To study the effect of anti–vascular endothelial growth factor (VEGF) therapy on subretinal hyperreflective exudation detected by spectral-domain optical coherence tomography (SD OCT) in myopic choroidal neovascularization (CNV).
Retrospective consecutive observational cohort study.
Thirty-one eyes of 31 consecutive highly myopic patients with CNV and showing a subretinal hyperreflective exudation on SD OCT were included. Morphologic changes were assessed before and after anti-VEGF therapy, based on the subretinal hyperreflective exudation thickness, retinal thickness at the level of the CNV, and central macular thickness.
After anti-VEGF treatment (mean follow-up of 1.9 ± 0.8 months, mean number of injections 1.8 ± 0.6), the subretinal hyperreflective exudation regressed completely in 29 of 31 eyes (93.5%) and partially in 2 of 31 eyes (6.5%). Mean subretinal hyperreflective exudation thickness, mean retinal thickness at the level of the CNV, and mean central macular thickness significantly decreased from 102 ± 50 μm to 2.6 ± 10.2 μm ( P < .0001), from 419 ± 99 μm to 312 ± 64 μm ( P < .0001), and from 361 ± 69 μm to 326 ± 72 μm ( P = .0008), respectively.
The subretinal hyperreflective exudation was an SD OCT finding that correlated with signs of active myopic CNV (either subretinal fluid/intraretinal cysts on SD OCT or dye leakage on fluorescein angiography) and responded to treatment with anti-VEGF agents. The presence of a subretinal hyperreflective exudation on SD OCT could help in making decisions on the need to perform or not perform fluorescein angiography, and regarding treatment or retreatment.
Myopia is a common eye disorder. The progressive and excessive elongation of the eyeball can lead to pathologic myopia (or high myopia). Pathologic myopia is defined by an axial length of the eye longer than 26 mm or a refractive error of at least -6 diopters (D). It is considered a significant cause of visual impairment and blindness worldwide. Choroidal neovascularization (CNV) developing between the retinal pigment epithelium (RPE) and the neurosensory retina (type 2 CNV) is a common vision-threatening complication associated with high myopia, accounting for 5%–11% of cases.
The development of anti–vascular endothelial growth factor (anti-VEGF) therapy over the last few years has revolutionized the management of exudative age-related macular degeneration (AMD). Its use has rapidly been extended to other retinal diseases complicated by CNV. Thus, the use of intravitreal bevacizumab or ranibizumab in myopic CNV appears as a favorable option, the majority of patients experiencing visual gain after treatment. Intravitreal injections of ranibizumab resulted in better visual outcomes than verteporfin photodynamic therapy in a randomized clinical trial. Myopic CNV has typically small dimensions and may be difficult to diagnose on clinical examination owing to associated fundus lesions, including posterior pole staphyloma, lacquer cracks, pigmentary changes, and chorioretinal atrophy. The gold-standard examination method for the diagnosis of type 2 CNV associated with myopic maculopathy is fluorescein angiography (FA). Spectral-domain optical coherence tomography (SD OCT) is a helpful tool for the diagnosis of myopic CNV, showing a hyperreflective lesion located beneath the neurosensory retina, sometimes associated with slight retinal changes including intraretinal cysts or subretinal fluid (SRF). However, previous studies have suggested that the FA sensitivity may be greater than that of SD OCT for the diagnosis of exudation in new-onset myopic CNV as it shows the CNV network on early frames and dye staining and leakage on late frames. On the other hand, some studies have recently reported the presence of subretinal hyperreflective exudations, also called gray hyperreflective subretinal lesions, in exudative AMD and suggested that they could be considered as a qualitative criterion for active CNV. In AMD, they represent an early sign of active, exudative neovascular lesions and should lead to prompt treatment. The presence of a subretinal hyperreflective exudation has been observed, in our clinical setting, on SD OCT scans of patients diagnosed with exudative myopic CNV. We hypothesized that these lesions could correspond to an early stage of the myopic CNV activity and regress after treatment with intravitreal anti-VEGF, as in exudative AMD. The aim of this study was to analyze SD OCT features of the subretinal hyperreflective exudation in eyes with active myopic CNV attested on FA and to study its evolution after anti-VEGF treatment.
We reviewed the charts of all consecutive highly myopic patients with history of myopic CNV seen at the University Eye Clinic of Créteil, Créteil, France, between November 4, 2013 and May 2, 2014. These patients had been treated with intravitreal injections of anti-VEGF (ranibizumab 0.5 mg or bevacizumab 1.25 mg) between November 2008 and March 2014. This retrospective consecutive observational cohort study was performed according to the French bioethical legislation and the Declaration of Helsinki for research involving human subjects. The French Society of Ophthalmology ethics committee approved the conduct of this study.
Inclusion criteria were as follows: high myopia (defined by an axial length of the eyeball longer than 26 mm or a refractive error of at least -6 D), the presence of a subretinal hyperreflective exudation on SD OCT, and the presence of active juxtafoveal or subfoveal myopic CNV confirmed by either FA (hyperfluorescent CNV network on early frames and fluorescein leakage on late frames) or SD OCT (SRF and/or intraretinal cysts). Non-naïve patients had not received any anti-VEGF injection in the 3 months preceding the inclusion. Patients with AMD, adult-onset foveomacular vitelliform dystrophy, or CNV attributable to causes other than myopia were excluded.
All patients underwent a clinical examination of both eyes at baseline, including best-corrected visual acuity (BCVA) measured on an Early Treatment Diabetic Retinopathy Study (ETDRS) chart converted into logarithm of the minimal angle of resolution (logMAR), 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 using a confocal scanning laser system (Spectralis HRA+OCT; Heidelberg Engineering, Heidelberg, Germany). Diagnosis of CNV activity was confirmed by FA and eyes were included when a subretinal hyperreflective exudation was detected on SD OCT scans, in the area corresponding to the CNV on FA. Subretinal hyperreflective exudation appeared as a moderately hyperreflective deposit with fuzzy edges located between the RPE and the ellipsoid zone. The SD OCT scan protocol used was a dense macular cube corresponding to a 49-line raster with 20 μm interline spacing, centered on the macula and on the CNV when needed. The scans could be oriented either vertically or horizontally or obliquely, depending on the topography of the staphyloma. The presence of other SD OCT signs associated with the CNV activity was assessed, including retinal thickening, intraretinal cysts, SRF, or lesions referred to as type 2 CNV: highly hyperreflective lesions located above the RPE–Bruch membrane complex. All patients were scheduled for 1, 2, or 3 intravitreal injections of anti-VEGF with bevacizumab 1.25 mg (Avastin; Genentech Inc, South San Francisco, California, USA) or ranibizumab 0.5 mg (Lucentis; Genentech Inc) injected monthly by a retina specialist. The number of anti-VEGF intravitreal injections scheduled (1, 2, or 3 injections) was decided by a senior ophthalmologist, according to the intensity of the neovascular activity, based on FA and SD OCT findings.
One month after the last intravitreal injection and at 6 months from baseline, each patient underwent an ophthalmologic examination including assessment of the BCVA using ETDRS chart converted into logMAR, slit-lamp biomicroscopy, fundus biomicroscopy, and SD OCT examination using the “follow-up” protocol of the Spectralis HRA+OCT. FA was performed at the discretion of the examiner.
We retrospectively assessed the morphologic changes induced by anti-VEGF therapy on the subretinal hyperreflective exudation between baseline and 1 month after the last intravitreal injection and between baseline and the final visit at 6 months: (1) qualitative evolution of the subretinal hyperreflective exudation (complete regression, partial regression, stabilization, or progression); and (2) subretinal hyperreflective exudation thickness, measured by vertically quantifying the highest subretinal hyperreflective exudation thickness in the area of the lesion. Subretinal hyperreflective exudation thickness was measured using the caliper tool of the Spectralis HRA+OCT viewer from the well-defined inner edge of the RPE or inner edge of the neovascular lesion to the inner fuzzy edge of the subretinal hyperreflective exudation. We also recorded BCVA changes and other SD OCT findings, including retinal thickness at the level of the CNV, central macular thickness, presence of SRF, intraretinal cysts, intraretinal hyperreflective dots, and leakage on FA. Qualitative grading and manual caliper measurements of the subretinal hyperreflective exudation and retinal thickness at the level of the CNV were recorded by 2 independent retina specialists (E.B. and V.C.). SRF was defined as a serous retinal detachment: hyporeflective space located between the ellipsoid zone and the RPE. Myopic CNV is usually small in size and displays few exudative features on SD OCT. We considered that an SRF was associated with the subretinal hyperreflective exudation when it was found around or above the subretinal hyperreflective exudation on SD OCT.
All values are presented as mean ± SD with a normal distribution detected using the Kolmogorov-Smirnov test. Statistical analyses were performed using GraphPad Prism Version 5.0 (GraphPad software, Inc, La Jolia, CA, USA). Comparisons of the mean BCVA, mean subretinal hyperreflective exudation thickness, mean retinal thickness at the level of the CNV, and mean central macular thickness measured between baseline and the first follow-up visit and between baseline and the 6-month visit were performed using the Student paired t test. Statistical significance was set at P < .05. The Bland-Altman method was used to measure the intergrader agreement for the measurements.
Baseline Clinical Characteristics of the Study Population
A total of 31 eyes of 31 patients (27 women, 4 men) were included in the analysis, with a mean age of 63 ± 12.5 years. The mean spherical equivalent refractive error was −12.8 ± 5.0 D. At baseline, 6 of 31 eyes (19.4%) were naïve of treatment whereas 25 of 31 eyes (80.6%) had previously received anti-VEGF intravitreal injections (bevacizumab or ranibizumab) for myopic CNV. The mean number of previous injections was 4.5 ± 2.4. The mean BCVA at baseline was 0.42 ± 0.27 logMAR. Twenty-seven patients (87.1%) presented visual symptoms such as visual loss (12/31, 38.7%), scotoma (11/31, 35.5%), and/or metamorphopsia (17/31, 54.8%). Four patients (12.9%) were completely asymptomatic. Epidemiologic and morphologic characteristics of the study population are listed in Table 1 .
|Patient Characteristics (N = 31)|
|Age, y (± SD)||63 ± 12.5|
|Eye Characteristics (N = 31)|
|Mean refractive error (± SD), D||−12.8 ± 5.0|
|Localization of CNV|
|Naïve, n (%)||6 (19.4)|
|Intravitreal injections, n (%)||25 (80.6)|
|Number of injections a (± SD)||4.5 (± 2.4)|
|Visual symptoms at baseline, n (%)||27 (87.1)|
|Decrease of visual acuity, n (%)||12 (38.7)|
|Metamorphopsia, n (%)||17 (54.8)|
|Scotoma (%)||11 (35.5)|
|Asymptomatic (%)||4 (12.9)|
Baseline Spectral-Domain Optical Coherence Tomography and Fluorescein Angiography Findings
At baseline, a subretinal hyperreflective exudation was present in all eyes (31/31, 100%). The subretinal hyperreflective exudation appeared as a moderately hyperreflective deposit with fuzzy edges located between the RPE and the ellipsoid zone. The mean subretinal hyperreflective exudation thickness was 102 ± 50 μm, the mean central macular thickness was 361 ± 69 μm, and the mean retinal thickness at the neovascular lesion was 419 ± 99 μm. The Bland-Altman analysis found a bias of 0.22 ± 10.98 μm, showing a good agreement between the 2 graders for the measurements of the subretinal hyperreflective exudation thickness and retinal thickness at the level of the CNV. FA was performed in 27 of 31 eyes (87.1%) at baseline, whereas it was not performed in 4 of 31 eyes that presented obvious exudative signs on SD OCT, such as SRF or intraretinal cysts. Signs of active CNV (visualization of the CNV network on early frames, dye staining and leakage on late frames) were observed in all 27 eyes in which FA was performed. Fourteen of these eyes (51.9%) also presented SD OCT exudative findings, including intraretinal cysts in 13 of 31 eyes (41.9%) and SRF in 8 of 31 eyes (25.8%). In these 8 eyes, the SRF was adjacent to the subretinal hyperreflective exudation (between the RPE and the ellipsoid zone) in 5 eyes, above it (between the inner surface of the subretinal hyperreflective exudation and the ellipsoid zone) in 1 eye, and both adjacent to and above it in 2 eyes. In 13 of 31 eyes (41.9%), the subretinal hyperreflective exudation was not associated with any other SD OCT exudative sign. Among these 13 eyes, 4 (30.8%) were visually asymptomatic.
Functional and Morphologic Changes After Anti–Vascular Endothelial Growth Factor Therapy: First Follow-up Visit
This study was conducted in “real-life” conditions and the first visit after anti-VEGF treatment took place approximately 1 month after the last scheduled injection, corresponding to a mean time of 1.9 ± 0.8 months from baseline. We called this visit “first follow-up visit.” The mean number of intravitreal injections administered was 1.8 ± 0.6 (28 of 31 eyes were treated with ranibizumab only and 3 of 31 with bevacizumab only). The mean BCVA did not change significantly between baseline and the first follow-up visit (0.42 ± 0.27 logMAR and 0.36 ± 0.26 logMAR, P = .067, respectively). At the time of the first follow-up visit, SD OCT showed a complete regression of the subretinal hyperreflective exudation in 29 of 31 eyes (93.5%) and a partial regression in 2 of 31 eyes (6.5%) ( Figure 1 ). These 2 eyes with partial regression of the subretinal hyperreflective exudation were retreated with intravitreal anti-VEGF injection (ranibizumab) and monitored monthly until complete regression of the subretinal hyperreflective exudation. In these 2 eyes, the subretinal hyperreflective exudation had completely disappeared 1 month after a single additional anti-VEGF injection. Finally, a complete regression of the subretinal hyperreflective exudation was observed in all cases. The regression of the subretinal hyperreflective exudation resulted in a well-delineated, highly hyperreflective lesion whose edges appeared well defined, like a shell, and it most likely corresponded to fibrosis. We considered that the regression was complete when the subretinal hyperreflective exudation (fuzzy-edged, moderately hyperreflective subretinal lesion) was no longer measurable using the caliper above the RPE or above the resulting shell-like, well-delineated, highly hyperreflective lesion. The mean subretinal hyperreflective exudation thickness significantly decreased from 102 ± 50 μm at baseline to 2.6 ± 10.2 μm ( P < .0001) at the time of the first follow-up visit. The mean central macular thickness changed significantly between baseline and the first follow-up visit: 361 ± 69 μm and 326 ± 72 μm ( P = .0008), respectively. The mean retinal thickness at the neovascular lesion significantly decreased from 419 ± 99 μm at baseline to 312 ± 64 μm at the time of the first follow-up visit ( P < .0001). At the time of the first follow-up visit, the SRF completely regressed in all eyes and intraretinal cysts persisted in 1 of 31 eyes (3.2%). The subretinal hyperreflective exudation had partially regressed in the eye with persistent intraretinal cysts at the time of the first follow-up visit. Functional and morphologic changes at baseline and at the time of the first follow-up visit are presented in Table 2 .
|Before Treatment (Baseline)||First Visit After Treatment (First Follow-up Visit)||P||Final Visit (6-Month Follow-up)||P|
|Presence of subretinal hyperreflective exudation, n (%)||31/31 (100)||2/31 (6.5)||5/31 (16.1)|
|Presence of other SD OCT exudative features, n (%)||18/31 (58.1)||1/31 (3.2)||1/31 (3.2)|
|Subretinal fluid only, n (%)||5/31 (16.1)||0/31||0/31|
|Intraretinal cysts only, n (%)||10/31 (32.3)||1/31 (3.2)||1/31 (3.2)|
|Subretinal fluid + intraretinal cysts, n (%)||3/31 (10.0)||0/31||0/31|
|SHE thickness, μm (mean ± SD)||102 ± 50||2.6 (± 10.2)||<.0001||16.3 (± 40.9)||<.0001|
|Retinal thickness at the level of the CNV, μm (mean ± SD)||419 ± 99||312 (± 64)||<.0001||279 (± 67)||<.0001|
|Central macular thickness, μm (mean ± SD)||361 ± 69||326 (± 72)||.0008||288 (± 91)||.0003|
|Visual acuity, logMAR (mean ± SD)||0.42 ± 0.27||0.36 (± 0.26)||.0670||0.43 (± 0.31)||.6931|
|Follow-up, mo (mean ± SD)||1.9 (± 0.8)||6.6 (± 1.4)|
|Number of intravitreal injections (mean ± SD)||1.8 (± 0.6)||0.2 (± 0.5) a|