 |
|
Incident Choroidal Neovascularization in Fellow Eyes of Patients With Unilateral Subfoveal Choroidal Neovascularization Secondary to Age-Related Macular DegenerationSST Report No. 20 From the Submacular Surgery Trials Research Group
Submacular Surgery Trials Research Group*
Arch Ophthalmol. 2007;125(10):1323-1330.
ABSTRACT
| |
Objective To describe incident choroidal neovascular lesions in fellow eyes of participants in the Submacular Surgery Trials who had age-related macular degeneration (AMD).
Methods Review of baseline fluorescein angiograms confirmed the absence of neovascular AMD in fellow eyes of 364 participants at risk. Subjects were eligible for a minimum of 2 years of follow-up with angiograms of eyes at risk reevaluated to estimate incidence rates of choroidal neovascularization (CNV) and to characterize these lesions.
Main Outcome Measures Incidence of CNV during follow-up, characteristics of the incident lesion (composition, size, and location), and visual acuity at the time of incidence.
Results Incident lesions were confirmed in 98 fellow eyes of participants, yielding 2- and 4-year cumulative incidence rates of 22% and 37%. Incident lesions were predominantly CNV in 87 fellow eyes (90%), extrafoveal in 29 fellow eyes (30%), and juxtafoveal in 9 fellow eyes (9%). Occult without classic CNV lesions were found in 64 eyes (67%), minimally classic CNV and predominantly classic CNV lesions in 12 eyes (13%) each, and predominantly blood lesions in 4 eyes (4%). Nearly two-thirds of all incident lesions were 3 disc areas or smaller in size. Median visual acuity decreased from 20/25 at baseline to 20/250 at the 4-year follow-up in fellow eyes with incident CNV.
Conclusions and Application to Clinical Practice Frequent angiographic follow-up of fellow eyes at risk for CNV may lead to earlier detection and treatment of neovascular AMD and better visual acuity outcomes.
Trial Registration clinicaltrials.gov Identifier: NCT00000150
INTRODUCTION
The development of choroidal neovascularization (CNV) in eyes with age-related macular degeneration (AMD) typically leads to compromise of central vision. It has been estimated that approximately 8 million individuals aged 55 years or older in the United States alone are at risk of progressing to the advanced forms of AMD, notably CNV or foveal geographic atrophy (GA).1 The incidence of neovascular AMD vs foveal GA is approximately 2:1 in these high-risk eyes that manifest intermediate AMD or that are the fellow eyes of individuals with unilateral advanced AMD. Early detection of CNV likely will result in a better level of visual acuity at time of diagnosis of neovascular AMD and less vision impairment at the conclusion of treatment.2 In this era of rapid evolution of pharmacologic therapies that have the potential to stop progression of CNV secondary to AMD,2-3 intervention at the earliest sign of incident CNV may improve vision in some and abolish further moderate or severe vision loss in nearly all individuals.
Early detection of CNV in the community may be facilitated by awareness of characteristics of early neovascular lesions in eyes at risk. Eighteen eyes of 18 individuals (7%) among 276 participants monitored prospectively in the Choroidal Neovascularization Prevention Trial (CNVPT) developed neovascular AMD between 2 and 21 months after enrollment.4 Although relatively few events occurred during a limited period of observation (up to 24 months in fewer than half of all participants), the incident CNV lesions consisted primarily of occult without classic CNV, the median lesion size was 2 to 3.5 disc areas (DAs) as defined in the Macular Photocoagulation Study (MPS), and the majority of lesions (61%) were subfoveal.4 At the time of the development of incident CNV, mean and median level of visual acuity was 20/32 with 9 of 18 eyes manifesting at least a 2-line loss of visual acuity from baseline.
In a case series by Chang et al,5 713 consecutive AMD patients with newly diagnosed unilateral occult CNV or bilateral acute involvement with occult CNV but a history of sequential development were evaluated to describe CNV lesion characteristics in the fellow eyes that manifested CNV subsequently. Choroidal neovascularization developed in the second eyes of 115 patients (11 had bilateral involvement at presentation). The lesion composition of the first affected eye appeared highly predictive of the lesion composition in the second eye.5 Median visual acuity in the second eyes newly affected by CNV was 20/50.
The MPS Group observed rates of incident CNV in the fellow eyes of AMD participants at risk of 12% and 19% at 2 years and 22% and 36% at 4 years of follow-up among participants in the extrafoveal MPS trial and the juxtafoveal or subfoveal MPS trials, respectively.6-7 Nearly one-third of the incident lesions in the extrafoveal trial were not subfoveal when first identified. Symmetry between study eye and incident fellow eye lesion composition was evaluated among the subfoveal MPS trial participants, but no relationship was found.7
The Age-Related Eye Disease Study (AREDS) demonstrated that micronutrient supplements containing vitamin C, vitamin E, betacarotene, and zinc significantly decreased the risk of progression to advanced AMD in individuals at risk.8 While it is assumed that implementation of the AREDS recommendations has reduced the incidence of neovascular AMD in eyes at risk of progression, neovascular maculopathy still develops in many individuals.
Therapies to manage CNV continue to evolve and currently include thermal laser photocoagulation, photodynamic therapy with or without triamcinolone, and intravitreal anti-VEGF (vascular endothelial growth factor) agents. Information describing the common presentations of incident CNV, with regard to lesion location, composition, and size, is useful to guide development and evaluation of appropriate therapies and for monitoring second eyes at risk to provide treatment when it is likely to be most effective. The purpose of this article is to report incidence rates of CNV lesions in second eyes at risk, to describe incident neovascular lesions, and to report the effect of incident CNV on visual acuity during follow-up of the Submacular Surgery Trials (SST) participants who had only nonneovascular AMD in their fellow (nonstudy) eye at time of enrollment in the SST.
METHODS
The design and methods of the SST have been described in detail in earlier articles.9-10 Methods pertinent to this analysis are summarized here.
PATIENT SELECTION AND FOLLOW-UP
Participants included in this investigation were enrolled in the SST randomized trials of surgical removal vs observation of subfoveal CNV lesions secondary to AMD. Eligibility criteria for the 454 participants in the SST trial for new (previously untreated) subfoveal CNV secondary to AMD, some component of which was classic CNV, included lesions of the size 3.5 to 9 DAs and best-corrected visual acuity (BCVA) of 20/100 to 20/800 in the study eye.9 Eligibility criteria for the 336 participants in the SST trial for subfoveal hemorrhagic lesions ( 50% of lesion area occupied by blood) included lesion size larger than 3.5 DAs with or without visible CNV and BCVA of 20/100 to light perception in the study eye.10 The only criterion for the nonstudy fellow eye of SST participants was visual acuity of at least light perception.
At baseline, BCVA of each eye was measured by an SST-certified vision examiner using Early Treatment Diabetic Retinopathy Study (ETDRS) charts. Stereoscopic film-based color photographs of the disc and macula of each eye, a stereoscopic film-based fluorescein angiogram with images of the study eye during the dye transit phase, and mid- and late-phase images of the fellow eye were taken by SST-certified photographers who followed a standard protocol. These images were reviewed independently by personnel at the SST Reading Center (Wilmer Reading Center, Baltimore, Maryland) to document adherence to eligibility criteria and to characterize baseline fundus characteristics of study and nonstudy eyes.
Best-corrected visual acuity measurements, funduscopic examinations, stereoscopic color photography, and fluorescein angiography were repeated at 3, 6, 12, and 24 months after SST enrollment. Annual follow-up examinations continued for 4 years after enrollment or until September 30, 2003, at which time all patients were eligible for at least 2 years of follow-up.
BASELINE ASSESSMENT OF FELLOW EYES
Personnel at the reading center were masked to all demographic and clinical information relating to SST participants. They initially recorded features characterizing the fellow eyes of participants with AMD as they enrolled as no AMD, questionable AMD, nonneovascular AMD, neovascular AMD, or indeterminate. The 399 fellow eyes initially assessed as having no AMD, questionable AMD, or nonneovascular AMD (<63 µm drusen, 63 µm drusen, or geographic atrophy) were reassessed to confirm absence of CNV at baseline. Fellow eyes in the indeterminate category were excluded because they were known to have poor-quality photographs and angiograms such that a determination of the presence or absence of CNV at baseline could not be made with confidence.
QUALITY ASSURANCE MEASURES
Two 20% samples of SST fellow eyes were randomly selected for independent re-evaluation by 2 graders. One sample was from all fellow eyes originally graded as free of CNV at baseline and throughout follow-up and another from all fellow eyes originally graded as having neovascular lesions at baseline. Within each sample, the first 10 cases, all cases that generated questions by the graders, and cases in which the original classification of CNV present or absent was reversed also were reviewed by the reading center ophthalmologists before a final judgment was rendered. The data from these samples were compared with the original gradings. Because the comparison yielded only a small number of discrepancies (<10% in the baseline no CNV group and <5% in the baseline CNV group), the original classification of fellow eyes as having neovascular or non neovascular lesions present at baseline or during follow-up was judged to be acceptable for selection of the fellow eyes believed to be at risk of incident CNV (399 eyes) and the fellow eyes that had incident CNV during follow-up (118 eyes).
During re-review of fellow eye photographs, 13 of 399 fellow eyes were reclassified as having neovascular AMD at baseline and 4 additional eyes were removed from the cohort due to the presence of pattern dystrophy, which interfered with assessment of either neovascular or nonneovascular AMD characteristics. Another eye originally in the no AMD category also was removed because of a large congenital posterior pole lesion that greatly diminished the possibility of neovascular AMD developing; 1 eye was reclassified to indeterminate AMD status because of poor-quality images. Therefore, 380 fellow eyes were confirmed to be free of CNV at baseline and thus at risk of development of CNV.
INCIDENT CASES OF CNV
Reading center personnel, who remained masked to all demographic and clinical information during follow-up, characterized the AMD features of the fellow eyes at all follow-up examinations using the same categories as at baseline. As noted earlier, these original gradings indicated that 118 fellow eyes had incident CNV during follow-up in the SST.
Two reading center graders independently reassessed the fundus photographs and fluorescein angiograms taken of the fellow eyes beginning with the date of incident CNV originally reported to confirm the presence of CNV in the fellow eye at that time. To ensure that this date corresponded to the earliest presentation of CNV in the fellow eye, the prior set of fundus photographs and fluorescein angiograms was reviewed to confirm the absence of CNV at that examination. Whenever CNV was identified at the examination immediately preceding the one originally recorded, the photographs taken at earlier examinations were reviewed sequentially until the date of earliest observation of CNV was identified. Whenever CNV was not confirmed to be present in the fundus photographs and fluorescein angiograms taken at the earliest incidence date originally recorded, the graders reviewed photographs taken at succeeding examinations from that date forward until the presence of CNV was confirmed angiographically or until they had evaluated the last set of available photographs for the participant. All fellow eyes reclassified by the graders as no CNV ever, CNV at baseline, or with a change in the incident CNV date were re-reviewed with the reading center ophthalmologists before a final judgment was made. Once the earliest set of follow-up photographs in which CNV was observed had been identified and confirmed, characteristics of the incident CNV, including CNV location, composition, and size, were graded independently by 2 readers and adjudicated, as necessary, with the supervising ophthalmologists (S.D.S. and S.B.B.).
During reassessment of images for this analysis, 9 eyes were removed from the incident CNV group: 8 eyes (8 of 13 eyes mentioned earlier) that were noted to have CNV present since enrollment and 1 eye with a large congenital lesion that may have been responsible for CNV development. Eleven more eyes could not be confirmed to have had evidence of CNV at any follow-up visit; therefore, 98 fellow eyes were confirmed to have incident CNV.
DATA ANALYSIS AND STATISTICAL METHODS
Percentage agreement and the  statistic11 (weighted if more than 2 categories) were calculated for a sample of 25 participants who had repeat masked gradings of the incident CNV lesion characteristics for quality-assurance purposes. Percentage agreement (range, 58%-100%) and estimated statistics (range, 0.41-1.00) indicated moderate to substantial reproducibility of the CNV lesion assessments. Thus, the initial characterization of CNV lesions prepared for this report was used in all analyses. Kaplan-Meier product-limit estimates12 of the proportion of fellow eyes in which CNV developed during follow-up were used to estimate cumulative incidence rates. Eyes were censored following the examination at which the last available photographs were taken. The Pearson  2 test13 was used to assess potential associations between categorical variables. The 2 test for trend in proportions among quantitatively ordered categories14 was used to compare length of follow-up of fellow eyes at risk between the 2 SST clinical trials. The McNemar test15 was used to explore whether the lesions in the fellow eye and the paired study eye were similar with respect to composition or size. For all tests, a difference was considered statistically significant when the P value was less than .05; P values were not adjusted for multiple comparisons. All analyses were performed using SAS version 8.2 software (SAS, Inc, Cary, North Carolina).
RESULTS
Nearly one-half of the SST AMD participants were confirmed at baseline to have fellow eyes at risk for incident CNV: 224 of 454 participants whose study (first) eyes had predominantly CNV subfoveal lesions and 156 of 336 participants whose first eyes had predominantly hemorrhagic subfoveal lesions. Of the 380 patients who had a fellow eye at risk of CNV, 16 did not have any gradable photographs available after baseline. Fundus photographs and fluorescein angiograms taken at the 12-month or a later examination were available for review in 91% of all 380 eyes at risk and for 95% of 364 fellow eyes with any gradable follow-up photographs. Corresponding percentages were 85% and 56% of 380 fellow eyes at risk and 89% and 58% for the 364 eyes at the 24- and 36-month examinations. Participant deaths and censored follow-up due to study termination accounted for most of the missing data and for longer follow-up in patients whose first eye had predominantly CNV compared with those with predominantly hemorrhagic lesions (P = .04).
INCIDENCE RATES OF CNV LESIONS IN FELLOW EYES AT RISK
Incident CNV lesions developed in 98 fellow eyes during follow-up of up to 4 years. Figure 1 shows the cumulative percentage of fellow eyes with incident CNV lesions at each follow-up examination time. By months 3, 6, and 12, the cumulative incidence of fellow eye CNV was 3%, 7%, and 14%. Lower incidence rates, 8% and 9%, were observed in the second and third years, respectively, than in the first year (14%). Cumulative incidence of fellow eye CNV was 37% by 48 months. Incidence rates were similar through month 12 regardless of the type of neovascular lesion initially present in the first eye. However, during subsequent follow-up, more fellow eyes progressed to CNV among patients whose first eye had predominantly CNV than those with predominantly hemorrhagic lesions (log-rank test, P = .02) (available in eFigure 1).
|
|
|
|
Figure 1. Cumulative percentage of all fellow eyes that had incident choroidal neovascularization (CNV) by follow-up examination time. Solid line indicates estimated percentage; bars, 95% confidence intervals. Number of fellow eyes at risk at each examination time, number of eyes with incident CNV observed at that examination, and number censored following that examination are shown below the horizontal axis.
|
|
|
|
|
|
|
eFigure 1. Cumulative percentage of fellow eyes in each Submacular Surgery Trials (SST) group that had incident choroidal neovascularization (CNV) by follow-up examination time. Solid line indicates fellow eyes of patients whose study eyes had predominantly hemorrhagic lesions at enrollment in the SST; broken line, fellow eyes of patients whose study eyes had predominantly CNV at time of SST enrollment. Numbers of fellow eyes at risk at each examination time, numbers of eyes with incident CNV observed at that examination, and numbers censored following that examination are shown below the horizontal axis.
|
|
|
BASELINE CHARACTERISTICS
Demographic and ocular characteristics of the fellow eye cohort at risk of CNV are described elsewhere (available in eTable 1). Median age of the subjects was 77 years, one-half were women, and the baseline visual acuity was 20/40 or better in 321 fellow eyes (88%). None of the demographic characteristics or characteristics of the SST study eye or fellow eye mentioned in eTable 1 was associated with incident CNV in the fellow eye during follow-up.
|
|
|
|
eTable 1. Baseline Characteristics of Patients at Risk of Choroidal Neovascularization in Fellow Eyes During Follow-up
|
|
|
INCIDENT CNV LESION CHARACTERISTICS
Incident lesions identified on fluorescein angiograms (n = 97) were predominantly CNV (CNV occupied 50% of the lesion area) in 87 eyes (90%). No substantive differences were found between lesions in fellow eyes by the type of neovascular lesion present at baseline in first eyes; therefore, only composite data are provided in Table 1 and subsequent data displays. The most common lesion composition at incidence was occult CNV without a classic component, which was observed in 64 eyes (67%). Minimally classic CNV or predominantly classic CNV lesions were each found in 12 eyes (13%). Occasional lesions were predominantly blood or scar when first recognized on photographs taken late in follow-up.
|
|
|
|
Table 1. Characteristics of Choroidal Neovascular Lesions and Visual Acuity at Incidence
|
|
|
Lesion components other than CNV were identified in 32 eyes (34%). The most frequent lesion component identified in addition to CNV was blood (blood which obscures the ability to detect CNV posterior to it) in 18 eyes (19%). In contrast, blood associated with the lesion (not necessarily considered a lesion component) was more common (39%), even among the occult no-classic-CNV lesions (33%, Table 2). Other lesion components such as blocked fluorescence (12%) or scar (6%) were infrequent, and no eye had a serous pigment epithelial detachment at CNV incidence. The majority (91%) of CNV lesions had little or no fibrosis (subretinal fibrosis occupying 0%-25% of the lesion area). Features associated with a retinal lesion anastomosis (RLA or RAP [retinal angiomatous proliferation]) such as intraretinal hemorrhage overlying a retinal vessel anterior to CNV or a retinal vessel diving posteriorly into the CNV lesion were seldom apparent (4%).
|
|
|
|
Table 2. Combinations of Characteristics of Choroidal Neovascular Lesions and Visual Acuity of Fellow Eyes at Incidence
|
|
|
The majority of lesions (61%) were subfoveal and nearly all (91%) subfoveal lesions had CNV underlying the fovea. However, a substantial proportion of the incident CNV lesions spared the fovea: 29 eyes (30%) had extrafoveal lesions and 9 eyes (9%) had juxtafoveal lesions.
Lesions tended to be small when first observed on angiography; 61% were 3 DAs or smaller and nearly half were 2 DAs or smaller. Lesions larger than 9 DAs at incidence were infrequent (9%). The median greatest linear dimension (GLD) at incidence was 3160 µm (1.76 disc diameter). Lesions rarely (14%) had surrounding atrophy; even when atrophy was present and included in the total lesion size, most lesions were 5 DAs or smaller.
ASSOCIATIONS BETWEEN FEATURES OF INCIDENT CNV
Occult with no classic CNV lesions tended to be small (73% were  3 DAs; median GLD, 2860 µm); the majority of minimally classic CNV lesions (67%; median GLD, 4680 µm) and predominantly classic CNV lesions (73%; median GLD, 4520 µm) were larger than 3 DAs (Table 2). Lesions that spared the fovea also tended to be small (92%) compared with lesions that were subfoveal (40%).
RELATIONSHIP BETWEEN CNV FEATURES AND TIME CNV FIRST OBSERVED
Three study examinations were performed during the first year of follow-up, but subsequent examinations, photography, and angiography were performed only at annual intervals. Fifty of all 98 incident cases of CNV in fellow eyes were identified within the first year of follow-up. Of the 74 incident CNV cases identified during the first 2 years of follow-up when all patients were eligible for examinations, 50 of 74 were observed during the first year and only 24 during the second year when fewer fellow eyes were at risk. With more frequent monitoring during the first year, a substantially higher proportion of incident lesions spared the fovea (P = .008), had occult without classic composition (P < .001), and were 3 DAs or smaller (P < .001) (Figure 2A-C) vs the proportion of lesions of similar type at later annual examinations. Many of these lesion characteristics were correlated with one another; thus, a higher proportion of incident CNV lesions identified in the first year had all 3 favorable characteristics (outside the fovea, occult CNV without classic, and small) as compared with subsequent years (P = .009) (Figure 2D).
|
|
|
|
Figure 2. Characteristics of incident choroidal neovascularization (CNV) in fellow eyes by year of follow-up when first observed. Each bar represents 100% of the incident cases identified during each year; the number of incident cases during the year that could be classified for each feature is shown at the top of each bar. A, Lesion location by year (1 case indeterminate). B, Lesion composition by year (3 cases indeterminate). "Other" includes cases that are predominantly blood or scar. C, Lesion size by year (2 cases indeterminate). D, Combination of lesion location, composition, and size (4 cases indeterminate). MPS indicates Macular Photocoagulation Study; DAs, disc areas.
|
|
|
EFFECT ON VISUAL ACUITY OF INCIDENT CNV
The distribution of visual acuity at first observation of CNV in the fellow eye is shown in Table 1 and median acuity by lesion composition and location is shown in Table 2. Visual acuity was 20/40 or better in 58 eyes (59%); only one-quarter of fellow eyes had visual acuity 20/100 or worse when CNV was first noted. Median visual acuity at incidence was 20/25 among the 29 eyes with extrafoveal lesions, 20/32 among the 9 eyes with juxtafoveal lesions, and 20/50 among the 57 eyes with subfoveal lesions at incidence. Irrespective of lesion location, occult lesions tended to have the best visual acuity and predominantly classic CNV lesions the poorest visual acuity. Visual acuity was not greatly affected by lesion size within the occult without classic CNV subgroup; larger incident lesions that contained a classic component had worse visual acuity than smaller ones.
We examined both the distribution of change in visual acuity from baseline to the time CNV was first observed and the median change in vision by lesion location and composition at incidence (available in eTable 2). The majority of eyes (39 of 64) with occult without classic CNV lesions maintained vision within 2 lines of baseline; the median loss of visual acuity was 0.6 lines (3 letters). In contrast, the majority of fellow eyes that had minimally classic CNV (9 of 12) or predominantly classic CNV (7 of 11) at incidence lost 4 or more lines of acuity; the median loss was 4.8 lines and 10.2 lines, respectively. The amount of visual acuity loss was dependent on lesion location: median loss for extrafoveal lesions was 0.0 lines; for juxtafoveal lesions, 0.8 lines; and for subfoveal lesions, 3.2 lines.
|
|
|
|
eTable 2. Change in Visual Acuity From Baseline to Incidence of Choroidal Neovascular Lesion by Lesion Location and Composition
|
|
|
Eyes in which incident CNV developed lost visual acuity from a median of 20/25 at baseline to a median of 20/250 at year 4 (available in eFigure 2) in contrast with fellow eyes that did not develop incident CNV or foveal GA and fellow eyes that already had CNV at study entry in which median visual acuity after 4 years was within 1 line of the baseline level.
|
|
|
|
eFigure 2. Median visual acuity in fellow eyes by presence or absence of choroidal neovascularization (CNV) and foveal geographic atrophy (GA). Open circles indicate fellow eyes free of CNV and foveal GA at baseline in whom no CNV developed during follow-up; solid circles, fellow eyes free of CNV and foveal GA at baseline in whom CNV developed over time; and solid squares, fellow eyes with neovascular age-related macular degeneration at baseline and throughout follow-up.
|
|
|
COMPARISON OF CHARACTERISTICS OF CNV LESIONS IN SECOND EYES VS FIRST EYES
A comparison of the study (first) eye lesion composition at SST enrollment and the fellow eye lesion composition at the incident examination is outlined in Table 3. The most common lesion composition in fellow eyes was occult CNV without classic CNV. Because study eyes could not have a lesion composed of occult CNV only, only 24 pairs of eyes could be classified as both predominantly classic CNV, minimally classic CNV, or predominantly blood. The same lesion composition developed in the fellow eye as in the first (study) eye in 19% of participants with minimally classic CNV, 9% of those with predominantly classic CNV, and 4% of those with predominantly blood in the first eye.
|
|
|
|
Table 3. Comparison of Composition of Incident Neovascular Lesions in Fellow Eyes With Composition of Subfoveal Neovascular Lesions in Study Eyes at Time of SST Enrollment
|
|
|
COMMENT
During follow-up of 364 patients with AMD and a neovascular lesion in only the study eye at time of enrollment in the SST, CNV developed in the second eye of 98 patients (27%), yielding a cumulative incidence rate of 37% by 48 months. The annual incidence rate for CNV varied from 7% to 14%, consistent with other prospective studies of fellow eye CNV.6-7 The annual incidence rate was highest during the first year of follow-up. Although not all patients were followed up beyond 2 years, there was no evidence that the incidence rate of second-eye CNV was decreasing during later follow-up.
The emphasis of this report has been to describe the incident CNV lesions in the second eyes of patients with AMD because relatively little information has been published detailing the angiographic features of new-onset CNV that develops in a cohort that has been monitored prospectively. Increased familiarity with features of new-onset CNV may lead to earlier detection and treatment of CNV in eyes at risk. The majority of incident neovascular lesions that developed during follow-up among the SST participants were predominantly CNV (90%) as opposed to predominantly scar or predominantly blood, and the majority had little or no fibrosis or surrounding atrophy. This observation suggests that the interval between CNV development and detection of this event on a study angiogram was quite short and increases the likelihood that lesion characteristics reported here represent early-onset CNV lesions. Furthermore, within the first year of SST follow-up, the protocol required 3 angiograms (months 3, 6, and 12), and more than half of the lesions described in this report were identified during this period of close monitoring. Thus, it is even more likely that the features of these lesions represent those typical of new CNV.
Lesions containing a classic CNV component were found in only 25% of eyes with half having predominantly classic and half minimally classic CNV composition. The higher frequency of lesions that had occult CNV without a classic component (67%) compared with lesions containing classic CNV is consistent with clinical impression and the high rate of occult CNV without classic CNV in study eyes of patients enrolled in AMD trials of treatment of predominantly CNV lesions2, 16 and incident CNV lesions in the CNVPT. The 50:50 split between predominantly classic and minimally classic CNV composition among the classic CNV–containing lesions is consistent with the distribution of these lesions in trials that enrolled patients with classic CNV lesions1, 17
Choroidal neovascular lesions consist of CNV (classic and/or occult) and other features, such as thick contiguous blood, fibrous tissue, hyperpigmentation, serous pigment epithelial detachment, or scar. Approximately one-third of the incident lesions in this study had 1 or more of these non-CNV features, which are included when determining lesion size. In this series, the most frequent lesion component other than CNV was blood (19%). The presence of any blood (preretinal, intraretinal, subretinal, subretinal pigment epithelial) associated with the lesion was twice as common as blood considered to be a lesion component. The presence of blood is an important prognostic feature for eyes with occult lesions without a classic component; 21 of 64 cases (33%) of incident occult CNV without classic CNV lesions were associated with blood. These eyes are more apt to have a poor natural course than those without blood.18
Although the majority of new CNV lesions were subfoveal, many were not, including 30% that were outside the foveal avascular zone. Our findings confirm this same observation among 18 incident cases in the CNVPT. Incident lesions tended to be small; smaller lesions tend to have better visual acuity at presentation and to lose less vision over time (with or without treatment).2, 17, 19 Thus, there is an opportunity to identify small lesions in the majority of patients in whom CNV develops in the second eye and, by doing so, to affect the long-term vision outcomes favorably in these patients.
Characteristics of incident CNV tended to cluster. Lesions with occult CNV and no classic CNV component more often were small and spared the fovea; minimally or predominantly classic CNV lesions tended to be larger and subfoveal. During the first year of follow-up, ie, the period of more intense monitoring in the SST, a higher proportion of incident CNV lesions spared the fovea, had occult CNV without a classic component, and were small in total area (Figure 2). Occult lesions and lesions outside the fovea were associated with better BCVA at time of detection. The presenting features of CNV lesions (lesion size and composition) and visual acuity influence vision outcomes after therapy.19-20 Our findings suggest that careful and close monitoring of eyes at risk of CNV development with sequential fluorescein angiography may lead to earlier detection of new-onset CNV and CNV lesions with characteristics that portend a more favorable outcome after treatment.
The demographic features of patients (median age, 77 years; 50% women) and visual acuity of second eyes at risk (median acuity, 20/25) were similar to those of other AMD cohorts at risk (S. B. Bressler, MD; T. E. Clemons, PhD; F. L. Ferris, MD; N. M. Bressler, MD; AREDS Research Group; unpublished data; 2006).6-7 Contrary to the work of Chang and associates,5 but in keeping with the findings of the MPS,6-7 there was no apparent correlation between the lesion composition of incident CNV lesions among the fellow eyes and the CNV lesions in the study eyes of these same individuals when they enrolled in the SST. The poor correlation may in part be due to a comparison that is limited by the timing of incident CNV in first vs second eyes. The characteristics of CNV lesions described in fellow (second) eyes represent incident CNV. The characteristics of the CNV lesions described for study (first) eyes represent those features present at the time of SST enrollment, which may not correspond to the lesion characteristics at the time of incidence.
This analysis of incident CNV in the fellow eyes of AMD participants in the SST suggests that the majority of lesions tend to have occult CNV without a classic CNV component and are small (3 DAs); they are frequently in a nonfoveal location and are associated with minimal loss of visual acuity. Furthermore, in this cohort of patients, many of these second eye lesions developed within the first 12 months of follow-up. Thus, careful and frequent follow-up (eg, every 3-6 months) of second eyes at risk with fluorescein angiography during at least the first year after detection of CNV in the first AMD eye may lead to early detection and treatment of neovascular AMD, ultimately providing a better visual prognosis for these eyes.
AUTHOR INFORMATION
Correspondence: Sharon D. Solomon, MD, Johns Hopkins University School of Medicine, 550 N Broadway, Ste 115, Baltimore, MD 21205-2005 (nmboffice{at}jhmi.edu).
Submitted for Publication: January 18, 2007; final revision received May 24, 2007; accepted June 5, 2007.
Financial Disclosure: None reported.
Funding/Support: The Submacular Surgery Trials were sponsored by the National Eye Institute, National Institutes of Health, US Department of Health and Human Services, Bethesda, Maryland (cooperative agreements U10 EY11547, EY11557, and EY11558 with the Johns Hopkins University, Baltimore, Maryland). Participating clinical centers were supported by contracts with the Johns Hopkins University.
*Authors/Writing Committee: This article was prepared by the following members of the Submacular Surgery Trials Research Group: Sharon D Solomon, MD; Barbara S Hawkins, PhD; Joan L Jefferys, ScM; Neil M. Bressler, MD; Susan B Bressler, MD; Johns Hopkins University School of Medicine and Wilmer Eye Institute, Baltimore, Maryland.
Group Information: A list of the members of the Submacular Surgery Trials (SST) Research Group was published in Arch Ophthalmol. 2004;122(11):1597-1611.
REFERENCES
| |
1. Age-Related Eye Disease Study Research Group. Potential public health impact of Age-Related Eye Disease Study results: AREDS report No. 11. Arch Ophthalmol. 2003;121(11):1621-1624.
FREE FULL TEXT
2. Rosenfeld PJ, Brown DM, Heier JS; et al. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med. 2006;355(14):1419-1431.
FREE FULL TEXT
3. Brown DM, Kaiser PK, Michels M; et al. Ranibizumab versus verteporfin for neovascular age-related macular degeneration. N Engl J Med. 2006;355(14):1432-1444.
FREE FULL TEXT
4. The Choroidal Neovascularization Prevention Trial Research Group. Choroidal neovascularization in the Choroidal Neovascularization Prevention Trial. Ophthalmology. 1998;105(8):1364-1372.
FULL TEXT
|
ISI
| PUBMED
5. 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):1380-1386.
ISI
| PUBMED
6. Macular Photocoagulation Study Group. Five-year follow-up of fellow eyes of patients with age-related macular degeneration and unilateral extrafoveal choroidal neovascularization. Arch Ophthalmol. 1993;111(9):1189-1199.
FREE FULL TEXT
7. 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(6):741-747.
FREE FULL TEXT
8. Age-Related Eye Disease Study Research Group. A randomized, placebo-controlled, clinical trial of high-dose supplementation with vitamins C and E, beta carotene, and zinc for age-related macular degeneration and vision loss: AREDS report No. 8. Arch Ophthalmol. 2001;119(10):1417-1436.
FREE FULL TEXT
9. Submacular Surgery Trials (SST) Research Group. Surgery for subfoveal choroidal neovascularization in age-related macular degeneration: ophthalmic findings: SST report No. 11. Ophthalmology. 2004;111(11):1967-1980.
FULL TEXT
|
ISI
| PUBMED
10. Submacular Surgery Trials (SST) Research Group. Surgery for hemorrhagic choroidal neovascular lesions of age-related macular degeneration: ophthalmic findings: SST report No. 13. Ophthalmology. 2004;111(11):1993-2006.
FULL TEXT
|
ISI
| PUBMED
11. Fleiss JL. Statistical Methods for Rates and Proportions. 2nd ed. New York, NY: John Wiley & Sons; 1981:217-225.12. Kaplan EL, Meier P. Nonparametric estimation from incomplete observations. J Am Stat Assoc. 1958;53:457-481.
FULL TEXT
|
ISI
13. Pearson K. On the criterion that a given system of deviations from the probable in the case of a correlated system of variables is such that it can be reasonably supposed to have arisen from random sampling. Philos Mag. 1900;50(5):157-175.
14. Armitage P, Berry G. Statistical Methods in Medical Research. 2nd ed. Boston, MA: Blackwell Scientific Publications Inc; 1987:372-374.15. McNemar Q. Note on the sampling error of the difference between correlated proportions or percentages. Psychometrika. 1947;12:153-157.
FULL TEXT
|
ISI
16. Gragoudas ES, Adamis AP, Cunningham ET, Feinsod M, Guyer DR. Pegaptanib for neovascular age-related macular degeneration. N Engl J Med. 2004;351(27):2805-2816.
FREE FULL TEXT
17. 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 1. Arch Ophthalmol. 1999;117(10):1329-1345.
FREE FULL TEXT
18. Stevens TS, Bressler NM, Maguire MG; et al. Occult choroidal neovascularization in age-related macular degeneration: a natural history study of morphologic changes. Arch Ophthalmol. 1997;115(3):345-350.
FREE FULL TEXT
19. Treatment of Age-Related Macular Degeneration With Photodynamic Therapy and Verteporfin in Photodynamic Therapy Study Groups. 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):407-418.
FULL TEXT
|
ISI
| PUBMED
20. Macular Photocoagulation Study Group. Visual outcome after laser photocoagulation for subfoveal choroidal neovascularization secondary to age-related macular degeneration: the influence of initial lesion size and initial visual acuity. Arch Ophthalmol. 1994;112(4):480-488.
FREE FULL TEXT
SECTION EDITOR: ROY W. BECK, MD, PhD
 CiteULike  Connotea  Del.icio.us  Digg  Reddit  Technorati  Twitter
What's this?
|
