- Email: [email protected]
Effects of Vitreomacular Adhesion on Anti-Vascular Endothelial Growth Factor Treatment for Exudative Age-Related Macular Degeneration
Effects of Vitreomacular Adhesion on Anti-Vascular Endothelial Growth Factor Treatment for Exudative Age-Related Macular Degeneration Sung Jun Lee, MD,1 Hyoung Jun Koh, MD2 Objective: To evaluate the effect of posterior vitreomacular adhesion (VMA), documented by optical coherence tomography (OCT), on the outcome of anti-vascular endothelial growth factor (VEGF) treatment for exudative age-related macular degeneration (AMD). Design: Retrospective comparative series. Participants: A total of 148 patients (148 eyes) who were newly diagnosed with exudative AMD and were treated by anti-VEGF in 1 eye from 2005 to 2008 with a minimum of 12 months follow-up. Methods: We retrospectively reviewed OCT and medical records of 148 patients with exudative AMD and categorized them according to the presence of posterior VMA into 2 subgroups: VMA (⫹) group (38 eyes) and VMA (⫺) group (110 eyes). Best-corrected visual acuity (BCVA) and central retinal thickness (CRT) after anti-VEGF treatment were compared between the 2 groups at baseline; at 1, 3, 6, and 12 months; and at the last visit (mean ⫽ 21 months). Main Outcome Measures: Mean changes in BCVA, which was converted to logarithm of the minimum angle of resolution (logMAR) values and CRT after anti-VEGF treatment. Results: Mean BCVA significantly decreased over time in the VMA (⫹) group compared with the VMA (⫺) group (P ⫽ 0.039). At the last follow-up, mean BCVA had deteriorated from 0.87 logMAR (20/149 Snellen equivalent; baseline) to 0.98 logMAR (20/189 Snellen equivalent) in the VMA (⫹) group, but improved from 0.82 logMAR (20/132 Snellen equivalent, baseline) to 0.72 logMAR (20/104, Snellen equivalent) in the VMA (⫺) group (P ⫽ 0.028). In paired comparisons of BCVA between baseline and each follow-up visit, the VMA (⫺) group showed significant improvement of BCVA at every follow-up visit (P⬍0.05); however, the VMA (⫹) group did not show significant visual improvement at any follow-up visit despite anti-VEGF treatment (P⬎0.05). Comparison of mean CRT between baseline and each follow-up visit showed a statistically significant decrease at every follow-up in both groups (P⬍0.05). Conclusions: Posterior VMA was associated with an inferior visual outcome after intravitreal anti-VEGF treatment for exudative AMD. Our results suggest that chronic tractional forces may antagonize the effect of anti-VEGF treatment, resulting in poor response to anti-VEGF treatment with patients with VMA. Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Ophthalmology 2011;118:101–110 © 2011 by the American Academy of Ophthalmology.
Age-related macular degeneration (AMD) is the leading cause of severe visual loss in industrialized countries1; however, despite intensive basic and clinical research, the pathogenesis and risk factors for AMD are incompletely characterized.2 In recent years, the advent of anti-vascular endothelial growth factor (VEGF) therapies, such as ranibizumab and bevacizumab, has revolutionized neovascular AMD treatment,3,4 and anti-VEGF has become the standard treatment for choroidal neovascularization (CNV) with a better visual outcome than previous therapies, such as photodynamic therapy.4,5 However, one study reported that up to 45% of cases (20/44 eyes) were nonresponders showing resistance to anti-VEGF.6 A current focus of anti-VEGF treatment is how to determine which eyes will respond to treatment. To date, 3 genetic studies on the response to © 2011 by the American Academy of Ophthalmology Published by Elsevier Inc.
treatment for wet AMD have shown that specific genotypes for complement factor H and LOC genes are associated with treatment response.7–9 Previous studies have described the relationship between the posterior vitreous and the macula in AMD and have suggested that vitreomacular adhesion (VMA) plays an important role in the development of exudative AMD.10 –15 The incidence of posterior VMA is higher in eyes with exudative AMD compared with eyes with non-exudative AMD and normal controls10; however, because previous studies did not compare the normal eye and the eye with exudative AMD in individual patients with unilateral AMD, their results may be affected by genetic and environmental factors. In a recent paired eye study, we controlled confounding variables by selecting only patients with unilateral ISSN 0161-6420/11/$–see front matter doi:10.1016/j.ophtha.2010.04.015
101
Ophthalmology Volume 118, Number 1, January 2011 exudative AMD and showed that eyes with exudative AMD had a significantly higher incidence of posterior VMA than paired normal eyes (P ⫽ 0.0007).16 This result indicates that VMA is a possible risk factor for exudative AMD. In another recent study, Mojana et al11 reported improvement in visual acuity after 25-gauge trans pars plana vitrectomy with hyaloid removal in 5 patients who had a history of demonstrable VMA and poorly responsive CNV despite aggressive anti-VEGF therapy. We postulated that a subpopulation of exudative AMD cases do not respond to anti-VEGF therapy and that VMA may play a role in this resistance to therapy. This study compares the visual outcome of anti-VEGF treatment for exudative AMD with or without posterior VMA documented by optical coherence tomography (OCT).
Patients and Methods We retrospectively reviewed data for consecutive patients who were diagnosed with exudative AMD at the vitreoretinal service clinic of the Yonsei University Eye and Ear, Nose, and Throat Hospital (Seoul, Korea) from October 2005 to December 2007. This retrospective study was approved by the institutional review board of the Yonsei University College of Medicine and adhered to the tenets of the Declaration of Helsinki. We used the hospital clinical database (CDR software, Yonsei University Severance Hospital, Seoul, Korea) to identify patients with exudative AMD who underwent intravitreal anti-VEGF injections. Data were abstracted from the charts of 251 consecutive patients, and 148 patients (148 eyes) who met the eligibility criteria were included in the analysis.
Inclusion and Exclusion Criteria Inclusion criteria were (a) age ⬎50 years; (b) exudative AMD proven by fundus photograph and fluorescein angiography; (c) treatment with intravitreal anti-VEGF monotherapy using ranibizumab or bevacizumab in 1 eye; (d) follow-up for a minimum of 12 months; (e) availability of results from an OCT examination at baseline, 1, 3, 6, and 12 months, and last follow-up; and (f) availability of best-corrected visual acuity (BCVA) records at baseline, 1, 3, 6, and 12 months, and last follow-up after treatment. Exclusion criteria were (a) intraocular surgery, including cataract extraction at baseline or during follow-up; (b) history of treatment for AMD, such as verteporfin photodynamic therapy and intravitreal anti-VEGF therapy; (c) myopia of more than 3 diopters; (d) CNV due to any cause other than AMD, such as angioid streak, choroidal rupture, and ocular histoplasmosis syndrome; (e) evidence of diabetic retinopathy; (f) presence of inflammatory ocular disease; and (g) presence of comorbid ocular conditions that might affect visual acuity.
Diagnosis of Age-Related Macular Degeneration Fluorescein angiograms obtained using the Heidelberg Retina Angiograph (Heidelberg Engineering, Heidelberg, Germany) were reviewed to verify diagnoses. The location of the neovascular complex was classified as subfoveal, juxtafoveal, or extrafoveal. The lesion composition was classified into 2 categories: predominant classic and occult (occult with minimally classic or pure occult).
102
Anti-Vascular Endothelial Growth Factor Treatment Intravitreal anti-VEGF injections were performed using bevacizumab (1.25 mg/0.05 ml; Avastin, Genentech Inc., South San Francisco, CA) or ranibizumab (0.5 mg/0.05 ml; Lucentis, Genentech Inc.). After the 3 standard monthly intravitreal injections of anti-VEGF, additional injections were administered at the physician’s discretion if any of the following were observed: (1) visual acuity loss, (2) evidence of persistent fluid on OCT at least 1 month after the previous injection, (3) newly developed macular hemorrhage. The total number of injections within 12 months after the initial treatment and until the end of the follow-up was recorded.
Detection of Vitreomacular Adhesion Using Optical Coherence Tomography A commercially available OCT instrument (Stratus OCT3, Zeiss Humphrey, San Leandro, CA) was used for analysis. The macula was scanned in the horizontal and vertical meridians using the cross-hair pattern or 6 radial lines at internals of 30 degrees with a scan length of 4 to 6 mm centered through the fovea and at any sites of suspected macular lesion, as determined by simultaneous evaluation of the red-free image on the OCT computer monitor. The presence of VMA was assessed on the OCT computer monitor and defined as attachment of the posterior hyaloid line to the inner retinal surface (Fig 1). The presence of VMA was determined by 2 independent blinded observers (SJL, HJK), and any discrepancies were resolved by open discussion.
Optical Coherence Tomography Central Retinal Thickness Measurement Central retinal thickness (CRT) was defined as the mean value of retinal thickness of the central points in the horizontal and vertical scanning lines. We used automated measurements provided by the Stratus OCT software to obtain the thickness (Fig 2A). In case of segmentation error, manual measurements were performed using the electronic calipers of the Stratus OCT software (Fig 2B).
Visual Acuity Measurement Best-corrected visual acuity was measured with a Snellen chart at baseline; at 1, 3, 6, and 12 months; and at last follow-up after initial anti-VEGF treatment. The Snellen BCVA was converted to logarithm of the minimum angle of resolution (logMAR) values. The logMAR value for counting fingers visual acuity was assigned as ⫹2.0 logMAR (0.01) according to methods published by Holladay.17
Statistical Analysis Statistical analysis was performed using SPSS software version 13.0 (SPSS Inc., Chicago, IL). The change of BCVA and CRT over time was compared between the 2 subgroups using repeatedmeasures analysis of variance (ANOVA). Categoric data were assessed using the chi-square test, and continuous variables were compared with the Student t test. In addition, paired t tests were used to compare follow-up and baseline data within a treatment group. The association between the change in CRT and BCVA outcomes at different time points during the study were assessed using the Pearson correlation analysis. P⬍0.05 was considered statistically significant.
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF
Figure 1. Optical coherence tomography images showing eyes with exudative AMD with (A, C) and without VMA (B, D) (arrows). AMD ⫽ age-related macular degeneration; VMA ⫽ vitreomacular adhesion.
Results Baseline Characteristics A total of 148 patients (65 male, 83 female) with exudative AMD who underwent treatment with intravitreal anti-VEGF in 1 eye were enrolled in the study. The average age for the entire study
group was 68.2⫾7.4 years. The mean follow-up time was 21.3⫾7.9 months (range, 12– 40 months). Subjects received a mean of 3.66 injections within 12 months and 4.48 injections until the end of the follow-up. Ranibizumab was administered in 28.2% of the total injections (187/663). After OCT analysis, we categorized eyes with exudative AMD into 2 subgroups according to the presence of posterior VMA (Table 1). The VMA (⫹) group contained 38 eyes, and the VMA
Table 1. Baseline Characteristics of Patients with and without Vitreomacular Adhesion VMA (ⴙ) VMA (ⴚ) P Group (n ⴝ 38) Group (n ⴝ 110) Value Age, yrs Gender (M/F) Follow-up time (mos) Baseline BCVA (logMAR) (Snellen equivalent) Baseline CRT (m) Treatment No. ⬍12 mos Total CNV location (%) Subfoveal Juxtafoveal Extrafoveal CNV type (%) Classic Occult
Figure 2. Example of correct segmentation by automated measurement (A), CRT (arrow), and manual measurement of retinal thickness using calipers because of segmentation error (arrow) (B). CRT ⫽ central retinal thickness.
68.1⫾8.5 19/19 19.8⫾8.4 0.87⫾0.58 (20/149) 278.2⫾150.4
68.3⫾7.1 46/64 21.8⫾7.8 0.82⫾0.50 (20/132) 304.1⫾130.9
0.871* 0.381† 0.193* 0.605* 0.593* 0.313*
3.87⫾1.77 4.21⫾1.74
3.58⫾2.39 4.57⫾3.53
0.437* 0.411* 0.199†
25 (65.8) 9 (23.7) 4 (10.5)
54 (49.1) 41 (37.3) 15 (13.6)
11 (40.7) 27 (59.3)
26 (23.6) 84 (76.4)
0.515†
BCVA ⫽ best-corrected visual acuity; CNV ⫽ choroidal neovascularization; CRT ⫽ central retinal thickness; logMAR ⫽ logarithm of the minimum angle of resolution; VMA ⫽ vitreomacular adhesion. *Student t test. † Chi-square test.
103
Ophthalmology Volume 118, Number 1, January 2011
Figure 3. Mean BCVA given as logMAR units after treatment with anti-VEGF. Mean BCVA significantly decreased over time in the VMA (⫹) group compared with the VMA (⫺) group (P ⫽ 0.039, repeatedmeasures ANOVA). BCVA ⫽ best corrected visual acuity; logMAR ⫽ logarithm of the minimum angle of resolution; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
(⫺) group contained 110 eyes. Baseline BCVA was 0.87⫾0.58 logMAR (Snellen equivalent, 20/149) in the VMA (⫹) group and 0.82⫾0.50 logMAR (Snellen equivalent, 20/132) in the VMA (⫺) group. There was no significant difference in baseline BCVA between the groups (t test, P ⫽ 0.593). Baseline CRT was 278.2⫾150.4 m in the VMA (⫹) group and 304.1⫾130.9 m in the VMA (⫺) group, with no significant difference between the 2 groups (t test, P ⫽ 0.313). There were also no significant differences between the 2 groups with respect to age, gender, follow-up time, treatment numbers, CNV location, and CNV type (Table 1).
Visual and Optical Coherence Tomography Outcomes Mean changes in BCVA after anti-VEGF treatment are shown in Figure 3. Mean BCVA significantly decreased over time in the VMA (⫹) group compared with the VMA (⫺) group (P ⫽ 0.039, repeated-measures ANOVA). Between baseline and the last follow-up, the mean BCVA worsened from 0.87 logMAR (20/149, Snellen equivalent) to 0.98 logMAR (20/189, Snellen equivalent) in the VMA (⫹) group, but improved from 0.82 logMAR (20/132,
Snellen equivalent) to 0.72 logMAR (20/104, Snellen equivalent) in the VMA (⫺) group (P ⫽ 0.028, repeated-measures ANOVA). In paired comparisons of mean BCVA between baseline and each follow-up visit, the VMA (⫺) group showed significant improvement in BCVA at every follow-up visit (P⬍0.05, paired t test; Table 2). In contrast, the VMA (⫹) group did not show significant visual improvement at any follow-up visit (P⬎0.05, paired t test; Table 2), and although there was a slight trend for improvement over the initial 12 months after treatment, the mean BCVA was actually worse at the final follow-up. The mean CRT significantly decreased compared with baseline at every follow-up in both groups (P⬍0.05, paired t test; Table 3 and Fig 4). At the last follow-up, the mean CRT decreased from baseline by 60.7 and 94.1 m in the VMA (⫹) and VMA (⫺) groups (P ⫽ 0.003 and P⬍0.001), respectively. There was no significant difference of the changes in mean CRT over time between the 2 groups (P ⫽ 0.862, repeated-measures ANOVA; Fig 4), and the changes in mean CRT from baseline were significantly different between the 2 groups only at 6 months (P ⫽ 0.297, 0.147, 0.047, 0.246, and 0.175 at 1, 3, 6, and 12 months, and at the last visit, respectively; repeated-measures ANOVA). The mean changes from baseline in BCVA and CRT are shown in Figure 5. We investigated the correlation between the decrease in CRT and the improvement in BCVA (logMAR) using Pearson correlation analysis. There were significant correlations between the decrease in CRT and improvement in BCVA at all follow-up times in the VMA (⫺) group (r⫽0.322, P ⫽ 0.001; r⫽0.335, P⬍0.001; r⫽0.317, P ⫽ 0.001; r⫽0.399, P⬍0.01; r⫽0.29, P ⫽ 0.01 for 1, 3, 6, and 12 months, and last visit, respectively). In contrast, there was no statistically significant correlation between the decrease in CRT and the improvement in BCVA at 1 and 3 months in the VMA (⫹) group (r⫽0.215, P ⫽ 0.195; r⫽0.282, P ⫽ 0.086, respectively), although there were significant correlations at subsequent follow-up times (r⫽0.405, P ⫽ 0.012; r⫽0.413, P ⫽ 0.010; r⫽0.386, P ⫽ 0.017 for 6 months, 12 months, and last visit, respectively). Visual outcome was divided into 3 groups for analysis. Decrease of more than 3 lines was defined as “worse.” Increase of 3 lines was defined as “improved.” The rest were allocated to “stable.” At month 12, 23.7% (9/38) of VMA (⫹) group patients were in the worsened group, 50% (19/38) were in the stable group, and 26.3% (10/38) were in the improved group. In the VMA (⫺) group, the rates were 13.6% (15/110), 46.4% (51/110), and 40% (44/110), respectively. The VMA (⫹) group showed a tendency of more patients in the worsened group and less improved group, but there was no statistical significance (P ⫽ 0.193, chi-square test).
Table 2. Best-Corrected Visual Acuity of Eyes with and without Vitreomacular Adhesion in Exudative Age-Related Macular Degeneration Treated with Anti-Vascular Endothelial Growth Factor
VMA (⫹) group (n⫽38) Mean ⫾ SD logMAR Snellen equivalent P value* VMA (⫺) group (n⫽110) Mean ⫾ SD logMAR Snellen equivalent P value*
Baseline
1 mo
3 mos
6 mos
12 mos
Last Visit
0.87⫾0.58 20/149
0.81⫾0.58 20/130 0.274
0.82⫾0.55 20/137 0.320
0.77⫾0.50 20/117 0.117
0.83⫾0.56 20/136 0.590
0.98⫾0.57 20/189 0.251
0.82⫾0.50 20/132
0.72⫾0.50 20/105 0.001
0.70⫾0.53 20/101 ⬍0.001
0.69⫾0.55 20/99 0.001
0.67⫾0.55 20/93 ⬍0.001
0.72⫾0.58 20/104 0.024
logMAR ⫽ logarithm of the minimum angle of resolution; SD ⫽ standard deviation; VMA ⫽ vitreomacular adhesion. *Paired Student t test.
104
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF Table 3. Central Retinal Thickness of Eyes with and without Vitreomacular Adhesion in Exudative Age-Related Macular Degeneration Treated with Anti-Vascular Endothelial Growth Factor in Optical Coherence Tomography
VMA (⫹) group (n⫽38) Mean ⫾ SD P value* VMA (⫺) group Mean ⫾ SD (n⫽110) P value*
Baseline
1 mo
3 mos
6 mos
12 mos
Last Visit
278.2⫾150.4
231.9⫾107.0 0.008
215.9⫾89.1 0.006
219.7⫾89.9 0.011
220.9⫾102.6 0.012
217.5⫾100.9 0.003
304.1⫾130.9
233.3⫾111.5 ⬍0.001
206.4⫾79.3 ⬍0.001
197.7⫾83.1 ⬍0.001
216.2⫾115.3 ⬍0.001
210.0⫾89.7 ⬍0.001
SD ⫽ standard deviation; VMA ⫽ vitreomacular adhesion. *Paired Student t test.
At last follow-up, the rates for the worsened, stable, and improved groups were 42.1% (16/38), 31.6% (12/38), and 26.3% (10/38) for the VMA (⫹) group, and 20% (22/110), 41.8% (46/110), and 38.2% (42/110) for the VMA (⫺) group, respectively. There was a statistically significant difference between the 2 groups (P ⫽ 0.026, chi-square test). Representative fundus photographs, fluorescein angiography, and OCT images are shown for 1 patient from each group (Figs 6 and 7). After anti-VEGF treatment, the patient without VMA showed a good response (Fig 6), whereas the patient with VMA showed a poor response (Fig 7). The patient without VMA had a visual acuity of 20/400 at baseline and received 6 intravitreal injections of bevacizumab over 23 months. After treatment, her visual acuity continually improved through subsequent follow-up periods (20/400, 20/800, 20/63, 20/40, and 20/32 at 1, 3, 6, 12, and 23 months, respectively). Baseline CRT was 566 m and decreased in parallel with the improved visual acuity (343, 324, 312, 195, and 154 m, respectively). The patient with VMA received 5 intravitreal injections of bevacizumab over 14 months. His visual acuity was 20/50 at baseline. After treatment, his visual acuity was slightly improved at 1 month (20/40), but decreased progressively through subsequent follow-up periods (20/100, 20/160, and 20/200 at 3, 6, and 12 months, respectively). Baseline CRT was 178 m
Figure 4. Mean CRT determined by OCT after treatment with antiVEGF. A significant decrease in CRT was detected after the first antiVEGF treatment; however, there was no significant difference between the 2 groups over time (P ⫽ 0.862, repeated-measures ANOVA). CRT ⫽ central retinal thickness; OCT ⫽ optical coherence tomography; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
and increased in association with decreased visual acuity (266, 304, 352, and 302 m at 1, 3, 6, and 12 months, respectively). Notably, his visual acuity was slightly improved and CRT decreased after the release of VMA at 14 months (20/160 and 268 m, respectively).
Figure 5. Mean changes in BCVA (A) and CRT (B) of eyes with exudative AMD after treatment with intravitreal anti-VEGF, according to the presence of posterior VMA. There were significant correlations between the decrease in CRT and the improvement in BCVA, with the exception of the VMA (⫹) group at 1 and 3 months. AMD ⫽ age-related macular degeneration; BCVA ⫽ best-corrected visual acuity; CRT ⫽ central retinal thickness; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
105
Ophthalmology Volume 118, Number 1, January 2011
Figure 6. Case 1: A 56-year-old woman with exudative AMD and no VMA in her right eye. Color fundus image (A) with early- and late-phase fluorescein angiogram images (B, C) shown at baseline. Horizontal (left) and vertical (right) OCT scans shown at baseline and at 1, 3, 6, 12, and 23 months after treatment with bevacizumab. AMD ⫽ age-related macular degeneration; OCT ⫽ optical coherence tomography; VMA ⫽ vitreomacular adhesion.
106
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF
Figure 7. Case 2: A 72-year-old man with exudative AMD and VMA in the right eye. Color fundus image (A) with early- and late-phase fluorescein angiogram images (B, C) shown at baseline. Horizontal (left) and vertical (right) OCT scans shown at baseline and at 1, 3, 6, 12, and 14 months after treatment with bevacizumab. AMD ⫽ age-related macular degeneration; OCT ⫽ optical coherence tomography; VMA ⫽ vitreomacular adhesion.
107
Ophthalmology Volume 118, Number 1, January 2011
Discussion This study identified a negative association between VMA and the visual outcome to treatment for exudative AMD with anti-VEGF. Visual prognosis after anti-VEGF treatment for exudative AMD was worse in eyes with posterior VMA. Our results indicate that there may be a subgroup of exudative AMD eyes that do not respond to anti-VEGF therapy, and that VMA may be a morphologic characteristic of resistance to such therapy. Although many studies have focused on determining the optimal dosing and treatment intervals of anti-VEGF therapies through several clinical trials, such as Anti-VEGF antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR); Minimally Classic/Occult trial of the AntiVEGF Antibody Ranibizumab In the Treatment of Neovascular AMD (MARINA); Phase IIIb, Multi-center, Randomized, Double-Masked, Sham Injection-Controlled Study of the Efficacy and Safety of Ranibizumab (PIER); Prospective OCT Imaging of Patients with Neovascular AMD Treated with intra-Ocular Ranibizumab (PrONTO); Safety Assessment of Intravitreal Lucentis for AMD (SAILOR); Study of Ranibizumab in Patients with Subfoveal Choroidal Neovascularization Secondary to AMD (SUSTAIN); and Exploring the value of optical coherence tomography for the management of ranibizumab therapy in AMD (EXCITE),4,18 –21 we suggest that other factors, such as anatomic and genetic characteristics, should be assessed in future studies in patients with exudative AMD who are resistant to anti-VEGF treatment in addition to changing the intervals of anti-VEGF treatment. Several previous studies have suggested an association between vitreous traction and AMD.10,13 Weber-Krause and Eckardt15 reported that posterior vitreous detachment (PVD) is rare in AMD, relative to the age of the patients. Our previous study16 showed that eyes with exudative AMD had a significantly higher incidence of posterior VMA than paired eyes that did not have exudative AMD (P ⫽ 0.0007). Because we compared both eyes in 1 patient, these results were not influenced by genetic and environmental factors. The results of our study are consistent with those of a recent multicenter study in which the incidence of PVD in eyes with nonexudative AMD was 69% (20/29), compared with 21% (6/29) in eyes with active exudative AMD (P ⫽ 0.0002). Vitreomacular adhesion was present in 38% (11/29) of eyes with exudative AMD and in only 10% (3/29) of eyes with nonexudative AMD (P ⫽ 0.008).14 In another study using spectral OCT, hyaloid adhesion was present in 17 eyes with exudative AMD (27.8%), 15 eyes with nonexudative AMD (25.4%), and 8 control eyes (16%), with significant differences between the groups (P ⫽ 0.002). Among the eyes with hyaloid adhesions, vitreomacular traction (VMT) was present in 10 eyes (59%) with exudative AMD, 2 eyes (13%) with nonexudative AMD, and 1 control eye (12%). The presence of VMT was significantly associated with the severity of AMD (P ⫽ 0.0082).11 Several groups have hypothesized that persistent VMA influences the development of CNV through the induction of chronic low-grade inflammation, prevention of diffusion of oxygen and nutrients to the macula, or confinement of
108
proangiogenic cytokines in the macula.22–24 Such mechanisms may also explain the poor response to anti-VEGF treatment for CNV. Our results showed no significant correlation between the decrease in CRT and the improvement in BCVA at 1 and 3 months in the VMA (⫹) group, suggesting that VMA may influence CNV through other mechanisms in addition to increasing retinal thickness by mechanical traction. It is not yet evident whether VMA is a cause or consequence of AMD, although several reports indicate that VMA may be a cause of AMD. Lambert et al25 found an attached vitreous in 8 of 10 patients (80%) with CNV and speculated that VMA induces mechanical traction that contributes to CNV progression. Schmidt et al26 observed that intraoperative findings during vitrectomy showed little liquefaction of the vitreous gel, an incomplete PVD, and remarkably firm attachments at the macula in all cases (10/10). Furthermore, there are reports of CNV regression after surgery for exudative AMD. Vitrectomy was performed in 12 eyes of 11 patients with highly active CNV in whom the posterior vitreous surface remained attached. Six months after surgery, CNV disappeared completely in 2 eyes and showed regression in 6 eyes.27 In another recent study,11 5 patients with evidence of VMT had 25-gauge trans pars plana vitrectomy with hyaloid removal. Four patients experienced an improvement in visual acuity after surgery, as assessed at the end of the follow-up period, and the visual acuity remained stable in 1 eye. These results suggest that surgery may be considered as a treatment option in such cases, although we do not know whether the visual improvement was secondary to the release of VMT or caused by other factors, such as improved retinal oxygenation and reduced inflammation. In regard to the number of injections, in the PrONTO trials featuring OCT-guided variable-dose ranibizumab regimens, patients received an average of 5.6 injections over 12 months.20 In the present study, subjects received a mean of 3.66 injections within 12 months and 4.48 injections until the end of follow-up, relatively few injections compared with the PrONTO trials. In some cases the patients could not follow our recommendation because of personal difficulties, such as economic problems. Also, we may not have followed up on the patients as frequently as in prospective clinical trials. However, in our study, all patients were treated with identical therapeutic protocols, and there was no significant difference in the number of injections between the 2 groups in our study (3.87 vs. 3.58 at 12 months, P ⫽ 0.437).
Study Limitations This study has several limitations, including the retrospective nature of the work, the lack of formal grading of lens opacity changes using a standardized grading system, and the lack of ultrasound examination to detect PVD. However, because we excluded patients with comorbid diseases that affect visual acuity, including significant lens opacity, the presence of cataract is not likely to have influenced differences in visual acuity between the 2 groups. The OCT examination alone does not detect complete PVD or
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF completely attached vitreous. However, we believe that focal VMT is more important in the pathogenesis of exudative AMD, in agreement with a concept of “anomalous PVD” proposed by Sebag.28,29 Because OCT is recognized as the best tool to evaluate focal vitreoretinal adhesion at present, the absence of an ultrasound examination is unlikely to influence our results. Future prospective studies using spectral domain OCT may be helpful for more accurate detection of VMA and analysis of configuration.30 We also excluded patients who had previously been treated for exudative AMD, allowing us to study only patients who were newly diagnosed. In pharmacokinetic studies of bevacizumab, detectable, but small, amounts of bevacizumab were detected in the serum and the paired uninjected eye.31,32 By including patients who were treated with intravitreal anti-VEGF monotherapy in only 1 eye, we could exclude any effects of anti-VEGF injections into the fellow eye. As the study by Diloreto et al33 showed that 58% (113/195) of the articles reviewed reported only the final visual acuity, this study also used final vision for analysis of visual outcome. This is one of the well-known limitations of retrospective studies,33,34 and we are well aware that it is ideal to report vision at predetermined time points. However, in our study, the vision showed improvement for 6 months and then showed the difference between 2 groups after 12 months. Therefore, we believe that it is meaningful to analyze and present the results after 12 months. In addition, this study was not a small case series, and the follow-up period was at least 12 months for both groups. The mean follow-up time for both groups was 19.8 months and 21.8 months, respectively, which showed no statistically significant difference (P ⫽ 0.193, Table 1). In addition, the results might be influenced by the fact that the VMA (⫹) group had fewer patients, although this was determined by the incidence of VMA. In conclusion, the results of our study indicate that posterior VMA was associated with an inferior visual outcome after intravitreal anti-VEGF treatment for exudative AMD. Our results suggest that chronic tractional forces associated with VMA may antagonize the effect of anti-VEGF treatment and cause this poor response. Further controlled prospective studies are required to assess the relationship between VMA and exudative AMD.
References 1. Klein R, Peto T, Bird A, Vannewkirk MR. The epidemiology of age-related macular degeneration. Am J Ophthalmol 2004; 137:486 –95. 2. Nowak JZ. Age-related macular degeneration (AMD): pathogenesis and therapy. Pharmacol Rep 2006;58:353– 63. 3. Avery RL, Pieramici DJ, Rabena MD, et al. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006;113:363–72. 4. Rosenfeld PJ, Brown DM, Heier JS, et al, MARINA Study Group. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006;355:1419 –31. 5. Freeman WR, Falkenstein I. Avastin and new treatments for AMD: where are we? Retina 2006;26:853– 8.
6. Lux A, Llacer H, Heussen FM, Joussen AM. Non-responders to bevacizumab (Avastin) therapy of choroidal neovascular lesions. Br J Ophthalmol 2007;91:1318 –22. 7. Brantley MA Jr, Edelstein SL, King JM, et al. Association of complement factor H and LOC387715 genotypes with response of exudative age-related macular degeneration to photodynamic therapy. Eye (Lond) 2009;23:626 –31. 8. Brantley MA Jr, Fang AM, King JM, et al. Association of complement factor H and LOC387715 genotypes with response of exudative age-related macular degeneration to intravitreal bevacizumab. Ophthalmology 2007;114:2168 –73. 9. Goverdhan SV, Hannan S, Newsom RB, et al. An analysis of the CFH Y402H genotype in AMD patients and controls from the UK, and response to PDT treatment. Eye (Lond) 2008;22: 849 –54. 10. Krebs I, Brannath W, Glittenberg C, et al. Posterior vitreomacular adhesion: a potential risk factor for exudative agerelated macular degeneration? Am J Ophthalmol 2007;144: 741– 6. 11. Mojana F, Cheng L, Bartsch DU, et al. The role of abnormal vitreomacular adhesion in age-related macular degeneration: spectral optical coherence tomography and surgical results. Am J Ophthalmol 2008;146:218 –27. 12. Ondes F, Yilmaz G, Acar MA, et al. Role of the vitreous in age-related macular degeneration. Jpn J Ophthalmol 2000;44: 91–3. 13. Quaranta-El Maftouhi M, Mauget-Faysse M. Anomalous vitreoretinal adhesions in patients with exudative age-related macular degeneration: an OCT study. Eur J Ophthalmol 2006; 16:134 –7. 14. Robison CD, Krebs I, Binder S, et al. Vitreomacular adhesion in active and end-stage age-related macular degeneration. Am J Ophthalmol 2009;148:79 – 82. 15. Weber-Krause B, Eckardt U. Incidence of posterior vitreous detachment in eyes with and without age-related macular degeneration: an ultrasonic study [in German]. Ophthalmologe 1996;93:660 –5. 16. Lee SJ, Lee CS, Koh HJ. Posterior vitreomacular adhesion and risk of exudative age-related macular degeneration: paired eye study. Am J Ophthalmol 2009;147:621– 6. 17. Holladay JT. Proper method for calculating average visual acuity. J Refract Surg 1997;13:388 –91. 18. Brown DM, Kaiser PK, Michels M, et al, ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular agerelated macular degeneration. N Engl J Med 2006;355:1432– 44. 19. Brown DM, Michels M, Kaiser PK, et al, ANCHOR Study Group. Ranibizumab versus verteporfin photodynamic therapy for neovascular age-related macular degeneration: two-year results of the ANCHOR Study. Ophthalmology 2009;116: 57– 65. 20. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An optical coherence tomography-guided, variable dosing regimen with intravitreal ranibizumab (Lucentis) for neovascular age-related macular degeneration. Am J Ophthalmol 2007; 143:566 – 83. 21. Regillo CD, Brown DM, Abraham P, et al, PIER Study Group. Randomized, double-masked, sham-controlled trial of ranibizumab for neovascular age-related macular degeneration: PIER Study year 1. Am J Ophthalmol 2008; 145:239 – 48. 22. Anderson DH, Mullins RF, Hageman GS, Johnson LV. A role for local inflammation in the formation of drusen in the aging eye. Am J Ophthalmol 2002;134:411–31.
109
Ophthalmology Volume 118, Number 1, January 2011 23. Donoso LA, Kim D, Frost A, et al. The role of inflammation in the pathogenesis of age-related macular degeneration. Surv Ophthalmol 2006;51:137–52. 24. Zarbin MA. Current concepts in the pathogenesis of agerelated macular degeneration. Arch Ophthalmol 2004;122: 598 – 614. 25. Lambert HM, Capone A Jr, Aaberg TM, et al. Surgical excision of subfoveal neovascular membranes in age-related macular degeneration. Am J Ophthalmol 1992;113:257– 62. 26. Schmidt JC, Mennel S, Horle S, Meyer CH. High incidence of vitreomacular traction in recurrent choroidal neovascularisation after repeated photodynamic therapy. Br J Ophthalmol 2006;90:1361–2. 27. Ikeda T, Sawa H, Koizumi K, et al. Pars plana vitrectomy for regression of choroidal neovascularization with age-related macular degeneration. Acta Ophthalmol Scand 2000;78:460 – 4. 28. Sebag J. Anomalous posterior vitreous detachment: a unifying concept in vitreo-retinal disease. Graefes Arch Clin Exp Ophthalmol 2004;242:690 – 8.
29. Sebag J. Vitreoschisis. Graefes Arch Clin Exp Ophthalmol 2008;246:329 –32. 30. Chang LK, Fine HF, Spaide RF, et al. Ultrastructural correlation of spectral-domain optical coherence tomographic findings in vitreomacular traction syndrome. Am J Ophthalmol 2008;146:121–7. 31. Bakri SJ, Snyder MR, Reid JM, et al. Pharmacokinetics of intravitreal bevacizumab (Avastin). Ophthalmology 2007;114: 855–9. 32. Nomoto H, Shiraga F, Kuno N, et al. Pharmacokinetics of bevacizumab after topical, subconjunctival and intravitreal administration in rabbits. Invest Ophthalmol Vis Sci 2009;50: 4807–13. 33. DiLoreto DA Jr, Bressler NM, Bressler SB, Schachat AP. Use of best and final visual acuity outcomes in ophthalmological research. Arch Ophthalmol 2003;121:1586 –90. 34. Jabs DA. Improving the reporting of clinical case series. Am J Ophthalmol 2005;139:900 –5.
Footnotes and Financial Disclosures Originally received: September 30, 2009. Final revision: April 5, 2010. Accepted: April 8, 2010. Available online: August 3, 2010.
This study was presented as a free paper at: the Asia Pacific Academy of Ophthalmology–American Academy of Ophthalmology Joint Congress, May 16, 2009, Bali, Indonesia. Manuscript no. 2009-1360.
1
Department of Ophthalmology, Dongguk University Ilsan Hospital, Dongguk University School of Medicine, Gyeonggido, Korea. 2
Department of Ophthalmology, Institute of Vision Research, Yonsei University College of Medicine, Seoul, Korea.
110
Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Hyoung Jun Koh, MD, Department of Ophthalmology, Yonsei University College of Medicine, No. 134 Shinchon-dong, Seodaemun-gu, 120-752, Seoul, Korea. E-mail: [email protected].
Age-related macular degeneration (AMD) is the leading cause of severe visual loss in industrialized countries1; however, despite intensive basic and clinical research, the pathogenesis and risk factors for AMD are incompletely characterized.2 In recent years, the advent of anti-vascular endothelial growth factor (VEGF) therapies, such as ranibizumab and bevacizumab, has revolutionized neovascular AMD treatment,3,4 and anti-VEGF has become the standard treatment for choroidal neovascularization (CNV) with a better visual outcome than previous therapies, such as photodynamic therapy.4,5 However, one study reported that up to 45% of cases (20/44 eyes) were nonresponders showing resistance to anti-VEGF.6 A current focus of anti-VEGF treatment is how to determine which eyes will respond to treatment. To date, 3 genetic studies on the response to © 2011 by the American Academy of Ophthalmology Published by Elsevier Inc.
treatment for wet AMD have shown that specific genotypes for complement factor H and LOC genes are associated with treatment response.7–9 Previous studies have described the relationship between the posterior vitreous and the macula in AMD and have suggested that vitreomacular adhesion (VMA) plays an important role in the development of exudative AMD.10 –15 The incidence of posterior VMA is higher in eyes with exudative AMD compared with eyes with non-exudative AMD and normal controls10; however, because previous studies did not compare the normal eye and the eye with exudative AMD in individual patients with unilateral AMD, their results may be affected by genetic and environmental factors. In a recent paired eye study, we controlled confounding variables by selecting only patients with unilateral ISSN 0161-6420/11/$–see front matter doi:10.1016/j.ophtha.2010.04.015
101
Ophthalmology Volume 118, Number 1, January 2011 exudative AMD and showed that eyes with exudative AMD had a significantly higher incidence of posterior VMA than paired normal eyes (P ⫽ 0.0007).16 This result indicates that VMA is a possible risk factor for exudative AMD. In another recent study, Mojana et al11 reported improvement in visual acuity after 25-gauge trans pars plana vitrectomy with hyaloid removal in 5 patients who had a history of demonstrable VMA and poorly responsive CNV despite aggressive anti-VEGF therapy. We postulated that a subpopulation of exudative AMD cases do not respond to anti-VEGF therapy and that VMA may play a role in this resistance to therapy. This study compares the visual outcome of anti-VEGF treatment for exudative AMD with or without posterior VMA documented by optical coherence tomography (OCT).
Patients and Methods We retrospectively reviewed data for consecutive patients who were diagnosed with exudative AMD at the vitreoretinal service clinic of the Yonsei University Eye and Ear, Nose, and Throat Hospital (Seoul, Korea) from October 2005 to December 2007. This retrospective study was approved by the institutional review board of the Yonsei University College of Medicine and adhered to the tenets of the Declaration of Helsinki. We used the hospital clinical database (CDR software, Yonsei University Severance Hospital, Seoul, Korea) to identify patients with exudative AMD who underwent intravitreal anti-VEGF injections. Data were abstracted from the charts of 251 consecutive patients, and 148 patients (148 eyes) who met the eligibility criteria were included in the analysis.
Inclusion and Exclusion Criteria Inclusion criteria were (a) age ⬎50 years; (b) exudative AMD proven by fundus photograph and fluorescein angiography; (c) treatment with intravitreal anti-VEGF monotherapy using ranibizumab or bevacizumab in 1 eye; (d) follow-up for a minimum of 12 months; (e) availability of results from an OCT examination at baseline, 1, 3, 6, and 12 months, and last follow-up; and (f) availability of best-corrected visual acuity (BCVA) records at baseline, 1, 3, 6, and 12 months, and last follow-up after treatment. Exclusion criteria were (a) intraocular surgery, including cataract extraction at baseline or during follow-up; (b) history of treatment for AMD, such as verteporfin photodynamic therapy and intravitreal anti-VEGF therapy; (c) myopia of more than 3 diopters; (d) CNV due to any cause other than AMD, such as angioid streak, choroidal rupture, and ocular histoplasmosis syndrome; (e) evidence of diabetic retinopathy; (f) presence of inflammatory ocular disease; and (g) presence of comorbid ocular conditions that might affect visual acuity.
Diagnosis of Age-Related Macular Degeneration Fluorescein angiograms obtained using the Heidelberg Retina Angiograph (Heidelberg Engineering, Heidelberg, Germany) were reviewed to verify diagnoses. The location of the neovascular complex was classified as subfoveal, juxtafoveal, or extrafoveal. The lesion composition was classified into 2 categories: predominant classic and occult (occult with minimally classic or pure occult).
102
Anti-Vascular Endothelial Growth Factor Treatment Intravitreal anti-VEGF injections were performed using bevacizumab (1.25 mg/0.05 ml; Avastin, Genentech Inc., South San Francisco, CA) or ranibizumab (0.5 mg/0.05 ml; Lucentis, Genentech Inc.). After the 3 standard monthly intravitreal injections of anti-VEGF, additional injections were administered at the physician’s discretion if any of the following were observed: (1) visual acuity loss, (2) evidence of persistent fluid on OCT at least 1 month after the previous injection, (3) newly developed macular hemorrhage. The total number of injections within 12 months after the initial treatment and until the end of the follow-up was recorded.
Detection of Vitreomacular Adhesion Using Optical Coherence Tomography A commercially available OCT instrument (Stratus OCT3, Zeiss Humphrey, San Leandro, CA) was used for analysis. The macula was scanned in the horizontal and vertical meridians using the cross-hair pattern or 6 radial lines at internals of 30 degrees with a scan length of 4 to 6 mm centered through the fovea and at any sites of suspected macular lesion, as determined by simultaneous evaluation of the red-free image on the OCT computer monitor. The presence of VMA was assessed on the OCT computer monitor and defined as attachment of the posterior hyaloid line to the inner retinal surface (Fig 1). The presence of VMA was determined by 2 independent blinded observers (SJL, HJK), and any discrepancies were resolved by open discussion.
Optical Coherence Tomography Central Retinal Thickness Measurement Central retinal thickness (CRT) was defined as the mean value of retinal thickness of the central points in the horizontal and vertical scanning lines. We used automated measurements provided by the Stratus OCT software to obtain the thickness (Fig 2A). In case of segmentation error, manual measurements were performed using the electronic calipers of the Stratus OCT software (Fig 2B).
Visual Acuity Measurement Best-corrected visual acuity was measured with a Snellen chart at baseline; at 1, 3, 6, and 12 months; and at last follow-up after initial anti-VEGF treatment. The Snellen BCVA was converted to logarithm of the minimum angle of resolution (logMAR) values. The logMAR value for counting fingers visual acuity was assigned as ⫹2.0 logMAR (0.01) according to methods published by Holladay.17
Statistical Analysis Statistical analysis was performed using SPSS software version 13.0 (SPSS Inc., Chicago, IL). The change of BCVA and CRT over time was compared between the 2 subgroups using repeatedmeasures analysis of variance (ANOVA). Categoric data were assessed using the chi-square test, and continuous variables were compared with the Student t test. In addition, paired t tests were used to compare follow-up and baseline data within a treatment group. The association between the change in CRT and BCVA outcomes at different time points during the study were assessed using the Pearson correlation analysis. P⬍0.05 was considered statistically significant.
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF
Figure 1. Optical coherence tomography images showing eyes with exudative AMD with (A, C) and without VMA (B, D) (arrows). AMD ⫽ age-related macular degeneration; VMA ⫽ vitreomacular adhesion.
Results Baseline Characteristics A total of 148 patients (65 male, 83 female) with exudative AMD who underwent treatment with intravitreal anti-VEGF in 1 eye were enrolled in the study. The average age for the entire study
group was 68.2⫾7.4 years. The mean follow-up time was 21.3⫾7.9 months (range, 12– 40 months). Subjects received a mean of 3.66 injections within 12 months and 4.48 injections until the end of the follow-up. Ranibizumab was administered in 28.2% of the total injections (187/663). After OCT analysis, we categorized eyes with exudative AMD into 2 subgroups according to the presence of posterior VMA (Table 1). The VMA (⫹) group contained 38 eyes, and the VMA
Table 1. Baseline Characteristics of Patients with and without Vitreomacular Adhesion VMA (ⴙ) VMA (ⴚ) P Group (n ⴝ 38) Group (n ⴝ 110) Value Age, yrs Gender (M/F) Follow-up time (mos) Baseline BCVA (logMAR) (Snellen equivalent) Baseline CRT (m) Treatment No. ⬍12 mos Total CNV location (%) Subfoveal Juxtafoveal Extrafoveal CNV type (%) Classic Occult
Figure 2. Example of correct segmentation by automated measurement (A), CRT (arrow), and manual measurement of retinal thickness using calipers because of segmentation error (arrow) (B). CRT ⫽ central retinal thickness.
68.1⫾8.5 19/19 19.8⫾8.4 0.87⫾0.58 (20/149) 278.2⫾150.4
68.3⫾7.1 46/64 21.8⫾7.8 0.82⫾0.50 (20/132) 304.1⫾130.9
0.871* 0.381† 0.193* 0.605* 0.593* 0.313*
3.87⫾1.77 4.21⫾1.74
3.58⫾2.39 4.57⫾3.53
0.437* 0.411* 0.199†
25 (65.8) 9 (23.7) 4 (10.5)
54 (49.1) 41 (37.3) 15 (13.6)
11 (40.7) 27 (59.3)
26 (23.6) 84 (76.4)
0.515†
BCVA ⫽ best-corrected visual acuity; CNV ⫽ choroidal neovascularization; CRT ⫽ central retinal thickness; logMAR ⫽ logarithm of the minimum angle of resolution; VMA ⫽ vitreomacular adhesion. *Student t test. † Chi-square test.
103
Ophthalmology Volume 118, Number 1, January 2011
Figure 3. Mean BCVA given as logMAR units after treatment with anti-VEGF. Mean BCVA significantly decreased over time in the VMA (⫹) group compared with the VMA (⫺) group (P ⫽ 0.039, repeatedmeasures ANOVA). BCVA ⫽ best corrected visual acuity; logMAR ⫽ logarithm of the minimum angle of resolution; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
(⫺) group contained 110 eyes. Baseline BCVA was 0.87⫾0.58 logMAR (Snellen equivalent, 20/149) in the VMA (⫹) group and 0.82⫾0.50 logMAR (Snellen equivalent, 20/132) in the VMA (⫺) group. There was no significant difference in baseline BCVA between the groups (t test, P ⫽ 0.593). Baseline CRT was 278.2⫾150.4 m in the VMA (⫹) group and 304.1⫾130.9 m in the VMA (⫺) group, with no significant difference between the 2 groups (t test, P ⫽ 0.313). There were also no significant differences between the 2 groups with respect to age, gender, follow-up time, treatment numbers, CNV location, and CNV type (Table 1).
Visual and Optical Coherence Tomography Outcomes Mean changes in BCVA after anti-VEGF treatment are shown in Figure 3. Mean BCVA significantly decreased over time in the VMA (⫹) group compared with the VMA (⫺) group (P ⫽ 0.039, repeated-measures ANOVA). Between baseline and the last follow-up, the mean BCVA worsened from 0.87 logMAR (20/149, Snellen equivalent) to 0.98 logMAR (20/189, Snellen equivalent) in the VMA (⫹) group, but improved from 0.82 logMAR (20/132,
Snellen equivalent) to 0.72 logMAR (20/104, Snellen equivalent) in the VMA (⫺) group (P ⫽ 0.028, repeated-measures ANOVA). In paired comparisons of mean BCVA between baseline and each follow-up visit, the VMA (⫺) group showed significant improvement in BCVA at every follow-up visit (P⬍0.05, paired t test; Table 2). In contrast, the VMA (⫹) group did not show significant visual improvement at any follow-up visit (P⬎0.05, paired t test; Table 2), and although there was a slight trend for improvement over the initial 12 months after treatment, the mean BCVA was actually worse at the final follow-up. The mean CRT significantly decreased compared with baseline at every follow-up in both groups (P⬍0.05, paired t test; Table 3 and Fig 4). At the last follow-up, the mean CRT decreased from baseline by 60.7 and 94.1 m in the VMA (⫹) and VMA (⫺) groups (P ⫽ 0.003 and P⬍0.001), respectively. There was no significant difference of the changes in mean CRT over time between the 2 groups (P ⫽ 0.862, repeated-measures ANOVA; Fig 4), and the changes in mean CRT from baseline were significantly different between the 2 groups only at 6 months (P ⫽ 0.297, 0.147, 0.047, 0.246, and 0.175 at 1, 3, 6, and 12 months, and at the last visit, respectively; repeated-measures ANOVA). The mean changes from baseline in BCVA and CRT are shown in Figure 5. We investigated the correlation between the decrease in CRT and the improvement in BCVA (logMAR) using Pearson correlation analysis. There were significant correlations between the decrease in CRT and improvement in BCVA at all follow-up times in the VMA (⫺) group (r⫽0.322, P ⫽ 0.001; r⫽0.335, P⬍0.001; r⫽0.317, P ⫽ 0.001; r⫽0.399, P⬍0.01; r⫽0.29, P ⫽ 0.01 for 1, 3, 6, and 12 months, and last visit, respectively). In contrast, there was no statistically significant correlation between the decrease in CRT and the improvement in BCVA at 1 and 3 months in the VMA (⫹) group (r⫽0.215, P ⫽ 0.195; r⫽0.282, P ⫽ 0.086, respectively), although there were significant correlations at subsequent follow-up times (r⫽0.405, P ⫽ 0.012; r⫽0.413, P ⫽ 0.010; r⫽0.386, P ⫽ 0.017 for 6 months, 12 months, and last visit, respectively). Visual outcome was divided into 3 groups for analysis. Decrease of more than 3 lines was defined as “worse.” Increase of 3 lines was defined as “improved.” The rest were allocated to “stable.” At month 12, 23.7% (9/38) of VMA (⫹) group patients were in the worsened group, 50% (19/38) were in the stable group, and 26.3% (10/38) were in the improved group. In the VMA (⫺) group, the rates were 13.6% (15/110), 46.4% (51/110), and 40% (44/110), respectively. The VMA (⫹) group showed a tendency of more patients in the worsened group and less improved group, but there was no statistical significance (P ⫽ 0.193, chi-square test).
Table 2. Best-Corrected Visual Acuity of Eyes with and without Vitreomacular Adhesion in Exudative Age-Related Macular Degeneration Treated with Anti-Vascular Endothelial Growth Factor
VMA (⫹) group (n⫽38) Mean ⫾ SD logMAR Snellen equivalent P value* VMA (⫺) group (n⫽110) Mean ⫾ SD logMAR Snellen equivalent P value*
Baseline
1 mo
3 mos
6 mos
12 mos
Last Visit
0.87⫾0.58 20/149
0.81⫾0.58 20/130 0.274
0.82⫾0.55 20/137 0.320
0.77⫾0.50 20/117 0.117
0.83⫾0.56 20/136 0.590
0.98⫾0.57 20/189 0.251
0.82⫾0.50 20/132
0.72⫾0.50 20/105 0.001
0.70⫾0.53 20/101 ⬍0.001
0.69⫾0.55 20/99 0.001
0.67⫾0.55 20/93 ⬍0.001
0.72⫾0.58 20/104 0.024
logMAR ⫽ logarithm of the minimum angle of resolution; SD ⫽ standard deviation; VMA ⫽ vitreomacular adhesion. *Paired Student t test.
104
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF Table 3. Central Retinal Thickness of Eyes with and without Vitreomacular Adhesion in Exudative Age-Related Macular Degeneration Treated with Anti-Vascular Endothelial Growth Factor in Optical Coherence Tomography
VMA (⫹) group (n⫽38) Mean ⫾ SD P value* VMA (⫺) group Mean ⫾ SD (n⫽110) P value*
Baseline
1 mo
3 mos
6 mos
12 mos
Last Visit
278.2⫾150.4
231.9⫾107.0 0.008
215.9⫾89.1 0.006
219.7⫾89.9 0.011
220.9⫾102.6 0.012
217.5⫾100.9 0.003
304.1⫾130.9
233.3⫾111.5 ⬍0.001
206.4⫾79.3 ⬍0.001
197.7⫾83.1 ⬍0.001
216.2⫾115.3 ⬍0.001
210.0⫾89.7 ⬍0.001
SD ⫽ standard deviation; VMA ⫽ vitreomacular adhesion. *Paired Student t test.
At last follow-up, the rates for the worsened, stable, and improved groups were 42.1% (16/38), 31.6% (12/38), and 26.3% (10/38) for the VMA (⫹) group, and 20% (22/110), 41.8% (46/110), and 38.2% (42/110) for the VMA (⫺) group, respectively. There was a statistically significant difference between the 2 groups (P ⫽ 0.026, chi-square test). Representative fundus photographs, fluorescein angiography, and OCT images are shown for 1 patient from each group (Figs 6 and 7). After anti-VEGF treatment, the patient without VMA showed a good response (Fig 6), whereas the patient with VMA showed a poor response (Fig 7). The patient without VMA had a visual acuity of 20/400 at baseline and received 6 intravitreal injections of bevacizumab over 23 months. After treatment, her visual acuity continually improved through subsequent follow-up periods (20/400, 20/800, 20/63, 20/40, and 20/32 at 1, 3, 6, 12, and 23 months, respectively). Baseline CRT was 566 m and decreased in parallel with the improved visual acuity (343, 324, 312, 195, and 154 m, respectively). The patient with VMA received 5 intravitreal injections of bevacizumab over 14 months. His visual acuity was 20/50 at baseline. After treatment, his visual acuity was slightly improved at 1 month (20/40), but decreased progressively through subsequent follow-up periods (20/100, 20/160, and 20/200 at 3, 6, and 12 months, respectively). Baseline CRT was 178 m
Figure 4. Mean CRT determined by OCT after treatment with antiVEGF. A significant decrease in CRT was detected after the first antiVEGF treatment; however, there was no significant difference between the 2 groups over time (P ⫽ 0.862, repeated-measures ANOVA). CRT ⫽ central retinal thickness; OCT ⫽ optical coherence tomography; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
and increased in association with decreased visual acuity (266, 304, 352, and 302 m at 1, 3, 6, and 12 months, respectively). Notably, his visual acuity was slightly improved and CRT decreased after the release of VMA at 14 months (20/160 and 268 m, respectively).
Figure 5. Mean changes in BCVA (A) and CRT (B) of eyes with exudative AMD after treatment with intravitreal anti-VEGF, according to the presence of posterior VMA. There were significant correlations between the decrease in CRT and the improvement in BCVA, with the exception of the VMA (⫹) group at 1 and 3 months. AMD ⫽ age-related macular degeneration; BCVA ⫽ best-corrected visual acuity; CRT ⫽ central retinal thickness; VEGF ⫽ vascular endothelial growth factor; VMA ⫽ vitreomacular adhesion.
105
Ophthalmology Volume 118, Number 1, January 2011
Figure 6. Case 1: A 56-year-old woman with exudative AMD and no VMA in her right eye. Color fundus image (A) with early- and late-phase fluorescein angiogram images (B, C) shown at baseline. Horizontal (left) and vertical (right) OCT scans shown at baseline and at 1, 3, 6, 12, and 23 months after treatment with bevacizumab. AMD ⫽ age-related macular degeneration; OCT ⫽ optical coherence tomography; VMA ⫽ vitreomacular adhesion.
106
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF
Figure 7. Case 2: A 72-year-old man with exudative AMD and VMA in the right eye. Color fundus image (A) with early- and late-phase fluorescein angiogram images (B, C) shown at baseline. Horizontal (left) and vertical (right) OCT scans shown at baseline and at 1, 3, 6, 12, and 14 months after treatment with bevacizumab. AMD ⫽ age-related macular degeneration; OCT ⫽ optical coherence tomography; VMA ⫽ vitreomacular adhesion.
107
Ophthalmology Volume 118, Number 1, January 2011
Discussion This study identified a negative association between VMA and the visual outcome to treatment for exudative AMD with anti-VEGF. Visual prognosis after anti-VEGF treatment for exudative AMD was worse in eyes with posterior VMA. Our results indicate that there may be a subgroup of exudative AMD eyes that do not respond to anti-VEGF therapy, and that VMA may be a morphologic characteristic of resistance to such therapy. Although many studies have focused on determining the optimal dosing and treatment intervals of anti-VEGF therapies through several clinical trials, such as Anti-VEGF antibody for the Treatment of Predominantly Classic Choroidal Neovascularization in AMD (ANCHOR); Minimally Classic/Occult trial of the AntiVEGF Antibody Ranibizumab In the Treatment of Neovascular AMD (MARINA); Phase IIIb, Multi-center, Randomized, Double-Masked, Sham Injection-Controlled Study of the Efficacy and Safety of Ranibizumab (PIER); Prospective OCT Imaging of Patients with Neovascular AMD Treated with intra-Ocular Ranibizumab (PrONTO); Safety Assessment of Intravitreal Lucentis for AMD (SAILOR); Study of Ranibizumab in Patients with Subfoveal Choroidal Neovascularization Secondary to AMD (SUSTAIN); and Exploring the value of optical coherence tomography for the management of ranibizumab therapy in AMD (EXCITE),4,18 –21 we suggest that other factors, such as anatomic and genetic characteristics, should be assessed in future studies in patients with exudative AMD who are resistant to anti-VEGF treatment in addition to changing the intervals of anti-VEGF treatment. Several previous studies have suggested an association between vitreous traction and AMD.10,13 Weber-Krause and Eckardt15 reported that posterior vitreous detachment (PVD) is rare in AMD, relative to the age of the patients. Our previous study16 showed that eyes with exudative AMD had a significantly higher incidence of posterior VMA than paired eyes that did not have exudative AMD (P ⫽ 0.0007). Because we compared both eyes in 1 patient, these results were not influenced by genetic and environmental factors. The results of our study are consistent with those of a recent multicenter study in which the incidence of PVD in eyes with nonexudative AMD was 69% (20/29), compared with 21% (6/29) in eyes with active exudative AMD (P ⫽ 0.0002). Vitreomacular adhesion was present in 38% (11/29) of eyes with exudative AMD and in only 10% (3/29) of eyes with nonexudative AMD (P ⫽ 0.008).14 In another study using spectral OCT, hyaloid adhesion was present in 17 eyes with exudative AMD (27.8%), 15 eyes with nonexudative AMD (25.4%), and 8 control eyes (16%), with significant differences between the groups (P ⫽ 0.002). Among the eyes with hyaloid adhesions, vitreomacular traction (VMT) was present in 10 eyes (59%) with exudative AMD, 2 eyes (13%) with nonexudative AMD, and 1 control eye (12%). The presence of VMT was significantly associated with the severity of AMD (P ⫽ 0.0082).11 Several groups have hypothesized that persistent VMA influences the development of CNV through the induction of chronic low-grade inflammation, prevention of diffusion of oxygen and nutrients to the macula, or confinement of
108
proangiogenic cytokines in the macula.22–24 Such mechanisms may also explain the poor response to anti-VEGF treatment for CNV. Our results showed no significant correlation between the decrease in CRT and the improvement in BCVA at 1 and 3 months in the VMA (⫹) group, suggesting that VMA may influence CNV through other mechanisms in addition to increasing retinal thickness by mechanical traction. It is not yet evident whether VMA is a cause or consequence of AMD, although several reports indicate that VMA may be a cause of AMD. Lambert et al25 found an attached vitreous in 8 of 10 patients (80%) with CNV and speculated that VMA induces mechanical traction that contributes to CNV progression. Schmidt et al26 observed that intraoperative findings during vitrectomy showed little liquefaction of the vitreous gel, an incomplete PVD, and remarkably firm attachments at the macula in all cases (10/10). Furthermore, there are reports of CNV regression after surgery for exudative AMD. Vitrectomy was performed in 12 eyes of 11 patients with highly active CNV in whom the posterior vitreous surface remained attached. Six months after surgery, CNV disappeared completely in 2 eyes and showed regression in 6 eyes.27 In another recent study,11 5 patients with evidence of VMT had 25-gauge trans pars plana vitrectomy with hyaloid removal. Four patients experienced an improvement in visual acuity after surgery, as assessed at the end of the follow-up period, and the visual acuity remained stable in 1 eye. These results suggest that surgery may be considered as a treatment option in such cases, although we do not know whether the visual improvement was secondary to the release of VMT or caused by other factors, such as improved retinal oxygenation and reduced inflammation. In regard to the number of injections, in the PrONTO trials featuring OCT-guided variable-dose ranibizumab regimens, patients received an average of 5.6 injections over 12 months.20 In the present study, subjects received a mean of 3.66 injections within 12 months and 4.48 injections until the end of follow-up, relatively few injections compared with the PrONTO trials. In some cases the patients could not follow our recommendation because of personal difficulties, such as economic problems. Also, we may not have followed up on the patients as frequently as in prospective clinical trials. However, in our study, all patients were treated with identical therapeutic protocols, and there was no significant difference in the number of injections between the 2 groups in our study (3.87 vs. 3.58 at 12 months, P ⫽ 0.437).
Study Limitations This study has several limitations, including the retrospective nature of the work, the lack of formal grading of lens opacity changes using a standardized grading system, and the lack of ultrasound examination to detect PVD. However, because we excluded patients with comorbid diseases that affect visual acuity, including significant lens opacity, the presence of cataract is not likely to have influenced differences in visual acuity between the 2 groups. The OCT examination alone does not detect complete PVD or
Lee and Koh 䡠 Vitreomacular Adhesion in AMD Treatment with Anti-VEGF completely attached vitreous. However, we believe that focal VMT is more important in the pathogenesis of exudative AMD, in agreement with a concept of “anomalous PVD” proposed by Sebag.28,29 Because OCT is recognized as the best tool to evaluate focal vitreoretinal adhesion at present, the absence of an ultrasound examination is unlikely to influence our results. Future prospective studies using spectral domain OCT may be helpful for more accurate detection of VMA and analysis of configuration.30 We also excluded patients who had previously been treated for exudative AMD, allowing us to study only patients who were newly diagnosed. In pharmacokinetic studies of bevacizumab, detectable, but small, amounts of bevacizumab were detected in the serum and the paired uninjected eye.31,32 By including patients who were treated with intravitreal anti-VEGF monotherapy in only 1 eye, we could exclude any effects of anti-VEGF injections into the fellow eye. As the study by Diloreto et al33 showed that 58% (113/195) of the articles reviewed reported only the final visual acuity, this study also used final vision for analysis of visual outcome. This is one of the well-known limitations of retrospective studies,33,34 and we are well aware that it is ideal to report vision at predetermined time points. However, in our study, the vision showed improvement for 6 months and then showed the difference between 2 groups after 12 months. Therefore, we believe that it is meaningful to analyze and present the results after 12 months. In addition, this study was not a small case series, and the follow-up period was at least 12 months for both groups. The mean follow-up time for both groups was 19.8 months and 21.8 months, respectively, which showed no statistically significant difference (P ⫽ 0.193, Table 1). In addition, the results might be influenced by the fact that the VMA (⫹) group had fewer patients, although this was determined by the incidence of VMA. In conclusion, the results of our study indicate that posterior VMA was associated with an inferior visual outcome after intravitreal anti-VEGF treatment for exudative AMD. Our results suggest that chronic tractional forces associated with VMA may antagonize the effect of anti-VEGF treatment and cause this poor response. Further controlled prospective studies are required to assess the relationship between VMA and exudative AMD.
References 1. Klein R, Peto T, Bird A, Vannewkirk MR. The epidemiology of age-related macular degeneration. Am J Ophthalmol 2004; 137:486 –95. 2. Nowak JZ. Age-related macular degeneration (AMD): pathogenesis and therapy. Pharmacol Rep 2006;58:353– 63. 3. Avery RL, Pieramici DJ, Rabena MD, et al. Intravitreal bevacizumab (Avastin) for neovascular age-related macular degeneration. Ophthalmology 2006;113:363–72. 4. Rosenfeld PJ, Brown DM, Heier JS, et al, MARINA Study Group. Ranibizumab for neovascular age-related macular degeneration. N Engl J Med 2006;355:1419 –31. 5. Freeman WR, Falkenstein I. Avastin and new treatments for AMD: where are we? Retina 2006;26:853– 8.
6. Lux A, Llacer H, Heussen FM, Joussen AM. Non-responders to bevacizumab (Avastin) therapy of choroidal neovascular lesions. Br J Ophthalmol 2007;91:1318 –22. 7. Brantley MA Jr, Edelstein SL, King JM, et al. Association of complement factor H and LOC387715 genotypes with response of exudative age-related macular degeneration to photodynamic therapy. Eye (Lond) 2009;23:626 –31. 8. Brantley MA Jr, Fang AM, King JM, et al. Association of complement factor H and LOC387715 genotypes with response of exudative age-related macular degeneration to intravitreal bevacizumab. Ophthalmology 2007;114:2168 –73. 9. Goverdhan SV, Hannan S, Newsom RB, et al. An analysis of the CFH Y402H genotype in AMD patients and controls from the UK, and response to PDT treatment. Eye (Lond) 2008;22: 849 –54. 10. Krebs I, Brannath W, Glittenberg C, et al. Posterior vitreomacular adhesion: a potential risk factor for exudative agerelated macular degeneration? Am J Ophthalmol 2007;144: 741– 6. 11. Mojana F, Cheng L, Bartsch DU, et al. The role of abnormal vitreomacular adhesion in age-related macular degeneration: spectral optical coherence tomography and surgical results. Am J Ophthalmol 2008;146:218 –27. 12. Ondes F, Yilmaz G, Acar MA, et al. Role of the vitreous in age-related macular degeneration. Jpn J Ophthalmol 2000;44: 91–3. 13. Quaranta-El Maftouhi M, Mauget-Faysse M. Anomalous vitreoretinal adhesions in patients with exudative age-related macular degeneration: an OCT study. Eur J Ophthalmol 2006; 16:134 –7. 14. Robison CD, Krebs I, Binder S, et al. Vitreomacular adhesion in active and end-stage age-related macular degeneration. Am J Ophthalmol 2009;148:79 – 82. 15. Weber-Krause B, Eckardt U. Incidence of posterior vitreous detachment in eyes with and without age-related macular degeneration: an ultrasonic study [in German]. Ophthalmologe 1996;93:660 –5. 16. Lee SJ, Lee CS, Koh HJ. Posterior vitreomacular adhesion and risk of exudative age-related macular degeneration: paired eye study. Am J Ophthalmol 2009;147:621– 6. 17. Holladay JT. Proper method for calculating average visual acuity. J Refract Surg 1997;13:388 –91. 18. Brown DM, Kaiser PK, Michels M, et al, ANCHOR Study Group. Ranibizumab versus verteporfin for neovascular agerelated macular degeneration. N Engl J Med 2006;355:1432– 44. 19. Brown DM, Michels M, Kaiser PK, et al, ANCHOR Study Group. Ranibizumab versus verteporfin photodynamic therapy for neovascular age-related macular degeneration: two-year results of the ANCHOR Study. Ophthalmology 2009;116: 57– 65. 20. Fung AE, Lalwani GA, Rosenfeld PJ, et al. An optical coherence tomography-guided, variable dosing regimen with intravitreal ranibizumab (Lucentis) for neovascular age-related macular degeneration. Am J Ophthalmol 2007; 143:566 – 83. 21. Regillo CD, Brown DM, Abraham P, et al, PIER Study Group. Randomized, double-masked, sham-controlled trial of ranibizumab for neovascular age-related macular degeneration: PIER Study year 1. Am J Ophthalmol 2008; 145:239 – 48. 22. Anderson DH, Mullins RF, Hageman GS, Johnson LV. A role for local inflammation in the formation of drusen in the aging eye. Am J Ophthalmol 2002;134:411–31.
109
Ophthalmology Volume 118, Number 1, January 2011 23. Donoso LA, Kim D, Frost A, et al. The role of inflammation in the pathogenesis of age-related macular degeneration. Surv Ophthalmol 2006;51:137–52. 24. Zarbin MA. Current concepts in the pathogenesis of agerelated macular degeneration. Arch Ophthalmol 2004;122: 598 – 614. 25. Lambert HM, Capone A Jr, Aaberg TM, et al. Surgical excision of subfoveal neovascular membranes in age-related macular degeneration. Am J Ophthalmol 1992;113:257– 62. 26. Schmidt JC, Mennel S, Horle S, Meyer CH. High incidence of vitreomacular traction in recurrent choroidal neovascularisation after repeated photodynamic therapy. Br J Ophthalmol 2006;90:1361–2. 27. Ikeda T, Sawa H, Koizumi K, et al. Pars plana vitrectomy for regression of choroidal neovascularization with age-related macular degeneration. Acta Ophthalmol Scand 2000;78:460 – 4. 28. Sebag J. Anomalous posterior vitreous detachment: a unifying concept in vitreo-retinal disease. Graefes Arch Clin Exp Ophthalmol 2004;242:690 – 8.
29. Sebag J. Vitreoschisis. Graefes Arch Clin Exp Ophthalmol 2008;246:329 –32. 30. Chang LK, Fine HF, Spaide RF, et al. Ultrastructural correlation of spectral-domain optical coherence tomographic findings in vitreomacular traction syndrome. Am J Ophthalmol 2008;146:121–7. 31. Bakri SJ, Snyder MR, Reid JM, et al. Pharmacokinetics of intravitreal bevacizumab (Avastin). Ophthalmology 2007;114: 855–9. 32. Nomoto H, Shiraga F, Kuno N, et al. Pharmacokinetics of bevacizumab after topical, subconjunctival and intravitreal administration in rabbits. Invest Ophthalmol Vis Sci 2009;50: 4807–13. 33. DiLoreto DA Jr, Bressler NM, Bressler SB, Schachat AP. Use of best and final visual acuity outcomes in ophthalmological research. Arch Ophthalmol 2003;121:1586 –90. 34. Jabs DA. Improving the reporting of clinical case series. Am J Ophthalmol 2005;139:900 –5.
Footnotes and Financial Disclosures Originally received: September 30, 2009. Final revision: April 5, 2010. Accepted: April 8, 2010. Available online: August 3, 2010.
This study was presented as a free paper at: the Asia Pacific Academy of Ophthalmology–American Academy of Ophthalmology Joint Congress, May 16, 2009, Bali, Indonesia. Manuscript no. 2009-1360.
1
Department of Ophthalmology, Dongguk University Ilsan Hospital, Dongguk University School of Medicine, Gyeonggido, Korea. 2
Department of Ophthalmology, Institute of Vision Research, Yonsei University College of Medicine, Seoul, Korea.
110
Financial Disclosure(s): The author(s) have no proprietary or commercial interest in any materials discussed in this article. Correspondence: Hyoung Jun Koh, MD, Department of Ophthalmology, Yonsei University College of Medicine, No. 134 Shinchon-dong, Seodaemun-gu, 120-752, Seoul, Korea. E-mail: [email protected].