Purpose
To explore whether monthly intravitreal ranibizumab injections are associated with a lower rate of new choroidal neovascularization (CNV) in fellow eyes of patients with unilateral neovascular age-related macular degeneration.
Design
Retrospective data analysis of randomized, controlled clinical trials.
Methods
Incidence of new CNV in fellow eyes was calculated at 12 and 24 months from 2 clinical trials (the Minimally Classic/Occult Trial of the Anti-VEGF Antibody Ranibizumab in the Treatment of Neovascular Age-Related Macular Degeneration [MARINA] study and the Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in Age-Related Macular Degeneration [ANCHOR] study), based on fluorescein angiographic reading center criteria and investigator evaluation. Patients treated with monthly ranibizumab (0.3 and 0.5 mg) were compared with those receiving a sham injection (MARINA) or photodynamic therapy (ANCHOR).
Results
In MARINA, new CNV developed in fellow eyes in 20.3% of the 0.3-mg ranibizumab group by 12 months and in 30.4% by 24 months. The conversion rate in the 0.5-mg ranibizumab group was 21.1% and 38.0% by 12 and 24 months, respectively. In the sham group, 26.4% converted by 12 months and 36.3% converted by 24 months. In ANCHOR, fellow eyes in 15.9% of the 0.3-mg ranibizumab group converted by 12 months and fellow eyes in 23.8% converted by 24 months. The conversion rate in the 0.5-mg ranibizumab group was 24.3% and 35.1% by 12 and 24 months, respectively. In the photodynamic therapy group, 25.4% converted by 12 months and 38.8% converted by 24 months. Differences in conversion rates at 12 and 24 months between the 0.3-mg or 0.5-mg ranibizumab groups and respective controls (sham or photodynamic therapy) were not statistically significant.
Conclusions
Results of this study do not support the hypothesis that monthly ranibizumab injections reduce the rate of CNV development in untreated fellow eyes.
Several epidemiologic studies have suggested that patients with unilateral neovascular age-related macular degeneration (AMD) have a substantial risk of developing choroidal neovascularization (CNV) in their fellow eye. The reported incidences of CNV in the fellow eyes of these unilateral cases varies from 12% to 22% at 2 years, 37% at 4 years, 22% to 38.7% at 5 years, or an annual rate of 9% to 10% for predominantly white study populations. More recently in the Age-Related Eye Disease Study (AREDS), of the 788 participants in the unilateral advanced AMD group (n = 714 with neovascular AMD in their fellow eye at baseline), in 278 (35%), neovascular disease developed in the second eye (median follow-up, 6.3 years).
For patients with unilateral neovascular AMD, proven treatments to prevent conversion to neovascular AMD in the fellow eye are limited to the intake of high-dose vitamins and zinc based on findings of AREDS. Recent animal studies and case reports, however, have raised the possibility that treatment with anti-vascular endothelial growth factor (anti-VEGF) agents such as bevacizumab (Avastin; Genentech, Inc, South San Francisco, California, USA) may have a therapeutic prophylactic effect on the untreated fellow eye.
Bakri and associates reported that, in a rabbit model, after an intravitreal injection of bevacizumab, small amounts of the drug were detected in the serum as well as in the fellow eye. Consistent with these findings, Avery and associates reported decreased fluorescein angiography (FA) leakage from retinal neovascularization in fellow eyes of patients with proliferative diabetic retinopathy, who were unilaterally treated with intravitreal bevacizumab. In the same report, the authors demonstrated that therapeutic effects of intravitreal bevacizumab could be observed in eyes receiving as little as 0.5% of the commonly used 1.25-mg dose. However, other studies failed to show such a systemic or fellow-eye effect. Studying diabetic patients before vitrectomy, Sawada and associates reported that aqueous VEGF levels obtained from untreated fellow eyes 1 week after injection of 1.25 mg intravitreal bevacizumab did not differ from baseline measurements in the treated eye before injection.
If an anti-VEGF has a measurable effect on neovascularization in the fellow eye, one could hypothesize that treatment with other anti-VEGF inhibitors could have a similar effect. The aim of this study was to explore whether repetitive intraocular injections of the anti-VEGF agent ranibizumab (Lucentis; Genentech, Inc) in eyes with neovascular AMD may lower the rate of development of CNV in the fellow eyes of patients with unilateral neovascular AMD.
Methods
A retrospective analysis was conducted of patients enrolled in 2 large, controlled clinical trials, the Anti-VEGF Antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in Age-Related Macular Degeneration (ANCHOR) study and the Minimally Classic/Occult Trial of the Anti-VEGF Antibody Ranibizumab in the Treatment of Neovascular Age-Related Macular Degeneration (MARINA) study, evaluating ranibizumab for the treatment of CNV secondary to AMD.
Study Population
Of all MARINA and ANCHOR patients, only subjects whose fellow eyes were defined as lacking signs of neovascular AMD at baseline were included in this analysis. The absence of neovascular AMD in the fellow eye was determined by both investigator evaluation as noted on the screening case report form and by central reading center evaluation of FA characteristics (University of Wisconsin Fundus Photograph Reading Center, Madison, Wisconsin, USA). The fellow eye had to show the absence of all of the following angiographic characteristics: neovascular or exudative characteristics, classic CNV, fibrovascular pigment epithelial detachment, other occult CNV or late leakage of undetermined source, blood, serous pigment epithelial detachment, fibrous tissue, serous sensory retinal detachment or subretinal fluid, and photocoagulation scar. Patients whose FA results could not be graded or were indeterminate regarding any of the above-mentioned characteristics were excluded from the analysis.
Assessments
Fellow eyes in both the MARINA and ANCHOR studies were examined at screening, days 0 and 7 as well as monthly from month 1 through month 24 using Early Treatment Diabetic Retinopathy protocol-based best-corrected visual acuity, slit-lamp examination, intraocular pressure measurement, and dilated binocular indirect and high magnification ophthalmoscopy. Recorded at each of these visits were concomitant medications, concurrent ocular procedures (except at screening), and adverse events (except at screening). Contrast sensitivity was carried out at screening, day 7, and months 1, 2, 3, 6, 12, and 24. FA results and fundus photography were examined at screening and at months 3, 6, 12, and 24 in MARINA and every 3 months starting at screening up to 24 months in ANCHOR (months 3, 6, 9, 12, 15, 18, 21, and 24). Optical coherence tomography was carried out at select sites at days 0 and 7 as well as months 1 and 12.
Criteria
Fellow eye conversion rates to neovascular AMD through 12 and 24 months of follow-up were calculated using 3 separate criteria: (1) FA only, (2) investigator determined only, and (3) either FA or investigator determined (FA/investigator). Because the aim was to capture every possible conversion, the FA/investigator criterion that would include all FA or investigator-determined conversions was considered primary for this analysis.
Fluorescein angiography conversion criterion
Patients were considered to have a neovascular AMD conversion via FA if, after their baseline visit, the reading center found any of the characteristics considered to be consistent with neovascular AMD listed in Table 1 . As soon as the presence of neovascular AMD was confirmed at any point, a patient was considered to have converted at all later points (first conversion carried forward).
| Presence of any of the following: |
| • Neovascular or exudative characteristics |
| • Classic choroidal neovascularization |
| • Fibrovascular pigment epithelial detachment |
| • Other occult choroidal neovascularization or late leakage of undetermined source |
| • Blood |
| • Serous pigment epithelial detachment |
| • Fibrous tissue |
| • Serous sensory retinal detachment |
| • Photocoagulation scar |
Investigator-determined conversion criterion
Independently, the authors (K.B.F., I.B., N.S.) determined fellow eye conversions to neovascular AMD by evaluation of adverse events, concurrent ocular procedures, and concomitant medications reported through month 24. The authors were blinded to the treatment assignments of the patients included in this analysis. Based on fellow eye listings recorded on case report forms, only entries that indicated a definite neovascular AMD conversion were selected ( Table 2 ). After the first such determination of a neovascular AMD conversion, a patient was considered to have converted at all later time points (first conversion carried forward).
| Conversion Criteria Determined by | Examples Include | Associated Diagnoses Include |
|---|---|---|
| Investigator-recorded adverse events |
| Not applicable |
| Investigator-recorded concurrent ocular procedures a |
|
|
| Investigator-recorded concomitant medications a | Medication classes of
|
|
a Concomitant medications and concurrent ocular procedures included only if associated diagnosis indicated conversion to neovascular AMD.
Statistical Analysis
The percentage of patients with neovascular AMD conversion in the fellow eye through months 12 and 24 was examined by treatment group and by study and was pooled across ranibizumab treatments and studies. Chi-square tests for contingency tables were used to compare conversion rates between patients whose study eyes were randomized to control (sham or PDT) versus 0.3 mg and 0.5 mg ranibizumab separately and pooled. In both the MARINA and ANCHOR studies, control patients could crossover to ranibizumab treatment in the second year, with the earliest crossover occurring at 18 months. All analyses were conducted using an intent-to-treat analysis.
The Kaplan-Meier method was used to estimate time to first conversion using the FA/investigator conversion criterion, and the log-rank test was used to compare the time course among treatment groups. For converters, time to first conversion was calculated as the number of days between the first treatment date and the first conversion date defined as the FA examination date for FA, onset date for adverse event, procedure date for concurrent ocular procedure, and start date for concomitant medication. For censored observations (nonconverters), follow-up time was calculated as days from first treatment date to last study date. If first treatment date was missing, the initial randomization date was used. For plotting purposes only, 2 conversions classified as having occurred by month 24 (defined as up to 730 days) are not shown (1 0.3-mg patient from the MARINA study at 735 days and 1 PDT patient from the ANCHOR study at 811 days). Statistical significance was defined as P < .05. All analyses were conducted using SAS software version 9.1 (SAS Institute, Inc, Cary, North Carolina, USA).
Results
A total of 716 patients enrolled in the MARINA study were assigned randomly to sham control (n = 238), monthly intravitreal ranibizumab injections of 0.3 mg (n = 238), and monthly intravitreal ranibizumab injections of 0.5 mg (n = 240). The ANCHOR study included a total of 423 patients assigned randomly to the verteporfin PDT control (n = 143) or to monthly intravitreal injections of 0.3 mg (n = 140) or 0.5 mg (n = 140) ranibizumab.
At baseline, 241 (34%) fellow eyes in the MARINA study and 204 (48%) fellow eyes in the ANCHOR study were found to be without evidence of neovascular AMD based on the criteria stated above, and therefore were included in this analysis. Baseline characteristics of fellow eyes and study eyes for these selected patients, including the AREDS group classifications by the reading center, are summarized in Table 3 .
| MARINA | ANCHOR | |||||||
|---|---|---|---|---|---|---|---|---|
| Sham (n = 91) | 0.3 mg (n = 79) | 0.5 mg (n = 71) | Pooled Ranibizumab (n = 150) | PDT (n = 67) | 0.3 mg (n = 63) | 0.5 mg (n = 74) | Pooled Ranibizumab (n = 137) | |
| Age (yrs), mean (SD) | 76.4 (6.6) | 77.1 (6.8) | 75.1 (8.1) | 76.1 (7.5) | 77.5 (8.4) | 76.8 (7.1) | 75.1 (8.4) | 75.9 (7.9) |
| Gender, n (% female) | 59 (64.8) | 48 (60.8) | 36 (50.7) | 84 (56.0) | 43 (64.2) | 29 (46.0) | 30 (40.5) | 59 (43.1) |
| Fellow eye VA (letters), mean (SD) | 77.5 (14.0) | 74.8 (13.4) a | 78.1 (14.1) | 76.4 (13.8) b | 76.2 (12.7) | 79.6 (13.5) | 78.5 (11.8) | 79.0 (12.6) |
| Study eye VA (letters), mean (SD) | 52.5 (16.2) | 52.5 (12.0) | 54.9 (12.9) | 53.6 (12.4) | 44.1 (13.8) | 48.3 (13.7) | 46.7 (13.4) | 47.4 (13.5) |
| AREDS Group, n (%) | ||||||||
| Group 1 | 6 (6.6) | 3 (3.8) | 8 (11.3) | 11 (7.3) | 7 (10.4) | 8 (12.7) | 9 (12.2) | 17 (12.4) |
| Group 2 | 22 (24.2) | 20 (25.3) | 12 (16.9) | 32 (21.3) | 20 (29.9) | 26 (41.3) | 19 (25.7) | 45 (32.8) |
| Group 3 | 59 (64.8) | 50 (63.3) | 43 (60.6) | 93 (62.0) | 37 (55.2) | 28 (44.4) | 45 (60.8) | 73 (53.3) |
| Group 4a | 4 (4.4) | 6 (7.6) | 8 (11.3) | 14 (9.3) | 2 (3.0) | 1 (1.6) | 1 (1.4) | 2 (1.5) |
| Could not grade | 0 (0.0) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (1.5) | 0 (0.0) | 0 (0.0) | 0 (0.0) |
The conversion rates of the fellow eyes based on the FA/investigator criterion are shown in Table 4 . In MARINA at month 12, the 0.3-mg and 0.5-mg ranibizumab groups had slightly lower conversion rates (20% to 21%) compared with sham (26%). This trend was not maintained at month 24, where the 0.3-mg and 0.5-mg ranibizumab groups had 30% and 38% conversions versus 36% in the sham group. In the ANCHOR study, at both months 12 and 24, the PDT and 0.5-mg ranibizumab groups had similar conversion rates (24% to 25% at month 12 and 35% to 39% at month 24), whereas the 0.3-mg ranibizumab group had a somewhat lower conversion rate (16% at month 12 and 24% at month 24). All comparisons between control and ranibizumab groups were not statistically significant.
| First Conversion Criteria | MARINA: Number (%) Converted | |||||||
|---|---|---|---|---|---|---|---|---|
| Month 12 | Month 24 | |||||||
| Sham (n = 91) | 0.3 mg (n = 79) | 0.5 mg (n = 71) | Pooled Ranibizumab (n = 150) | Sham (n = 91) | 0.3 mg (n = 79) | 0.5 mg (n = 71) | Pooled Ranibizumab (n = 150) | |
| FA b | 19 (20.9%) | 15 (19.0%) | 13 (18.3%) | 28 (18.7%) | 29 (31.9%) | 22 (27.8%) | 21 (29.6%) | 43 (28.7%) |
| Chi-square P value vs sham | .76 | .68 | .67 | .57 | .75 | .60 | ||
| Investigator determination c | 14 (15.4%) | 8 (10.1%) | 9 (12.7%) | 17 (11.3%) | 23 (25.3%) | 13 (16.5%) | 17 (23.9%) | 30 (20.0%) |
| Chi-square P value vs sham | .31 | .62 | .36 | .16 | .85 | .34 | ||
| FA/investigator determination d | 24 (26.4%) | 16 (20.3%) | 15 (21.1%) | 31 (20.7%) | 33 (36.3%) | 24 (30.4%) | 27 (38.0%) | 51 (34.0%) |
| Chi-square P value vs sham | .35 | .44 | .31 | .42 | .82 | .72 | ||
| First Conversion Criteria | ANCHOR: Number (%) Converted | |||||||
|---|---|---|---|---|---|---|---|---|
| Month 12 | Month 24 | |||||||
| PDT (n = 67) | 0.3 mg (n = 63) | 0.5 mg (n = 74) | Pooled Ranibizumab (n = 137) | PDT (n = 67) | 0.3 mg (n = 63) | 0.5 mg (n = 74) | Pooled Ranibizumab (n = 137) | |
| FA b | 12 (17.9%) | 9 (14.3%) | 13 (17.6%) | 22 (16.1%) | 22 (32.8%) | 12 (19.0%) | 23 (31.1%) | 35 (25.5%) |
| Chi-square P value vs PDT | .57 | .96 | .74 | .074 | .82 | .28 | ||
| Investigator determination c | 14 (20.9%) | 7 (11.1%) | 14 (18.9%) | 21 (15.3%) | 21 (31.3%) | 12 (19.0%) | 18 (24.3%) | 30 (21.9%) |
| Chi-square P value vs PDT | .13 | .77 | .32 | .11 | .35 | .14 | ||
| FA/investigator determination d | 17 (25.4%) | 10 (15.9%) | 18 (24.3%) | 28 (20.4%) | 26 (38.8%) | 15 (23.8%) | 26 (35.1%) | 41 (29.9%) |
| Chi-square P value vs PDT | .18 | .89 | .42 | .066 | .65 | .20 | ||
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
