Purpose
To analyze the contribution of fluorescein angiography (FA) and spectral-domain optical coherence tomography (SD OCT) to the diagnosis of recent choroidal neovascularization (CNV) associated with high myopia.
Design
Retrospective, observational case series.
Methods
Ninety eyes of 73 highly myopic patients (refractive error ≥−6 diopters) with CNV in 1 or both eyes were included. Epidemiologic features, refractive error, fundus examination, fluorescein angiography, and SD OCT findings at onset of CNV were analyzed.
Results
Mean age at onset of CNV was 54.4 ± 14 years. CNV was bilateral in 17 of 73 cases. Mean refractive error was −13.9 ± 5.2 diopters. Myopic CNV was associated more frequently with patchy or geographic atrophy ( P = .019). CNV was associated with exudative features on fluorescein angiography in 82% of cases (64/78), and on SD OCT in 48.6% of cases (36/74). There was no agreement about signs of active CNV between these 2 imaging methods (κ = 25.7 ± 10%; P = .0044). CNV area was significantly smaller in younger patients (<55 years) than in older patients (0.57 mm 2 vs 1.21 mm 2 , respectively; P = .023).
Conclusions
Exudative features of myopic CNV are more obvious on FA than on SD OCT, suggesting that fluorescein angiography should be performed when new-onset myopic CNV is suspected. Myopic CNV occurring in older patients (≥55 years) is larger than those seen in younger patients and resembles CNV associated with age-related macular degeneration. This suggests an overlap between myopic CNV in older patients and age-related macular degeneration.
High myopia is defined by an axial length of the eye higher than 26 mm or by a refractive error of −6 diopters (D) or more. Choroidal neovascularization (CNV) growing between the retinal pigment epithelium and neurosensory retina (type 2 CNV) is the most common central vision-threatening complication in patients with high myopia, accounting for 4% to 11% of cases. New-onset myopic CNV may be difficult to diagnose purely by fundus examination for several reasons: choroidal pallor frequently is present in association with other anomalies such as posterior pole staphyloma, pigmentary changes, lacquer cracks, and chorioretinal atrophic lesions. Exudative features, macular edema, or subretinal fluid, frequently associated with other causes of CNV, such as exudative age-related macular degeneration (AMD), are observed less commonly in myopic CNV. Retinal hemorrhages frequently are seen with recent myopic CNV, but also are a common finding in cases of recent lacquer cracks associated with high myopia. Fluorescein angiography (FA) is an important tool that facilitates the diagnosis of recent myopic CNV by showing the CNV network on early frames and staining and leakage of the dye on late frames. Spectral-domain optical coherence tomography (SD OCT) is another important diagnostic tool for myopic CNV, showing a hyperreflective lesion located beneath the neurosensory retina, sometimes associated with discrete retinal changes including edema or neurosensory serous retinal detachment.
However, neither leakage on FA or exudative signs on SD OCT has 100% sensitivity for genuine new-onset myopic CNV. The aim of this study was to compare the contribution of FA or SD OCT versus both of them for the diagnosis of new-onset myopic CNV in a retrospective series of patients with high myopia. Age of onset-related features of FA and SD OCT in myopic CNV also were analyzed in this study.
Methods
Patients with high myopia complicated by new-onset subfoveal or juxtafoveal CNV in 1 or both eyes were analyzed retrospectively at the University Eye Clinic of Créteil, Créteil, France. The research followed the tenets of the Declaration of Helsinki. Retrospective review of patient records was approved by the Ethics Committee of the French Society of Ophthalmology.
Exclusion criteria were (1) mild and moderate myopia (refractive error ≥ −6 D), (2) CNV associated with other retinal diseases such as AMD, (3) clinical features of age-related maculopathy such as soft or hard drusen and pigmentary alterations, (4) other macular disease (ie, retinal dystrophies), and (5) media opacity preventing detailed imaging.
All patients underwent a clinical examination of both eyes including best-corrected visual acuity measured on an Early Treatment Diabetic Retinopathy Study chart, slit-lamp examination, fundus examination, FA, and SD OCT with scans centered on the foveal area and on the CNV lesion. Both FA and SD OCT were obtained with a confocal scanning laser system (Spectralis HRA+OCT; Heidelberg Engineering, Heidelberg, Germany). Horizontal and vertical SD OCT scans of 6 or 3 mm were centered on the fovea and on the suspected myopic CNV lesions.
Diagnostic criteria for high myopia were a refractive error of −6 D or more and various clinical features observed on fundus examination usually associated with high myopia, including posterior pole staphyloma, diffuse or patchy chorioretinal atrophy, peripapillary atrophy, tilted disc, lacquer cracks, and depigmentation of the fundus allowing good visualization of the choroidal vessels. Because the occurrence of CNV usually does not appear at the same time in both eyes, the analysis of SD OCT scans and FA was performed retrospectively independently for each eye at initial presentation.
The diagnosis of recent CNV was confirmed by the following examinations: fundus examination (retinal hemorrhage associated with a grayish macular lesion with a pigmented ring), FA (hyperfluorescent CNV network on early frames and leakage of the dye on late frames), and SD OCT (hyperreflective lesion located above the retinal pigment epithelium–Bruch membrane complex associated with retinal thickening, retinal cysts, or subretinal accumulation of fluids). To be included, eyes were required to show exudative signs on at least 1 of the 2 imaging examinations: FA or SD OCT. For each eye, surface area of CNV was measured by circling manually the CNV’s hyperfluorescence on early FA frames using a specific tool of the Spectralis HRA+OCT. The presence of lacquer cracks, large atrophic areas, or patchy atrophic lesions located in the posterior pole also was analyzed in both eyes of each patient by fundus examination, infrared frames, and FA. When the diagnosis of myopic CNV was doubtful at the first examination by FA and SD OCT because of retinal hemorrhage, potentially associated with an isolated lacquer crack or a CNV, these latter exams were performed again within few days (7 to 15 days) after the first clinical examination.
In each case, the contribution of FA and SD OCT to the diagnosis of new-onset myopic CNV was analyzed at the same time by 2 clinicians together (N.L. and V.C.). Analysis also was performed to compare FA and SD OCT characteristics of CNV in 2 age groups.
Statistical Analysis
Patient and eye characteristics were described and compared by the presence or absence of CNV. Multilevel logistic modeling was used first because of the hierarchical structure of the data (2 eyes nested within patients). Because no random effect was observed at the patient level (variance, 10 to 8), we considered eyes as independent variables and used classical tests. Categorical data are reported as a number (%) and were compared using chi-square or Fisher exact tests as appropriate. Continuous data are reported as mean ± standard deviation or median (interquartile range) as appropriate and were compared using the nonparametric Kruskal-Wallis test. We also compared eye characteristics according to patient age, categorizing age according to the median value. Agreement between angiographic and SD OCT data was assessed using a chi-square test and the κ statistic, the latter allowing estimation of the level of agreement taking into account agreement obtained by chance. All comparisons were 2-sided with P ≤ .05 being considered significant. Data were analyzed using Stata statistical software (Stata/SE 12; Stata Corp, College Station, Texas, USA).
Results
Ninety eyes of 73 patients complicated with new-onset myopic CNV (59 women and 14 men) were included in this study, with a mean age at onset of CNV in the first eye of 54.4 ± 14 years. The CNV was unilateral in 56 (76.7%) patients (affecting the right eye in 26 patients and the left eye in 30 patients) and involved both eyes in 17 (23.3%) patients, with 90 eyes affected by CNV represented in this study. Among the 56 patients with unilateral CNV, 3 (5.4%) were monophthalmus; thus, 53 (94.6%) fellow eyes without CNV were analyzable. Median refractive error was −14 D (interquartile range, 9 to 17 D) in the right eye and −13 D (interquartile range, 9 to 16 D) in the left eye ( Table 1 ).
| Patient characteristics | |
| Median age (IQR) a | 55 (22 to 85) |
| Sex ratio (M/F) | 14/59 |
| Eye characteristics (n = 143) | |
| Median refractive error (IQR), right eye/left eye | −14 (9 to 17)/−13 (9 to 16) |
| Atrophy, n (%) b | |
| None | 84 (61.3) |
| Patchy | 31 (22.6) |
| Geographic | 22 (16.1) |
| Lacquer cracks, n (%) c | 44 (31.6) |
| Peripapillary atrophy, n (%) d | 123 (89.1) |
| CNV, n (%) | 90 (62.9) |
| Localization of CNV in the 73 patients | |
| Right eye | 26 (35.6) |
| Left eye | 30 (41.1) |
| Both eyes | 17 (23.3) |
| Median area of CNV (IQR), mm 2 | 0.7 (0.37 to 1.55) |
| Diffusion during FA, n (%) e | 64 (82) |
| Exudation by SD OCT, n (%) f | 36 (48.6) |
| CNV associated to atrophy, n (%) g | 33 (37.5) |
| CNV associated with lacquer cracks, n (%) h | 20 (22.7) |
| Isolated CNV, n (%) i | 35 (39.8) |
a Seventy-one patients; 2 missing data.
e Seventy-eight patients; 12 missing data.
f Seventy-four patients; 16 missing data.
g Eighty-eight patients; 2 missing data.
h Eighty-eight patients; 2 missing data.
Where CNV was identified, leakage of dye on late frames of FA was observed in 82% of eyes (64/78; 12 missing data), whereas exudative signs such as intraretinal hyporeflective cysts or subretinal fluid were observed by SD OCT in 48.6% of eyes (36/74; 16 missing data). Missing data for FA and SD OCT resulted from poor-quality images because of high myopia preventing reliable analysis of FA or SD OCT. Among the eyes affected with CNV, diagnosis was made by FA alone in 60% of cases, by OCT alone in 28.9% of cases, and by both OCT and FA in 11.1% of cases. When considering the group of eyes affected with CNV, for which data from both FA and SD OCT was interpretable and reliable (thus excluding missing data: eyes for which one or the other examination was not interpretable or analyzable because of insufficient quality attributed to the high myopia), diagnosis of CNV was made by FA alone in 61.3% of cases (38/62), by OCT alone in 22.6% of cases (14/62), and by both OCT and FA in 16.1% of cases (10/62). When comparing leakage on FA with occurrence of exudative signs on SD OCT, there was no agreement about signs of active CNV between the 2 imaging methods (FA and SD OCT): κ value of 25.7 ± 10% ( P = .0044). Examples of leakage observed by FA and exudation observed by SD OCT are presented in Figures 1, 2, and 3 .
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
