induced destruction. The subcellular localization of motexafin lutetium (Lu-Tex) was determined in human umbilical vein endothelial cells using fluorescence microscopy. Lu-Tex exhibits a fluorescence emission profile at 750 nm and this signature fluorescence marker is used to characterize and quantify sensitizer concentrations within the tissues. Lu-Tex was found to localize within the lysosomes and endoplasmic reticulum as evidenced by co-staining with organelle-specific fluoroprobes. Following illumination, some relocalization of the sensitizer occurred with partitioning being observed in the mitochondria, suggesting that the primary subcellular localization site could not possibly fully account for all of the PDT-induced damages. Sensitizer-alone and light administrationalone treatment groups did not induce any changes in the cell viability. Significant cell death due to Lu-Tex-mediated PDT was observed in endothelial cells producing a steep dose response.

Vascular occlusion following PDT is marked by the release of vasoactive molecules, vasoconstriction, blood cell aggregation, endothelial cell damage, blood flow stasis, and hemorrhage. The response is dependent on sensitizer type, concentration, and the time interval between administration and treatment. Benzoporphyrin derivative monoacid ring A (BPD-MA)-induced

PDT resulted in selective destruction of tumor microvasculature in a chrondosarcoma rodent model when compared with the surrounding normal microvasculature; illumination was applied within 30 minutes following sensitizer administration (5). However, no acute change in vascular status was observed when illumination occurred at three hours. The vascular shutdown results correlated with the anti-tumor effect since tumor-bearing animals treated at five minutes responded more positively than those treated at three hours.

LIGHT APPLICATION

The light used for ophthalmic applications is nonthermal monochromatic laser light matched to the sensitizer’s far-red (infrared) absorbance profile. Infrared light possesses greater transmission through both blood and tissue than light at lower wavelengths thereby enabling the treatment of pigmented or hemorrhagic lesions. The energy at which light is delivered is a product of the radiant power (expressed in milliwatts per square centimeter, mW/cm2) and the time of illumination. The radiant energy, often termed fluence, is expressed as joules per square centimeter (J/cm2). Therefore, to deliver a fluence of 50 J/cm2 light at a power density of 600 mW/cm2, an illumination time of 83 seconds is required.

Upon illumination, photons (hy) interact with the ground singlet state sensitizer (1Sensitizer) causing it to undergo an electronic transition to an activated short-lived excited singlet state (1Sensitizer*). The singlet state can then either convert back to the ground state causing fluorescence or undergo intersystem crossing to generate the longer-lived excited triplet state sensitizer (3Sensitizer*). From the triplet state, a photon can be emitted causing phosphorescence with conversion to the ground state or the triplet state can interact with oxygen or biological substrates leading to microvascular damage (6,7). Two photooxidation processes can occur between the triplet state and molecular oxygen (3O2) causing irreversible damage to vascular components. The direct interaction of the excited triplet state with biomolecular substrates is termed the type-I mode and is favored in areas with low oxygen concentrations. Biomolecular radicals are generated and react with oxygen-forming cytotoxic oxidizing products. The type-II mechanism entails interaction from the excited triplet state sensitizer to ground state oxygen-producing singlet oxygen ð1OÞ with theoretical regeneration of the ground state sensitizer. However, photobleaching and photoproduct formation can deplete the ground state sensitizer concentration.

Singlet oxygen is highly electrophilic, oxidizing biological substrates and initiating a cascade of radical

chain reactions that damage cellular components. Singlet oxygen production is thought to be responsible for most of the damage induced by PDT. Singlet oxygen possesses a reactive path length of less than 0.02 mm so that any effect has a limited potency (2). The photochemical processes involved are complex and are different for each sensitizer and are also subject to the microenvironment. Intersystem crossing is kinetically important for the formation of the excited triplet state and for PDT potency. Molecules with high fluorescence quantum yields will generate lower triplet quantum yields and are more likely to be used as diagnostic agents. Conversely, molecules with low fluorescence quantum yields will generate high triplet quantum yields and therefore should produce a high yield of cytotoxic species.

PDT AGENTS

The ideal photosensitizer should be chemically pure and possess the appropriate physical and biologic properties that make it inherently non-toxic until activated by light. The agent should possess strong absorption properties in the far-red spectral region (660–780 nm) where light has greatest penetration into blood and tissue and possess efficient photophysical properties for destroying neovascular endothelial cells. The sensitizer should also localize selectively in the neovasculature while being rapidly cleared from the blood and overlying photoreceptors. In addition, rapid cutaneous clearance would limit cutaneous photosensitivity. Several photosensitizers were explored and underwent different stages of preclinical and clinical development.

Photosensitizing candidate molecules are generally related to porphyrins. Porphyrins are fused tetrapyrrolic macrocycles that are omnipresent in nature as major biological pigments. Protoporphyrin IX, a typical porphyrin molecule, forms the nonprotein portion of hemoglobin. Reduction, oxidation, or expansion of the macrocyclic ring leads to different molecular subclasses. A reduction at one of the four pyrrole rings in the porphyrin macrocycle yields a chlorin molecule. The electronic conjugation system is altered causing further absorption into the far-red wavelength region, from 630 to approximately 660 to 690 nm. Increasing the macrocycle conjugation system further, by the formation of a pentadentate, metallophotosensitizer yields a texaphyrin molecule and results in even further absorption in the far-red spectral region (700–760 nm). Phthalocyanines are tetrapyrrolic structures fused together by nitrogen atoms instead of carbon bridges; absorption is exhibited in the 650to 700-nm wavelength region. Purpurins possess a reduced pyrrole ring and also

an extended ring conjugation system; the absorption maxima is between 650 and 690 nm.

BENZOPORPHYRIN DERIVATIVE MONOACID (VERTEPORFIN, VISUDYNEe, BPD-MA)

BPD-MA consists of equal amounts of two regioisomers that differ in the location of the carboxylic acid and methyl ester on the lower pyrrole rings of the chlorin macrocycle. BPD-MA, due to its hydrophobicity, is formulated with liposomes. The monoacid analogues were developed because they produced greater PDT responses compared with the diacids (8). The monoacid regioisomers are converted, in the liver, to the diacids. The regioisomers responded similarly in experimental efficacy settings; however, the pharmacokinetic properties were different in the rat, dog, and monkey but not in humans, where the plasma half-life was five to six hours (9,10). It is thought the latter may be due to differences in plasma esterases or lipoprotein profiles.

PDT studies undertaken using experimentally induced CNV in primates resulted in closure of the neovasculature and choriocapillaris, but not the retinal vasculature.

Liposomal BPD-MA was infused at a dose of 0.375 mg/kg for 10 to 32 minutes. Illumination with infrared light at a fluence of 150 J/cm2 (689–692 nm laser light at 600 mW/cm2) occurred 30 to 55 minutes following the start of the infusion (7). When the same treatment parameters were performed on normal primate eyes, some retinal pigment epithelium (RPE) damage and choriocapillaris closure occurred locally with little damage in contiguous tissues. When light was delivered within 30 to 45 minutes following sensitizer delivery, sensitizer administration rates had little effect on vascular occlusion rates. BPD-MA localization in the choroid and RPE was confirmed using fluorescence microscopy in rabbits. Retention occurred within five minutes with progression to the outer segments within 20 minutes. No BPD-MA was detected within the choroid or photoreceptors at two hours; however, a small trace was detected in the RPE at 24 hours (11). A similar pharmacokinetic pattern was observed in monkeys using in vivo fluorescence imaging (12).

The long-term effects on the retina and choroid were evaluated in cynomolgus monkeys with experimental CNV (13). Fundus photography and angiography analyses were performed at 24 hours and then weekly for four to seven weeks following a treatment with 0.375 mg BPD-MA/kg and a fluence of 150 J/cm2. Eyes were examined histologically at the end of the follow-up period. CNV closure also resulted in the closure of the choriocapillaris with damage occurring to RPE cells. However, these areas appeared

to regenerate somewhat in the four to seven weeks study period. Of 28 CNV lesions followed for four weeks, 72% remained closed.

However, lesion retreatment was necessary to sustain vascular closure. The effect of three different dosing treatments was evaluated in disease-free primate eyes (14). Treatments, using sensitizer doses of 6, 12, or 18 mg/m2, 20 minutes after drug infusion and a fluence of 100 J/cm2 were performed every two weeks. A cumulative dose response was observed. Damage to the retina, choroid, and optic nerve was limited in the 6 mg/m2 sensitizer subgroup. The higher dose groups exhibited severe choriocapillaris and photoreceptor damage at six weeks.

Many other photosensitizer agents [i.e., tin ethyl etiopurpurin (Purlytine, SnET2), (Optrine, lutetium texaphyrin, Lu-Tex), mono-L-aspartyl chlorin e6 (NPe6 or MACE), chloroaluminum sulfonated phthalocyanine (AlPcS4), and ATX-S10] have been explored for the treatment of exudative AMD and other retinal conditions; however, they have not been used in clinical practice (15–27). Verteporfin PDT has emerged as the dominant therapeutic option for exudative AMD since the publication of the previous edition of this book, and we will focus most of our discussion to the clinical results with this photosensitizer.

LIGHT CONSIDERATIONS

Generally any light source that is matched to the photosensitizer’s absorption profile can be used for PDT. For ophthalmology, fiberoptic delivery of a laser source is required to permit focusing on the retina with a slit lamp system. Lasers are needed because highenergy monochromatic collimated light can be coupled efficiently to fiber optics allowing delivery within an acceptable time frame. Diode lasers that are stable, compact, and relatively inexpensive in the 630to 730nm wavelength range are readily available.

CLINICAL OUTCOMES

PDT is a superior alternative to laser photocoagulation for subfoveal CNV. Using preclinical CNV models, the neovascularization and normal choriocapillaris can be closed while preserving the outer and inner retina. In contrast, during the process of destroying neovascularization lying beneath the RPE and sensory retina with laser photocoagulation, thermal conductance to the retina results in acute necrosis of all layers of the retina. This later results in atrophy leading to loss of vision. However, with PDT treatment, visual acuity generally remains stable immediately after treatment and has been shown, in a minority of patients, to improve immediately.

This suggests that the photoreceptors and inner retinal elements are generally preserved (28).

Verteporfin Human Trials

The safety and efficacy of verteporfin (BPD-MA, Visudynee) have been confirmed (Table 1) in phase I, II, and III clinical trials (28,34,35). The phase I and phase II studies proved that a single treatment of verteporfin PDT could occlude CNV vessels for one to four weeks following administration, as measured by fluorescein angiography (34). The maximal tolerated light dose, defined by retinal closure, was 150 J/cm2. The minimal light dose required to achieve closure of the vessels was 25 J/cm2.

Treatment of Age-Related Macular Degeneration with Photodynamic Therapy Trial

The one-year results of the Treatment of Age-Related Macular Degeneration with Photodynamic Therapy

(TAP) Study were published in 1999 (28). The study consisted of two multicenter, double-masked, placebocontrolled randomized trials with identical protocols. Eligible AMD patients had subfoveal CNV whose greatest linear dimension was up to 5400 mm and best-corrected visual acuity ranged from 20/40 to 20/200. Verteporfin at 6 mg/m2 was infused intravenously for 10 minutes. Then, a diode laser was used to activate the dye (689-nm diode laser, 50 J/cm2, 600 mW/cm2, 83-second duration, spot size 1000 mm larger than greatest linear diameter of the CNV lesion) 15 minutes after the start of infusion. Patients were evaluated by clinical examination and fluorescein angiography approximately every three months, and retreated at the discretion of the treating ophthalmologist. Of the 609 eyes enrolled in the study (402 treatment and 207 placebo), 94% completed the 12 months follow-up. In the treatment group, 246 (61%) of 402 eyes lost fewer than 15 letters of visual acuity

from baseline, compared with 96 (47%) of 207 placebo eyes, a difference that was statistically significant (p!0.001) (28). Subgroup analysis demonstrated the greatest benefit (67% vs. 39% losing less than 15 letters of visual acuity, p!0.001) for those eyes with predominantly classic CNV (greater than 50% of the entire lesion being classic CNV at baseline before treatment). No significant lasting adverse effects were reported (28). The results at various time points following enrollment are summarized in Table 1 (28–33). The average number of verteporfin PDT treatments was 3.4 by 12 months and 5.6 by 24 months (28,29). The treatment effect for verteporfin PDT of predominantly classic subfoveal CNV persisted at 24 months (29). Additionally, for the subgroup of predominantly classic CNV patients, those treated with PDT were more likely to have visual acuity greater than 20/200 at the 24-month follow-up (30).

Verteporfin in Photodynamic Therapy Trial

In the Verteporfin in Photodynamic Therapy (VIP) Study, patients with pathologic myopia, occult CNV, and classic CNV (with visual acuity better than 20/40)

were evaluated (Table 2) (36–39). For pathologic myopia with subfoveal CNV, patients treated with verteporfin PDT were less likely than placebo to lose 8 and 12 Early Treatment Diabetic Retinopathy Study (ETDRS) letters at the 12-month follow-up (36). Additionally, at the 24-month follow-up, the distribution of change in visual acuity favored the PDT group over placebo (38). One arm of the VIP trial evaluated verteporfin for treatment of occult-only CNV with at least 50 letters on the ETDRS scale or some classic component with at least 70 letters (better than 20/40) on the ETDRS scale (37).

For the subgroup of occult-only CNV, the PDT group was less likely than placebo to lose 15 and 30 ETDRS letters at the 24-month follow-up. For the subgroup of patients with a visual acuity score of less than 65 ETDRS letters or lesion size less than or equal to four disc areas, verteporfin PDT-treated patients were less likely than placebo to lose 15 and 30 ETDRS letters at the 24-month follow-up. While the overall safety profile was favorable, 4.4% of PDT-treated patients lost at least 20 ETDRS letters within seven days of treatment (37). This loss of at

228 JAIN ET AL.

least 20 ETDRS letters from baseline visual acuity within seven days was termed acute severe visual decrease (40).

TAP and VIP Trials

The TAP and VIP trial data were combined and analyzed in a series of reports (Table 3) (40–43). The most significant data to be gleaned from these reports was that baseline lesion size was the most important predictor of visual acuity following verteporfin PDT, regardless of lesion composition (41). Size was a significant factor for patients with predominantly classic lesions greater than one disc area, minimally classic lesions less than four disc areas, and occult-only lesions less than five disc areas (42).

Verteporfin PDT was also evaluated for the treatment of subfoveal CNV secondary to pathologic myopia, ocular histoplasmosis, angioid streaks, and idiopathic causes (44,45). The main findings of these

papers were that verteporfin was well tolerated, effective in decreasing fluorescein leakage, and broadly applicable to subfoveal CNV, regardless of etiology. Unfortunately, both studies are limited by their small numbers and lack of controls.

The Photodynamic Therapy of Ocular Histoplasmosis Study trial has yielded two reports to date (46,47). In a non-comparative, prospective study, 56% and 45% of patients gained at least seven ETDRS letters following verteporfin for subfoveal CNV secondary to ocular histoplasmosis at the 12and 24-month follow-up periods, respectively.

Evolution of PDT Treatment

Spaide and colleagues popularized the use of concomitant intravitreal triamcinolone acetonide and PDT (48). They demonstrated that combination therapy resulted in improved visual acuity and lack of fluorescein leakage following therapy, with the greatest

effect seen in treatment-naı¨ve patients (48). These results were durable, lasting out to 12 months, and the most frequent side effect was increased intraocular pressure in 38.5%. These results have been supported by similar work, which suggests that the results are broadly applicable to all sub-types of AMD (49–55).

Other combination therapies include other steroids and anti-VEGF drugs. Recently, retinal specialists have been utilizing dexamethasone in combination with PDT. The data from this and other steroid combination studies should be published in the near future. The FOCUS Study studied the combination of ranibizumab (Lucentis, Genentech, South San Francisco, California, U.S.A.) and PDT versus PDT alone as a treatment for AMD patients with predominantly classic CNV. The FOCUS Study showed that the combined use of PDT with ranibizumab (Lucentis) was better than PDT alone (56). Further details on the FOCUS Study are found in Chapter 8 of this book.

SUMMARY POINTS

&Verteporfin PDT has proven itself useful in the treatment of subfoveal CNV of several etiologies.

&There are many exciting reports indicating that the efficacy of verteporfin PDT can be enhanced with the concomitant intravitreal triamcinolone acetonide for all types of subfoveal and non-subfoveal CNV (46,48,50,53,57–60).

&While recent anti-VEGF therapies hold much promise for the treatment of CNV, there are initial reports that verteporfin PDT might play a significant role as an adjunctive treatment used in conjunction with these new therapies (56,61–65).

&There is a need for long-term prospective studies to quantify and validate these initial reports.

&Verteporfin PDT has been shown to be useful in the treatment of subfoveal CNV for AMD as well as other etiologies.

&Use of verteporfin PDT can be enhanced with the concomitant use of intravitreal triamcinolone acetonide.

&Many studies are currently underway evaluating the use of verteporfin PDT with other steroids and anti-VEGF agents with initially promising initial results.

&It may be that, in the future, the role of verteporfin PDT will be as an adjunct in combination with antiVEGF or other intravitreal therapies.

REFERENCES

1.Henderson BW, Dougherty TJ. How does photodynamic therapy work? Photochem Photobiol 1992; 55(1):145–57.

2.Dougherty TJ, Gomer CJ, Hender BW, et al. Photodynamic therapy. J Natl Cancer Inst 1998; 90(12):889–905.

3.Hausmann W. Die sensibilisierende wirkung des hamatoporphyrins. Biochem Z 1911; 30:276–316.

4.Castellani A, Page G, Concioli M. Photodynamic effect of haematoporphyrin on blood microcirculation. J Pathol Bacteriol 1963; 86:99–102.

5.Fingar VH, Kik PK, Haydon PS, et al. Analysis of acute vascular damage after photodynamic therapy using benzoporphyrin derivative (BPD). Br J Cancer 1999; 79(11–12): 1702–8.

6.Fingar VH. Vascular effects of photodynamic therapy. J Clin Laser Med Surg 1996; 14(5):323–8.

7.Husain D, Miller JW, Kenney AG, et al. Photodynamic therapy and digital angiography of experimental iris neovascularization using liposomal benzoporphyrin derivative. Ophthalmology 1997; 104(8):1242–50.

8.Richter AM, Waterfield E, Jain AK, et al. Photosensitising potency of structural analogues of benzoporphyrin derivative (BPD) in a mouse tumour model. Br J Cancer 1991; 63(1):87–93.

9.Levy J, Chan A, Strong A. The clinical status of benzoporphyrin derivative. Proc SPIE 1995; 2625:86–95.

10.Richter AM, Jain AK, Canaan AJ, et al. Photosensitizing efficiency of two regioisomers of the benzoporphyrin derivative monoacid ring A (BPD-MA). Biochem Pharmacol 1992; 43(11):2349–58.

11.Haimovici R, Kramer M, Miller JW, et al. Localization of lipoprotein-delivered benzoporphyrin derivative in the rabbit eye. Curr Eye Res 1997; 16(2):83–90.

12.Husain D, Miller J. Photodynamic therapy of exudative age-related macular degeneration. Semin Ophthalmol 1997; 12:14–25.

13.Husain D, Kramer M, Kenny AG, et al. Effects of photodynamic therapy using verteporfin on experimental choroidal neovascularization and normal retina and choroid up to 7 weeks after treatment. Invest Ophthalmol Vis Sci 1999; 40(10):2322–31.

14.Reinke MH, Canakis C, Husain D, et al. Verteporfin photodynamic therapy retreatment of normal retina and choroid in the cynomolgus monkey. Ophthalmology 1999; 106(10):1915–23.

15.Mori K, Yoneya S, Ohta M, et al. Angiographic and histologic effects of fundus photodynamic therapy with a hydrophilic sensitizer (mono-L-aspartyl chlorin e6). Ophthalmology 1999; 106(7):1384–91.

16.Kilman GH, Puliafito CA, Grossman GA, et al. Retinal and choroidal vessel closure using phthalocyanine photodynamic therapy. Lasers Surg Med 1994; 15(1):11–8.

17.Kilman GH, Puliafito CA, Stern D, et al. Phthalocyanine photodynamic therapy: new strategy for closure of choroidal neovascularization. Lasers Surg Med 1994; 15(1):2–10.

18.Asrani S, Zeimer R. Feasibility of laser targeted photo-occlu- sion of ocular vessels. Br J Ophthalmol 1995; 79(8):766–70.

19.Obana A, Gohto Y, Kanai M, et al. Selective photodynamic effects of the new photosensitizer ATX-S10(Na) on choroidal neovascularization in monkeys. Arch Ophthalmol 2000; 118(5):650–8.

20.Mang TS, Allison R, Hewson G, et al. A phase II/III clinical study of tin ethyl etiopurpurin (Purlytin)-induced photodynamic therapy for the treatment of recurrent cutaneous metastatic breast cancer. Cancer J Sci Am 1998; 4(6):378–84.

21.Primbs GB, Casey R, Wamser K, et al. Photodynamic therapy for corneal neovascularization. Ophthalmic Surg Lasers 1998; 29(10):832–8.

22.Blumenkranz MS, Woodburn KW, Quing F, et al. Lutetium texaphyrin (Lu-Tex): a potential new agent for ocular

fundus angiography and photodynamic therapy. Am J Ophthalmol 2000; 129(3):353–62.

23.Rockson SG, Lorenz DP, Cheong WF, et al. Photoangioplasty: an emerging clinical cardiovascular role for photodynamic therapy. Circulation 2000; 102(5):591–6.

24.Kereiakes DJ, Szyniszewski AM, Whar D, et al. Phase I drug and light dose-escalation trial of motexafin lutetium and far red light activation (phototherapy) in subjects with coronary artery disease undergoing percutaneous coronary intervention and stent deployment: procedural and longterm results. Circulation 2003; 108(11):1310–5.

25.Verigos K, Stripp H, Mick R, et al. Updated results of a phase I trial of motexafin lutetium-mediated interstitial photodynamic therapy in patients with locally recurrent prostate cancer. J Environ Pathol Toxicol Oncol 2006; 25(1–2):373–87.

26.Graham KB, Arbor JD, Coonnolly EJ, et al. Digital angiography using lutetium texaphyrin in a monkey model of choroidal neovascularization. Invest Ophthalmol Vis Sci 1999; 40:402 (Abstract).

27.Arbor JD, Connolly EJ, Graham K, et al. Photodynamic therapy of experimental choroidal neovascularization using intravenous infusion of lutetium texaphyrin. Invest Ophthalmol Vis Sci 1999; 40:401 (Abstract).

28.Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: one-year results of 2 randomized clinical trials-TAP Report. Arch Ophthalmol 1999; 117(10):1329–45.

29.Bressler NM. Photodynamic therapy of subfoveal choroidal neovascularization in age-related macular degeneration with verteporfin: two-year results of 2 randomized clinical trials-TAP Report 2. Arch Ophthalmol 2001; 119(2): 198–207.

30.Bressler NM, Arnold J, Benchaboune M, et al. Verteporfin therapy of subfoveal choroidal neovascularization in patients with age-related macular degeneration: additional information regarding baseline lesion composition’s impact on vision outcomes-TAP Report No. 3. Arch Ophthalmol 2002; 120(11):1443–54.

31.Rubin GS, Bressler NM. Effects of verteporfin therapy on contrast on sensitivity: results from the treatment of agerelated macular degeneration with photodynamic therapy (TAP) investigation-TAP Report No. 4. Retina 2002; 22(5): 536–44.

32.Blumenkranz MS, Bressler NM, Bressler SB, et al. Verteporfin therapy for subfoveal choroidal neovascularization in age-related macular degeneration: three-year results of an open-label extension of 2 randomized clinical trials-TAP Report No. 5. Arch Ophthalmol 2002; 120(10):1307–14.

33.Bressler SB, Pieramici DJ, Koester JM, et al. Natural history of minimally classic subfoveal choroidal neovascular lesions in the treatment of age-related macular degeneration with photodynamic therapy (TAP) investigation: outcomes potentially relevant to management-TAP Report No. 6. Arch Ophthalmol 2004; 122(3):325–9.

34.Miller JW, Schmidt-Erfurth U, Sickenberg M, et al. Photodynamic therapy with verteporfin for choroidal neovascularization caused by age-related macular degeneration: results of a single treatment in a phase 1 and 2 study. Arch Ophthalmol 1999; 117(9):1161–73.

35.Schmidt-Erfurth U, Miller JW, Sickenberg M, et al. Photodynamic therapy with verteporfin for choroidal neovascularization caused by age-related macular degeneration: results of retreatments in a phase 1 and 2 study. Arch Ophthalmol 1999; 117(9):1177–87.

36.Verteporfin in Photodynamic Therapy (VIP) Study Group. Photodynamic therapy of subfoveal choroidal

neovascularization in pathologic myopia with verteporfin. 1-year results of a randomized clinical trial-VIP Report No. 1. Ophthalmology 2001; 108(5):841–52.

37.Verteporfin in Photodynamic Therapy (VIP) Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization—Verteporfin in Photodynamic Therapy Report 2. Am J Ophthalmol 2001; 131(5):541–60.

38.Blinder KJ, Blumenkranz MS, Bressler NM, et al. Verteporfin therapy of subfoveal choroidal neovascularization in pathologic myopia: 2-year results of a randomized clinical trial-VIP Report No. 3. Ophthalmology 2003; 110(4):667–73.

39.Pieramici DJ, Bressler SB, Koester JM, et al. Occult with no classic subfoveal choroidal neovascular lesions in agerelated macular degeneration: clinically relevant natural history information in larger lesions with good vision from the Verteporfin in Photodynamic Therapy (VIP) Trial: VIP Report No. 4. Arch Ophthalmol 2006; 124(5):660–4.

40.Arnold JJ, Blinder KJ, Bressler NM, et al. Acute severe visual acuity decrease after photodynamic therapy with verteporfin: case reports from randomized clinical trialsTAP and VIP Report No. 3. Am J Ophthalmol 2004; 137(4):683–96.

41.Blinder KJ, Bradley S, Bressler NM, et al. Effect of lesion size, visual acuity, and lesion composition on visual acuity change with and without verteporfin therapy for choroidal

neovascularization secondary to age-related macular degeneration: TAP and VIP Report No. 1. Am J Ophthalmol 2003; 136(3):407–18.

42.Barbazetto I, Burdan A, Bressler NM, et al. Photodynamic therapy of subfoveal choroidal neovascularization with verteporfin: fluorescein angiographic guidelines for evaluation and treatment-TAP and VIP Report No. 2. Arch Ophthalmol 2003; 121(9):1253–68.

43.Azab M, Benchabourne M, Blinder KJ, et al. Verteporfin therapy of subfoveal choroidal neovascularization in agerelated macular degeneration: meta-analysis of 2-year safety results in three randomized clinical trials: treatment of age-related macular degeneration with photodynamic therapy and verteporfin in photodynamic therapy study Report No. 4. Retina 2004; 24(1):1–12.

44.Sickenberg M, Schmidt-Erfurth U, Miller JW, et al. A preliminary study of photodynamic therapy using verteporfin for choroidal neovascularization in pathologic myopia, ocular histoplasmosis syndrome, angioid streaks, and idiopathic causes. Arch Ophthalmol 2000; 118:327–36.

45.Lim JI, Flaxel CJ, Labree L. Photodynamic therapy for choroidal neovascularisation secondary to inflammatory chorioretinal disease. Ann Acad Med Singapore 2006; 35(3):198–202.

46.Rosenfeld PJ, Saperstein DA, Bressler NM, et al. Photodynamic therapy with verteporfin in ocular histoplasmosis: uncontrolled, open-label 2-year study. Ophthalmology 2004; 111(9):1725–33.

47.Saperstein DA, Rosenfeld PJ, Bressler NM, et al. Photodynamic therapy of subfoveal choroidal neovascularization with verteporfin in the ocular histoplasmosis syndrome: one-year results of an uncontrolled, prospective case series. Ophthalmology 2002; 109(8):1499–505.

48.Spaide RF, Sorenson J, Maranan L. Combined photodynamic therapy with verteporfin and intravitreal triamcinolone acetonide for choroidal neovascularization. Ophthalmology 2003; 110(8):1517–25.

49.Augustin AJ, Schmidt-Erfurth U. Verteporfin and intravitreal triamcinolone acetonide combination therapy for occult choroidal neovascularization in age-related macular degeneration. Am J Ophthalmol 2006; 141(4):638–45.

50.Augustin AJ, Schmidt-Erfurth U. Verteporfin therapy combined with intravitreal triamcinolone in all types of choroidal neovascularization due to age-related macular degeneration. Ophthalmology 2006; 113(1):14–22.

51.Chan WM, Lai TY, Wong AL, et al. Combined photodynamic therapy and intravitreal triamcinolone injection for the treatment of subfoveal choroidal neovascularisation in age related macular degeneration: a comparative study. Br J Ophthalmol 2006; 90(3):337–41.

52.Ergun E, Maar N, Ansari-Shahrezaei S, et al. Photodynamic therapy with verteporfin and intravitreal triamcinolone acetonide in the treatment of neovascular age-related macular degeneration. Am J Ophthalmol 2006; 142(1):10–6.

53.Nicolo M, Ghiglione D, Lai S, et al. Occult with no classic choroidal neovascularization secondary to age-related macular degeneration treated by intravitreal triamcinolone and photodynamic therapy with verteporfin. Retina 2006; 26(1):58–64.

54.Ruiz-Moreno JM, Montero JA, Barile S. Triamcinolone and PDT to treat exudative age-related macular degeneration and submacular hemorrhage. Eur J Ophthalmol 2006; 16(3):426–34.

55.Schmidt-Erfurth U, Michels S, Augustin A. Perspectives on verteporfin therapy combined with intravitreal corticosteroids. Arch Ophthalmol 2006; 124(4):561–3.

56.Heier JS, Boyer DS, Ciulla TA, et al. Ranibizumab combined with verteporfin photodynamic therapy in neovascular age-related macular degeneration: year 1 results of the FOCUS Study. Arch Ophthalmol 2006; 124(11):1532–42.

57.Marticorena J, Gomez-Ulla F, Fernandez M, et al. Combined photodynamic therapy and intravitreal triamcinolone acetonide for the treatment of myopic subfoveal choroidal neovascularization. Am J Ophthalmol 2006; 142(2):335–7.

58.Marticorena J, Gomez-Ulla F, Fernandez M, et al. Photodynamic therapy and high-dose intravitreal triamcinolone to treat exudative age-related macular degeneration: 1-year outcome. Retina 2006; 26(6):602–12.

59.Spaide RF, Sorenson J, Maranan L. Combined photodynamic therapy and intravitreal triamcinolone for nonsubfoveal choroidal neovascularization. Retina 2005; 25(6):685–90.

60.Spaide RF, Sorenson J, Maranan L. Photodynamic therapy with verteporfin combined with intravitreal injection of triamcinolone acetonide for choroidal neovascularization. Ophthalmology 2005; 112(2):301–4.

61.Aggio FB, Melo GB, Hofling-Lima AL, et al. Photodynamic therapy with verteporfin combined with intravitreal injection of bevacizumab for exudative age-related macular degeneration. Acta Ophthalmol Scand 2006; 84(6):831–3.

62.Brown DM, Kaiser PK, Michaels M, et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 2006; 355(14):1432–44.

63.Kim IK, Husain D, Michaud N, et al. Effect of intravitreal injection of ranibizumab in combination with verteporfin PDT on normal primate retina and choroid. Invest Ophthalmol Vis Sci 2006; 47(1):357–63.

64.Moshfeghi AA, Rosenfeld PJ, Pulifito CA, et al. Systemic bevacizumab (Avastin) therapy for neovascular age-related macular degeneration: twenty-four-week results of an uncontrolled open-label clinical study. Ophthalmology 2006; 113(11):2002–11.

65.Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for

neovascular age-related macular degeneration. N Engl J Med 2006; 355(14):1419–31.

INTRODUCTION

Age-related macular degeneration (AMD) is a leading cause of rapid and severe visual loss and legal blindness in developed countries (1,2). Ten million Americans are visually disabled due to AMD and 10% of patients aged 66 to 74 show signs of AMD (3,4). Estimated prevalence is from 7% to 30% in persons aged 75 to 85 years (4–6). The “wet” form of AMD is responsible for the most severe and rapid vision loss. In North America, 200,000 to 400,000 people will develop this form of AMD each year. Wet AMD accounts for 12% of cases overall but 90% of cases of legal blindness (see chap 8 on Wet AMD) (4).

Vision loss due to neovascular (wet) AMD involves the growth of abnormal “new” vessels through breaks in Bruch’s membrane from the choroid and under the retinal pigment epithelium (RPE). Unfortunately these new vessels, called choroidal neovascular membranes (CNV), leak serum, blood, and other exudates, resulting in many of the problems related to the wet form of the disease (refer to the chap 1 on pathology of AMD). It is these new vessels that have been the primary target of most current therapies (see Chapters 13–18). Of these therapies, laser treatment, photodynamic therapy (PDT), intravitreal injections with pegaptanib sodium and ranibizumab have been proven by prospective randomized clinical trials to be effective in treating CNV (7–35). Other anti-angiogenic agents that are targeting these new vessels are undergoing clinical trials (36–40) (Chapter 17).

RATIONALE FOR RADIATION THERAPY FOR AMD

When compared with proven and experimental treatment methods, theoretical advantages of radiation therapy include absence of iatrogenic mechanical or thermal laser damage and systemic side effects (41).

An additional advantage is that eyes with primarily occult CNV are eligible for treatment (41) as are eyes with extensive subretinal hemorrhage. In addition, unlike PDT, pegaptanib or ranibizumab sodium treatment, radiation is known to exert a longer-term effect on tissues. Thus repeated treatments may not be necessary. Radiation therapy for AMD has been studied for the past 10 years. Eight randomized, controlled trials have evaluated the use of various radiation types and methods of delivery for treating AMD. Thus far, studies have shown varying degrees of benefit from radiation therapy, with a trend toward better results with higher radiation doses and fewer (larger) treatment fractions.

The scientific rationale for using radiation therapy for a benign disease characterized by neovascular growth is based on experimental and clinical evidence. Radiation is known to potentially destroy vascular tissue (42–45). Specifically, low-dose radiation has been shown to inhibit neovascularization (46–49). For example, in plaque-irradiated choroidal melanomas, a ring of chorioretinal atrophy is commonly found around the tumor’s base and decreased or absent blood flow are demonstrated by fluorescein angiography (Fig. 1). These findings demonstrate the ability of radiation to destroy normal and neovascular blood vessels, but the resultant chorioretinal atrophy is an unacceptable endpoint when treating macular degeneration (49,50).

If relatively low-dose radiotherapy could inhibit CNVand secondary disciform scars, this would lead to better visual outcomes (Fig. 2A,B). The main question persists: is there a therapeutic window in which the dose of radiation used is high enough to induce regression of CNV but low enough to spare the normal retina and choroid? Radiation specialists believe this is possible. Proliferating endothelium is more susceptible to radiation damage than nonproliferating capillary endothelial cells and larger vessels, and thus neovascular endothelial cells and

234 FLAXEL AND FINGER

Figure 1 Example of chorioretinal atrophy at the edge of a melanoma treated with radioactive plaque.

inflammatory cells are particularly radiosensitive (51). There is also the potential for radiotherapy to inhibit further neovascular growth and induce neovascular regression by inducing programmed cell death and modifying the growth factor profiles of the neovascular complexes (44,52). This has been shown by the regression of both benign intracerebral arteriovenous malformations and choroidal hemangiomas after radiation therapy (53–55). Finally, it is thought that inflammation may play a role in neovascularization and, as noted, radiation inhibits the inflammatory response (44,56).

Radiation Toxicity

Doses of ionizing radiation absorbed by the body are reported in conventional units called grays (Gy) or Systeme Internationale (SI) units called rads, representing a given quantity of energy delivered per gram of tissue. A rad is 100 ergs of energy per gram of tissue, while a gray (Gy) equals 100 rads [as does a Gray Equivalent, used when the type of radiation is not standard (i.e., the charged particles of proton beam irradiation)]. The potential toxicity of radiation is well known (45,47,49,57–61). However, studies have shown that the normal neural retina and choroid are relatively radiation resistant (60,47). It is also known that factors influencing the development of radiation retinopathy include total dose delivered, daily fraction size, preexisting microangiopathy, and diabetes or prior chemotherapy (50,51).

Fractionation of the radiation involves dividing the total amount of radiation into smaller doses and delivering these doses over an extended period of time. These small, frequent doses allow healthy cells time to grow back, repairing damage inflicted by the radiation. However, fraction size affects dose; for example, 400 cGy delivered in one over a five minutes period does not equal to 200 cGy per day for two days. The 400 cGy delivered over a short period of time will deliver a higher overall amount of radiation than the second dosage method.

While radiation-induced retinopathy has been reported at doses of 30 to 35 Gy, it is more commonly associated with doses of 45 to 60 Gy (Fig. 3). Radiation optic neuropathy is rare at doses below 50 Gy (Fig. 4A–C) (50,51,57–59). Fraction sizes greater than 2.5 Gy may predispose to toxicity, especially with total doses greater than 45 Gy (58,59). However, there is increasing evidence that fractionated doses

Figure 3 Radiation retinopathy one year following proton beam radiation using 14 gray equivalent (GE).

with larger daily fraction sizes are lower than standard overall doses can be delivered safely and effectively to small regions. Lens doses of 15 Gy or more will induce cataract and transient dry eye; keratitis and epiphora are expected complications (57–59,62). Other concerns with external beam therapy are radiation exposure of the brain and contralateral eye (Fig. 5) (56–59).

PRIOR STUDIES AND ALTERNATE DELIVERY METHODS FOR RADIATION TREATMENT

Initial reports regarding radiation for AMD began to appear in the literature in 1993 (63). Chakravarthy’s preliminary results described 19 patients who were treated with radiation therapy for subfoveal CNV due to AMD. The study also included seven matched control subjects. At one year, 63% of treated patients showed stabilization of vision, while there was deterioration of acuity in all control eyes over the same time period. By image analysis, this study also showed significant neovascular membrane regression in 77% of treated patients at one year, with concurrent progressive enlargement of the neovascular membranes in all control subjects (63). These results and those from other centers led to a prospective, randomized British trial of radiation therapy which reported results in 2002 and is discussed below in more detail (64). Multiple additional reports on external beam radiotherapy (EBRT) and plaque radiotherapy have showed promising but variable results.

In 1996, Finger and colleagues reported the results of low-dose EBRT of 12 to 15 Gy and plaque

(A)

(B)

(C)

Figure 4 (A) Post-radiation optic neuropathy after 14 gray equivalent (GE) proton beam treatment, one year after treatment. (B,C) Note in these early and late fluorescein angiography’s that the choroidal neovascular membrane (CNVM) appears dry.

236 FLAXEL AND FINGER

Figure 5 External beam radiotherapy (EBRT) dose overlay showing radiation delivered to other structures during EBRT. Source: From Ref. 65.

radiotherapy with equivalent dosage in 137 patients. They found decreased subretinal hemorrhages, exudates, and leakage of neovascular membranes with maintenance of visual acuity (65). Subsequently, Stalmans et al. reported failure to control CNV with radiation dosage of 20 Gy in 2 Gy fractions in 111 patients (66). Spaide and colleagues reported similar findings in 1997, when 10 Gy delivered in 5 Gy fractions that failed to control neovascular growth in AMD. This study never disclosed what percentage of treated patients had recurrent CNV (previously treated by laser photocoagulation) (67). Several further reports in 1998 and 1999 reported possible beneficial effects of radiation. Conducted in France, the Radiotherapy Study, conducted in France, reported potential benefit to 16 Gy delivered in foursessions of 4 Gy each with mean follow-up of 6.4 months (68). In 1999, a second French group from France also reported stabilization of visual acuity and anatomical outcome in eyes with AMD (69). However, this group also reported a significant rate of complications, including radiation retinopathy, optic neuropathy, choroidal vasculopathy, and branch retinal vein occlusion when patients received doses of either 20 Gy in five fractions via lateral beam (effectively, a 30 Gy dose) or 16 to 20 Gy in four to five fractions delivered via lateral arc (69). This study did not include a control group. Follow-up time ranged from 12 to 24 months (69).

Chakravarthy undertook a meta-analysis of Phase I clinical trials utilizing low-dose external beam radiotherapy. Results were published in 2000 (70). This report suggested that low-dose EBRT inhibited exudative AMD, but that higher doses were more effective in preventing severe vision loss

(O6 lines on the Snellen visual acuity chart) (70). In support of this conclusion, Berginks and colleagues reported good results with relatively high-radiation doses of 24 Gy and concluded that there was a dose– response effect, with more favorable effects at higher dosages (71,72).

The only published study to evaluate treatment of recurrent CNV with radiation was by Marcus and colleagues who reported the safety and visual outcome of radiation treatment (73). They treated 18 eyes consecutively with seven fractions of 2 Gy for a total dose of 14 Gy, then treated the next 16 eyes with five fractions of 3 Gy for a total dose of 15 Gy. They found no radiation toxicity, but also no significant differences in contrast sensitivity or fluorescein angiography stabilization rates, though they noted a trend for palliative benefit with higher fraction sizes of 4 Gy or higher (73).

Implant Radiation Therapy (Brachytherapy)

Several groups have used brachytherapy to deliver a relatively high dose of radiation to the involved macula with less irradiation of surrounding structures, using methods developed for localized treatment of ocular tumors (Fig. 6A,B) (74–76). Finger and his group employed plaque radiotherapy in eyes with neovascular AMD with no adverse effects (65,74). They found no sight-limiting complications in this Phase I clinical trial, in which they treated 23 eyes with palladium-103 plaques (Figs. 7 and 9) (74). Encouraged by these early results, they enrolled an additional eight eyes and treated all eyes with a mean dose of 17.62 Gy by palladium-103 plaque (76). Their seven year results were reported in 2003, with the conclusion that most patients experienced decreased exudation or stabilization with the dosage employed. They recommended a randomized clinical trial to evaluate brachytherapy for AMD treatment (Fig. 8A,B) (76). Since this report, the group has increased their dose (35 Gy–2 mm from the inner sclera) utilizing 10 mm eye plaques and pallidium-103 seeds. They have noted no adverse effects and everyone had promising results (personal communication).

Charged Particle Radiation Therapy

In June 2000, Friedrichsen and Flaxel published on the use of proton beam irradiation for subfoveal CNV in AMD along with their data from the Phase I/II planned dose-escalation clinical trial (77). This method of irradiation allows a higher dose (and dose rate) to be delivered to a specific volume of tissue. Like most forms of external beam radiation therapy, proton beam therapy requires an entry site and irradiates all the tissue in its path. However, proton H dose volumes are limited to a section of the eye, decreasing irradiation of normal tissues outside the

(B)

Figure 6 (A) Dose overlay of plaque radiation delivery. (B) Typical plaque used for brachytherapy treatment for AMD (10 mm). Source: From Ref. 65.

beam, and in the contralateral eye. Proton beam irradiation was delivered as a single dose, utilizing light field patient orientation with temporal beam entry, initially with 8 GE (Grey equivalent) beginning in March 1994 and increasing to 14 GE in March 1995. No acute radiation-related adverse effects were noted. Twenty-one eyes were treated with 8 GE followed by an initial stabilization of subretinal leakage on fluorescein angiography (FA) in 50% of eyes at 12-month follow-up but with regrowth in all but three eyes at 15-month follow-up. However, in the 14 GE-treated eyes, 83% showed no leakage after 12 months of follow-up and 78% of eyes had unchanged or improved vision. For those eyes followed for longer than nine months (in the 14 GE-treated group), 83% with 20 out of 100 or better vision prior to proton beam treatment showed improvement in vision. Also, severe visual loss

Figure 7 A palladium-103 plaque assembly with seeds prior to implantation.

increased up to 37% at two years with 8 GE-treated eyes, while with 14 GE, the incidence of severe visual loss was 3.7% throughout the follow-up period. There were no cases of cataract, dry eye, lash loss, or optic neuropathy in any of the study eyes and no radiation retinopathy in the 8-GE group; however, radiation retinopathy was found in 48% of eyes treated with 14 GE at a mean of 14 months. There was one case outside the study of severe proliferative radiation retinopathy and optic neuropathy within one year of treatment, with severe visual loss. The authors concluded that their preliminary data suggest that proton beam irradiation correlates with CNV regression, maintains visual function, is more effective at 14 GE, is less beneficial in larger lesions and that radiation complications are more common with longer follow-up but only in the 14-GE group (77). Because of the significant risk of complications, proton beam treatment is not recommended until further studies can be done regarding dose delivery and with consideration of fractionization of the dosage.

REVIEW OF CONTROLLED RADIATION

STUDIES FOR AMD

There are now several completed studies in the use of radiation in AMD (Table 1). Reports on randomized trials of radiation treatment include those from Holz’s RAD Study Group in Germany (78), Kobayashi and colleagues in Japan (79), the United Kingdom group

238 FLAXEL AND FINGER

(A)

(B)

Figure 8 (A) Fluorescein angiography of an eye with CNV before implantation of palladium-103 plaque. (B) Fluorescein angiography of the same eye following treatment with palladium-103 plaque. Abbreviation: CNV, choroidal neovascular membranes.

(64), Valmaggia and colleagues from Switzerland (80), Marcus et al. from the Medical College of Georgia (81,82), and the Age-related Macular Degeneration Radiation Trial (AMDRT) study group report from the United States (84,85).

The RAD study is a randomized, prospective, double-blind, placebo-controlled trial performed at nine centers throughout Germany (78). This study enrolled 205 patients who were treated with either eight fractions of 2 Gy (101 eyes), or eight fractions of 0 Gy (104 eyes). At one-year follow-up, no benefit was seen in either classic or occult subfoveal CNV due to AMD (approximately 50% of treated eyes had only

occult CNV, while the other half had a combination of classic and occult disease). There have been no serious complications relating to the radiation treatment to date (78).

A randomized, prospective, placebo-controlled trial was also carried out at a center in Japan (79). This study enrolled 101 patients and followed them for two years. They also reported no significant treat- ment-related side effects from a total dose of 20 Gy delivered in 10 divided doses over a period of 14 days, with irradiation through a single lateral port. The investigators concluded that radiotherapy showed a beneficial effect compared with no treatment, with favorable factors being smaller area of CNV, higher degree of occult CNV, and better initial visual acuity (79). Both groups are continuing follow-up on all patients.

Hart et al. reported the results of a large, multicenter randomized trial in the United Kingdom in 2002. This trial included 203 patients randomly assigned either to radiotherapy using 12 Gy of 6 mV photons (delivered in six fractions) or observation (64). They did not find a statistically significant benefit to radiation treatment and felt their results did not support the routine use of radiation treatment for AMD (64).

A Swiss group reported the results of 18-month follow-up in 161 patients with subfoveal CNV who were enrolled in a prospective study (80). The examiners treated the posterior pole of the affected eye with 1 Gy (4!0.25 Gy) in the control group and 8 Gy (4!2 Gy) or 16 Gy (4!4 Gy) in the treatment groups. They found that patients with classic CNV, or with initial distance visual acuity R20/100, benefited more from treatment. However, no significant difference was found between control and treatment groups in reading ability and size of CNV (80). They also reported no radiation treatment side effects in any group (80).

At Association for Research in Vision and Ophthalmology (ARVO) in 2003, Marcus and colleagues from the Medical College of Georgia reported the four-year results of a small, doublemasked clinical trial that included 42 observed and 41 treated eyes (81,82). They used low-dose external beam irradiation at 14 Gy in seven fractions of 2 Gy, and reported no benefit and possible detriment to vision in long-term (2–4 years) follow-up (81,82).

Another Japanese study published in 2004 with two-year follow-up utilized external-beam radiation therapy in 21 eyes of 18 patients, with a group of 15 non-treated controls (83). This group reported improved or maintained visual acuity rates of 81% in the treated group versus 40% in the control group (83). This study, however, was non-randomized. Another non-randomized trial from Japan also reported shortterm benefit to low-dose radiation in 68 eyes (86).

The problem of conflicting data from multiple studies led the National Eye Institute to sponsor a prospective randomized pilot study in the United States (84,85). This non-funded, multi-center pilot study included two groups of patients randomized

to either treatment or observation, and was called the AMDRT (84,85). Eligibility criteria for the new subfoveal CNV study included lesions not amenable to Macular Photocoagulation Study (MPS) laser treatment, classic, mixed or occult CNV by FA, blood

obscuring !50% of the lesion, visual acuity (VA)O20/ 320, and no contraindication to EBRT (i.e., prior chemotherapy, diabetes, or history of periorbital or ocular radiation). Randomization was to either EBRT (five daily sessions of 4 Gy for a total dose of 20 Gy) or observation. The primary outcome measure was a three-line or greater loss of visual acuity over the five-year follow-up period. There was also a recurrent CNV study arm with similar criteria (84,85). Eighty-

eight patients were enrolled through 10 sites and were randomized to either radiotherapy [20 Gy delivered in five daily fractions of 4 Gy each; 6 mV (NZ41)] or no radiotherapy (sham NZ22 or observation NZ25). The results were reported in 2004 and concluded that external beam radiation at 5!4 Gy may have a modest and short-lived (six-month) benefit in preserving visual acuity. There were no safety concerns (84,85).

Multimodality Treatment and Novel Methods for Radiation Delivery

Marcus and colleagues from the Medical College of Georgia submitted an ARVO abstract in 2002, updated in 2004, on the use of transpupillary thermotherapy (TTT) and radiotherapy of CNV in AMD (87,88). The initial report was a safety evaluation in which four eyes of four patients were treated with TTT at 810 nm for 60 seconds at a power of 360 to 1000 mW, followed within eight hours by administration of 6 mV photon beam to deliver 20 Gy in five fractions at 4 Gy per fraction over five days (87). They found no safety risk and proceeded with a prospective non-randomized case series including eight patients following the same protocol (88). They found mixed results with the treatment of occult subfoveal CNV, but again, there were no safety concerns (88). This study has, however, been halted due to little subject interest in undergoing a combination of two experimental therapies (personal communication).

In 2002, Tong and colleagues from the University of California at Davis reported on the use of stereotactic external beam radiation to treat eyes with AMD (89,90). This method allows radiation to be delivered to a smaller, better-defined area than standard EBRT. Patients treated with varying doses of radiation were followed for 24 months (89). They concluded that the method was safe at all studied dosages with stable visual acuity until 12 to 18 months posttreatment, at which time the effect of the radiation appeared to cease (89). At doses of 28–32 Gy, vision tended to stabilize for a longer period (at least 24 months) (89). The oneyear data from this pilot study werepublished in 2005 (90). The investigators found no significant acute side effects and no benefit in either VA or membrane size from increasing the radiation dosage. They concluded that their results were consistent with trends in palliative benefit and that there was no evidence that therapeutic effectiveness is dose-dependent. Therefore, they found no justification for potentially dangerous escalations in radiation dosage for treatment of neovascular AMD (90).

Other novel methods of delivering radiation are being studied, including a spoon-shaped device made by Theragenics. This is inserted through a conjunctival incision and traverses to an episcleral submacular position, where the radiation source is uncovered for a short (minutes) period of time (75,91). Hubbard reported results of preliminary work with the Theragenics device (Therasight Ocular Brachytherapy System) at ARVO 2005 and found the device to be well tolerated by patients and readily positioned and inserted by clinicians. No adverse events were reported after short post-operative follow-up (75). In a personal communication, the company has supplied the following device description and Figure 9: “The

TheraSightw Ocular Brachytherapy System (TheraSight System) is a radiation device that primarily consists of a sealed palladium–103 source on the distal end of an insertion applicator. The device delivers high dose rate, low energy X rays of 21–23 keV in a minimally invasive procedure where the source is inserted in the retrobulbar space behind the eye. The energy is deposited locally to the target tissue, consisting of new choroidal blood vessels intruding into the subretinal space. The radiation therapy is intended to reduce neovascularization.” This device delivers about 14 Gy to the inner retina.

Other applicators include Immonen’s applicator and Freire’s applicator, both for strontium-90 (92). Immonen’s applicator was calculated to deliver 15 Gy to the inner retina (Fig. 10A) and Freire’s applicator allows a dose at 1.5 mm depth of 6 cGy/second, thus allowing the total dose delivered to be altered based on exposure time (Fig. 10B). Finger has pointed out that low-energy photo-emitting palladium-103 will deposit less radiation to the subjacent sclera, choroids, and retina than the beta-particle emitting 90Sr, possibly explaining why Immonen et al. noted increased and earlier chorioretinal atrophy within the targeted zone (Fig. 11A,B) (92).

(A)

(B)

Figure 11 (A) Immonen pre-treatment fluorescein angiography (FA) utilizing strontium-90. (B) Immonen posttreatment FA.

Fujii et al. evaluated the feasibility and initial safety of retinal-sparing subretinal delivery of strontium beta-radiation using a novel-selective subretinal brachytherapy system (Neovista, Atlanta, Georgia) on 90 rabbit and 4 dog eyes (92). The surgery involved vitrectomy, creation of a subretinal bleb, and introduction of the probe, which was calculated to deliver a radiation dose of 0–246 Gy into the subretinal space (92). Lim and co-workers from Los Angeles presented further work with the Neovista device in 10 patients at the 2005 ARVO meeting (91). This was a tolerability and safety study that compared two probe designs delivering 26 Gy to the CNV over a period of two to three minutes, reportedly sparing the overlying retina. They found no retinal detachments or endophthalmitis complications; however, there were three adverse events that led to further device modifications (91). Figure 12 was supplied by the Neovista company demonstrating the results in one of their initial trial eyes. The group has since switched to an epiretinal radiation delivery device.

CONCLUSIONS

Several well-organized, multi-center clinical trials conducted in the United States and Europe have shown no benefit to EBRT (64,78–84). Most of these studied doses or dose rates less than those used in brachytherapy or proton irradiation studies. Three of these studies did show evidence of some benefit in limiting lesion size and vision loss, mainly with radiation dosages of 20 Gy and higher (Table 1) (79,80,83).

In addition, it is possible that radiation treatment might be of benefit when combined with PDT or antiangiogenic drugs, or with TTT as described by Marcus and colleagues (87,88). Combined treatment would potentially allow complete closure of the neovascular complex, with PDT or injection of an anti-angiogenic agent followed by radiation to extend the effects of treatment. Similarly, groups studying low-dose proton beam radiation combined with PDT hope that this will limit CNV recurrence. This approach might avoid the complications seen with higher doses of radiation using the proton beam, or of multiple laser or pharmacologic treatments. Finally, other groups are evaluating different ways to deliver radiation in order to limit toxicity and allow higher radiation doses (75,91).

This review has found significant evidence that radiation can halt the growth of choroidal neovascularization. However, the prospective randomized evidence-based studies reported to date do not support the widespread treatment of patients. Further prospective randomized studies are needed

to actually determine whether a different method of delivering the radiation will offer longer-term benefit with less chance of toxicity, or whether a more efficacious method might involve combining radiation with another treatment modality.

SUMMARY POINTS

&Proliferating endothelium is more susceptible to radiation damage than are non-proliferating capillary endothelial cells and larger vessels, and thus neovascular endothelial cells and inflammatory cells are particularly radiosensitive.

&Radiation may induce programmed cell death

and modify the growth factor profiles of the neovascular complexes as well as limit the inflammatory response.

&Factors influencing the development of radiation retinopathy include total dose delivered, daily fraction size, preexisting microangiopathy, and diabetes or prior chemotherapy.

&While radiation-induced retinopathy has been reported at doses of 30–35 Gy, it is more commonly associated with doses of 45–60 Gy (Fig. 5). Radiation optic neuropathy is rare at doses below 50 Gy.

REFERENCES

1.Ferris FL, Fine SL, Hyman L. Age-related macular degeneration and blindness due to neovascular maculopathy. Arch Ophthalmol 1984; 102:1640–2.

2.American Academy of Ophthalmology. Age-related macular degeneration, Preferred Practice Pattern. San Francisco: American Academy of Ophthalmology, 1998.

3.Leibowitz HM, Krueger DE, Maunder LR, et al. The Framingham Eye Study monograph: an ophthalmological and epidemiological study of cataract, glaucoma, diabetic retinopathy, macular degeneration, and visual acuity in a general population of 2631 adults, 1973–1975. Surv Ophthalmol 1980; 24(Suppl.):335–610.

4.Kahn HA, Leibowitz HM, Ganley JP, et al. The Framingham eye study II. Association of ophthalmic pathology with single variables previously measured in the Framingham heart study. Am J Epidemiol 1977; 106:33–41.

5.Klein BE, Klein R. Cataracts and macular degeneration in older Americans. Arch Ophthalmol 1982; 100:571–3.

6.Vinding T. Age related macular degeneration, macular changes, prevalence and sex ratio. An epidemiological study of 1000 aged individuals. Acta Ophthalmol 1989; 67:609–16.

7.Macular Photocoagulation Study Group. Argon laser photocoagulation for AMD: results of a randomized clinical trial. Arch Ophthalmol 1982; 100:912–8.

8.Macular Photocoagulation Study Group. Krypton laser photocoagulation for neovascular lesions of AMD: results of a randomized clinical trial. Arch Ophthalmol 1990; 108:816–24.

244 FLAXEL AND FINGER

9.Macular Photocoagulation Study Group. Subfoveal neovascular lesions in AMD: guidelines for evaluation and treatment in the Macular Photocoagulation Study. Arch Ophthalmol 1991; 109:1242–57.

10.Macular Photocoagulation Study Group. Recurrent CNV after argon laser treatment for neovascular maculopathy. Arch Ophthalmol 1986; 104:503–12.

11.Macular Photocoagulation Study Group. Occult choroidal neovascularization. Influence on visual outcome in patients with age-related macular degeneration. Arch Ophthalmol 1996; 114:400–12.

12.Macular Photocoagulation Study Group. Laser photocoagulation of subfoveal neovascular lesions in AMD: results of a randomized clinical trial. Arch Ophthalmol 1991; 109:1220–31.

13.Macular Photocoagulation Study Group. Visual outcome after laser photocoagulation for subfoveal CNV secondary to AMD: the influence of initial lesion size and initial visual acuity. Arch Ophthalmol 1994; 112:480–8.

14.Freund KB, Yanuzzi LA, Sorenson JA. Age-related macular

15.Algvere PV, Berglin L, Gouras P, Sheng Y. Transplantation of fetal retinal pigment epithelium in age-related macular degeneration with subfoveal neovascularization. Graefes Arch Clin Exp Ophthalmol 1994; 232:707–16.

16.Rezai KA, Kohen L, Wiedermann P, Heimann K. Iris pigment epithelium transplantation. Graefes Arch Clin Exp Ophthalmol 1997; 235:558–62.

17.Thomas MA, Grand MG, Williams DF, Lowe MA. Surgical management of subfoveal choroidal neovascularization. Ophthalmology 1992; 99:952–68.

18.Scheider A, Gundisch O, Kampik A. Surgical extraction of subfoveal choroidal new vessels and submacular haemorrhage in age-related macular degeneration; results of a prospective study. Graefes Arch Clin Exp Ophthalmol 1999; 237:10–5.

19.Machemer R, Steinhorst UH. Retinal separation, retinotomy, and macular relocation: II. A surgical approach for age-related macular degeneration? Graefes Arch Clin Exp Ophthalmol 1993; 231:635–41.

20.Ninomiya Y, Lewis JM, Hasegawa T, Tana Y. Retinotomy and foveal translocation for surgical management of subfoveal choroidal neovascular membranes. Am J Ophthalmol 1996; 122:613–21.

21.Eckardt C, Eckardt U, Conrad HG. Macular rotation with and without counter-rotation of the globe in patients with age-related macular degeneration. Graefes Arch Clin Exp Ophthalmol 1999; 237:313–25.

22.Lewis H, Kaiser PK, Lewis S, Estafanous M. Macular translocation for subfoveal choroidal neovascularization in age-related macular degeneration: a prospective study. Am J Ophthalmol 1999; 128:135–46.

23.Husain D, Miller JW, Michaud N, Connolly E, Flotte TJ,

Gragoudas ES. Intravenous infusion of liposomal benzoporphyrin derivative for photodynamic therapy of experimental choroidal neovascularization. Arch Ophthalmol 1996; 114:978–85.

24.Kliman GH, Puliafito CA, Stern D, Borirakchanyavat S, Gregory WA. Phthalocyanine photodynamic therapy: new strategy for closure of choroidal neovascularization. Lasers Surg Med 1994; 15:2–10.

25.Lin SC, Lin CP, Feld JR, Duker JS, Puliafito CA. The photodynamic occlusion of choroidal vessels using benzoporphyrin derivative. Curr Eye Res 1994; 13:513–22.

26.Miller JW, Walsh AW, Kramer M, et al. Gragoudas ESPhotodynamic therapy of experimental choroidal neovascularization using lipoprotein-delivered benzoporphyrin. Arch Ophthalmol 1995; 113:810–8.

27.Peyman GA, Moshfeghi DM, Moshfeghi A, et al. Photodynamic therapy for choriocapillaris using tin ethyl etiopurpurin (SnET2). Ophthalmic Surg Lasers 1997; 28:409–17.

28.Schmidt-Erfurth U, Miller J, Sickenberg M, et al. Photodynamic therapy of subfoveal choroidal neovascularization: clinical and angiographic examples. Graefes Arch Clin Exp Ophthalmol 1998; 236:365–74.

29.Treatment of Age-Related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: one-year results of 2 randomized clinical trials—TAP report 1. Arch Ophthalmol 1999; 117:1329–45.

30.Verteporfin in Photodynamic Therapy (VIP) Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in AMD 2 year results of a randomized clinical trial including lesions with occult with no classic choroidal neovascularization-VIP Report No. 2. Am J Ophthalmol 2001; 131:541–60.

31.Verteporfin in Photodynamic Therapy (VIP) Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in pathologic myopia, 2 year results of a randomized clinical trial—VIP Report No. 3. Ophthalmology 2003; 110:667–73.

32.Gragoudas ES, Adamis AP, Cunningham ET, et al. Pegaptanib for neovascular AMD. N Engl J Med 2004; 351:2805–16.

33.Mainster MA, Reichel E. Transpupillary thermotherapy for age-related macular degeneration: long-pulse photocoagulation, apoptosis and heat-shock proteins. Ophthalmic Surg Lasers 2000; 31:359–73.

34.Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for

neovascular age-related macular degeneration. N Engl J Med 2006; 355:1419–31.

35.Brown DM, Kaiser PK, Michels M, et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 2006; 355:1432–44.

36.Gillies MC, Simpson JM, Luo W, et al. A randomized clinical trial of a single dose of intravitreal triamcinolone acetonide for neovascular age-related macular degeneration (one-year results). Arch Ophthalmol 2003; 121:667–73.

37.Amato RJ, Loughan MS, Flynn E, Folkman J. Thalidomide is an inhibitor of angiogenesis. Proc Natl Acad Sci USA 1994; 91:4082–5.

38.Pharmacological Therapy for Macular Degeneration Study Group. Interferon-2a is ineffective for patients with choroidal neovascularization secondary to age-related macular degeneration. Results of a prospective randomized placebo-controlled clinical trial. Arch Ophthalmol 1997; 115:865–72.

39.Thomas MA, Ibanez HE. Interferon alfa-2a in the treat-

ment of subfoveal choroidal neovascularization. Am J Ophthalmol 1993; 115:563–8.

40.Ciulla TA, Danis RP, Harris A. Age-related macular degeneration: review of experimental treatments. Surv Ophthalmol 1998; 32:134–46.

41.Finger PT, Augsberger JJ. Controversies, radiotherapy and the treatment of age-related macular degeneration. Arch Ophthalmol 1998; 116:1507–11.

42.Baker DG, Krochak RJ. The response of the microvascular system to radiation: a review. Cancer Invest 1989; 7:287–94.

43.Krishnan L, Krishnan EC, Jewell WR. Immediate effect of irradiation on microvasculature. Int J Radiat Oncol Biol Phys 1988; 15:147–50.

44.Finger PT, Immonen I, Freire J, Brown G. Brachytherapy for macular degeneration associated with subretinal neovascularization. In: Alberti WE, Richard G, Sagerman RH, eds. Age-Related Macular Degeneration. Germany: Springer, 2001:167–72.

45.Chakravarthy U, Gardiner TA, Archer DB, Maguire CJF. A light microscopic and autoradiographic study of nonirradiated and irradiated ocular wounds. Curr Eye Res 1989; 8:337–48.

46.Archambeau JO, Mao XW, Yonemoto LT, et al. What is the role of radiation in the treatment of subfoveal membranes: review of radiobiologic, pathologic, and other considerations to initiate a multimodality discussion. Int J Radiat Oncol Biol Phys 1998; 40:1125–36.

47.De Gowin RL, Lewis JL, Hoak JC, Mueller AL, Gibson DP.

Radiosensitivity of human endothelial cells in culture. J Lab Clin Med 1974; 84:42–8.

48.Hosoi Y, Yamamoto M, Ono T, Sakamoto K. Prostacyclin production in cultured endothelial cells is highly sensitive to low doses of ionizing radiation. Int J Radiat Biol 1993; 63:631–8.

49.Johnson LK, Longenecker JP, Fajardo LF. Differentialradiation response of cultured endothelial cells and smooth myocytes. Anal Quant Cytol 1982; 4:188–98.

50.Finger PT. Radiation therapy for choroidal melanoma. Surv Ophthalmol 1997; 42:215–32.

51.Sagerman RH, Chung CT, Alberti WE. Radiosensitivity of ocular and orbital structures. In: Alberti WE, Sagerman RH, eds. Radiotherapy of Intraocular and Orbital Tumors. Berlin: Springer, 1993:375–85.

52.Langley RE, Bune EA, Quartuccio SG, Medeiras D, Braunhut SJ. Radiation induced apoptosis in microvascular endothelial cells. Br J Cancer 1997; 75:666–72.

53.Perez CA, Brady LW. Principles and Practice of Radiation Oncology. 2nd ed. Philadelphia, PA: JB Lippincott Co., 1992.

54.Schilling H, Sauerwein W, Lommatzsch A, et al. Long-term results after low dose ocular irradiation for choroidal haemamgiomas. Br J Ophthalmol 1997; 81:267–73.

55.Engenhart R, Wowra B, Debus J, Kimmig BN, et al. The role of high-dose, single-fraction irradiation in small and large intracranial arteriovenous malformations. Int J Radiat Oncol Biol Phys 1994; 30:521–9.

56.Finger PT, Chakravarthy Y. External beam radiation therapy is effective in the treatment of age-related macular degeneration. Arch Ophthalmol 1998; 116:1507–9.

57.Scott TA, Augsburger JJ, Brady LW, Hernandez C, Woodleigh R. Low dose ocular irradiation for diffuse choroidal hemangiomas associated with bullous nonrhegmatogenous retinal detachment. Retina 1991; 11:389–93.

58.Brown GC, Shields JA, Sanborn G, Augsburger JJ, Savino PJ, Schatz NJ. Radiation retinopathy. Ophthalmology 1982; 89(12):1494–501.

59.Chan RC, Shukovsky LJ. Effects of irradiation on the eye. Radiology 1976; 120:673–5.

60.Parsons JT, Fitzgerald CR, Hood CI, Ellingwood KE, Bova FJ, Million RR. The effects of irradiation on the eye and optic nerve. Int J Radiat Oncol Biol Phys 1983; 9:609–22.

61.Archer DB, Amoaku SMK, Gardinier TA. Radiation retinopathy, clinical, histological and ultrastructural correlations. Eye 1991; 5:239–51.

62.Plowman PN, Harnett AN. Radiotherapy in benign orbital disease. I: complicated ocular angiomas. Br J Ophthalmol 1988; 72:286–8.

63.Chakravarthy U, Houston RF, Acher DB. Treatment of agerelated subfoveal neovascular membranes by teletherapy: a pilot study. Br J Ophthalmol 1993; 77:265–73.

64.Hart P, Chakravarthy U, Mackenzie G, et al. Visual outcomes in the subfoveal radiotherapy study a randomized controlled trial of teletherapy for age-related macular degeneration. Arch Ophthalmol 2002; 120:1029–38.

65.Finger PT, Berson A, Sherr D, Riley R, Balkin RA, Bosworth JL. Radiation therapy for subretinal neovascularization. Ophthalmology 1996; 103:878–89.

66.Stalmans P, Leys A, Van Limbergen E. External beam

radiation therapy (20 Gy, 2 Gy fractions) fails to control the growth of choroidal neovascularization in age-related macular degeneration: a review of 111 cases. Retina 1997; 17:481–92.

67.Spaide RF, Guyer DR, McCormick B, et al. External beam radiation therapy for choroidal neovascularization. Ophthalmology 1998; 105:24–30.

68.Donati G, Soubrane D, Quaranta M, Coscas G, Soubrane G. Radiotherapy for isolated occult subfoveal neovascularisation in age-related macular degeneration: a pilot study. Br J Ophthalmol 1999; 83:646–51.

69.Mauget-Faysse M, Chiquet C, Milea D, et al. Long term results of radiotherapy for subfoveal choroidal neovascu-

larisation in age related macular degeneration. Br J Ophthalmol 1999; 83:923–8.

70.Chakravarthy U. External beam radiotherapy in exudative age-related macular degeneration: a pooled analysis of phase-1 data. Br J Radiol 2000; 73:305–13.

71.Bergink GJ, Deutman AF, van den Broek JFCM, van Daal WA, van der Maazen RW. Radiation therapy for agerelated subfoveal choroidal neovascular membranes, a pilot study. Doc Ophthalmol 1995; 90:67–74.

72.Bergink GJ, Hoyng CB, van der Maazen RWM. A randomized controlled trial on efficacy of radiation therapy in the control of subfoveal choroidal neovascularization in age-related macular degeneration: radiation versus observation. Graefes Arch Clin Exp Ophthalmol 1998; 236:321–5.

73.Marcus DM, Sheils WC, Young JO, et al. Radiotherapy for recurrent choroidal neovascularization complicating agerelated macular degeneration. Br J Ophthalmol 2004; 88:114–9.

74.Finger PT, Berson A, Ng T, Szechter A. Ophthalmic plaque radiotherapy for age-related macular degeneration associated with subretinal neovascularization. Am J Ophthalmol 1999; 127:170–7.

75.Hubbard G, Ciulla T, Marcus D, et al. A new ocular brachytherapy system for the treatment of exudative AMD. Invest Ophthalmol Vis Sci 2005; 46 (E-abstract 2425).

76.Finger PT, Gelman YP, Berson AM, Szechter A. Pallidium103 plaque radiation therapy for macular degeneration: results of a 7 year study. Br J Ophthalmol 2003; 87:1497–503.

77.Flaxel CJ, Friedrichsen EJ, Smith JO, et al. Proton beam irradiation of subfoveal choroidal neovascularization in age-related macular degeneration. Eye 2000; 14:155–64.

78.The Radiation Therapy for Age-related Macular Degeneration (RAD) Study Group. A prospective, randomized, double-masked trial on radiation therapy for neovascular age-related macular degeneration (RAD Study). Ophthalmology 1999; 106:2239–47.

79.Kobayashi H, Kobayashi K. Age-related macular degeneration: long-term results of radiotherapy for subfoveal neovascular membranes. Am J Ophthalmol 2000; 130:617–35.

246 FLAXEL AND FINGER

80.Valmaggia C, Reis G, Ballinari P. Radiotherapy for subfoveal choroidal neovascularization in age-related macular degeneration: a randomized clinical trial. Am J Ophthalmol 2002; 133:521–9.

81.Marcus DM, Sheils C, Johnson MH, et al. External beam irradiation of subfoveal choroidal neovascularization complicating age-related macular degeneration. One-year results of a prospective, double masked randomized clinical trial. Arch Ophthalmol 2001; 119:171–80.

82.Lott M, Marcus D, Sheils W, Johnson M, Samy C. External beam irradiation of subfoveal CNV complicating AMD: 4 year results of a prospective, double masked, randomized clinical trial. Invest Ophthalmol Vis Sci 2003; 44 (E-abstract 5007).

83.Churei H, Ohkubo K, Nakajo M, et al. External-beam radiation therapy for age-related macular degeneration: two years’ follow-up results at a total dose of 20 Gy in 10 fractions. Radiat Med 2004; 22:398–404.

84.Marcus D, Peskin E, Alexander J, et al. The age-related macular degeneration radiotherapy trial (AMDRT): 1—year results. Invest Ophthalmol Vis Sci 2003; 44 (E-abstract 3158).

85.Marcus D, Peskin E, AMDRT Study Group. The age-related macular degeneration radiotherapy trial (AMDRT): one year results from a pilot study. Am J Ophthalmol 2004; 138:818–28.

86.Tamai M. The results of randomized controlled trial (RCT) of low-dose radiation for wet-AMD on 1 year term basis. Invest Ophthalmol Vis Sci 2003; 44 (E-abstract 3157).

87.Lee S, Sheils W, Redd J, Samy C, Marcus D. Multimodality transpupillary thermotherapy and radiotherapy of choroidal neovascular membranes in age-related macular degeneration: a phase I safety study. Invest Ophthalmol Vis Sci 2002; 43 (E-abstract 4411).

88.Ying M, Fuller J, Alexander J, Sheils W, Lee Y, Marcus D. Multimodality transpupillary thermotherapy and radiotherapy of occult subfoveal choroidal neovascular membranes in AMD. Invest Ophthalmol Vis Sci 2004; 45 (E-abstract 5136).

89.Tong A, Hauser D, Barak A, et al. Stereotactic external beam radiation treatment in eyes with AMD CNV—a 2-year follow-up. Invest Ophthalmol Vis Sci 2002; 43 (E-abstract 1221).

90.Barak A, Hauser D, Yipp P, et al. A phase I trial of stereotactic external beam radiation for subfoveal choroidal neovascular membranes in age-related macular degeneration. Br J Radiol 2005; 78:827–31.

91.Lim JI, DeJuan E, Sadda V, et al. Subretinal radiation treatment of occult CNV due to AMD. Invest Ophthalmol Vis Sci 2005; 46 (E-abstract 1384).

92.Finger PT, Immonen I, Freire J, Brown G. Implant radiotherapy for exudative macular degeneration associated with subretinal neovascularization. In: Peyman G, Meffert S, Conway MD, Chou F, eds. Vitreoretinal Surgical Techniques. London: Martin Dunitz, 2001:555–60.

Phillip J. Rosenfeld

Bascom Palmer Eye Institute, Miami, Florida, U.S.A.

INTRODUCTION

In 1989, Ferrara and Henzel (1) isolated a diffusible protein from bovine pituitary follicular cells that showed cell-specific mitogenic activity for vascular endothelium. They named this protein vascular endothelial growth factor (VEGF). Further research showed that VEGF was in fact Michelson’s factor X, which was the postulated diffusible angiogenesis factor (2). As discussed in Chapter 5, VEGF was then shown to have a major role in choroidal neovascularization (CNV) (3,4).

The human VEGF-A gene, located on chromosome 6p21.3, consists of eight exons and seven introns. Alternative splicing produces mRNA transcripts that code for at least six different protein isoforms: 121, 145, 165, 183, 189, and 206 amino acids in length (5). These different isoforms vary in their affinity for heparin binding, and as such, in their affinity for the extracellular matrix. The larger isoforms, such as VEGF189 and VEGF206, bind heparin with high affinity, and are therefore almost completely sequestered in the extracellular matrix. The smaller isoform, VEGF121, does not bind heparin and is freely diffusible. All VEGF isoforms contain a plasmin cleavage site. Cleavage at this site creates a freely diffusible, 110 kD, bioactive form of VEGF (VEGF110). Plasmin-mediated extracellular proteolysis may therefore be an important regulator of VEGF bioavailablility (6).

CURRENT ANTI-VEGF THERAPIES

Aptamers: Pegaptanib Sodium (Macugen, New York)

The first anti-VEGF therapy to undergo clinical testing was a VEGF aptamer. Approved by the Food and Drug Administration (FDA) in 2004, Pegaptanib (Macugen—OSI/Eyetech Pharmaceuticals, New York) was the first anti-VEGF agent with proven efficacy for the treatment of CNV secondary to

age-related macular degeneration (AMD). Pegaptanib is an aptamer—a short single-stranded oligonucleotide sequence that functions as a high affinity inhibitor of a specific protein target. Aptamers are created by a form of in vitro evolution called systematic evolution of ligands by exponential enrichment (SELEX) (7).

Pegaptanib is a 28-base RNA oligonucleotide that is covalently linked to two 20 kD polyethylene glycol moieties to extend the half-life. Pegaptanib selectively binds to the heparin-binding domain of VEGF165 and larger isoforms, preventing ligand-receptor binding. The smaller VEGF isoforms and proteolytic fragments are therefore not inhibited by pegaptanib (7).

Safety and efficacy of pegaptanib for the treatment of neovascular AMD was established through the VEGF Inhibition Study in Ocular Neovascularization (VISION) study (8). VISION consisted of two phase III prospective, multicenter, randomized, controlled, double-masked trials comparing intravitreal injections of pegaptanib with sham injections. Patients (1186 total) were randomized to receive pegaptanib (at a dose of 0.3, 1.0, or 3.0 mg) or sham injection (usual care), every six weeks for a total of 54 weeks. The primary end point of the study was the number of patients losing less than 15 letters of Early Treatment Diabetic Retinopathy Study (ETDRS) visual acuity at 54 weeks. Patients with all CNV lesion subtypes with sizes up to and including 12 disc areas in size were included. Concomitant photodynamic therapy (PDT) with verteporfin (Visudynew, Novartis, East Hanover, New Jersey, U.S.A.) was allowed at the physician’s discretion. Twenty-five percent of the VISION patients received PDT during the study period.

In the pooled analysis, efficacy was demonstrated for all three doses, without a dose–response relationship. Seventy percent of pegaptanib-treated patients lost less than 15 letters, compared with 55% of usual care patients. More pegaptanib-treated patients maintained or gained visual acuity (33%) at 54 weeks than usual care patients (23%). In addition, the usual

248 KLESERT ET AL.

care group was twice as likely to experience severe vision loss (R30 letters) during the study period than pegaptanib-treated patients. However, only 6% of pegaptanib-treated patients in the study gained R15 letters at 54 weeks (compared with 2% of usual care controls), and as a group, the pegaptanib-treated patients lost an average of eight letters over the study period (compared with 15 letters in the usual care group). Adverse ocular events in the VISION trial resulted in severe vision loss in 0.1% of patients. These adverse events included endophthalmitis (1.3%), traumatic lens injury (0.6%), and retinal detachment (0.6%).

For year 2 of the VISION study, patients were re-randomized to the treatment and usual care arms

(9). The results indicated that those patients continuing with pegaptanib treatment for a second year did better than those reassigned to the usual care control arm at 54 weeks, and better than those assigned to the usual care arm for the entire two years. The percentage of pegaptanib-treated patients who progressed to moderate visual loss (from baseline) during the second year of treatment was half (7%) that of those reassigned to the control group at 54 weeks (14%), and those who continued in the control group for the second year (14%). Of note, however, patients who had benefited from their year 1 treatment assignment (defined as %0 letters of vision loss from baseline), and who subsequently lost R10 letters of vision after re-randomization at 54 weeks, were allowed to receive “salvage therapy” (a reassignment back to their original year 1 treatment arm). Year 2 safety data continue to show that pegaptanib is a relatively safe drug. Non-ocular hemorrhagic events were not significantly different from the usual care group (10).

Studies with pegaptanib continue. The Verteporfin Intravitreal Triamcinolone Acetonide Study (VERITAS) is a phase III prospective, multicenter, randomized, double-masked trial comparing PDT combined with one of two doses of intravitreal triamcinolone (1 mg, 4 mg) versus PDT combined with 0.3 mg of intravitreal pegaptanib. Approximately 100 patients have been enrolled, including all CNV lesion subtypes (11).

Studies are also ongoing at OSI/Eyetech to create a sustained-release form of pegaptanib, with the goal of reducing the frequency of intravitreal injections required for treatment, and thereby reducing the risk of serious adverse events associated with intravitreal injections, such as endophthalmitis and retinal detachment. Preliminary animal work with poly(lactic-co-glycolic) acid (PLGA)-based microsphere encapsulation suggests that sustained-release of pegaptanib for greater than six months is possible with a single intravitreal injection (12).

Monoclonal Antibodies: Ranibizumab (Lucentis)

In June 2006, ranibizumab (Lucentis—Genentech, South San Francisco, California, U.S.A.) became the second VEGF inhibitor approved by the FDA for use in the treatment of CNV secondary to AMD. Ranibizumab is a humanized, affinity-maturated Fab fragment of a murine monoclonal antibody directed against human VEGF-A. Ranibizumab is a potent, non-selective inhibitor of all VEGF-A isoforms and bioactive proteolytic products. Ranibizumab was specifically designed as a molecule smaller than its parent full-size precursor anti-VEGF antibody, because it was felt that the full-sized antibody was unable to cross the inner retina and choroid, as suggested by a histologic study of the Herceptin antibody (13). More recent histologic analysis of bevacizumab in rabbits by Sharar et al. (14) however, suggests that the full-length antibody actually does penetrate all layers of the retina quite effectively. Because ranibizumab is missing the Fc region, it is also felt that the molecule will be less likely to incite an immune response, as it can no longer bind to complement C1q or Fc gamma receptors (15).

Efficacy and safety of ranibizumab has thus far been established through two large prospective, multicenter, randomized, double-masked, controlled clinical trials: Minimally Classic/Occult Trial of AntiVEGF Antibody Ranibizumab in the Treatment of Neovascular Age-Related Macular Degeneration (MARINA) (16) and Anti-VEGF Antibody for the Treatment of Predominantly Classic CNV in AMD (ANCHOR) (17). The MARINA trial was limited to patients with subfoveal occult or minimally classic CNV, either primary or recurrent, with evidence of recent disease progression. In MARINA, 716 patients were randomized 1:1:1 to receive monthly intravitreal injections of ranibizumab (either 0.3 or 0.5 mg) or sham injections. The primary outcome measure was the proportion of patients losing less than 15 ETDRS letters at 12 months. 94.5% of patients assigned to the 0.3 mg group and 94.6% of patients assigned to the 0.5 mg ranibizumab treatment arms, compared with 62.2% in the sham-treatment arm, met this endpoint. More eyes gained 15 or more letters of visual acuity by month 12 in the ranibizumab treatment arms than the control arms: 24.8% in the 0.3 mg group, 33.8% in the 0.5 mg group, 5.0% in the sham-treated group. Mean visual acuity increased by 6.5 letters in the 0.3 mg group and 7.2 letters in the 0.5 mg group at 12 months. In contrast, mean visual acuity dropped by 10.4 letters in the sham-treated group. In general, vision gains were maintained throughout year two of the MARINA trial in ranibizumab-treated patients, whereas vision continued to decline in the shamtreated patients; mean loss was 14.9 letters in the sham group.

There was also a difference in the lesion size outcomes between the ranibizumab and control groups. While lesion size on average remained stable in the ranibizumab-treated patients, lesion size increased by about 50% in the sham-treated patients at 12 months. The area of leakage in the ranibizumab-treated lesions decreased on average by approximately 50%.

Adverse ocular events in ranibizumab-treated patients in the MARINA trial over 24 months included presumed endophthalmitis in 1.0% of patients and serious uveitis in 1.3% of patients. No retinal detachments were observed in the ranibizumabtreated patients, although retinal tears were identified in two patients (0.4%). Lens damage as a result of intravitreal injection was seen in one patient (0.2%). No statistically significant difference in serious systemic adverse events was observed between the treatment and control arms of the study, although there was a trend toward the increased rate of serious (1.3% in 0.3 mg group; 2.1% in 0.5 mg group; 0.8% in sham group) and non-serious (9.2% in 0.3 mg group; 8.8% in 0.5 mg group; 5.5% in sham group) non-ocular hemorrhages.

The ANCHOR trial—has likewise demonstrated efficacy of ranibizumab for the treatment of predominantly classic CNV lesions secondary to AMD. ANCHOR was designed as a head-to-head comparison between ranibizumab and PDT with verteporfin (Visudyne), which was then the standard of care for subfoveal CNV. 423 patients were randomized 1:1:1 to receive monthly intravitreal injections with ranibizumab 0.3 mg and sham PDT, ranibizumab 0.5 mg with sham PDT or monthly sham injections plus active verteporfin PDT. The primary end point was the number of patients losing fewer than 15 letters of baseline visual acuity at 12 months. This end point was achieved in 94.3% of the patients receiving 0.3 mg ranibizumab and 96.4% of patients receiving 0.5 mg ranibizumab versus 64.3% of the verteporfin group. The percentage of patients experiencing an improvement over baseline visual acuity of at least 15 letters was 35.7% and 40.3% respectively, in the ranibizumabtreated patients, versus only 5.6% in the verteporfintreated patients. Mean visual acuity increased by 8.5 letters in the 0.3 mg ranibizumab group and 11.3 letters in the 0.5 mg ranibizumab group at 12 months. In contrast, mean visual acuity dropped by 9.5 letters in the verteporfin PDT group at 12 months.

Measurement of CNV lesion size throughout the ANCHOR study revealed positive morphologic effects similar to those observed in the MARINA study. In general, average total lesion size remained relatively stable in the ranibizumab-treated patients over one year, while increasing significantly in the verteporfintreated patients. Moreover, average total area of

leakage and average total area of classic CNV leakage both decreased significantly at one year in the ranibizumab-treated patients, while they increased in the verteporfin-treated group.

No statistically significant difference in serious systemic adverse events was observed between the ranibizumab and verteporfin arms of the study, but as in the MARINA trial, there was a trend toward an increased rate of serious (1.5% in 0.3 mg group; 2.1% in 0.5 mg group; 0% in PDT group) and non-serious (5.1% in 0.3 mg group; 6.4% in 0.5 mg group; 2.1% in PDT group) non-ocular hemorrhages. Serious adverse ocular events in the ranibizumab-treated ANCHOR trial patients over 12 months included presumed endophthalmitis in 0.7% of patients and significant uveitis in 0.4% of patients. One patient each developed a retinal detachment (0.4%) or vitreous hemorrhage (0.4%). There were no cases of lens damage as a result of the intravitreal injection. The most common adverse event (12% patients) was mild post-injection inflammation.

The PIER study is a phase IIIb, prospective, multicenter, randomized, double-masked, controlled study of 184 patients with predominantly classic or occult CNV randomized to receive ranibizumab or sham injections monthly for the first three months, followed by once every three months for a total of 24 months. The purpose of PIER is to help determine the optimal dosing schedule for ranibizumab. The oneyear results of the PIER study showed that 83% (0.3 mg) and 90% (0.5 mg) of ranibizumab-treated eyes lost less than 15 letters of visual acuity, compared to 49% of sham eyes. However, the percentage of eyes improving 15 or more letters was only 12% (0.3 mg) and 13% (0.5 mg) in ranibizumab-treated eyes, compared with 10% of sham eyes (18).

Prospective optical coherence tomography (PRONTO) imaging of patients with neovascular AMD treated with intraocular ranibizumab is a two-year, single site, open-label, uncontrolled study of 40 patients designed to evaluate the durability of response of ranibizumab and whether optical coherence tomography (OCT) can be used to guide treatment of neovascular AMD (19). As in the PIER study, patients receive monthly injections of ranibizumab for the first three months. Thereafter, re-treatment with ranibizumab is performed if one of the following changes were observed between visits: a loss of 5 letters in vision in conjunction with fluid on OCT, increase in OCT central retinal thickness of at least 100 mm, new onset classic CNV, new macular hemorrhage, or persistent macular fluid detected by OCT at least 1 month after the previous injection of ranibizumab. At 12 months, mean visual acuity improved by 9.3 letters (p!0.001) and the mean OCT central retinal thickness decreased by 178 mm

250 KLESERT ET AL.

(p!0.001). Visual acuity improved 15 or more letters in 35% of patients. These visual acuity and OCToutcomes were achieved with an average of 5.6 injections over 12 months. Once a fluid-free macula was achieved, the mean injection-free interval was 4.5 months before another reinjection was necessary. Unlike the PIER study, visual acuity gains did occur despite the less frequent dosing scheme. PRONTO outcomes suggest that OCT can be useful for guiding re-treatment with intravitreal ranibizumab in neovascular AMD, and that use of an OCT-guided variable-dosing regimen could decrease the injection burden without sacrificing improvements in visual acuity.

Monoclonal Antibodies: Bevacizumab (Avastin)

Bevacizumab (Avastinw, Genentech, South San Francisco, California, U.S.A.) is a full-length humanized murine monoclonal antibody directed against human VEGF-A. It was FDA approved in 2004 for the intravenous treatment of metastatic colorectal cancer. Its potential for use in the treatment of CNV was first tested by Michels et al. (20) via intravenous infusion in a 12-week open-label uncontrolled study. Striking effects were observed on both visual acuity, and the OCT and angiographic characteristics of the neovascular lesions. However, patients experienced a mean increase of 12 mmHg in systolic blood pressure, which was felt to be a deterrent to its common use.

This systemic side effect, combined with the promising visual and anatomic results from the intravenous infusion of bevacizumab, led investigators to consider intravitreal injection of bevacizumab (21). Since then, several retrospective, uncontrolled, openlabel case series have been published regarding the use of intravitreal bevacizumab for the treatment of CNV secondary to AMD (22–25). As with ranibizumab, the effect of bevacizumab has been impressive.

Avery and colleagues (22) treated 79 patients with 1.25 mg of intravitreal bevacizumab monthly and reported the early results at three months follow-up. Many of these patients had prior failed treatment with verteporfin or pegaptanib. At three months, median Snellen visual acuity improved from 20/200 at baseline to 20/80. Mean central retinal thickness by OCT decreased by 67 mm at 3 months. No ocular or systemic adverse events were observed.

Spaide and colleagues treated (23) 266 patients with monthly 1.25 mg of intravitreal bevacizumab. By three months, Snellen visual acuity improved from a mean of 20/184 at baseline to 20/109, with 38.3% of patients experiencing some improvement in visual acuity. Mean central retinal thickness by OCT improved from 340 mm at baseline to 213 mm at 3 months. Again, no adverse ocular or systemic adverse events were observed.

In contrast to the intravenous administration of bevacizumab, intravitreal injection of bevacizumab did not result in the systemic side effect of hypertension in any of these studies. The systemic concentration of bevacizumab when given intravenously is obviously several times larger than the systemic concentrations seen after intravitreal injections, and no elevation in blood pressure has yet been reported in patients treated with intravitreal bevacizumab.

Animal and in vitro studies published thus far have failed to identify any specific toxicity associated with bevacizumab use. Luthra et al. (26) demonstrated that viability of human RPE cells, rat neurosensory cells and human microvascular endothelial cells in culture was normal after exposure to bevacizumab at concentrations of up to 1 mg/mL. Rabbit studies by Manzano et al. (27) found no changes in the electroretinogram (ERG) patterns of eyes injected with intravitreal bevacizumab at doses up to 5.0 mg. Mild vitreous inflammation was seen at 5.0-mg dose, but not at lower doses. Bakri et al. (28) looked at retinal histology of rabbit eyes injected with bevacizumab, and again found no histologic changes compared with control eyes.

One important aspect in which ranibizumab and bevacizumab may differ is their pharmacokinetics. Because of its larger molecular weight, it is assumed that bevacizumab has a significantly longer half-life in the vitreous, and possibly systemically as well. A longer half-life may allow for less frequent injections to achieve the same biologic effect. Recent unpublished data, however, indicate that the half-lives of the two drugs may actually be quite similar. Per the package insert for Lucentis, the half-life of ranibizumab in the vitreous is approximately 3 days based on animal studies. Pharmacokinetic studies in rabbits reveal that the half-life of bevacizumab in the vitreous is only marginally longer at 4.3 days (29).

Although the limited data thus far suggest that bevacizumab is highly effective and safe for the treatment of CNV secondary to AMD, without a head-to-head prospective clinical trial, the relative efficacy and safety of bevacizumab compared with ranibizumab will remain unknown. Fortunately, The National Eye Institute has agreed to sponsor a trial comparing bevacizumab with ranibizumab in AMD patients with subfoveal CNV. This study, the Comparison of Treatment Trial (CATT) study, will randomize patients into one of four treatment arms: monthly intravitreal injection of ranibizumab, monthly injection of bevacizumab, monthly injection of ranibizumab followed by as-needed treatment, and monthly injection of bevacizumab followed by as-needed treatment. Until the results of the CATT study are available, bevacizumab is nonetheless

an attractive treatment option due to its cost advantage over ranibizumab, especially for those patients without drug insurance coverage or with large drug co-payment requirements.

Because bevacizumab has been FDA approved only for the intravenous treatment of metastatic colon cancer, intravitreal injection of bevacizumab is an off-label use of drug by an altered route of administration. This makes documentation of the informed consent process especially important when using bevacizumab. During informed consent, the physician should explain to patients that the safety and efficacy of bevacizumab have not been establishedwith certainty, and that there may be unknown risks with its use. A bevacizumab-specific consent form is recommended, and can be found on the website of the Ophthalmic Mutual Insurance Company (OMIC) (30,31).

Bevacizumab comes in preservative-free 100 mg vials, containing 4 cc of a 25 mg/cc solution, intended for one-time use only for treatment of a single cancer patient. A single vial can theoretically be aliquoted out to provide up to eighty individual 0.05 cc intravitreal doses in 1 cc tuberculin syringes. The pharmacy should confirm the dose and sterility, provide proper storage instructions, and mark all aliquots with an expiration date. Although bevacizumab is a very stable drug with a shelf-life of many months, compounded aliquots will usually have an expiration date due to sterility concerns.

Combination Therapy with PDT

The RhuFab V2 Ocular Treatment Combining the Use of VISUDYNEw to Evaluate Safety (FOCUS) study (32) is a two-year, phase I/II, multicenter, randomized, single-masked, controlled study of 162 patients with predominantly classic CNV. FOCUS compared the safety and efficacy of intravitreal ranibizumab (0.5 mg) combined with verteporfin PDT versus verteporfin PDT alone (combined with sham injection). Patients received monthly ranibizumab (0.5 mg) (nZ106) or sham (nZ56) injections. The PDT was performed seven days before initial ranibizumab or sham treatment and then quarterly as needed. The primary outcome measure was the proportion of patients who lost fewer than 15 letters from baseline at 12 months. At 12 months, 90.5% of the ranibizumabtreated patients and 67.9% of the control patients lost fewer than 15 letters (p!0.001).

The most frequent ranibizumab-associated serious ocular adverse events were intraocular inflammation (11.4%) and endophthalmitis (1.9%; 4.8% if including presumed cases). On average, patients with serious inflammation had better visual acuity outcomes at 12 months than did controls. Key serious non-ocular adverse events included myocardial infarctions in the PDT-alone group (3.6%) and

cerebrovascular accidents in the ranibizumab-treated group (3.8%). Notably, ranibizumab-treated patients experiencing intraocular inflammation still had better visual acuity outcomes at 12 months than the control patients. Thus, ranibizumab combined with PDT was more efficacious than PDT alone for treating neovascular AMD.

In addition, the FOCUS study showed that despite a history of prior PDT therapy, a significant proportion of these patients were able to gain visual acuity when treated with ranibizumab and PDT. The need for additional PDT was 27.6% for the combined group but 91.1% for the PDT group. A difference in the rate of PDT re-treatment was seen by the 3-month follow-up period and maintained for the study.

The FOCUS study however did not compare the ranibizumab plus PDT combination to ranibizumab alone. The DENALI study is a randomized, controlled, multicenter clinical trial that will perform the comparison study. The study will gauge the safety, efficacy and impact on re-treatment rates of Visudyne (verteporfin, Novartis) and Lucentis (ranibizumab, Genentech, South San Francisco, California, U.S.A.) as a combination therapy against wet AMD. DENALI is expected to enroll 300 wet AMD patients at 45 centers in the United States and five centers in Canada. The two-year study will investigate whether patients receiving the combination therapy require fewer re-treatments than control patients treated with Lucentis monotherapy.

Intravitreal Injection Technique

It appears that the greatest risks associated with the use of current anti-VEGF therapies for the treatment of AMD (endophthalmitis, retinal detachment, lens trauma) come from the intravitreal injection itself. Therefore, proper injection technique and careful antiseptic practices are important.

Supplies that are recommended for prepping the eye include 5% povidine-iodine solution, povidineiodine sticks, and a sterile lid speculum. At our center, we use sterile gloves, a sterile drape, and an empty sterile 1 cc tuberculin syringe to mark the sclera. Alternatively, one can use a caliper to mark the location for the injection procedure. The drug is drawn from the drug vial using a filtered needle attached to a tuberculin syringe. The needle is then changed to a sterile 30-gauge needle prior to the injection. Preinjection prophylactic antibiotic drops may also be used, although no benefit of antibiotic prophylaxis has been established.

The eye should first be anesthetized. In our hands, topical anesthesia appears to work just as well as subconjunctival injection of lidocaine, but either method can be used. For topical anesthesia, a cotton tip applicator is soaked with tetracaine and

252 KLESERT ET AL.

placed under the upper or lower eyelid in the conjuntival fornix, so that it rests against the superotemporal or inferotemporal bulbar conjuctiva at the site where the injection is planned. The patient should be instructed to look in the opposite direction and remain that way, so as not to scratch the cornea on the cotton tip applicator. After three to five minutes, the applicator can be removed and the eye prepped with 5% povidine-iodine solution placed directly on the eye, and povidine sticks used to clean the eyelids, lashes and periocular skin. Gloves are worn and a sterile lid speculum is inserted between the eyelids. A sterile drape may be used over the eye if desired, but is not necessary. The patient is then asked to fix his or her gaze in the direction opposite to where the injection is planned, so as to provide the best possible exposure. Providing the patients with an object to fixate upon, such as their own raised thumb, can improve stability of the eye during the injection.

A sterile 1 cc syringe hub or a sterile caliper can be used to mark the site of injection. The safest point of injection in phakic patients is 4 mm posterior to the limbus, and the round tip of the tuberculin 1 cc syringe happens to be 4 mm in diameter. The drug is then injected into the vitreous cavity through the pars plana using a 30-gauge needle (0.05 cc total volume in the case of ranibizumab or bevacizumab, 0.1 cc total volume in the case of pegaptanib). The needle is withdrawn and a dry cotton tip applicator is immediately applied over the injection site for a few seconds to help prevent prolapse and incarceration of vitreous in the wound, which can serve as a possible wick for the introduction of bacteria into the eye. Antibiotic drops are then placed in the eye and the lid speculum is removed. The eye pressure is monitored following the injection to confirm that it returns to normal. Finally, the patient is sent home with prophylactic antibiotic drops to be used for three days.

Most compliant patients do not need to be rechecked in the clinic until they are due for their next injection four to six weeks later, presuming you give them clear instructions on the signs and symptoms of infection or retinal detachment and are confident that they will call you immediately if they were to develop these symptoms. Povidine-iodine can be quite irritating to the corneal epithelium. It is therefore normal for patients to have some degree of irritation, burning and tearing following their injection, in addition to varying amounts of subconjuctival hemorrhage. The wise physician will warn their patients of these possibilities at the time of injection in order to prevent the inevitable after-hours telephone call. However, any antiseptic-associated discomfort should resolve by the following day.

Therefore, any pain or decreased vision reported by the patient on post-injection day one or later should be taken very seriously.

Safety Considerations

The observation that injection of intravitreal bevacizumab (33) or pegaptanib (34) for the treatment of proliferative diabetic retinopathy results in regression of neovascularization in the fellow eye provides compelling evidence that these molecules are indeed absorbed systemically to levels that are clinically relevant. Although no serious systemic concerns were raised by the MARINA, ANCHOR or VISION studies, it should be remembered that studies of this size are powered to detect only relatively large differences in rare events between the study groups. A modest increase in the risk of heart attack or stroke, for example, might not be detected by these studies. In this regard, both the MARINA (16) and ANCHOR (17) trials revealed a non-statistically significant trend toward an increased risk of serious systemic hemorrhage. In MARINA, the incidence of such events was 1.3% in 0.3 mg group, 2.1% in 0.5 mg group, versus 0.8% in sham group at 24 months. In ANCHOR, the incidence of such events was 1.5% in 0.3 mg group, 2.1% in 0.5 mg group, versus 0% in sham group at 12 months. A similar trend was observed for non-serious systemic hemorrhages. No such trend was observed in the VISION trial (10) of pegaptanib, in which the incidence of serious systemic hemorrhage was 0.5% in the treatment arm, versus 1.9% in the sham arm.

These data simply underscore the fact that antiVEGF agents are potent drugs, and they should always be used with due caution and consideration.

FUTURE ANTI-VEGF THERAPIES

VEGF Trap

Pegaptanib, ranibizumab and bevacizumab all act through inhibition of VEGF-A; they do not bind other members of the VEGF family. VEGF Trap (Regeneron, Tarrytown, New York, U.S.A.) is an experimental new drug designed to inhibit all members of the VEGF family: VEGF-A, -B, -C, -D, and Placental growth factors (PlGF-1 and PlGF -2). VEGF Trap is a recombinant chimeric VEGF receptor fusion protein in which the binding domains of VEGF receptors 1 and 2 are combined with the Fc portion of immunoglobulin G to create a stable, soluble, highaffinity inhibitor. VEGF Trap also binds VEGF-A with higher affinity (kD!1 pmol/L) than any of the currently available anti-VEGF drugs (35). Whether the broader spectrum and higher affinity of VEGF Trap equates to improved efficacy in the treatment of CNV secondary to AMD remains to be determined.

The CLEAR-AMD 1 study is a randomized, multicenter, placebo-controlled, dose-escalation study designed to assess the safety, tolerability and bioactivity of VEGF Trap (35). The study enrolled 25 patients with CNV secondary to AMD with lesions %12 disc areas is size and with R50% active leakage, and with ETDRS visual acuity of 20/40 or worse. Patients were randomized to receive either placebo or one of three doses of VEGF Trap (0.3, 1.0, or 3.0 mg/kg). The VEGF trap was given as a single intravenous dose, followed by a four-week observation period, followed by three additional doses two weeks apart. Dose-limiting toxicity was observed for two of the five patients treated with the 3.0 mg/kg dose: one patient developed grade 4 hypertension and the other developed grade 2 proteinuria. Although reduced leakage on fluorescein angiography and reduced retinal thickening on OCT was observed in the treated patients, there was no corresponding reduction in CNV lesion size or improvement in visual acuity observed in these patients over the short 71-day study period. It was concluded that the maximum tolerated IV dose of VEGF Trap was 1.0 mg/kg.

The CLEAR-IT 1 study is similarly designed to assess the safety, tolerability and bioactivity of VEGF Trap through the intravitreal route of administration (36). The study enrolled 21 patients using the same inclusion criteria as CLEAR-AMD 1, and randomized them to receive one of six doses of VEGF Trap as single intravitreal injection: 0.05, 0.15, 0.5, 1.0, 2.0 or 4.0 mg. After 43 days of follow-up, no adverse ocular or systemic events were observed. Mean decrease in excess foveal thickness for all patients was 72%. The mean increase in ETDRS visual acuity was 4.75 letters and visual acuity remained stable or improved in 95% of patients. Notably, 3 out of 6 patients treated with the higher doses (2.0 or 4.0 mg) gained R3 lines of visual acuity by day 43. Clearly, VEGF Trap given intravitreally shows promise as a novel treatment for CNV in AMD patients.

Small Interfering RNAs (siRNAs)

The therapeutic potential of RNA interference was born in 1998, when Fire and Mello (37) discovered that injection of gene-specific double stranded RNA into cells resulted in potent silencing of that gene’s expression. They had discovered one of fundamental mechanisms by which the cell regulates gene expression and protects itself against viral infection: RNA interference. Fire and Mello were awarded the Nobel Prize in Physiology and Medicine for 2006.

The components of the RNA interference machinery have since been identified. Doublestranded RNA binds to a protein complex called Dicer, which cleaves it into multiple smaller fragments. A second protein complex called RNA induced

silencing complex (RISC) then binds these RNA fragments and eliminates one of the strands. The remaining strand stays bound to RISC, and serves as a probe that recognizes the corresponding messenger RNA transcript in the cell. When the RISC complex finds a complementary messenger RNA transcript, the transcript is cleaved and degraded, thus silencing that gene’s expression (38).

Small interfering RNAs (siRNAs) have quickly become important tools in genetic research, and their potential as therapeutic agents is being explored in many areas of medicine. Reich and Tolentino (38,39) were the first to apply siRNA technology toward the treatment of CNV. Bevasiranib/Cand5 (Acuity Pharmaceuticals, Philadelphia, Pennsylvania, U.S.A.) is a siRNA inhibitor of VEGF, which is given as an intravitreal injection. A phase I, open-label, dose escalation study of 15 patients revealed no serious ocular or systemic adverse effects at a dose up to 3.0 mg.

The CARE study (Cand5 Anti-VEGF RNA: Evaluation) is a phase II multicenter, randomized, double-masked, trial of bevasiranib/Cand5 in patients with CNV secondary to AMD (40). 127 Patients with predominantly classic, minimally classic, or retinal angiomatous proliferation lesions (occult no classic lesions excluded) were randomized to receive one of three doses of the drug (0.2, 1.5, and 3.0 mg) at baseline and at 6 weeks. The primary endpoint was the mean change in ETDRS visual acuity from baseline at 12 weeks, which was 4 letters (0.2 mg), 7 letters (1.5 mg), and 6 letters (3.0 mg). The authors have theorized that these disappointing results stem from the fact that bevaciranib/Cand5 only blocks the production of new VEGF–VEGF already present at the time of injection was not inhibited. The investigators postulated that a baseline combination treatment with a VEGF protein blocker may be required to “mop up” the preexisting VEGF load. However, the half-life of VEGF is short and it does not explain why the results were not seen by 12 weeks time point with siRNA treatment. Efficacy for the proposed treatment combination remains to be shown. The investigators envision a role of bevasiranib/Cand5 as a long-term “maintenance” drug. The CARE trial raised no safety concerns, with only one patient developing uveitis.

Other therapeutic targets for siRNAs are being investigated. siRNAs directed against the VEGFR-1 receptor have shown promise in a mouse model of CNV (41), and are currently in clinical development (Sirna-027, Sirna Therapeutics, Boulder, Colorado, U.S.A.).

Receptor Tyrosine Kinase Inhibitors

Non-RNA inhibitors of VEGF receptor tyrosine kinase activity have been identified, and their anti-angiogenic

254 KLESERT ET AL.

properties are being investigated for use in the treatment of systemic malignancy, as well as CNV. One advantage of this class of drugs over those discussed thus far in this chapter is the possibility of an oral route of administration, thereby avoiding the ocular complications associated with frequent intravitreal injections.

One promising compound is PTK787, which is a non-selective inhibitor of all known VEGF receptors (42). PTK787 has been shown to inhibit retinal neovascularization in a hypoxic mouse model (43,44). Phase I/II clinical trials of PTK787 (Vatalanib, Novartis, East Hanover, New Jersey, U.S.A.), have been done in patients with both solid and hematologic malignancies, such as the randomized, doublemasked, multicenter, phase I/II study of the safety of vatalanib administered in conjunction with photodynamic therapy with verteporfin to patients with predominantly classic, minimally classic or occult with no classic subfoveal CNV secondary to AMD. A multicenter phase I trial of PTK787/Vatalanib in patients with AMD is the ADVANCE study. Patients with all CNV lesion types will receive PDT with Visudyne at baseline, and will be randomized to receive concurrent treatment with either 500 or 1000 mg of oral PTK787/Vatalanib or placebo, once daily for three months (45). ADVANCE is designed to assess the safety and efficacy of the drug.

AG-013958 (Pfizer, San Diego, California, U.S.A.) is a selective VEGFR and PDGFR inhibitor that is currently in phase I/II testing. The route of administration being examined is subtenon injection. Preliminary results of 21 patients with subfoveal CNV indicated that adverse events were mild (15).

Anti-VEGF treatment has enabled a sizeable proportion of treated patients to attain significant visual improvement or to maintain vision. Future research will hopefully continue to build on these advances and make restoration of vision a reality for the majority of these patients.

SUMMARY POINTS

&The only two anti-VEGF agents currently approved by the FDA for treatment of CNV are pegaptanib (Macugen, New York, U.S.A.) and ranibizumab (Lucentis).

&Pegaptanib, an aptamer (short oligonucleotide) that specifically binds and inhibits VEGF isoforms containing at least 165 amino acids, was shown to slow the rate of vision loss in a large, prospective, randomized clinical trial.

&Ranibizumab, an antigen binding fragment of a humanized monoclonal antibody directed against all the biologically active forms of VEGF, including the known active proteolytic breakdown products,

was shown to slow the rate of vision loss in two large, prospective, randomized clinical trials.

&Bevacizumab is a full-sized humanized monoclonal antibody with VEGF binding characteristics similar to ranibizumab, is approved by the FDA for systemic treatment of metststic colorectal cancer and lung cancer, but is used off-label for the treatment of neovascular AMD.

&Efficacy and safety of bevacizumab for the treatment of neovascular AMD have been reported in several retrospective case series and two small prospective studies, but a large, prospective, randomized, controlled clinical trial has not yet been performed.

&Additional anti-VEGF drugs are in different stages of development but none have yet entered phase III trials as of January 2007.

REFERENCES

1.Ferrara N, Henzel WJ. Pituitary follicular cells secrete a novel heparin-binding growth factor specific for vascular endothelial cells. Biochem Biophys Res Commun 1989; 161:851–8.

2.Michaelson IC. The mode of development of the vascular system of the retina with some observations on its significance for certain retinal disorders. Trans Ophthalmol Soc UK 1948; 68:137–80.

3.Kvanta A, Algvere PV, Berglin L, et al. Subfoveal fibrovascular membranes in age-related macular degeneration express vascular endothelial growth factor. Invest Ophthalmol Vis Sci 1996; 37:1929–34.

4.Wells JA, Murthy R, Chibber R, et al. Levels of vascular endothelial growth factor are elevated in the vitreous of patients with subretinal neovascularisation. Br J Ophthalmol 1996; 80:363–6.

5.Robinson C, Stinger S. The splice variants of vascular

endothelial growth factor (VEGF) and their receptors. J Cell Sci 1991; 114:853–65.

6.Keyt BA, Berleau LT, Nguyen HV, et al. The carboxylterminal domain (111–165) of vascular endothelial growth factor is critical for its mitogenic potency. J Biol Chem 1996; 271:7788–95.

7.Ng EW, Shima DT, Calias P, et al. Pegaptanib, a targeted anti-VEGF aptamer for ocular vascular disease. Nat Rev Drug Discov 2006; 5:123–32.

8.Gragoudas ES, Adamis AP, Cunningham ET, Jr., et al. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med 2004; 351:2805–16.

9.Chakravarthy U, Adamis AP, Cunningham ET, Jr., et al. Year 2 efficacy results of 2 randomized controlled clinical trials of pegaptanib for neovascular age-related macular degeneration. Ophthalmology 2006; 113:1508–21.

10.D’Amico DJ, Masonson HN, Patel M, et al. Pegaptanib sodium for neovascular age related macular degeneration: two-year safety results of the two prospective, multicenter, controlled clinical trials. Ophthalmology 2006; 113: 992–1001.

11.Kaiser PK, VERITAS Study Group. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

12.Adamis AP, Cook G, Shima D, et al. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

13.Mordenti J, Cuthbertson RA, Ferrara N, et al. Comparisons of the intraocular tissue distribution, pharmacokinetics, and safety of 125I-labeled full-length and fab antibodies in rhesus monkeys following intravitreal administration. Toxicol Pathol 1999; 27:536–44.

14.Shahar J, Avery RL, Heilweil G, et al. Electrophysiologic and retinal penetration studies following intravitreal injection of bevacizumab (Avastin). Retina 2006; 26:262–9.

15.Kaiser PK. Antivascular endothelial growth factor agents and their development: therapeutic implications in ocular diseases. Am J Ophthalmol 2006; 142:660e1–10.

16.Rosenfeld PJ, Brown DM, Heier JS, et al. Ranibizumab for

neovascular age-related macular degeneration. N Engl J Med 2006; 355:1419–31.

17.Brown DM, Kaiser PK, Michels M, et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med 2006; 355:1432–44.

18.Regillo CD, Brown DM, Abraham H, Kaiser PK, Mieler WF. Randomized, double-masked, sham-controlled trial of ranibizumab for neovascular age-related macular degeneration: PIER study year 1. Am J Ophthalmol 2007 (in press).

19.Fung AE, Lalwani GA, Rosenfeld PJ, et al. An OCT guided, variable dosing regimen with intravitreal ranibizumab (Lucentis) for Neovascular age-related macular degeneration. Am J Ophthalmol (in press).

20.Michels S, Rosenfeld JR, Puliafito CA, et al. Systemic bevacizumab (Avastin) therapy for neovascular agerelated macular degeneration: twelve-week results of an uncontrolled open-label clinical study. Ophthalmology 2005; 112:1035–47.

21.Rosenfeld PJ, Moshfeghi AA, Puliafito CA. Optical coherence tomography findings after an intravitreal injection of bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmic Surg Lasers Imaging 2005; 36:331–5.

22.Avery RL, Pieramici DJ, Rabena MD, et al. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006; 113:363–72.

23.Spaide RF, Laud K, Fine HF, et al. Intravitreal bevacizumab treatment of choroidal neovascularization secondary to age-related macular degeneration. Retina 2006; 26:383–90.

24.Rich RM, Rosenfeld PJ, Puliafito CA, et al. Short-term safety and efficacy of intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Retina 2006; 26:495–511.

25.Bashshur ZF, Bazarbachi A, Schakal A, et al. Intravitreal

bevacizumab for the management of choroidal neovascularization in age-related macular degeneration. Am J Ophthalmol 2006; 142:1–9.

26.Luthra S, Narayanan R, Marques LE, et al. Evaluation of in vitro effects of bevacizumab (Avastin) on retinal pigment epithelial, neurosensory retinal, and microvascular endothelial cells. Retina 2006; 26:512–8.

27.Manzano RP, Peyman GA, Khan P, et al. Testing intravitreal toxicity of bevacizumab (Avastin). Retina 2006; 26:257–61.

28.Bakri SJ, Cameron JD, McCannel CA, et al. Absence of histologic retinal toxicity of intravitreal bevacizumab in a rabbit model. Am J Ophthalmol 2006; 142:162–4.

29.Bakri SJ, Snyder MR, Pulido JS, et al. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

30.www.omic.com (Accessed on July 15, 2007).

31.Klesert TR. So you want to try intravitreal Avastin. Retina Times 2006; 14:18–21.

32.Heier JS, Boyer DS, Ciulla TA, et al. Ranibizumab combined with verteporfin photodynamic therapy in neovascular age-related macular degeneration: year 1 results of the FOCUS study. Arch Ophthalmol 2006; 124:1532–42.

33.Avery RL, Pearlman J, Pieramici DJ, et al. Intravitreal bevacizumab (Avastin) in the treatment of proliferative diabetic retinopathy. Ophthalmology 2006; 113:1695e1–15.

34.Adamis AP, AltaweelM, Bressler NM, et al. Changes in retinal

neovascularization after pegaptanib (Macugen) therapy in diabetic individuals. Ophthalmology 2006; 113:23–8.

35.Nguyen QD, Shah SM, Hafiz G, et al. A phase I trial of an IV-administered vascular endothelial growth factor trap for treatment in patients with choroidal neovascularization due to age-related macular degeneration. Ophthalmology 2006; 113:1522–38.

36.Nguyen QD, Hariprasad S, Shar SM, et al. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

37.Fire A, Xu S, Montgomery MK, et al. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature 1998; 391:806–11.

38.Reich SJ, Fosnot J, Kuroki A, et al. Small interfering RNA (siRNA) targeting VEGF effectively inhibits ocular neovascularization in a mouse model. Mol Vis 2003; 9:210–6.

39.Tolentino MJ, Brucker AJ, Fosnot J, et al. Intravitreal injection of vascular endothelial growth factor small interfering RNA inhibits growth and leakage in a nonhuman primate, laser-induced model of choroidal neovascularization. Retina 2004; 24:132–8.

40.Brucker AJ, The Cand5 Study Group. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

41.Shen J, Samul R, Silva RL, et al. Suppression of ocular neovascularization with siRNA targeting VEGF receptor 1. Gene Ther 2006; 13:225–34.

42.Wood JM, Bold G, Buchdunger E, et al. PTK787/ZK 222584, a novel and potent inhibitor of vascular endothelial growth factor receptor tyrosine kinases, impairs vascular endothelial growth factor-induced responses and tumor growth after oral administration. Cancer Res 2000; 60: 2178–89.

43.Maier P, Unsoeld AS, Junker B, et al. Intravitreal injection of specific receptor tyrosine kinase inhibitor PTK787/ ZK222 584 improves ischemia-induced retinopathy in mice. Graefes Arch Clin Exp Ophthalmol 2005; 243:593–6.

44.Ozaki H, Seo MS, Ozaki K, et al. Blockade of vascular endothelial cell growth factor receptor signaling is sufficient to completely prevent retinal neovascularization. Am J Pathol 2000; 156:697–707.

45.Joondeph BC, Szczesny P, Sforzolini B. Abstract of Papers, Combined Meeting of Club Jules Gonin and The Retina Society, Cape Town, South Africa, October 15–20, 2006.

INTRODUCTION

Age-related macular degeneration (AMD) is the leading cause of visual loss in people older than 65 years in the United States (1–6). Approximately 200,000 Americans per year lose central vision due to AMD and 50,000 will lose vision in both eyes. Currently, there are an estimated 38 million American seniors with a projected 88 million by 2030, which will lead to a proportional increase in the population at risk from vision loss due to AMD. Ninety percent of the severe visual loss from AMD results from choroidal neovascularization (CNV) (2,7,8). Although thermal laser photocoagulation, photodynamic therapy, and various drug therapies to treat neovascular AMD are available or on the horizon, they have only proven to be moderately effective and applicable to a subset of patients (9–18). As a result, the development of preventive strategies for patients at high risk of developing CNV is very desirable. Even a modestly effective bilateral preventive treatment can have a substantial impact on the development of late AMD (geographic atrophy and/or CNV) and the rate of legal blindness caused by CNV. According to one estimate, an intervention that reduced the risk of developing CNV by just 30% in eyes of patients with bilateral large drusen could eventually halve the rate of bilateral blindness from AMD (19).

Several natural history studies have identified the presence of large, soft drusen as a significant risk factor for the development of late complications of AMD (20–22). In 1973, Gass first described the disappearance of drusen after laser photocoagulation (23). Subsequently, laser photocoagulation to promote drusen resorption has been examined in numerous studies as a potential prophylaxis against late complications of AMD.

ANATOMY AND PATHOPHYSIOLOGY

In order to rationalize the potential therapeutic role of prophylactic laser photocoagulation for drusen resorption, it is necessary to define drusen and

understand the anatomy and pathophysiology of the outer retina, retinal pigment epithelium (RPE), Bruch’s membrane and choriocapillaris. The RPE, a monolayer of hexagonal-shaped cells external to the neurosensory retina and internal to Bruch’s membrane, is intrinsically involved in the outer retina’s metabolism. Its functions include phagocytosis of photoreceptor outer segments, maintenance of the blood–retinal barrier and the transportation of nutrients and waste products (24–26). Bruch’s membrane is not a true membrane but a five-layered connective tissue sheet (27). The basal lamina of the RPE is the most internal layer. The inner collagenous layer, elastic lamina and outer collagenous layer comprise the middle elements. The basal lamina of the choriocapillaris is the final structure. The choriocapillaris is the innermost layer of the choroid and is comprised of an anastomosing sheet of large, fenestrated capillaries. The blood flow in the choroid is one of the highest in the body, largely to meet the high metabolic needs of the outer retina and RPE. Nutrients and waste products pass through the fenestrations of the choriocapillaris. Typically, Bruch’s membrane is not a barrier to these molecules and the RPE transports them to and from the outer retina via active and passive mechanisms (28).

Druse (plural drusen) is a German-derived word that means “nodule.” Literally, drusen are crystalline nodules found in stones. In the ophthalmic literature, there have been numerous clinical and histopathologic definitions of drusen (27). The lack of standard terminology for drusen makes interpretation of the literature difficult. Recently, a clinical classification and grading for AMD was developed. In this system, drusen are whitish-yellow spots external to the retina or RPE (29). Hard drusen are less than 63 mm, well defined and yellow–white. Soft drusen are greater than 63 mm and are often also referred to as large drusen. They can have indistinct and distinct borders, may coalesce to form larger, confluent drusen and typically are white–yellow in color. Pathologically, three types of soft drusen have been described: (i) localized detachments of RPE and basal linear deposit in eyes with

258 HSU AND HO

diffuse basal linear deposits; (ii) localized detachments of the RPE and basal laminar deposit in eyes with diffuse basal laminar deposits; and (iii) localized RPE detachments due to focal accumulation of basal linear deposit in eyes without diffuse basal linear deposits (30,31). Ultrastructurally, basal laminar deposits consist of membrane-bound vesicles, wide-spaced collagen and amorphous, granular material located between the plasma membrane and basal lamina of the RPE. Basal linear deposits are located external to the RPE’s basal lamina in the inner collagenous zone. They consist of vesicular and granular electron-dense material and small foci of wide-spaced collagen (30–35). Histochemically, drusen have been shown to consist of lipids, mucopolysaccharides, and glycoconjugates (36–38).

The RPE is a metabolically active tissue layer and, most likely, drusen are derived from RPE (39–41). Studies have demonstrated that RPE cells over time accumulate intracellular lipofuscin and other byproducts of the catabolism of photoreceptor outer segments (42). It has been shown that the RPE deposits cellular material into the sub-RPE space via evagination of its plasma membrane. This probably represents the deposition of the intracellular accumulation of its phagocytic by-products. These plasma membranebound vesicles break down into drusenoid material (41). With normal aging, Bruch’s membrane also undergoes ultrastructural and histochemical changes (43–46). Bruch’s membrane increases in thickness, accumulates lipids and develops protein crosslinking. The hydraulic conductivity (flow per unit pressure) of Bruch’s membrane in normal eyes decreases with age (45). Similar to drusen, these alterations in Bruch’s membrane may also represent the accumulation of waste products from the RPE. The basal linear and laminar deposits and the alterations in Bruch’s membrane may impair the flow of fluid to and from the choriocapillaris. The reduced flow of nutrients and oxygen and the impaired removal of waste products may impose a metabolic strain on the outer retina and RPE. The relative hypoxia of the RPE and outer retina from an enlarged, hydrophobic (lipidladen) Bruch’s membrane and drusen may induce the formation of angiogenic factors and may promote the formation of CNV (47).

DRUSEN AS A RISK FACTOR FOR CNV

Laser-induced drusen regression has generated investigation because soft drusen have been identified as risk factors for CNV and subsequent visual loss. In 1973, Gass noted that 9 of 49 patients (18%) with bilateral macular drusen developed visual loss in one eye due to “disciform detachment or degeneration” over an average of 4.5 years (23). Smiddy followed 71

patients with bilateral macular drusen for an average of 4.3 years (20). Eight eyes of seven patients (9.9%) developed exudative maculopathy. Severe visual loss (more than six lines) occurred in seven eyes and the five-year cumulative risk of developing severe visual loss was 12.7%. Holz prospectively followed 126 patients with bilateral drusen and “good visual acuity” (48). The three-year cumulative incidence of developing CNV or pigment epithelial detachment was 13.3%.

The risk for CNV is higher in patients with drusen in one eye and CNV in the other eye. In Gass’ study, 31 of 91 patients lost central vision from CNV in their fellow eye over an average of four years (23). The Macular Photocoagulation Study Group followed 127 patients who had extrafoveal CNV in one eye (21). In the fellow eye, the risk of developing CNV was 58% over five years if large drusen and RPE hyperpigmentation were present. The risk dropped to 10% if no drusen or hyperpigmentation was present. In another study, the Macular Photocoagulation Study Group verified that large drusen are a significant independent risk factor for CNV (49). In this same study, the risk for CNV jumped to 87% in eyes with five or more drusen, focal hyperpigmentation, one or more large drusen and systemic hypertension. In Sandberg’s study, 127 patients with unilateral CNV were followed for an average of 4.5 years (50). CNV developed in the fellow eyes at a rate of 8.8% per year. Macular appearance, which included large drusen, was significantly associated with CNV. One prospective study followed 101 patients with unilateral CNV and drusen in the fellow eye for up to nine years (51). The yearly incidence of CNV varied between 5% and 11%. Significant risk factors were the number, size and confluence of drusen.

Numerous pathologic studies have shown a correlation between drusen and AMD. Spraul and Grossniklaus examined 51 eyes with AMD and 40 age-matched control eyes (34). Soft, confluent and large drusen as well as basal (linear) deposits correlated with AMD. Curcio demonstrated that basal linear deposits and large drusen are 24 times more likely to be found in eyes with AMD than age-matched control eyes (32).

IMPACT OF LASER PHOTOCOAGULATION

ON PRESENCE OF DRUSEN

In order to understand how laser results in drusen resorption, it is necessary to examine the cellular effects of laser on the outer retina, RPE, Bruch’s membrane and choriocapillaris. Laser energy is largely absorbed by the melanin of the RPE and choroid with shorter wavelengths (e.g., 514 nm argon green laser) having better absorption compared to

longer wavelengths (e.g., 810 nm diode laser). Absorption of the laser light elevates the tissue temperature and causes denaturation of proteins. This thermal effect is called photocoagulation (52,53). The histopathologic characteristics of a laser burn depend on the power, spot size and duration. Smiddy examined the light microscopic changes to a human retina 24 hours after argon laser application (54). The juxtafoveal region was treated with laser spots 200 mm in size and 0.5 seconds in duration. The power ranged between 200 and 400 milliwatts (mW). Histopathologically, there was a choroidal infiltrate of mononuclear and polymorphonuclear cells. The choriocapillaris was acellular at the center of the burn. The RPE was disrupted and the outer and inner retinal nuclear layers were pyknotic. The ganglion and nerve fiber layers were also affected. Thomas conducted a similar study examining a human eye 24 hours after argon laser (55). One laser spot with a power of 310 mW, spot size of 100 mm and duration of 0.5 seconds was applied in the superonasal quadrant. Variable RPE necrosis and advanced choriocapillaris necrosis was seen. A second argon laser burn with a power of 210 mW, spot size of 500 mm and duration of 0.5 seconds in the peripapillary region demonstrated significant RPE disruption, choriocapillaris necrosis and Bruch’s membrane disruption.

A number of studies have been performed with laser on cynomologus monkeys whose fovea is similar to the human fovea. Smiddy placed a 13-spot burn in the juxtafoveal region of a cynomologus monkey with argon green laser and examined the histopathologic effects at one and seven days (56). He used a 200 mm spot size, 0.2 second duration and power between 100 and 200 mW. The desired reaction was a laser burn that turned the retina light gray. At day one, the ganglion cell layer was partially preserved but all deeper layers were necrotic with RPE hyperplasia. At day seven, there was disruption of the retina to the level of the ganglion cell layer. In a second study, Smiddy demonstrated that the RPE undergoes cellular proliferation after argon laser (57). Peyman examined the histopathologic effects of argon blue–green laser to the parafoveal area of cynomologus monkeys (58). He used a 100 mm spot size, 0.1 second duration and power of 100 mW. At day one, there was coagulative necrosis of the RPE, outer nuclear layer and outer plexiform layer. The choroid was minimally affected. At days 12 and 21, glial tissue had replaced the outer retina. There was an inflammatory infiltrate and the RPE was hyperplastic. If the power was increased to 320 mW, the basement membrane was ruptured and choroidal hemorrhages developed. Coscas treated the parafoveal region of adult baboons with argon green laser and examined the light and electron microscopic changes at one hour, three weeks and six weeks (59).

As in the above studies, they showed disruption of the outer retina, necrosis of the RPE and a macrophage response. Depending on the laser settings, there was variable involvement of the choriocapillaris. In a review of macular photocoagulation, Swartz found that the histologic characteristics of a moderate argon-green burn showed a typical cone-shaped lesion sparing the inner retina (60). The laser intensities of these studies exceed those in most human laser to drusen trials.

No histopathologic studies have been performed on human eyes examining the effects of laser on drusen. However, a limited number of experimental animal studies have been reported. Duvall and Tso applied argon green laser directly to drusen in two eyes of a rhesus monkey and noted the light microscopic and ultrastructural characteristics of drusen resorption (61). At zero to two days, outer segment retinal disruption, RPE necrosis and fibrin deposition were noted. The drusen were still present. At three to eight days, two types of macrophages were present. One type was in the outer retina and subretinal space and had an appearance that was consistent with blood-borne monocytes. The second type of macrophage contained cell processes that surrounded the drusen material. These cell processes were traced by serial sectioning to the pericytes of the choriocapillaris. At nine days and beyond, there was resorption of the drusen. Blood-borne monocytes were densely packed in the subretinal space. The cell processes of the choroidal pericytes contained drusenoid material. The authors postulated that the fibrin deposition from the laser photocoagulation initiated a phagocytic response, which resulted in clearance of the drusen by choroidal pericytes. Perry examined the choroidal microvascular response to argon laser in cats (62). He demonstrated activation of the endothelial cells in the choriocapillaris after laser photocoagulation. Della treated a rhesus monkey with soft large drusen (63). He used an argon laser to apply a grid pattern in the macula. Six weeks after laser, the directly treated drusen had disappeared.

THEORIES ON DRUSEN REDUCTION

AND CNV PREVENTION

Drusen disappearance after laser photocoagulation is clearly documented in the literature (64–76). However, the mechanism of drusen disappearance is not well understood. Several theories have been proposed: (i) phagocytosis of drusen; (ii) decreased deposits by removal of RPE; (iii) release of soluble mediators; (iv) thinning of Bruch’s membrane; and (v) mechanical alteration of the structure of Bruch’s membrane. It is clear from the above studies that laser induces an inflammatory response and the intensity

260 HSU AND HO

of the reaction depends on the laser settings as well as the laser subjects. The differences in these factors between various studies make interpretation difficult (25,54–57,59–62). Furthermore, in most of the clinical studies of laser to drusen, the calibrated intensity is minimal whitening. This is different from the above studies where stronger intensities were evaluated. However, despite these limitations, we can postulate that laser-induced phagocytosis of drusen occurs. Blood-borne inflammatory cells may ingest the drusen material. Studies certainly indicate their presence after laser. Duvall and Tso noted drusenoid material in cell processes after laser photocoagulation and attributed the origin of these cell processes to choroidal pericytes (61). Dysfunctional RPE, destroyed by laser, is replaced by proliferating RPE (57). The RPE has phagocytic ability and the proliferating RPE may be involved in drusen resorption (69). Also, the removal of dysfunctional RPE cells may halt further drusen development and allow removal of accumulated material. After laser-induced tissue damage, the RPE and other cells may produce soluble mediators. For instance, Glaser showed that RPE cells release an inhibitor of neovascularization (77). These soluble mediators may enhance the natural processes that result in spontaneous drusen resorption (23,70,78). They might also account for the observation that drusen distant from laser burns disappear after photocoagulation.

Bruch’s membrane in AMD eyes is diffusely thickened and hydrophobic. The structural effect on Bruch’s membrane by laser is variable. Thomas showed the integrity of Bruch’s membrane depended on the energy density of the laser (55). Photocoagulation may thin the abnormally thick Bruch’s membrane and, in theory, improve its hydraulic conductivity. The increased metabolic transport could improve drusen clearance and decrease drusen formation. The laser could also exert a mechanical effect on Bruch’s membrane, causing contraction of collagen and elastin (similar to laser trabeculoplasty) and improving egress of material through a more permeable Bruch’s membrane. Peyman showed that photocoagulation may improve perioxidase diffusion from the vitreous to the choroid (79).

However, it is important to note that drusen reduction seems to occur during the natural course of AMD. Soft drusen often progress to confluence, drusenoid PED, and fading which leads to RPE disturbances or atrophy in some cases. Over the course of five years, large drusen have been seen to disappear in 34% of eyes with very early changes consisting of one or a few large drusen (78). Also, among fellow eyes of patients enrolled in the Macular Photocoagulation Study with CNV in one eye, areas of large drusen disappeared from one or more areas and

new large drusen appeared in an additional 13% of eyes (80). Nevertheless, large spontaneous reductions of greater than 50% in drusen area are uncommon in patients with 10 or more large drusen (81).

Similar to drusen reduction, it is unclear how laser to drusen might prevent CNV. Some of the same theories on the mechanism of drusen reduction apply to CNV prevention. Improved transport of nutrients across Bruch’s membrane might reduce the metabolic strain on the RPE/outer retina and stop the production of angiogenic factors from the RPE. Indeed, laser might even induce the production of vasoinhibitory growth factors from the RPE. Gass postulated that laser “tacks” down the RPE to Bruch’s membrane, eliminating a potential cleavage plane for CNV (23). Proliferating RPE, induced by the laser, may envelop early CNV and prevent further growth.

UNCONTROLLED STUDIES AND CASE REPORTS

Since Gass first described the disappearance of drusen after laser photocoagulation, there have been a number of case reports and uncontrolled clinical studies that have examined the prophylactic treatment of drusen. Sigelman published a case report of a 58-year old woman with a disciform scar secondary to AMD in the right eye and confluent soft drusen in the left eye (82). The patient’s vision dropped to 20/40 with metamorphopsia in the left eye. There was no CNV but an increased density and size of foveal drusen. Using a wavelength of 576 nm (yellow), power of 180 mW, duration of 0.3 seconds, and spot size of 200 mm, he directly treated drusen and also applied a parafoveal grid for a total of 56 spots. Treated and untreated drusen disappeared and the vision returned to 20/20 one year after treatment.

Hyver reported laser photocoagulation in a patient with CNV in one eye and large, confluent soft drusen in the fellow eye (83). Using a wavelength of 630 nm, spot size of 200 mm, duration of 0.05 seconds and power of 200 mW, 24 burns were placed in the temporal macula with no direct drusen treatment. Burn intensity was calibrated to create barely visible whitening. Ten months after treatment the visual acuity had dropped from 20/25 to 20/60, which was felt to be due to the development of a granular subfoveal material. No CNV was noted on fluorescein angiography.

Cleasby treated 29 eyes in patients with “exudative senile maculopathy (ESM)” in the fellow eye (65). In addition, one eye of 25 patients with “nonexudative senile maculopathy (NSM)” in both eyes was treated. The criteria for NSM included the presence of drusen, retinal pigment epithelial atrophy and clumping and/or cholesterol deposits in the macula in individuals older than 50. Argon laser was used to

directly treat drusen “within a broad ring around the fovea.” The desired intensity was a minimally visible reaction in the retina. The laser parameters were a spot size of 50 to 100 mm, power between 100 and 150 mW and duration of 0.05 to 0.1 seconds. The number of applications was approximately 200 to 300 shots. In the group of 29 patients with ESM in one eye, three developed ESM in the treated eye over an average follow-up of 28.4 months. This represented a 4.4% yearly rate of ESM formation, which is less than the natural history of AMD. In the NSM group, neither the control eyes nor the treated eyes developed ESM over an average followup of 27.3 months. All 25 treated eyes and five control eyes showed a reduction in drusen. There were no reported complications from the laser. Despite a small number of patients, no control group for the ESM eyes and no randomization for NSM eyes, this study suggested prophylactic laser to drusen might be beneficial.

Wetzig treated 42 eyes of 27 patients with prophylactic laser in a retrospective, nonrandomized study (75). All patients had macular soft drusen and recent visual changes (visual loss or metamorphopsia). The vision ranged from 20/20 to 20/400. Only 25% of eyes had a best-corrected pre-laser visual acuity of 20/40 or better. The mean age at treatment was 69 years. Eyes with CNV or hemorrhagic exudative changes were excluded. Thirty-one eyes were treated with krypton red laser, one eye with a combination of xenon and krypton, eight eyes with argon laser and two eyes with a combination of argon and krypton laser. Both eyes were treated in some patients and several eyes were retreated. The desired intensity of the laser reaction was a faint, white gray spot. Approximately 50 to 75 spots were applied in a scatter pattern around the fovea. The vision improved, remained stable or worsened by only one line in 22 eyes (52%) over an average follow-up of 3.7 years. CNV developed in 12%. Drusen disappeared in these treated eyes, usually beginning at three months. Wetzig published a follow-up of these patients six years after the original publication (76). The average follow-up time was 120 months. Of the treated eyes, 33% remained stable or lost one line of visual acuity, 21% lost two to three lines and 46% lost three or more lines. CNV developed in 21% of treated eyes during the follow-up and several patients had progressive enlargement of the treatment scars. While no control group was designated, seven eyes with drusen had gone untreated. In this untreated group, three eyes retained 20/40 or better visual acuity, two eyes lost two or more lines and two eyes worsened to 20/400 or less. Overall, no clear beneficial effect of prophylactic laser was demonstrated. However, the retrospective, nonrandomized design with a small number of eyes

limits the conclusions that can be drawn from this study. Also, it included many patients with poor vision and selected patients with visual symptoms. These patients may have harbored subtle occult CNV.

Figueroa treated 20 patients with argon laser (66). Group 1 consisted of 14 patients with bilateral drusen with one eye randomly assigned to receive laser treatment. Group 2 consisted of six patients with CNV in one eye and drusen in the fellow eye. The patients ranged in age from 55 to 80 years and the average follow-up was 18 months. Drusen temporal to the fovea were directly treated with the argon green laser with a spot size of 100 mm, duration of 0.1 seconds, power of 100 mW, and mean number of laser spots of 30. The desired laser intensity was calibrated to achieve a light gray–white lesion. Treated drusen disappeared by approximately two months while surrounding, untreated drusen disappeared at a mean of 10 months. Visual acuity improved in 30% of eyes by one line or more. This was secondary to the resorption of untreated subfoveal drusen. The visual acuity remained unchanged in 65% of eyes and decreased in 5% (one eye). The one eye that worsened developed a choroidal neovascular membrane away from the laser scars. Figueroa updated these results and presented new data in a second publication with 30 patients in Group 1 and 16 patients in Group 2 (67). The laser settings were the same as described above. All treated drusen disappeared at an average of 3.5 months. In all but three patients, the untreated drusen resolved by an average of 8.6 months. The drusen disappearance progressed in a temporal to nasal direction. Superonasal drusen persisted for the longest amount of time. Two of the 30 control eyes in Group 1 (bilateral drusen) demonstrated spontaneous drusen resolution. After an average of three years, one control eye but no treated eyes developed CNV. Three fellow eyes (18%) in Group 2 developed CNV. In one eye, the CNV developed adjacent to the laser scars. Again, due to the small number of patients, interpretation of these results should be approached with caution.

Sarks treated 18 eyes of 16 patients with bilateral drusen and one eye of 10 patients with exudative changes in the other eye (74). Patients were 55 years or older and followed for a mean of 16.8 months. Inclusion criteria included visual acuity 20/40 or better and no evidence of atrophy or CNV. A ring of 40 to 50 non-confluent laser burns was applied approximately 1500 mm from the foveal center. Drusen were not directly targeted. Argon green laser was used with a spot size of 100 mm, duration of 0.05 to 0.1 seconds and power calibrated to produce “a barely discernable whitening of the RPE.” In 14 of the 16 patients with bilateral drusen, only one eye was treated. In these

262 HSU AND HO

treated eyes, the vision remained stable in 10 eyes and improved in four eyes. Vision decreased in four eyes and remained stable in 10 eyes in the untreated group. Overall, in the two treated groups, visual acuity improved in 12 eyes (40%), remained unchanged in 16 eyes (53%) and worsened in 2 eyes (7%). Visual improvement was related to foveal drusen resorption, which occurred in all treated eyes but none of the untreated eyes. Two treated patients developed CNV at seven and eight months post-treatment in retina adjacent to laser burns. Expansion of laser-induced atrophy was minimal in this study.

Guymer treated one eye of 12 patients at high risk for visual loss secondary to AMD (71). All 12 treated eyes demonstrated macular drusen and visual acuity of 20/40 or better. Ten patients had end-stage lesions in one eye and two patients had bilateral soft confluent drusen. Twelve laser spots were placed in a ring 750 to 1000 mm from the fovea. Argon green laser was used with a spot size of 200 mm, duration of 0.2 seconds and power calibrated to achieve faint blanching of the RPE (80–300 mW). The average follow-up was 16 months. Visual acuity remained the same or improved in 11 patients. Nine of the 11 patients had a reduction in drusen size, number and confluence. One patient lost four lines due to development of CNV that did not originate from a laser site. Two patients developed profound atrophy at the laser site and four others developed RPE pigmentary changes at the laser sites. This study showed that a small number of laser applications could promote drusen disappearance. It also showed no correlation between drusen resolution and improvement or deterioration of dark-adapted retinal thresholds.

Ruiz Moreno performed a prospective, nonrandomized clinical study of laser photocoagulation and looked at the development of macular atrophy in a consecutive series of patients with soft drusen who underwent argon green laser photocoagulation (84). Eyes had to have documented recent loss of visual acuity preceding treatment in order to be included in

Table 1 Controlled Bilateral Drusen Studies

the study. Fifty-two consecutive eyes of 52 patients received direct photocoagulation to drusen. Laser parameters included a spot size of 200 mm, duration of 0.2 seconds and power titrated to a light gray–white retinal reaction. Treatment was performed greater than 500 mm from the foveal center with a mean of 117 spots and was completed over two sessions. Macular atrophy occurred in nine eyes (17.7%) about 37.2 months after photocoagulation (range 7–75 months) and was associated with a significant decrease in visual acuity. There was no significant correlation between the areas of atrophy and the number of treatment spots (pZ0.97) or the intensity of treatment spots (pZ0.09). Due to the uncontrolled nature of this study, it is unclear if the macular atrophy is attributable to the laser treatment or related to the natural course of AMD.

CONTROLLED STUDIES

Information from the above studies confirmed that laser promoted drusen reduction. However, the visual benefit of this prophylactic laser was still unclear. These studies provided the impetus for more controlled studies and larger clinical trials. Tables 1 and 2 summarize the findings from these studies.

Frennesson conducted a randomized, prospective study of prophylactic laser treatment (68). One eye of 13 patients with bilateral soft drusen was treated. In a second group, the fellow eye of six patients with a disciform lesion in the other eye was treated. The control group consisted of 19 patients who had been randomized to observation. The groups were matched for age and visual acuity but there were more men in the treatment group. The visual acuity in all treated eyes was 20/25 or better. Patients with macular pigment clumping, atrophy, pigment epithelial detachments or exudative AMD were excluded. A horseshoe-shaped grid pattern with direct drusen treatment as well as scatter treatment was applied with argon green laser. Laser parameters included

a spot size of 200 mm, duration of 0.05 seconds and power of 100 to 200 mW. The number of laser spots varied from 51 to 154 spots with intensity calibrated to achieve a “grayish reaction.” Drusen area on color fundus photographs and fluorescein angiograms were calculated at baseline and follow-up for both groups. Follow-up results were published at 6 months, 12 months, 3 years, and 8 years (68–70,88). The mean drusen area significantly decreased in the treated eyes and significantly increased in the control eyes. Over three years, five eyes (33%) in the control group but none in the treatment group developed CNV. By eight years of follow-up, 29 patients including 13 treated and 16 controls remained in the study. Nine of the 16 controls (56%) developed CNV (five fellow eyes and four bilateral drusen eyes) and only 2 of 13 treated eyes (15%) developed CNV (two bilateral drusen eyes). While vision decreased significantly in both groups, the magnitude was greater in the control group with quadrupling of minimum angle of resolution in controls versus doubling of minimum angle of resolution in treated patients. This study demonstrated that laser treatment promotes drusen resorption, which had also been shown in the above studies. Importantly, it suggested that laser prophylaxis might prevent the exudative complications of AMD. However, as with the above studies, the sample size was small and the confidence interval large, making it difficult to draw valid conclusions.

Little randomized one eye of 27 patients with bilateral confluent soft drusen to prophylactic treatment (73). Mean age of patients was 69.7 years. The minimum visual acuity was 20/60 with a mean followup of 3.2 years. Foveal atrophy, pigment epithelial detachments and exudative changes were exclusionary criteria. Drusen were directly treated. Laser settings for the dye laser (577–620 nm) were a spot size of 100–200 mm, duration of 0.05 to 0.1 seconds and power of 100–200 mW calibrated to induce a slight lightening of the RPE/outer retina. No laser spots were applied within 300 mm of the foveal center and rarely within 500 mm. A total of 23 to 526 laser spots were applied, and 37% of eyes were treated over more than one session. Six treated eyes and no control eyes improved two or more lines, 16 treated and 17 control

eyes remained stable, and five treated and 10 control eyes lost two or more lines. Drusen resorption within 1500 mm of the fovea occurred more completely in the treated eyes than control eyes. In five eyes of both groups, there was equal drusen disappearance. Four control patients and two treated patients developed CNV. Laser scar enlargement occurred in three eyes. It is again difficult to draw conclusions from this study due to the small sample size but visual acuity and drusen resorption were significantly better in the treated eyes.

Olk studied the use of diode laser photocoagulation for 152 patients with macular drusen (bilateral drusen, 77 patients; fellow eye, 75 patients) (85). These investigators also compared the ability of subthreshold (invisible) diode laser photocoagulation with threshold (visible) laser photocoagulation to reduce the number of large drusen. Visual acuity was 20/63 or better at baseline. During the first 12 months of follow-up, threshold laser photocoagulation appeared to induce a more rapid disappearance of drusen compared with subthreshold laser. By 18 months, no difference was noted between the two groups. During the 24 months of follow-up, laser treated eyes had significant drusen reduction and improvement in visual acuity compared with observed eyes. CNV occurred at similar rates in both treated and observed eyes.

Scorolli compared using argon laser with subthreshold 810 nm diode-laser in 144 patients with bilateral macular drusen (78 eyes received argon, 66 eyes received diode laser) (86). One eye of each patient was treated with the second eye serving as control. During a mean of 18 months follow-up, best-corrected visual acuity was statistically significantly improved in both treatment groups compared with controls, with no significant difference between the argon and diode groups. Drusen reduction occurred in both treated groups as well compared with controls. CNV developed in three eyes receiving argon laser, one eye receiving diode laser, and four eyes in the untreated group. Visual field testing revealed minor but statistically significant reductions in the argon group but not in the diode group. A slight reduction in contrast sensitivity was also noted in the argon group but not seen in the diode group. However, it should be noted

264 HSU AND HO

that the treatment protocol in the argon group (green laser with power titrated to graying effect, 0.2 second duration, 100 mm spot size, and 200 spots placed 500 mm outside the foveal avascular zone) differed from the diode group (150 mW power, 0.1 second duration, 125 mm spot size, 48 spots placed 750 to 2250 mm outside the foveal avascular zone).

In 1994, the largest randomized pilot trial to date, the Choroidal Neovascularization Prevention Trial (CNVPT) was begun (64,72,81,89,90). A total of 312 eyes of 156 patients without exudative AMD and more than 10 large (more than 63 mm) drusen in each eye were enrolled in the Bilateral Drusen Study and 120 eyes of 120 patients with exudative AMD in one eye and more than 10 large drusen in the other eye were enrolled in the Fellow Eye Study. Study eyes were required to have visual acuity of 20/40 or better with no evidence of current or past CNV and progressive ocular disease. Fluorescein angiography was used to exclude CNV in the study eye at baseline. Patients in the bilateral drusen arm had one eye randomized to laser treatment with the second eye serving as the control. Patients in the fellow eye arm had the nonexudative AMD eye randomized to laser treatment or control. Laser parameters included a spot size of 100 mm, duration of 0.1 seconds, and power titrated to a light gray–white lesion. Figure 1 shows the treatment protocol for 85% of the patients, which consisted of 20 laser spots placed in three rows temporal to the fovea and greater than 750 mm from the center. Figure 2 shows the treatment protocol for the remaining patients, who received 24 laser spots of the same intensity placed in two rows temporal to the fovea and greater than 750 mm from the center.

The CNVPT protocol specified that eyes assigned to treatment be retreated at 6 months nasal to the fovea

100

300

750–1000 Microns

300 300

Fovea

Figure 1 Configuration of burns in the Laser 20 treatment protocol of the choroidal neovascularization prevention trial. Source: From Ref. 64.

Figure 2 Configuration of burns in the Laser 24 treatment protocol of the choroidal neovascularization prevention trial. Source: From Ref. 64.

in a mirror image of the first treatment if the area of drusen had not decreased by 50% from baseline. At six months, 28% of 78 eyes in the Bilateral Drusen Study and 41% of 37 eyes in the Fellow Eye Study had a 50% reduction in drusen and were exempt from retreatment. By 12 months, 54% of 35 eyes in the Bilateral Drusen Study and 27% of 11 eyes in the Fellow Eye Study had a 50% reduction. One eye in the observed group had a 50% reduction in drusen area. Less than 10% of treated eyes and more than 90% of observed eyes showed no reduction in the area of drusen at 12 months (64). Laser-treated eyes with a 50% or more reduction in drusen at this follow-up were more likely to have improved contrast sensitivity as well as oneand two-line increases in visual acuity compared with laser-treated eyes with less drusen reduction or observed eyes (pZ0.001) (72).

Enrollment was suspended early due to the apparent increase in CNV development within the first 12 months of follow-up in the Fellow Eye Study. CNV was seen in 10 of 59 treated eyes versus only 2 of 61 control eyes (pZ0.02). In the Bilateral Drusen Study, CNV developed in 4 of 156 treated eyes and 2 of 156 control eyes (pZ0.62). With additional follow-up, the significant increase in CNV incidence in treated fellow eyes compared with control eyes was maintained through 18 months but by 30 months the incidence of CNV was the same in both groups (91). Moreover, there were no statistically significant differences in these fellow eyes compared with controls in terms of change in visual acuity, contrast threshold, critical print size, or incidence of geographic atrophy.

Owens who reported the results of a randomized, controlled clinical trial, the Drusen Laser Study, saw similar findings (87,92,93). A total of 177 eyes of 177 patients with exudative AMD in one eye and drusen with or without pigment clumping were enrolled in the fellow eye group and 210 eyes of 105 patients with soft

drusen with or without pigment clumping were enrolled in the bilateral group. Baseline visual acuities of the study eyes were 20/40 or better, and fluorescein angiography was performed at baseline to exclude CNV. Argon green or yellow dye laser was used with a 200 mm spot size, 0.2 second duration, and 65 to 120 mW power. Twelve spots at a distance of 1000 mm from the center of the foveola were applied in a circular protocol pattern. Over three years of follow-up in the bilateral drusen group, CNV developed in 12 of 103 treated eyes (11.6%) and 7 of 103 observed eyes (6.8%, pZ0.23). Visual acuity decreased by 15 or more letters in 6 of 72 treated eyes (8.3%) and 10 of 72 observed eyes (13.9%, pZ0.39). During three years of follow-up in the fellow eye group, CNV developed in 27 of 91 treated eyes (29.7%) and only 15 of 85 observed eyes (17.7%, pZ0.06). Visual acuity decreased in 21 of 73 treated eyes (28.8%) and 13 of 66 observed eyes (19.7%, pZ0.21). Neither of these results was statistically significant, but the investigators felt compelled to halt recruitment into the trial due to concern for laser-induced CNV in the fellow eye group. In the final analysis, one of the most significant findings from the fellow eye group was that the incidence of CNV occurred six months earlier in the laser treated eyes compared with the no laser eyes (pZ0.05). This finding was maintained throughout the three years of follow-up.

The increased incidence of CNV in laser-treated fellow eyes has been somewhat unexpected. It is known that fellow eyes are at higher risk for CNV development than bilateral drusen eyes. One possibility is that some of the fellow eyes had undetected CNV at baseline that was stimulated by the laser treatment. These eyes may simply harbor more advanced AMD that is less amenable to prophylaxis. Differences in laser treatment strategy may also play a role. While some groups have specifically targeted macular drusen, the CNVPT and Drusen Laser Study Group followed a pattern that resulted in laser being directed either between or directly on drusen. The intensity of laser photocoagulation may also play a role. Using a computerized method of laser burn quantitation, Kaiser demonstrated that patients in the CNVPT who received more intense burns were more likely to have greater drusen resolution (90). However, a higher laser burn intensity seemed to correlate with increased risk of CNV development. Ultimately, one major challenge may be to deliver a sufficient amount of energy to promote a protective effect while limiting the risk of CNV stimulation.

FUTURE DIRECTIONS: MULTICENTERED

CLINICAL TRIALS

Based on the favorable data from Olk using diode laser, a larger multicenter, randomized, prospective

Figure 3 Artist’s illustration of 48 diode-laser lesions in a grid pattern of four concentric circles 750 to 2250 mm from the center of the foveal avascular zone. Source: From Ref. 86.

clinical trial known as the Prophylactic Treatment of AMD (PTAMD) Trial is currently in progress to compare subthreshold infrared (810 nm) diode laser photocoagulation with observation. Enrolled patients had visual acuities of 20/63 or better. Figure 3 demonstrates the laser protocol, which consisted of a grid of 48 sub-threshold 810-nm diode laser spots with a spot size of 125 mm applied in four concentric circles outside the foveal avascular zone. Only one laser treatment was applied throughout the study duration. Patients with at least five large drusen (more than 63 mm) within 2250 mm of the foveal avascular zone in both eyes were placed in the bilateral arm of the study with one eye being randomized to treatment and the other serving as control. Approximately 600 patients were enrolled into this arm by November 2001. A substudy of 100 eyes (50 patients) enrolled in this bilateral arm of the PTAMD revealed that the number of laser-induced lesions and the surface area of the laser-induced RPE changes on fluorescein angiography at three months post-treatment were strong predictors of major drusen reduction by 18 months post-treatment (94). This finding may explain the higher rate of drusen reduction in patients who were treated with threshold diode laser in the pilot study and echoes the findings of Kaiser from the CNVPT.

The PTAMD also enrolled patients with neovascular AMD in one eye and at least five large drusen in the fellow eye. These patients were placed in the unilateral arm with the eligible fellow eye being randomized to treatment or observation. Enrollment in the unilateral arm was suspended in April 2000 after 242 patients were enrolled due to an increased incidence of CNV and higher rates of worsening visual acuity in treated eyes (95). Follow-up of these patients is on-going.

Based on the findings of the CNVPT, the multi-center randomized clinical trial known as the Complications of Age-Related Macular Degeneration Prevention Trial (CAPT) was proposed and is being conducted with support from the National Eye Institute. The goal of CAPT is to determine whether prophylactic low-intensity laser treatment to the retina can prevent vision loss associated with the complications of advanced AMD (96,97). While fellow eyes in the CNVPT that were treated showed a higher rate of CNV development, in patients with bilateral drusen, the risk was found to be relatively low and similar between treated eyes and control eyes. As a result, only patients with bilateral high-risk drusen (10 or more drusen larger than 125 mm within 3000 mm of fovea) were incorporated into the CAPT design. Baseline visual acuity was 20/40 or better. The

laser treatment protocol also was modified based on the CNVPT findings. Given an apparent increased incidence of CNV in eyes that received more intense laser burns, the burn intensity was reduced to a barely visible lesion (90). The initial treatment consisted of 60 burns (100 mm spot size, 0.1 seconds duration) in a grid pattern within an annulus between 1500 and 2500 mm from the fovea. Retreatment could be performed at 12 months if 10 or more drusen 125 mm or larger remained within 1500 mm of the fovea. Figure 4 shows the follow-up treatment protocol, which consisted of 30 burns in the 1000 to 2000 mm annulus centered on the fovea with drusen being treated directly.

A total of 1052 patients were recruited by March 2001 with one eye being randomized to receive laser treatment and the other eye to observation. Patients

will be followed for a minimum of five years with the primary outcome measure being change in visual acuity. Secondary outcome measures include the incidence of advanced stage AMD changes (CNV, pigment epithelial detachments, geographic atrophy), contrast threshold, and critical print size. Sample fundus photographs at baseline and 12

months follow-up from two patients in the lasertreatment group are depicted in Figures 5 and 6. Drusen regression is demonstrated in Figure 5, and CNV development with subsequent disciform scarring is shown in Figure 6. While studies have demonstrated regression of drusen in laser-treated eyes, it is important to remember that drusen

268 HSU AND HO

regression can also occur spontaneously though typically at a lower rate compared to laser-treated eyes. Fundus photographs from one patient in the observation group demonstrating spontaneous drusen regression are shown in Figure 7.

Currently, many questions remain unanswered with regard to the use of laser photocoagulation in patients with high-risk drusen. Most of the studies reviewed support the fact that drusen number is reduced in patients who receive laser treatments. Furthermore, it seems that the greater the intensity of treatment, the faster the resolution. However, this increased intensity may also correlate with a higher risk of CNV (90). The clinical significance of drusen reduction is also unclear at this time. While several smaller studies have demonstrated a correlation between drusen reduction and improvement in visual acuity, this finding has yet to be confirmed by a large, randomized, controlled study.

Another important finding from the CNVPT, PTAMD, and Drusen Laser Study has been the increased risk of CNV in patients with neovascular AMD in one eye who underwent laser treatment in the fellow eye with high-risk drusen. Moreover, the Drusen Laser Study demonstrated a six-month earlier onset of CNV in the laser-treated eyes compared with controls. As a result, these fellow eyes should not be considered for laser prophylaxis using these protocols. At this point, clinical studies on laser prophylaxis seem best reserved for patients with bilateral high-risk drusen in the absence of neovascular complications.

Based on the multitude of laser treatment regimens, no single method has proven superior. Drusen reduction has been found with varying wavelengths, burn intensities, and treatment area of laser. Also in question is whether laser should be applied to drusen directly, indirectly, or both. As results become available from the CAPT and PTAMD trials as well as other ongoing studies, the effect of laser prophylaxis and drusen reduction on the natural course of AMD should become clarified.

SUMMARY POINTS

&A prophylactic treatment for AMD is highly desirable and would have a significant public health impact.

&Laser photocoagulation can induce drusen regression.

&The long-term effect of laser-induced drusen reduction on the natural history of AMD and visual function remains unclear.

&In patients with neovascular AMD in one eye, prophylactic laser appears to increase both the risk of CNV development as well as promote

earlier CNV development when performed in the fellow eye and should be avoided.

&Results from randomized clinical trials are necessary before laser prophylaxis for eyes with drusen should be recommended outside this context.

REFERENCES

1.Kini MM, Leibowitz HM, Colton T, et al. Prevalance of senile cataract, diabetic retinopathy, senile macular degeneration and open-angle glaucoma in the Framingham Eye Study. Am J Ophthalmol 1978; 85:28–34.

2.Leibowitz HM, Krueger DE, Maunder LR, et al. The Framingham Eye Study monograph. Surv Ophthalmol 1980; 24:335–610.

3.The Eye Diseases Prevalence Research Group. Causes and prevalence of visual impairment among adults in the United States. Arch Ophthalmol 2004; 122:477–85.

4.Sommer A, Tielsch JM, Katz J, et al. Racial differences in the cause-specific prevalence of blindness in East Baltimore. N Engl J Med 1991; 14:1412–7.

5.Klein R, Klein BE, Linton KLP. Prevalence of age-related maculopathy: the Beaver Dam Eye Study. Ophthalmology 1992; 99:933–45.

6.National Eye Institute. Age-related Macular Degeneration: Status of Research. Bethesda, MD: National Eye Institute/ National Institutes of Health/U.S. Department of Health and Human Services, 1997.

7.Ferris FL. Senile macular degeneration: review of epidemiologic features. Am J Ophthalmol 1983; 118:132–51.

8.Ferris FL, Fine SL, Hyman L. Age-related macular degeneration and blindness due to neovascular maculopathy. Arch Ophthalmol 1984; 102:1640–2.

9.Macular Photocoagulation Study Group. Argon laser photocoagulation for neovascular maculopathy after five years. Results from randomized clinical trials. Arch Ophthalmol 1991; 109:1109–14.

10.Macular Photocoagulation Study Group. Subfoveal neovascular lesions in age-related macular degeneration: guidelines for evaluation and treatment in the Macular Photocoagulation Study. Arch Ophthalmol 1991; 109: 1242–57.

11.Macular Photocoagulation Study Group. Laser photocoagulation for subfoveal neovascular lesions of age-related macular degeneration: updated findings from two clinical trials. Arch Ophthalmol 1993; 111:1200–9.

12.Macular Photocoagulation Study Group. Laser photocoagulation for juxtafoveal choroidal neovascularization: five year results from randomized clinical trials. Arch Ophthalmol 1994; 112:500–9.

13.Treatment of Age-related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: one-year results of 2 randomized clinical trials—TAP report 1. Arch Ophthalmol 1999; 117:1329–45.

14.Treatment of Age-related Macular Degeneration with Photodynamic Therapy (TAP) Study Group. Photodynamic therapy of subfoveal choroidal neovascularization in agerelated macular degeneration with verteporfin: two-year results of 2 randomized clinical trials—TAP report 2. Arch Ophthalmol 2001; 119:198–207.

15.Verteporfin in Photodynamic Therapy Study Group. Verteporfin therapy of subfoveal choroidal neovascularization in age-related macular degeneration: two-year results of a

randomized clinical trial including lesions with occult with no classic choroidal neovascularization—verteporfin in photodynamic therapy report 2. Am J Ophthalmol 2001; 131:541–60.

16.Margherio RR, Margherio AR, DeSantis ME. Laser treatments with verteporfin therapy and its potential impact on retinal practices. Retina 2000; 20:325–30.

17.Gragoudas ES, Adamis AP, Cunningham ET, et al. Pegaptanib for neovascular age-related macular degeneration. N Eng J Med 2004; 351:2805–16.

18.Spaide RF, Sorenson J, Maranan L. Photodynamic therapy with verteporfin combined with intravitreal injection of triamcinolone acetonide for choroidal neovascularization. Ophthalmology 2005; 112:301–4.

19.Lanchoney DM, Maguire MG, Fine SL. A model of the incidence and consequences of choroidal neovascularization secondary to age-related macular degeneration. Comparative effects of current treatment and potential prophylaxis on visual outcomes in high-risk patients. Arch Ophthalmol 1998; 116:1045–52.

20.Smiddy WE, Fine SL. Prognosis of patients with bilateral macular drusen. Ophthalmology 1984; 91:271–7.

21.Bressler SB, Maguire MG, Bressler NM, et al. The Macular Photocoagulation Study Group. Relationship of drusen and abnormalities of the retinal pigment epithelium to the prognosis of neovascular macular degeneration. Arch Ophthalmol 1990; 108:1442–7.

22.Klein R, Klein BEK, Jensen SC, et al. The five-year incidence and progression of age-related maculopathy. The Beaver Dam Eye Study. Ophthalmology 1997; 104:7–21.

23.Gass JD. Drusen and disciform macular detachment and degeneration. Arch Ophthalmol 1973; 90:206–17.

24.Bok D. Retinal photoceptor-pigment epithelium interactions. Invest Ophthalmol Vis Sci 1985; 26:1659–94.

25.Marshall J. Interactions between sensory cells, glial cells and the retinal pigment epithelium and their response to photocoagulation. Dev Ophthalmol 1981; 2:308–17.

26.Young R. The daily rhythm of shedding and degradation of rod and cone outer segment membranes in the chick retina. Invest Ophthalmol Vis Sci 1978; 17:105–16.

27.Loeffler KU, Lee WR. Terminology of sub-RPE deposits: do we all speak the same language? Br J Ophthalmol 1998; 82:1104–5.

28.Foulds W. The choroidal circulation and retinal metabolism: an overview. Eye 1990; 4:242–8.

29.Bird AC, Bressler NM, Bressler SB, et al. An international classification and grading system for age-related maculopathy and age-related macular degeneration. Surv Ophthalmol 1995; 39:367–74.

30.Bressler NM, Silva JC, Bressler SB, et al. Clinicopathologic correlation of drusen and retinal pigment epithelial abnormalities in age-related macular degeneration. Retina 1994; 14:130–42.

31.Green WR, Enger C. Age-related macular degeneration histopathologic studies: the 1992 Lorenz E. Zimmerman Lecture. Ophthalmology 1993; 100:1519–35.

32.Curcio C, Millican C. Basal linear deposit and large drusen are specific for early age-related maculopathy. Arch Ophthalmol 1999; 117:329–39.

33.Russell SR, Mullins RF, Schneider BL, et al. Location, substructure and composition of basal laminar drusen compared with drusen associated with aging and agerelated macular degeneration. Am J Ophthalmol 2000; 129:205–14.

34.Spraul C, Grossniklaus H. Characteristics of drusen and Bruch’s membrane in post-mortem eyes with age-related macular degeneration. Arch Ophthalmol 1997; 115:267–83.

35.van der Schaft TL, de Bruijn WC, Mooy CM, et al. Is basal laminar deposit unique for age-related macular degeneration. Arch Ophthalmol 1991; 109:420–5.

36.Farkas TG, Syvlester V, Archer D. The ultrastructure of drusen. Am J Ophthalmol 1971; 71:1196–205.

37.Farkas TG, Sylvester V, Archer D, et al. The histochemistry of drusen. Am J Ophthalmol 1971; 71:1206–15.

38.Mullins R, Johnson L, Anderson D. Characterization of drusen-associated glycoconjugates. Ophthalmology 1997; 104:288–94.

39.Sarks J, Sarks S, Killingsworth M. Evolution of soft drusen in age-related macular degeneration. Eye 1994; 8:269–83.

40.Ishibashi T, Sorgente N, Patterson R, et al. Pathogenesis of drusen in the primate. Invest Ophthalmol Vis Sci 1986; 27:184–93.

41.Ishibashi T, Patterson R, Ohnishi Y, et al. Formation of drusen in the human eye. Am J Ophthalmol 1986; 101:342–53.

42.Hogan MJ. Role of the retinal pigment epithelium in macular disease. Trans Am Acad Ophthalmol Otolaryngol 1972; 76:64–80.

43.Feeney-Burns L, Ellersieck M. Age-related changes in the ultrastructure of Bruch’s membrane. Am J Ophthalmol 1985; 100:686–97.

44.Pauleikhoff D, Zuels S, Sheraidah GS, et al. Correlation between biochemical composition and fluorescein binding of deposits in Bruch’s membrane. Ophthalmology 1992; 99:1548–53.

45.Moore D, Hussain A, Marshall J. Age-related variation in the hydraulic conductivity of Bruch’s membrane. Invest Ophthalmol Vis Sci 1995; 36:1290–7.

46.Pauleikhoff D, Harper CA, Marshall J, et al. Aging changes in Bruch’s membrane. A histochemical and morphologic study. Ophthalmology 1990; 97:171–8.

47.Zarbin M. Age-related macular degeneration: review of pathogenesis. Eur J Ophthalmol 1998; 8:199–206.

48.Holz FG, Wolfensberger TJ, Piguet B, et al. Bilateral macular drusen in age-related macular degeneration: prognosis and risk factors. Ophthalmology 1994; 101:1522–8.

49.Macular Photocoagulation Study Group. Risk factors for choroidal neovascularization in the second eye of patients with juxtafoveal or subfoveal choroidal neovascularization secondary to age-related macular degeneration. Arch Ophthalmol 1997; 115:741–7.

50.Sandberg MA, Weiner A, Miller S, et al. High-risk characteristics of fellow eye of patients with unilateral neovascular age-related macular degeneration. Ophthalmology 1998; 105:441–7.

51.Sarraf D, Gin T, Yu F, et al. Long-term drusen study. Retina 1999; 19:513–9.

52.Mainster M. Wavelength selection in macular photocoagulation. Ophthalmology 1986; 93:952–8.

53.Peyman G, Raichand M, Zeimer R. Ocular effects of various laser wavelenghts. Surv Ophthalmol 1984; 28:391–404.

54.Smiddy WE, Fine SL, Green WR, et al. Clinicopathologic correlation of krypton red, argon blue-green, and argon green laser photocoagulation in the human fundus. Retina 1984; 4:15–21.

55.Thomas EL, Apple DJ, Swartz M, Kavka-Van Norman D. Histopathology and ultrastructure of krypton and argon laser lesions in a human retina-choroid. Retina 1984; 4:22–39.

56.Smiddy WE, Fine SL, Quigley HA, et al. Comparison of krypton and argon laser photocoagulation. Arch Ophthalmol 1984; 102:1086–92.

57.Smiddy WE, Fine SL, Quigley HA, et al. Cell proliferation after laser photocoagulation in primate retina. Arch Ophthalmol 1986; 104:1065–9.

270 HSU AND HO

58.Peyman GA, Li M, Yoneya S, et al. Fundus photocoagulation with the argon and krypton lasers: a comparative study. Ophthalmic Surg 1981; 12:481–90.

59.Coscas G, Soubrane G. The effects of red krypton and green argon laser on the foveal region. Ophthalmology 1983; 90:1013–22.

60.Swartz M. Histology of macular photocoagulation. Ophthalmology 1986; 93:959–63.

61.Duvall J, Tso MO. Cellular mechanisms of resolution of drusen after laser coagulation. An experimental study. Arch Ophthalmol 1985; 103:694–703.

62.Perry D, Reddick R, Risco J. Choroidal microvascular repair after argon laser photocoagulation. Invest Ophthalmol Vis Sci 1984; 25:1019–26.

63.Della NG, Wilson DJ, Klein ML. Clinical and pathologic effects of grid macular laser in aged primate eyes containing drusen. Invest Ophthalmol Vis Sci 1997; 38(Suppl.):S18.

64.Choroidal Neovascularization Prevention Trial Research Group. Laser treatment in eyes with large drusen. Shortterm effects seen in a pilot randomized clinical trial. Ophthalmology 1998; 105:11–23.

65.Cleasby G, Nakanishi A, Norris J. Prophylactic photocoagulation of the fellow eye in exudative senile maculopathy. Mod Probl Ophthalmol 1979; 20:141–7.

66.Figueroa MS, Regueras A, Bertrand J. Laser photocoagulation to treat macular soft drusen in age-related macular degeneration. Retina 1994; 14:391–6.

67.Figueroa MS, Regueras A, Bertrand J. Laser photocoagulation for macular soft drusen. Updated results. Retina 1997; 17:378–84.

68.Frennesson CI, Nilsson SE. Effects of argon (green) laser treatment of soft drusen in early age-related maculopathy: a 6 month prospective study. Br J Ophthalmol 1995; 79:905–9.

69.Frennesson CI, Nilsson SE. Laser photocoagulation of soft drusen in early age-related maculopathy (ARM). The one-

year results of a prospective, randomised trial. Eur J Ophthalmol 1996; 6:307–14.

70.Frennesson CI, Nilsson SE. Prophylactic laser treatment in early age related maculopathy reduced the incidence of exudative complications. Br J Ophthalmol 1998; 82:1169–74.

71.Guymer RH, Gross-Jendroska M, Owens SL, et al. Laser treatment in subjects with high-risk clinical features of agerelated macular degeneration. Posterior pole appearance and retinal function. Arch Ophthalmol 1997; 115:595–603.

72.Ho AC, Maguire MG, Yoken J, et al. Laser-induced drusen reduction improves visual function at 1 year. Choroidal Neovascularization Prevention Trial Research Group. Ophthalmology 1999; 106:1367–74.

73.Little HL, Showman JM, Brown BW. A pilot randomized controlled study on the effect of laser photocoagulation of confluent soft macular drusen. Ophthalmology 1997; 104:623–31.

74.Sarks SH, Arnold JJ, Sarks JP, Gilles MC, Walter CJ. Prophylactic perifoveal laser treatment of soft drusen. Aust N Z J Ophthalmol 1996; 24:15–26.

75.Wetzig PC. Treatment of drusen-related aging macular degeneration by photocoagulation. Trans Am Ophthalmol Soc 1988; 86:276–90.

76.Wetzig PC. Photocoagulation of drusen-related macular degeneration: a long-term outcome. Trans Am Ophthalmol Soc 1994; 92:299–303.

77.Glaser BM, Campochiaro PA, Davis JL, Jr., et al. Retinal pigment epithelial cells release an inhibitor of neovascularization. Arch Ophthalmol 1985; 103:1870–5.

78.Bressler N, Munoz B, Maguire MG. Five-year incidence and disappearance of drusen and retinal pigment epithelial abnormalities: Waterman study. Arch Ophthalmol 1995; 113:301–8.

79.Peyman G, Spitznas M, Straatsma B. Chorioretinal diffusion of perioxidase before and after photocogulation. Invest Ophthalmol Vis Sci 1971; 10:489.

80.Javornik N, Hiner CJ, Marsh MJ, et al. Changes in drusen and RPE abnormalities in age-related macular degeneration. Invest Ophthalmol Vis Sci 1992; 33(Suppl.):1230.

81.Ho AC, Javornik N, Maguire MG, et al. The Choroidal Neovascularization Prevention Trial (CNVPT): assessment of macular risk factors for CNVat 1 year. Invest Ophthalmol Vis Sci 1997; 38(Suppl.):517.

82.Sigelman J. Foveal drusen resorption one year after perifoveal laser photocoagulation. Ophthalmology 1991; 98: 1379–83.

83.Hyver SW, Schatz H, McDonald HR, et al. A case of visual acuity loss following laser photocoagulation for macular drusen. Arch Ophthalmol 1997; 115:554–5.

84.Ruiz-Moreno JM, De La Vega C, Zarbin MA. Macular atrophy after photocoagulation of soft drusen. Retina 2003; 23:315–21.

85.Olk JR, Friberg TR, Stickney KL, et al. Therapeutic benefits of infrared (810-nm) diode laser macular grid photocoagulation in prophylactic treatment of nonexudative agerelated macular degeneration. Ophthalmology 1999; 106:2082–90.

86.Scorolli L, Corazza D, Morara M, et al. Argon laser vs. subthreshold infrared (810-nm) diode laser macular grid photocoagulation in nonexudative age-related macular degeneration. Can J Ophthalmol 2003; 38:489–95.

87.Owens SL, Guymer RH, Gross-Jendroska M, et al. Fluorescein angiographic abnormalities after prophylactic macular photocoagulation for high-risk age-related maculopathy. Am J Ophthalmol 1999; 127:681–7.

88.Frennesson CI. Prophylactic laser treatment in early agerelated maculopathy: an 8-year follow-up in a randomized pilot study shows a reduced incidence of exudative complications. Acta Ophthalmol Scand 2003; 81:449–54.

89.The Choroidal Neovascularization Prevention Trial Research Group. Choroidal neovascularization in the choroidal neovascularization prevention trial. Ophthalmology 1998; 105:1364–72.

90.Kaiser RS, Berger JW, Maguire MG, et al. Laser burn intensity and the risk for choroidal neovascularization in the CNVPT Fellow Eye Study. Arch Ophthalmol 2001; 119:826–32.

91.The Choroidal Neovascularization Prevention Trial Research Group. Laser treatment in fellow eyes with large drusen: updated findings from a pilot randomized clinical trial. Ophthalmology 2003; 110:971–8.

92.Owens SL, Bunce C, Brannon AJ, et al. Prophylactic laser treatment appears to promote choroidal neovascularization in high-risk ARM: results of an interim analysis. Eye 2003; 17:623–7.

93.Owens SL, Bunce C, Brannon AJ, et al. Prophylactic laser treatment hastens choroidal neovascularization in unilateral age-related maculopathy: final results of the Drusen Laser Study. Am J Ophthalmol 2006; 141:276–81.

94.Rodanant N, Friberg TR, Cheng L, et al. Predictors of drusen reduction after subthreshold infrared (810 nm) diode laser macular grid photocoagulation for nonexudative age-related macular degeneration. Am J Ophthalmol 2002; 134:577–85.

95.Friberg TR, Musch D. Prophylactic treatment of age-related macular degeneration (PTAMD): update on the clinical trial. Invest Ophthalmol Vis Sci 2002; 43 (E-abstract 2904).

96.Complications of Age-Related Macular Degeneration Prevention Trial study group. The complications of age-related macular degeneration prevention trial (CAPT): rationale, design and methodology. Clin Trials 2004; 1:91–107.

97.Complications of Age-Related Macular Degeneration Prevention Trial study group. Baseline characteristics, the 25-item National Eye Institute visual functioning questionnaire, and their associations in the Complications of AgeRelated Macular Degeneration Prevention Trial (CAPT). Ophthalmology 2004; 111:1307–16.