Selective occlusion of subfoveal choroidal neovascularization in pathologic myopia using a new...

Selective Occlusion of Subfoveal ChoroidalNeovascularization in Pathologic Myopia

Using a New Technique of IngrowthSite Treatment

ROGÉRIO A. COSTA, MD, DANIELA CALUCCI, COMT, LUIZ F. TEIXEIRA, MD,JOSE A. CARDILLO, MD, AND PEDRO P. BONOMO, MD

● PURPOSE: To evaluate the visual and angiographiceffects as well as optical coherence tomography findingsafter a new technique of ingrowth site treatment ofsubfoveal choroidal neovascularization (CNV) second-ary to pathologic myopia with the use of indocyaninegreen (ICG)-mediated photothrombosis.● DESIGN: Interventional, noncomparative case series.● METHODS: In the setting of a tertiary referral center,patients with pathologic myopia in whom fluorescein andconventional ICG angiography demonstrated distinctCNV vessels supplying the subfoveal neovascular com-plex were submitted to focal ingrowth site treatmentusing a new therapeutic modality termed ICG-mediatedphotothrombosis. Prospective evaluation including visualacuity assessment, fluorescein and ICG angiography, andoptical coherence tomography (OCT) evaluation wasperformed at weeks 1, 12, 24, and 48 after treatment.● RESULTS: Six consecutive patients (six eyes) had treat-ment using a single session of ICG-mediated photothrom-bosis at the CNV ingrowth site. Obliteration of the entireneovascular complex was achieved immediately aftertreatment in all patients. At last follow-up, visual acuityimprovement of 1 or more Early Treatment DiabeticRetinopathy Study lines was observed in five of sixpatients, and fluorescein angiography showed an absence(four eyes) or minimal leakage (two eyes) from CNV.

Indocyanine green angiography demonstrated selectiveobliteration of the neovascular complex. Accordingly,reduction of retinal edema was observed in the OCTevaluation in all patients. There was no significantcomplication related to the procedure.● CONCLUSIONS: The use of lower irradiances of 810-nm continuous light application and intravenous ICGinfusion for neovascular ingrowth site photothrombosisinduced selective CNV hypoperfusion, as demonstratedby fluorescein and ICG angiography in patients withpathologic myopia. These findings were either consistentwith the visual acuity improvement observed in five of sixpatients or with the partial restoration of the retinalarchitecture seen in OCT evaluation 12 months aftertreatment. (Am J Ophthalmol 2003;135:857–866.© 2003 by Elsevier Inc. All rights reserved.)

T HE VERTEPORFIN IN PHOTODYNAMIC THERAPY (VIP)

clinical trial is evaluating the use of the photosen-sitive dye verteporfin (Visudyne; Novartis Ophthal-

mics, Duluth, Georgia, USA) in photodynamic therapy forsubfoveal neovascular complications caused by diseasesother than age-related macular degeneration (AMD).Based on the results observed 1 year after treatment inpatients with choroidal neovascularization (CNV) due topathologic myopia, verteporfin therapy should be consid-ered to increase the chance of preservation or partialrestoration of the visual acuity1; however, improvement of3 or more lines of visual acuity was exceptional amongtreated patients. As myopia is the leading cause of CNV inpeople of working age2 and some studies confirmed theself-limited course of neovascular membranes in suchdisease,3,4 investigators have been seeking new therapeuticmodalities that can be more effectively used for treatingCNV in pathologic myopia patients and that are able toimprove the visual outcome of this disease.5–9

Accepted for publication Dec 11, 2002.InternetAdvance publication at ajo.com Dec 12, 2002.From the Instituto da Visao–IPEPO (R.A.C., D.C., L.F.T., J.A.C.,

P.P.B.), Department of Ophthalmology, Federal University of Sao Paulo,and Consultores de Retina e Associados (R.A.C., J.A.C.), Araraquara,Sao Paulo, Brazil.

Doctors Costa and Cardillo are consultants to Akorn, Inc., a UnitedStates manufacturer of indocyanine green. Any financial support washerein provided by the company as the consulting agreement has justbeen settled.

Inquiries to Rogerio Alves Costa, MD, Rua Italia 1905, Apto 74,Araraquara–SP, 14801-350 Brazil; fax: (�55) 16-3331-2197; e-mail:[email protected]

© 2003 BY ELSEVIER INC. ALL RIGHTS RESERVED.0002-9394/03/$30.00 857doi:10.1016/S0002-9394(02)02257-2

Among alternative treatments under evaluation forCNV treatment, the photocoagulation for CNV feedervessels theoretically has been expected to be an idealmethod. The feasibility of this method for subfovealneovascular complications was already demonstrated byShiraga and coworkers10 and by Staurenghi and cowork-ers,11 where 43% and 62% of the patients, respectively,had a final visual acuity of 20/100 or better despite theexpected poor visual outcome of such lesions in AMD.The assumptions of the use of dynamic indocyanine green(ICG) angiography and the poor results obtained in CNVwith “umbrella-like” feeder vessels, however, appear to bethe major limitations of this therapeutic modality.

In this article, we report on our initial experience ofICG-mediated photothrombosis, a new therapeutic modal-ity recently proposed by Costa and coworkers,12,13 toocclude the feeder vessels at the CNV ingrowth site inpatients with subfoveal CNV resulting from pathologicmyopia. Unlike previous reports, dynamic ICG angiogra-phy was not required and a single spot of continuouslow-intensity 810-nm light was applied directly to theCNV ingrowth site after intravenous ICG infusion.

DESIGN

A PROSPECTIVE, INTERVENTIONAL CASE SERIES WAS PER-

formed to evaluate the visual and angiographic effects aswell as optical coherence tomography (OCT) findings inpatients with subfoveal CNV caused by pathologic myopiato undergo neovascular ingrowth site photothrombosis.

METHODS

THE CLINICAL STUDY PROTOCOL WAS APPROVED BY THE

local Institutional Review Board before patient enrollmentat the retina and vitreous section of the OphthalmologyDepartment of the Federal University of Sao Paulo began,and all participants gave written informed consent beforeparticipating in the study. All fluorescein and ICG angiog-raphy studies and OCT evaluations were performed by anexperienced certified ophthalmic technician (D.C.). AllCNV ingrowth sites identification and treatments wereperformed by a single retinal specialist (R.A.C.).

The patient enrollment period was from April throughNovember 2001. Pathologic myopia was defined as an eyerequiring a distance correction of at least �6.0 diopters(spherical equivalent). An eye that had a spherical equiv-alent that was less myopic than �6.0 diopters was eligibleif there were retinal abnormalities consistent with patho-logic myopia (such as lacquer cracks) and if the axiallength of the eye was at least 26.5 mm.1 Candidates forinclusion were those with exudative and hemorrhagicmanifestations of subfoveal CNV diagnosed by fluoresceinangiography less than 72 hours before enrollment in the

study was offered and in whom a CNV ingrowth site wasidentifiable.

Patients were not offered enrollment in the study if theyhad drusen, traumatic choroidal rupture, peripapillarychanges with atrophic or pigmented “punched out” chori-oretinal lesions, uveitis, or any other ophthalmic disorderthat might affect visual function. Patients were also ex-cluded if they were unable to cooperate for the laserprocedure; were unable to comprehend the nature of theprocedure or the study to give an informed consent; had anallergy to fluorescein, ICG, or iodine; had previouslyundergone laser photocoagulation; or had opacity thatobstructed laser access to the macular area.

Initial evaluations (used as baselines for patients en-rolled in the study) were obtained within 72 hours ofenrollment in the study and included a medical history,vision testing, color photography, fluorescein angiography,conventional ICG angiography, and OCT.

Best-corrected visual acuity was measured using a ret-roilluminated Lighthouse for the Blind distance visualacuity test chart (using modified Early Treatment DiabeticRetinopathy Study [ETDRS] charts 1, 2, and R). Fundusphotography, fluorescein angiography, and ICG angiogra-phy were performed using a fundus camera (TRC-50IA/IMAGEnet; Topcon, Tokyo, Japan). For inclusion in thisstudy, it was mandatory that CNV caused by pathologicmyopia extend under the geometric center of the fovealavascular zone. The lesion could include occult and classicCNV or other features that could obscure the identifica-tion of CNV on fluorescein angiography, including blood,fluorescence not from blood, or serous detachment of theretinal pigment epithelium. The neovascular complex didnot need to be well-defined, subretinal fibrous tissue couldbe present, and the lesion could be of any size.

Conventional digital ICG angiography should demon-strate distinct neovascular vessels in early frames, enablingprecise identification of the CNV ingrowth site. Briefly,the procedure used for ingrowth site identification was asfollows: 1 ml of an aqueous solution containing 25 mg ofICG (IC-Green, Akorn, Buffalo Grove, Illinois, USA) wasadministered as a single intravenous (IV) bolus into acubital vein, followed by injection of a 10-ml saline flush.After 5 minutes of angiographic imaging, we performed asecond injection of 1 ml of the ICG-solution (25 mg/ml)followed by a 10-ml saline flush and continued angio-graphic imaging for 10 more minutes; as the result, at leastfour early frames were available at the end of the exami-nation, increasing the chance of detection of the neovas-cular nets. The neovascular vessels and CNV ingrowth sitewere identified by comparing early- and late-phase ICGangiographic digital images with each other and withfluorescein angiograms, looking for angiographic findingssimilar to those observed by Melberg and coworkers14 aswell as Shiraga and coworkers.15

Optical coherence tomography (OCT 2000; HumphreyInstruments, San Leandro, California, USA) evaluation

AMERICAN JOURNAL OF OPHTHALMOLOGY858 JUNE 2003

consisted of two linear scans 5 mm in length centered onthe foveal region at 0 degrees (horizontal scan) and 270degrees (vertical scan).

After evaluations, six patients met eligibility criteria forthe study. They were informed verbally and in writing ofthe potential benefits and risks of the procedure, and allsigned a written form stating that they understood andconsented to ICG-mediated photothrombosis of the neo-vascular ingrowth site.

Indocyanine green-mediated photothrombosis is a newmodality of laser-dye treatment that uses large-spot, low-intensity 810 nm light to apply laser energy continuouslyto an area where ICG has been highly concentrated.12,13

The laser settings used in this series were modified based onthe results of a preliminary study using the same techniqueof CNV ingrowth site treatment in patients with exudativeAMD (Costa and coworkers, Invest Ophthalmol Vis Sci2002;43(4 Suppl):2502) and angioid streaks.16

The laser delivery system used for ICG-mediated pho-tothrombosis consists of a modified infrared diode laser(Trimode-L; OPTO, Sao Carlos, Brazil) coupled to aslit-lamp biomicroscope (SL-3E; Topcon, Tokyo, Japan).The diode laser was tuned to 810 nm, and the beam width(spot size) was adjusted to 0.8 mm (800 �m), to safelycover the entire CNV ingrowth site.

For the ICG-mediated photothrombosis procedure, aloading dose of 2.0 ml of a highly concentrated solution ofICG (25 mg/mL, approximately 1 mg/kg body weight) wasadministered as an IV bolus, followed immediately by a5.0-ml saline flush. Twenty minutes after the loading dose,with the laser spot positioned in the center of the CNVingrowth site as seen through a fundus contact lens of1.02� magnification (OGFA-2; Ocular Instruments, Bel-levue, Washington, USA), a second IV injection of 50 mgof ICG in 2 ml of solution was administered, followed by a5.0-ml saline flush. Ten seconds after the second flushinfusion, 30 W/cm2 laser energy was applied to the CNVingrowth site for 80 continuous seconds. Ten minutes afterthe conclusion of this first application, 30 W/cm2 laserenergy was applied a second time for 80 seconds.

To evaluate CNV obliteration, fluorescein angiographywas performed 30 minutes after the second (and last)application of laser energy.

Patients were scheduled for follow-up examinations atweeks 1, 12, 24 and 48 after treatment. At these visits,patients’ best-corrected visual acuity was determined andthey underwent complete ophthalmic examinations, colorfundus photography, fluorescein and ICG angiography, andOCT using the same procedures as at baseline. Additionalexaminations were performed if an adverse event wassuspected.

Follow-up best-corrected visual acuity determinationswere used to evaluate the treatment response as well as anindicator of the safety of the procedure. A change in visualacuity of 1 line was considered relevant when rating the

treatment response, and a change of 3 or more lines wasconsidered clinically relevant.

Fluorescein angiography was used for CNV evaluationafter treatment. The evaluator, who was aware of thepatient’s visual acuity at baseline and follow-up, used thefollowing semiquantitative grading system1 to compare theextent (area) of leakage of fluorescein from CNV atfollow-up to the extent (area) of leakage from CNV atbaseline: progression (leakage due to CNV beyond the areanoted at baseline [grade 1]); moderate leakage (area ofleakage occupying 50% or more of the area of CNV notedat baseline and no progression [grade 2]); minimal leakage(area of leakage occupying less than 50% of the area ofCNV noted at baseline and no progression [grade 3]); andabsence of leakage (no leakage in the area affected by theoriginal CNV lesion [grade 4]). These gradings were basedonly on lesion area, not on other fluorescein features suchas the amount of fluorescence or the area of leakageextending beyond classic CNV or a fibrovascular pigmentepithelial detachment.

Optical coherence tomography was used to grade thetreatment response based on changes in the degree ofretinal edema from baseline. Based on these changes, themacular architecture was classified as having undergonepartial recovery (absence of or decrease in retinal edema),no recovery (unchanged retinal edema), or deterioration(increased retinal edema).

Treatment response was rated 48 weeks after ICG-mediated photothrombosis as remarkable (fluorescein leak-age grade 3 or 4, no increase in retinal edema, and any lossin visual acuity limited to less than 1 line); acceptable(fluorescein leakage grade 2, no increase in retinal edema,and any loss in visual acuity limited to less than 1 line); orunsatisfactory (fluorescein leakage grade 1 or any gradefluorescein leakage with increase in retinal edema or loss of1 line or more in visual acuity).

Data on adverse ocular events were obtained from caserecords forms and inspection of fluorescein and ICGangiography and OCT documentation.

RESULTS

ALL SIX PATIENTS WHO WERE ENROLLED COMPLETED THE

48-week study. The two women and four men ranged inage from 33 to 51 years (mean, 40.5 years). Fluorescein andICG angiography showed that at baseline, all study eyeshad active CNV extending under the geometric center ofthe foveal avascular zone, with OCT demonstrating thatall eyes had increased retinal thickness or retinal elevationdue to fluid accumulation. Best-corrected visual acuity atbaseline ranged from 20/125�1 to 20/400 (mean 20/160,logMAR equivalent 0.91 � 0.197). Baseline characteris-tics of the patients and CNV in the six study eyes aresummarized in Table 1.

NEW TECHNIQUE FOR OCCLUDING SUBFOVEAL CNVVOL. 135, NO. 6 859

The ingrowth site was identified in all six patients onbaseline angiograms. Early ICG angiography showed neo-vascular vessels as distinctly visible as the retinal vessels inthese patients. The filling of the neovascular vesselsappeared just before the filling of the retinal vessels andbranched out toward the surrounding neovascular lesion.In four patients the ingrowth site was located within thejuxtafoveal area; in the remaining two, the CNV ingrowthsite was extrafoveal. In four of six patients, early-phasefluorescein angiograms clearly demonstrated neovascularvessels or focal hyperfluorescent areas from which theneovascular lesion appeared to arise (Figure 1).

One interesting finding occurred during ICG-mediatedphotothrombosis of the CNV ingrowth site. Althoughlaser energy was applied only to the CNV ingrowth site,

the entire neovascular complex turned gray during thelaser applications. This effect was transient and resolvedspontaneously a few seconds after the end of each lasersession.

Indocyanine green-mediated photothrombosis of theCNV ingrowth site obliterated the entire neovascular netin all eyes, which appears as a lack of fluorescein CNVfilling in fluorescein angiography obtained 30 minutes aftertreatment. Minimal fluorescein staining of the retinal vesselswall or retinal capillary nonperfusion over the point of laserapplication was eventually observed (Figure 2).

At 1 week after treatment, best-corrected visual acuityhad improved from a mean of 20/160 at baseline to a meanof 20/125-1 (Table 2), representing a mean change invisual acuity from baseline of 0.9 � 0.6 lines. At the final

TABLE 1. Baseline Characteristics

Patient No., Age,

Sex, Study Eye

Refractive Error

(Diopters)

Fellow Eye

Status

Ingrowth Site

Location CNV Lesion Type

Area of Lesion

(MPS Disk Areas)

Neovascular

Lesion Fibrosis

1, 35, F, RE �14.00 Active CNV Juxtafoveal Purely classic �1 �25%

2, 33, M, LE �11.00 Fibrotic scar Juxtafoveal Purely occult �1–�2 25%–50%

3, 51, M, LE �08.25 Fibrotic scar Juxtafoveal Occult/classic �1–�2 �25%

4, 42, M, LE �09.50 Unremarkable Extrafoveal Purely classic �1–�2 25%–50%

5, 42, F, RE �09.25 Fibrotic scar Extrafoveal Occult/classic �3 25%–50%

6, 40, M, LE �08.00 Unremarkable Juxtafoveal Occult/classic �1–�2 25%–50%

CNV � choroidal neovascularization; F � female; LE � left eye; M � male; MPS � Macular Photocoagulation Study; RE � right eye.

FIGURE 1. Neovascular ingrowth site identification of Patient 1. (Top left). Before treatment, early-frame fluorescein angiographydemonstrated a distinct small vessel appearing from choroidal vessels (open arrow) and spreading out to surrounding neovascularnets (arrowheads). (Top right). Rapid filling of the entire neovascular net was observed in the next 1.8 second. (Bottom left). Adeep choroidal hyperfluorescence corresponding to the area where neovascular vessels spread out (open arrow) was seen inearly-phase indocyanine green angiography. (Bottom right). Note neovascular nets, probably efferent ones (arrowheads), in theindocyanine green angiogram taken 1.7 seconds later, just above of those previously identified (white lines).


visit 48 weeks after ICG-mediated photothrombosis of theCNV ingrowth site, mean visual acuity had improved by2.7 � 2.1 lines over baseline. All but one patient had atleast 2 lines of improvement in visual acuity.

Fluorescein angiography indicated that all patients hada marked decrease in the extent of CNV leakage 1 weekafter treatment, with all patients having a complete ab-sence of leakage at this follow-up examination. By 24weeks, two patients had minimal leakage within the area ofthe CNV identified at baseline and four had no leakage(Figure 3), with no further changes observed in lesionactivity up to week 48. Indocyanine green angiographyalso demonstrated that a complete CNV regression wasobtained in four eyes. In the remaining two eyes somehyperfluorescence within the CNV area could be evi-denced, indicating that partial reperfusion of the neovas-cular lesion occurred. Choroidal occlusion after ICG-mediated photothrombosis of the CNV ingrowth site

appeared to be restricted to the area of laser application,with minimal detectable angiographic alterations in thenormal choriocapillaris and choroidal vessels in the vicin-ity of the CNV ingrowth site. Normal choroidal vesselswithin the CNV area, which were obscured by neovascularlesion hyperfluorescence before treatment, were not af-fected by laser application. A hypofluorescent irregulararea slightly larger than the area in which laser was appliedcould be identified in two patients and may indicate someocclusion of choroidal vessels (Figure 4); conversely, theseirregular areas observed after the procedure might besecondary to blockage from ICG-fluorescence from thechoroidal layers caused by the presence of a nonperfusedfibrotic neovascular component (Figure 5).

Optical coherence tomography findings agreed withthose of fluorescein and ICG angiography in showing anabsence of optically clear spaces and thus an absence offluid in the retina due to a reduction in or absence of CNV

FIGURE 2. Fluorescein angiography (FA) of Patient 1. (Top). Baseline FA performed 48 hours before treatment demonstratedsignificant fluorescein leakage from a subfoveal classic choroidal neovascularization (CNV). (Bottom). Thirty minutes afterindocyanine green-mediated photothrombosis of the juxtafoveal CNV ingrowth site: the hypofluorescent area corresponds to thelocation of the former CNV. (Mobile black spot that appears inferiorly on angiograms was an artifact caused by the high refractiveerror of the patient [14.00 DE]).


leakage in all eyes at the final 48-week visit. On follow-upOCT of all eyes, the reflective band corresponding to theretinal pigment epithelium and choriocapillaris was moreeasily identified than at baseline, and a well-defined, highlyhyperreflective region of fusiform thickening consistentwith a quiescent neovascular complex or preexisting sub-retinal fibrosis could also be seen (Figure 5).

Using the criteria described for this study, a “remark-able” treatment response (minimal or absent leakage offluorescein dye in the baseline CNV area, no deteriorationof macular architecture and no loss of visual acuity) wasachieved in all patients by 48 weeks after ICG-mediatedphotothrombosis of the CNV ingrowth site. In addition, arelevant clinical improvement (�3 lines) in visual acuitywas observed in three treated eyes. The treatment did notcause any systemic photosensitivity complications, skinphotosensitivity reactions, or major ocular adverse eventssuch as nonperfusion in retinal vessels.

DISCUSSION

PATHOLOGIC MYOPIA HAS BEEN REPORTED TO BE A MAJOR

cause of blindness in the United States.17 The maincomplication leading to decreased central vision in theseeyes is CNV, which affects as many as 11% of patients withhigh myopia.18–20 The decision regarding treatment ofneovascular complications in such disease remains prob-lematic, because previous studies describing the naturalprogression of myopic CNV are somewhat conflict-ing3,4,21,22; for example, Yoshida and coworkers4 recentlyreported on a series of 73 eyes of 63 patients with untreatedCNV complicating pathologic myopia that vision re-mained stable or increased in 81% of eyes after a meanfollow-up of 82 months if patients were 40 years old or

younger at the time of CNV onset. Despite all informationthat disagreed regarding the natural progression of CNV inpathologic myopia, a common finding in these studies wasthat spontaneous visual acuity improvement of 3 or morelines from baseline examinations was infrequent during thefollow-up period. Furthermore, according to the first VIPreport, even among treated patients only 2% and 6%presented such an extent of visual acuity improvement 3months and 12 months, respectively, after PDT withverteporfin.1

Of the various treatments under investigation to managesubfoveal neovascular lesions, photocoagulation of CNVfeeder vessels would seem to be ideal because of its highrate of success in improving visual acuity, its minimaldamage induced by the treatment, and its lack of interfer-ence with other therapeutic procedures. There is noinformation in the available literature about the feasibilityof this method for CNV in pathologic myopia, however,probably because the results of previous studies indicatedthat such therapy would be less successful in lesions with“umbrella-like” feeder vessels11 in addition to requiring asophisticated high-speed ICG angiography for feeder vesselidentification.10,11

We believed we could overcome these limitations tolaser photocoagulation of CNV feeder vessels just discussedand probably extend its application to CNV complicatingpathologic myopia. First, we designed an initial study toassess the feasibility of ingrowth site identification withoutthe use of dynamic ICG angiography in patients withsubfoveal lesions due to pathologic myopia. We found thatthe results of fluorescein and ICG angiography are stronglycorrelated, indicating that the neovascular vessels detectedin early-phase angiograms arise from a single area at thedeeper choroidal layers. We call this region the CNVingrowth site. We found that this area was never larger

TABLE 2. Best-Corrected Visual Acuity (Snellen/LogMAR)

Patient No. Baseline 1 Week 12 Weeks 24 Weeks 48 Weeks

1 20/160 � 2 20/100 20/60 20/60 � 1 20/60

0.86 0.70 0.50 0.52 0.50

2 20/400 20/250 � 1 20/100 � 2 20/100 20/100 � 2

1.30 1.12 0.74 0.70 0.66

3 20/125 20/100 � 1 20/80 � 1 20/60 � 2 20/60 � 2

0.80 0.72 0.58 0.54 0.54

4 20/125 � 2 20/100 20/100 � 2 20/100 � 2 20/80

0.78 0.70 0.66 0.66 0.60

5 20/125 20/125 � 2 20/100 � 1 20/100 20/80 � 1

0.80 0.76 0.68 0.70 0.62

6 20/160 � 1 20/160 20/160 20/160 � 1 20/160

0.92 0.90 0.90 0.88 0.90

Mean � 1 standard

deviation

20/160 20/125 � 1 20/100 � 1 20/100 � 2 20/80 � 2

0.910 � 0.197 0.816 � 0.166 0.676 � 0.137 0.666 � 0.130 0.636 � 0.141

LogMAR � logarithm of the minimal angle of resolution.


than 750 �m in the eyes studied, regardless of the size ofthe neovascular lesion and that it was juxtafoveal orextrafoveal in all the patients we studied.

Second, we proposed a new approach to feeder vesselocclusion, to address the problem that almost every patienthas one or more of the factors, such as multiple or largeCNV feeder vessels or CNV with an umbrella-like pattern,which have been reported to have a negative impact onthe success of FV occlusion.10,11 We had already demon-strated that effective and relatively selective occlusion ofthe choriocapillaris12 and neovascular lesions complicatingAMD13 could be achieved by continuous application of810-nm laser to photoactivate areas/lesions with a highconcentration of ICG. The laser irradiance used (30W/cm2) in this series was chosen based on the results ofour previous studies using ICG-mediated photothrombosisfor ingrowth site treatment in exudative AMD16 and morerecently in angioid streaks (Costa and coworkers, Invest

Ophthalmol Vis Sci 2002;43(Suppl 4):2502) and morerecently in angioid streaks.16 This laser irradiance level ishigher than the level of approximately 1.8 to 5.0 W/cm2

used clinically for regular ICG-mediated photothrombosis,but it was considered necessary to achieve definitive andlocalized vascular occlusion of vessels in which blood flowvelocity can be as high as 2 mm/s.23 The level we selectedcould have caused side effects such as irreversible damageto the inner retinal layers of the fundus in the photoco-agulated area,24 but we believed that the risk of this effectwas lower with our technique because the intensity we usedwas at least 50 times lower than that proposed by Staurenghiand coworkers11 and Flower25 for feeder vessel treatment.Experimental observations and theoretical models have sug-gested that intensity (measured in W/cm2) is the main factorcontributing to temperature changes in laser-treated tissues,and this is borne out by the relative effects of various lasersused clinically. For example, a faster and more intense

FIGURE 3. Comparison of pretreatment, and 1- and 12-week posttreatment fluorescein angiography from Patient 2. (Top).Baseline fluorescein angiography demonstrated an active subfoveal choroidal neovascularization (CNV) surrounded by ahypofluorescent halo corresponding to subretinal hemorrhage that extended up to the temporal inferior retinal vessels. (Middle).One week after indocyanine green (ICG)-mediated photothrombosis of the CNV ingrowth site hypoperfusion of the neovascularlesion was observed. No damage to the retinal vessels could be identified. We believe that some increase in subretinal hemorrhageoccurred acutely because during laser application at the CNV ingrowth site the efferent feeder vessels occlusion might havehappened before the occlusion of the afferent ones. (Bottom). Three months after ingrowth site ICG-mediated photothrombosis,some staining but no leakage within the original CNV area was observed. Subretinal hemorrhage was completely resolved.


increase in tissue temperature occurs when conventionalphotocoagulation is used than when much higher energy(measured in J/cm2) is used for thermotherapy of AMD.26

Although this new technique may work in minimizing theadverse effect to surrounding tissues, further comparisons withsham treatment without ICG dye are still needed; addition-ally, even if the development of atrophic creep (enlargementof the laser scars) was not observed in our patients 12 monthsafter treatment, such possibility should be considered withlonger follow-up.

Definitive characterization for feeder vessels has not yetbeen established, and as proposed by Staurenghi and

Flower,27 a feeder vessel is a vessel that, when successfullytargeted, produces a favorable result, namely, reduction ofthe associated CNV blood flow. In our series, fluoresceinnonperfusion of the entire neovascular complex was con-firmed by angiography 30 minutes after the end of treat-ment in all patients. Proper identification of the CNVingrowth site and treatment led to obliteration of neovas-cular vessels, even occult nets that were poorly defined onbaseline fluorescein angiograms. Although minimal retinaldamage such as capillary non-perfusion was eventuallydetected in fluorescein angiography immediately aftertreatment over the site of laser application, no evidence of

FIGURE 4. Indocyanine green (ICG) angiography of Patient 2. (Top left). Before treatment, a subfoveal neovascular lesion withjuxtafoveal ingrowth site was observed. (Top right). Midphase angiogram demonstrated a well-defined hyperfluorescent areasurrounded by an area of blockage caused by blood. (Bottom left). Three months after ingrowth site ICG-mediated photothrombosis,complete regression of the neovascular lesion was observed. Choroidal occlusion was restricted to the area of laser application(dotted line), with normal choroidal vessels in the vicinity of the choroidal neovascularization ingrowth site, which were obscuredby the neovascular lesion before treatment, perfused and with no angiographic signs of damage. (Bottom right). Midphase angiogramrevealed an irregular well-defined hypofluorescent area probably corresponding to a nonperfused fibrotic neovascular lesioncomponent.


FIGURE 5. Optical coherence tomography (OCT) findings after neovascular ingrowth site photothrombosis. (Top). Patient 2(vertical scans): at baseline retinal elevation and subfoveal displacement of fluid was observed; best-corrected visual acuity (BCVA)was 20/400. One week after treatment OCT disclosed partial recovery of the macular architecture and diminished retinal edema(BCVA 20/250�1). Twelve weeks after ingrowth site photothrombosis, BCVA had improved further to 20/100�2. Retinalthickness and foveal height appeared practically normal, and there was minimal thickening of the reflective band corresponding tothe choriocapillaris and retinal pigment epithelium in the subfoveal region. The quiescent neovascular lesion was barely visible 24weeks after treatment, and the BCVA was 20/100. No significant changes were observed 48 weeks after the procedure, and BCVAat last follow-up was 20/100�2. (Bottom). Patient 1 (horizontal scans): at baseline, retinal elevation and increased retinal thicknesscaused by a subfoveal choroidal neovascularization (CNV) was seen (BCVA 20/160�2). One week after CNV ingrowth siteindocyanine green-mediated photothrombosis, OCT disclosed diminished retinal edema (BCVA 20/100). Twelve weeks after theprocedure BCVA had further improved to 20/60, caused by partial restoration of the macular architecture despite minimal leakageobserved on fluorescein angiography from partially perfused CNV. After 24 weeks the quiescent neovascular lesion was seen as aregion of fusiform hyperreflective region beneath the fovea, and vision remained at 20/60�1. No further significant changes wereobserved at 48 weeks.


damage to the nerve fiber layer was noted. Apparently, theretinal damage was closely related to the amount ofpigmentation in the patient’s fundus and the amount ofsubretinal fluid present at the site of laser application.

Optical coherence tomography was essential in evaluat-ing the response to treatment. With the advent of moderntherapeutic modalities to modulate CNV activity (feedervessel treatment, photodynamic therapy, transpupillarythermotherapy, anti–vascular endothelial growth factor[VEGF] aptamer therapy), fluorescein and ICG angiogra-phy probably will not be sufficient for accurate treatmentevaluation. In our study, even if there was some CNVhyperfluorescence at the last follow-up visit in two pa-tients, suggesting that some portion of the neovascularcomplex was reperfused, OCT evaluation in these patientsshowed a significant reduction in the amount of intrareti-nal and subretinal fluid compared with baseline, suggestingCNV staining rather than fluorescein leakage.

In the case series we report, the novel treatment we callingrowth site ICG-mediated photothrombosis led to con-siderable, rapid improvement in visual function, as docu-mented by change in best-corrected visual acuity, with nocomplications, in five of six patients with subfoveal CNVdue to pathologic myopia who underwent the procedure.Change in visual acuity was strongly related to selectiveCNV hypoperfusion, as documented by fluorescein andICG angiography leading to partial restoration of themacular architecture as demonstrated by OCT evaluations.Our study has several limitations, including the absence of acontrol group, the small number of patients, and the limitedfollow-up period. Nevertheless, the results presented in thisreport suggest that further studies are warranted to assess theclinical value of this new modality of treatment for themanagement of CNV in pathologic myopia.

REFERENCES

1. VIP Study Group. Photodynamic therapy of subfoveal cho-roidal neovascularization in pathologic myopia with verte-porfin. 1-year results of a randomized clinical trial—VIPreport no. 1. Ophthalmology 2001;108:841–852.

2. Cohen SY, Laroche A, Leguen Y, Soubrane G, Coscas GJ.Etiology of choroidal neovascularization in young patients.Ophthalmology 1996;103:1241–1244.

3. Avila MP, Weiter JJ, Jalkh AE, et al. Natural history ofchoroidal neovascularization in degenerative myopia. Oph-thalmology 1984;91:1573–1581.

4. Yoshida T, Ohno-Matsui K, Ohtake Y, et al. Long-termvisual prognosis of choroidal neovascularization in highmyopia. Ophthalmology 2002;109:712–719.

5. Thomas MA, Dickinson JD, Melberg NS, et al. Visual resultsafter surgical removal of subfoveal choroidal neovascularmembranes. Ophthalmology 1994;101:1384–1396.

6. Bottoni F, Airaghi P, Perego E, et al. Surgical removal ofidiopathic, myopic, and age-related subfoveal neovasculariza-tion. Graefes Arch Clin Exp Ophthalmol 1996;234(Suppl1):S42–S50.

7. Fujikado T, Ohji M, Hayashi A, Kusaka S, Tano Y. Ana-

tomic and functional recovery of the fovea after fovealtranslocation surgery without large retinotomy and simulta-neous excision of a neovascular membrane. Am J Ophthal-mol 1998;126:839–842.

8. Fujii GY, Humayun MS, Pieramici DJ, et al. Initial experi-ence of inferior limited macular translocation for subfovealchoroidal neovascularization resulting from causes other thanage-related macular degeneration. Am J Ophthalmol 2001;131:90–100.

9. Hamelin N, Glacet-Bernard A, Brindeau C, et al. Surgicaltreatment of subfoveal neovascularization in myopia: macu-lar translocation versus surgical removal. Am J Ophthalmol2002;133:530–536.

10. Shiraga F, Ojima Y, Matsuo T, et al. Feeder vessel photoco-agulation of subfoveal choroidal neovascularization second-ary to age-related macular degeneration. Ophthalmology1998;105:662–669.

11. Staurenghi G, Orzalesi N, La Capria A, Aschero M. Lasertreatment of feeder vessels in subfoveal choroidal neovascu-lar membranes. Ophthalmology 1998;105:2297–2305.

12. Costa RA, Farah ME, Freymuller E, et al. Choriocapillarisphotodynamic therapy using indocyanine green. Am J Oph-thalmol 2001;132:557–565.

13. Costa RA, Farah ME, Cardillo JA, Belfort R Jr. Photody-namic therapy with indocyanine green for occult subfovealchoroidal neovascularization caused by age related maculardegeneration. Curr Eye Res 2001;23:271–275.

14. Melberg NS, Thomas MA, Burgess DB. The surgical removalof subfoveal choroidal neovascularization. Ingrowth site as apredictor of visual outcome. Retina 1996;16:190–195.

15. Shiraga F, Shiragama C, Matsuo T, et al. Identification ofingrowth site of idiopathic subfoveal choroidal neovascular-ization by indocyanine green angiography. Ophthalmology2000;107:600–607.

16. Costa RA, Calucci D, Cardillo JA, Farah ME. Selectiveocclusion of subfoveal choroidal neovascularization in an-gioid streaks using a new technique of ingrowth site treat-ment. Ophthalmology 2003; forthcoming.

17. Curtin BJ. The myopias: basic science and clinical manage-ment. Philadelphia: Harper & Row, 1985:7–10..

18. Curtin BJ, Karlin DB. Axial length measurements and funduschanges of the myopic eye. Am J Ophthalmol 1971;71:42–50.

19. Grossniklaus HE, Green WR. Pathologic findings in patho-logic myopia. Retina 1992;12:127–133.

20. Levy JH, Pollock HM, Curtin BJ. The Fuchs’ spot: anophthalmoscopic and fluorescein angiographic study. AnnOphthalmol 1977;9:1433–1443.

21. Hotchkiss ML, Fine SL. Pathologic myopia and choroidalneovascularization. Am J Ophthalmol 1981;91:177–183.

22. Hampton GR, Kohen D, Bird AC. Visual prognosis ofdisciform degeneration in myopia. Ophthalmology 1983;90:923–926.

23. Flower RW. Variability in choriocapillaris blood flow distri-bution. Invest Ophthalmol Vis Sci 1995;36:1247–1258.

24. Flower RW. Experimental studies of indocyanine greendye-enhanced photocoagulation of choroidal neovasculariza-tion feeder vessels. Am J Ophthalmol 2000;129:501–512.

25. Flower RW. Optimizing AMD-related CNV feeder vesseltreatment. Am J Ophthalmol 2002;134:228–239.

26. Mainster MA, Reichel E. Transpupillary thermotherapy forage-related macular degeneration: long-pulse photocoagula-tion, apoptosis, and heat shock proteins. Ophthalmic SurgLasers 2000;31:359–373.

27. Staurenghi G, Flower R. Clinical observations supporting atheoretical model of the role of choriocapillaris blood flow intreatment of AMD-associated CNV. Am J Ophthalmol2002;133:801–808.


Selective occlusion of subfoveal choroidal neovascularization in pathologic myopia using a new...

Documents

Transcript of Selective occlusion of subfoveal choroidal neovascularization in pathologic myopia using a new...