Advertisement

Spectral-Domain Optical Coherence Tomography Analysis of Fibrotic Lesions in Neovascular Age-Related Macular Degeneration

Published:February 27, 2020DOI:https://doi.org/10.1016/j.ajo.2020.02.016

      Purpose

      To describe the spectral-domain optical coherence tomography (OCT) features of fibrotic lesions associated with neovascular age-related macular degeneration (nAMD) and to outline the progression pathways from initial macular choroidal neovascular lesions (CNVs) to fibrosis.

      Methods

      Patients with nAMD were retrospectively included when macular subretinal fibrosis was present. Fibrosis was categorized using spectral-domain OCT with respect to retinal pigment epithelium (RPE) in 836 spectral-domain OCT slices from 44 eyes of 39 patients. In addition, in 47 distinct eyes, 4181 spectral-domain OCT slices were retrospectively reviewed to longitudinally assess progression from the initial lesion to the final fibrosis.

      Results

      Cross-sectional analysis classified fibrosis on spectral-domain OCT slices, as type A if located underneath the RPE, as type B if located above the RPE, and as type C if the remaining RPE was undistinguishable. The longitudinal analysis series revealed 3 progression pathways from the original CNV: 1) progression to type A, followed by RPE erosion and subretinal hyperreflective material, then type B and type C fibroglial lesion (FGL; 17/47 eyes); 2) progression to type B then type C FGL (17/47 eyes); and 3) persistence of type A with development of a flat, fibroatrophic lesion (13/47 eyes). Subretinal hyperreflective material, macular hemorrhage, or RPE tear occurred in 14 of 47, 13 of 47, and 10 of 47 eyes, respectively.

      Conclusion

      This spectral-domain OCT analysis identified various patterns of macular fibrosis in eyes with nAMD. Three pathways of progression to fibrosis were described including the well-established pathway of type 2 CNV progression to FGL and the progression of type 1 fibrovascular CNV to FGL or fibroatrophic lesion.
      Age-related macular degeneration (AMD) is a leading cause of visual loss in developed countries.
      • Bressler N.M.
      Age-related macular degeneration is the leading cause of blindness.
      With a rising prevalence worldwide, an estimated 288 million people will be affected with AMD by 2040.
      • Wong W.L.
      • Su X.
      • Li X.
      • et al.
      Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis.
      Originating from the choroid or retina, aberrant angiogenesis, leading to the development of retinal pigment epithelial detachment (PED), subretinal fluid, intraretinal fluid, or hemorrhages, plays a central role in neovascular AMD (nAMD).
      • Zayit-Soudry S.
      • Moroz I.
      • Loewenstein A.
      Retinal pigment epithelial detachment.
      ,
      • Mrejen S.
      • Sarraf D.
      • Mukkamala S.K.
      • Freund K.B.
      Multimodal imaging of pigment epithelial detachment: a guide to evaluation.
      The advent of intravitreal vascular endothelial growth factor (VEGF) inhibitors has greatly improved the visual prognosis in nAMD. Several effective therapeutic strategies using anti-VEGF treatment have been established during the last decade.
      • Schmidt-Erfurth U.
      • Chong V.
      • Loewenstein A.
      • et al.
      Guidelines for the management of neovascular age-related macular degeneration by the European Society of Retina Specialists (EURETINA).
      Nevertheless, common complications of nAMD, including submacular hemorrhage, retinal pigment epithelium (RPE) tears, geographic atrophy, and the development of fibrotic scars, cannot be avoided despite anti-VEGF treatment, finally leading to a loss of central vision.
      • Cohen S.Y.
      • Oubraham H.
      • Uzzan J.
      • Dubois L.
      • Tadayoni R.
      Causes of unsuccessful ranibizumab treatment in exudative age-related macular degeneration in clinical settings.
      ,
      • Rosenfeld P.J.
      • Shapiro H.
      • Tuomi L.
      • et al.
      Characteristics of patients losing vision after 2 years of monthly dosing in the phase III ranibizumab clinical trials.
      As a result, despite the development of efficacious therapeutics, nAMD remains one of the most common vision-threatening diseases in the elderly.
      • Gemmy Cheung C.M.
      • Li X.
      • Cheng C.Y.
      • et al.
      Prevalence and risk factors for age-related macular degeneration in Indians: a comparative study in Singapore and India.
      ,
      • Kawasaki R.
      • Yasuda M.
      • Song S.J.
      • et al.
      The prevalence of age-related macular degeneration in Asians: a systematic review and meta-analysis.
      Fibrosis represents the wound healing response to choroidal neovascularization (CNV) in nAMD.
      • Keane P.A.
      • Patel P.J.
      • Liakopoulos S.
      • Heussen F.M.
      • Sadda S.R.
      • Tufail A.
      Evaluation of age-related macular degeneration with optical coherence tomography.
      Typically a later complication of nAMD, fibrotic scars respond minimally to anti-VEGF therapies. Risk factors for the development of fibrosis in treatment-naïve nAMD eyes treated with anti-VEGF therapy, according to the Comparison of Age-Related Macular Degeneration Treatment Trials (CATT), include classic CNV, blocked fluorescence with dye-based fluorescein angiography (FA), foveal retinal thickness >212 μm, subfoveal tissue thickness >275 μm, foveal subretinal fluid, and subretinal hyperreflective material (SHRM).
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      In fact, the presence of predominantly classic CNV (type 2 CNV) was associated with 4.5-fold risk for scarring compared with occult CNV (type 1 CNV).
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      Similarly, Bloch and associates
      • Bloch S.B.
      • Lund-Andersen H.
      • Sander B.
      • Larsen M.
      Subfoveal fibrosis in eyes with neovascular age-related macular degeneration treated with intravitreal ranibizumab.
      observed a 6-fold risk of macular fibrosis in eyes with type 2 CNV compared with occult type 1 CNV.
      Progression from a neovascular membrane to a fibrovascular lesion and finally to a fibrotic scar leads to destruction of the photoreceptors or the RPE and choriocapillaris layers, ultimately leading to visual acuity loss.
      • Cohen S.Y.
      • Oubraham H.
      • Uzzan J.
      • Dubois L.
      • Tadayoni R.
      Causes of unsuccessful ranibizumab treatment in exudative age-related macular degeneration in clinical settings.
      ,
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      However, the transition from angiogenesis to fibrosis, referred to as the “angiofibrotic switch,” in nAMD remains poorly described. Detailed imaging studies of fibrotic progression, especially of type 1 CNV, using spectral-domain OCT, is lacking.
      • Keane P.A.
      • Patel P.J.
      • Liakopoulos S.
      • Heussen F.M.
      • Sadda S.R.
      • Tufail A.
      Evaluation of age-related macular degeneration with optical coherence tomography.
      ,
      • Kuiper E.J.
      • de Smet M.D.
      • van Meurs J.C.
      • et al.
      Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
      ,
      • Kuiper E.J.
      • Van Nieuwenhoven F.A.
      • de Smet M.D.
      • et al.
      The angio-fibrotic switch of VEGF and CTGF in proliferative diabetic retinopathy.
      Currently the assessment of fibrotic lesions is based on multimodal imaging, including color fundus photography, spectral-domain OCT, FA, or indocyanine green angiography (ICGA).
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Bloch S.B.
      • Lund-Andersen H.
      • Sander B.
      • Larsen M.
      Subfoveal fibrosis in eyes with neovascular age-related macular degeneration treated with intravitreal ranibizumab.
      ,
      • Bloom S.M.
      • Singal I.P.
      The outer bruch membrane layer: a previously undescribed spectral-domain optical coherence tomography finding.
      More recently, OCT angiography (OCTA) has provided a noninvasive assessment of flow within fibrotic CNV, but no information was available concerning the surrounding fibrotic tissue, which generated signal voids on the choriocapillaris slab.
      • Miere A.
      • Semoun O.
      • Cohen S.Y.
      • et al.
      Optical coherence tomography angiography features of subretinal fibrosis in age-related macular degeneration.
      Roberts and associates
      • Roberts P.
      • Sugita M.
      • Deák G.
      • et al.
      Automated identification and quantification of subretinal fibrosis in neovascular age-related macular degeneration using polarization-sensitive OCT.
      have shown, using polarization-sensitive OCT, that subretinal fibrosis can be identified as an intrinsically birefringent structure, allowing a clear separation of neovascular tissue from the fibrous tissue. Even though tissue segmentation based on polarization-sensitive OCT and not signal intensity (spectral-domain OCT) has clear advantages,
      • Roberts P.
      • Sugita M.
      • Deák G.
      • et al.
      Automated identification and quantification of subretinal fibrosis in neovascular age-related macular degeneration using polarization-sensitive OCT.
      in a real-life clinical setting spectral-domain OCT remains the criterion standard imaging technique.
      Several groups have proposed classification systems of nAMD using OCT, including the fibrosis stage, in eyes treated with photodynamic therapy (PDT).
      • Rogers A.H.
      • Martidis A.
      • Greenberg P.B.
      • Puliafito C.A.
      Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
      Nonetheless, these classifications did not use high-resolution spectral-domain OCT, and therefore the morphologic details of fibrotic lesions were not thoroughly described.
      Compared to previous OCT generations, spectral-domain OCT delivers faster imaging, enabling depth-resolved visualization of the macular layers and adjacent tissues close to microscopic resolution.
      • Chong N.H.V.
      • Keonin J.
      • Luthert P.J.
      • et al.
      Decreased thickness and integrity of the macular elastic layer of Bruch’s membrane correspond to the distribution of lesions associated with age-related macular degeneration.
      With a resolution of 5 μm and layer-by-layer assessment of intraretinal microstructures, spectral-domain OCT has elucidated the broad phenotypically heterogeneous nature of AMD and its complex morphology.
      • de Boer J.F.
      • Cense B.
      • Park B.H.
      • Pierce M.C.
      • Tearney G.J.
      • Bouma B.E.
      Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography.
      ,
      • Toth L.A.
      • Stevenson M.
      • Chakravarthy U.
      Anti-vascular endothelial growth factor therapy for neovascular age-related macular degeneration: outcomes in eyes with poor initial vision.
      In the last few years, spectral-domain OCT has become an indispensable tool for the diagnosis and follow-up of patients with nAMD.
      • Keane P.A.
      • Liakopoulos S.
      • Ongchin S.C.
      • et al.
      Quantitative subanalysis of optical coherence tomography after treatment with ranibizumab for neovascular age-related macular degeneration.
      ,
      • Coscas F.
      • Coscas G.
      • Souied E.
      • Tick S.
      • Soubrane G.
      Optical coherence tomography identification of occult choroidal neovascularization in age-related macular degeneration.
      Despite the fast evolution of imaging techniques, the potential of spectral-domain OCT to evaluate lesions within the fibrosis spectrum secondary to AMD remains insufficiently explored. On spectral-domain OCT, fibrotic lesions can harbor various phenotypes, presenting a wide range of morphologic patterns, including fibrovascular PED vs. subretinal and/or sub-RPE fibrotic scars. In this cross-sectional study, we aim to describe the spectral-domain OCT morphologic features of a broad spectrum of fibrotic lesions associated with nAMD. Moreover, a distinct retrospective longitudinal analysis was performed, in a well-defined large series of patients, to delineate retrospectively the progression pathways from the initial neovascular lesion to the fibrotic scar observed on spectral-domain OCT. Our hypothesis is that spectral-domain OCT may be useful to define different types of fibrosis and that different patterns of progression from CNV to fibrosis exist. This may be useful in future investigations to treat and prevent the development of fibrosis.

      Methods

       Study Design and Study Population

      Informed consent was obtained from all subjects from our clinic in agreement with the Declaration of Helsinki for research involving human subjects. This study was institutional review board–approved and was carried out in compliance with local and national institutional review board guidelines.
      This study was a retrospective analysis of patients from the Department of Ophthalmology of the University Paris Est Creteil, who presented between March 2008 and March 2018 with fibrotic lesions caused by advanced nAMD, as diagnosed on fundus biomicroscopy or color photography. All included patients had previously received long-standing anti-VEGF therapy, using an as needed regimen.
      This study consisted of 2 separate analyses on 2 distinct series of eyes. The first analysis consisted of a description of the spectral-domain OCT morphologic features observed in the wide spectrum of fibrotic lesions associated with nAMD but without active exudative features at the time of analysis. The second analysis consisted of a retrospective study, in nAMD eyes, of the sequence of progression from the original CNV lesion to the fibrotic scar.
      In summary, 2 distinct series of eyes were retrospectively included in the present study: 1) a cross-sectional series of nAMD eyes with lesions within the fibrotic spectrum and 2) a series of eyes with fibrotic scars secondary to nAMD, for which a longitudinal long-term follow-up (over 5 years) was available.

       Inclusion and Exclusion Criteria

      Eligibility criteria included a diagnosis of long-standing nAMD, based on spectral-domain OCT, FA, and the international AMD classification
      • Bird A.C.
      • Bressler N.M.
      • Bressler S.B.
      • et al.
      An international classification and grading system for age-related maculopathy and age-related macular degeneration. The International ARM Epidemiological Study Group.
      and a history of no antiangiogenic treatment for ≥3 months. The absence of exudative signs on spectral-domain OCT—including subretinal and intraretinal fluid—was also required for both the cross-sectional and longitudinal cohorts. On spectral-domain OCT the lesion was required to fit within the borders of the 30° × 30° field of view scan raster. For the longitudinal series, follow-up of ≥5 years was mandatory.
      Diagnosis of fibrosis was based on fundus biomicroscopy and color pictures. Only patients that underwent both FA and spectral-domain OCT at diagnosis of nAMD, as well as spectral-domain OCT during follow-up, were included in the cross-sectional and longitudinal analysis.
      On fundus examination, a fibrotic lesion was defined as a well demarcated, elevated mound of yellowish-white tissue.
      • Toth L.A.
      • Stevenson M.
      • Chakravarthy U.
      Anti-vascular endothelial growth factor therapy for neovascular age-related macular degeneration: outcomes in eyes with poor initial vision.
      On FA, fibrosis caused by late AMD displayed staining, with minimal or no leakage in the late phase of the angiographic sequence.
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Bird A.C.
      • Bressler N.M.
      • Bressler S.B.
      • et al.
      An international classification and grading system for age-related maculopathy and age-related macular degeneration. The International ARM Epidemiological Study Group.
      ,
      • Kaiser P.K.
      • Blodi B.A.
      • Shapiro H.
      • Acharya N.R.
      MARINA Study Group
      Angiographic and optical coherence tomographic results of the MARINA study of ranibizumab in neovascular age-related macular degeneration.
      On spectral-domain OCT, the lesion was defined as fibrotic if >50% of its area was occupied by compact, sheet-like hyperreflective material, situated either above or underneath the RPE. In addition, in the longitudinal cohort only patients without scar at the initial examination were included.
      Exclusion criteria included cases of fibrosis secondary to causes other than AMD. Eyes with diseases other than AMD were excluded, such as cases of adult onset foveomacular vitelliform dystrophy, myopic CNV or pathologic myopia, and inherited macular dystrophies. Patients with media opacities preventing an adequate fundus view were also excluded.
      AMD eyes that progressed from nAMD to GA with no evidence of fibrosis, as well as eyes with sub-RPE scar/fibrosis/hyperreflective material that did not have a yellow-whitish aspect on fundus photography were excluded. In addition, patients lost to follow-up were excluded. For the cross-sectional series, patients harboring a fibroatrophic flat pattern on spectral-domain OCT (thickness of the fibrotic lesion <100 μm) were excluded.

       Procedures

      All patients in the study underwent a complete ophthalmic examination, consisting of best-corrected visual acuity (BCVA), fundus biomicroscopy, FA, and spectral-domain OCT. Color fundus photography (Canon CR-2 Retinal Camera; Canon, Tokyo, Japan) was also performed in a subset of eyes. MultiColor imaging (using 3 monochromatic laser sources: blue reflectance [wavelength 488 nm], green reflectance [wavelength 515 nm], and infrared reflectance [wavelength 820 nm]) was also performed in a subset of patients.
      • Tan A.C.S.
      • Fleckenstein M.
      • Schmitz-Valckenberg S.
      • Holz F.G.
      Clinical application of multicolor imaging technology.
      MultiColor, FA, and spectral-domain OCT were performed using Spectralis HRA+OCT (Heidelberg Engineering, Heidelberg, Germany). Each spectral-domain OCT image consisted of a 19-line horizonal volume scan, 30° × 30° field of view, centered on the fovea, with a minimum averaging of 9 frames.

       Cross-Sectional Series

      For the cross-sectional analysis, lesions at the last follow-up were assessed on spectral-domain OCT images of study eyes by 2 readers (M.A-R., A.S.). In cases of disagreement, a further assessment was performed by a third expert reader (E.H.S.). The cross-sectional analysis was performed on all spectral-domain OCT slices corresponding to the fibrotic lesion of each study eye.
      An additional exclusion criterion was applied for this series: lesions with features of fibrosis on color images and features of atrophy on spectral-domain OCT were classified as fibroatrophic lesions (FALs) and excluded. FALs exhibit a maximum height of <100 μm and display homogenous reflectivity with spectral-domain OCT.
      The presence/absence of the following features were evaluated by spectral-domain OCT: presence of abnormalities within the fibrotic lesion (ie, hyperreflective lamellae, hyporeflective lamellae, or hyporeflective spaces), presence of perilesional atrophy, presence of outer retinal tubulations or intraretinal degenerative spaces, and loss of adjacent RPE and ellipsoid zone bands.
      The borders of the lesions of the fibrotic scar were classified as well defined or ill-defined and the location of the fibrotic scar with respect to the RPE was also noted (Figure 1). Spectral-domain OCT measurements included: greatest fibrotic lesion diameter, maximal subfoveal lesion height, and maximal lesion height. If maximal lesion height <100 μm, the lesion was considered a FAL and thus excluded from this cross-sectional study. These measurements were performed manually using the spectral-domain OCT caliper tool. Greatest lesion diameter corresponded to the maximal horizontal diameter of the lesion on a spectral-domain OCT slice. Maximal subfoveal lesion height corresponded to the manual measurement of the fibrotic lesion in the subfoveal area, while maximal lesion height corresponded to the greatest lesion height regardless of its location (sub- or extrafoveal). Central macular thickness measurements were calculated using the automated spectral-domain OCT tool. To determine the location of the lesion (ie, sub-, juxta-, or extrafoveal), an Early Treatment Diabetic Retinopathy Study (ETDRS) macular subfield grid was placed manually at the center of the macula.
      Figure thumbnail gr1
      Figure 1Multimodal imaging of the right eye of an 84-year-old patient with a lesion within the fibrotic spectrum secondary to late neovascular age-related macular degeneration (nAMD). (Upper left) Color fundus photography shows a well-demarcated yellow, elevated mound of tissue in the macular area. (Upper middle) Early fluorescein angiography (FA) reveals a slight staining of the lesion. (Upper right) Late staining on FA with no accompanying leakage. (Lower panels) Horizontal slices of spectral-domain optical coherence tomography (spectral-domain OCT) guided by FA reveal that the location of the lesion within the fibrotic spectrum with respect to the RPE is different: on the lower left panel, the lesions is located underneath a clearly visible RPE, while in the lower right panel the fibrotic lesion is subretinal, with no clear visualization of the RPE.

       Longitudinal Series

      In addition to the cross-sectional description of the spectral-domain OCT features of fibrovascular lesions, the progression of the original lesion to the final fibrovascular phenotype was studied. To delineate retrospectively the progression pathway of fibrotic scars observed on spectral-domain OCT, we included a distinct series of patients with fibrotic scars secondary to advanced nAMD with long follow-up (≥5 years). In this series, only patients without fibrosis at first examination in our imaging records were included.
      For the longitudinal series, the retrospective analysis of spectral-domain OCT examinations was performed by 2 independent readers (E.H.S., A.M.). In cases of disagreement, a further assessment was performed by a third expert reader (G.Q.). A single spectral-domain OCT slice, passing through the center of the lesion, was analyzed for each study eye on each visit. The spectral-domain OCT line passing through the center of the lesion within the fibrotic spectrum was analyzed retrospectively at each visit, in order to detect the progression pathway from the baseline type of CNV to the fibrotic scar. Baseline type 1 CNV was defined on spectral-domain OCT as a neovascular lesion localized underneath the RPE, associated with a PED.
      • Mrejen S.
      • Sarraf D.
      • Mukkamala S.K.
      • Freund K.B.
      Multimodal imaging of pigment epithelial detachment: a guide to evaluation.
      ,
      • Mehta H.
      • Tufail A.
      • Daien V.
      • et al.
      Real-world outcomes in patients with neovascular age-related macular degeneration treated with intravitreal vascular endothelial growth factor inhibitors.
      On FA, type 1 CNV was characterized by an ill-defined hyperfluorescent lesion, with heterogenous or stippled late staining and leakage. Type 2 CNV was defined on spectral-domain OCT as a hyperreflective neovascular lesion situated above the RPE. The diagnostic criteria for well-defined CNV on FA included a well-demarcated lacy area of early hyperfluorescence with progressive leakage of dye in the overlying subsensory retinal space during the late phases of the angiogram. The margins of CNV were well-defined in the early phase.
      • Mehta H.
      • Tufail A.
      • Daien V.
      • et al.
      Real-world outcomes in patients with neovascular age-related macular degeneration treated with intravitreal vascular endothelial growth factor inhibitors.
      Type 3 NV was defined as an intraretinal hyperreflective lesion, situated above a (serous/drusenoid) PED. Angiographic features for diagnosis of type 3 NV included a discrete late focal area of hyperfluorescence (focal staining) or a typical “hot spot” (focal leakage).
      • Su D.
      • Lin S.
      • Phasukkijwatana N.
      • et al.
      An updated staging system of type 3 neovascularization using spectral domain optical coherence tomography.
      Aneurysmal type 1 neovascularization (polypoidal choroidal vasculopathy [PCV]) was defined as a peaked PED (corresponding to the aneurysmal lesion) adjacent to a shallow irregular PED (corresponding to type 1 neovascularization).
      • Dansingani K.K.
      • Gal-Or O.
      • Sadda S.R.
      • Yannuzzi L.A.
      • Freund K.B.
      Understanding aneurysmal type 1 neovascularization (polypoidal choroidal vasculopathy): a lesson in the taxonomy of “expanded spectra” - a review.
      On ICGA, aneurysmal type 1 CNV (PCV) showed the typical branching vascular network and hyperfluorescent polypoidal dilations. The presence of the following characteristics was assessed from baseline nAMD diagnosis to the last visit: type of baseline CNV, presence/absence of large macular hemorrhage (>1 disc diameter), presence/absence of SHRM, presence/absence of an RPE tear, integrity of the RPE band, presence/absence of RPE erosion, and presence/absence of accompanying perilesional atrophy. Focal RPE erosion was defined as a single limited erosion of RPE visible on a spectral-domain OCT scan, whereas a dashed aspect of the RPE was termed RPE erosions.

       Statistical Analysis

      Statistical analysis was performed using Stata software (v 13.0; Statacorp LP, College Station, Texas, USA) and included descriptive statistics for main clinical features. BCVA was converted from Snellen to ETDRS letters for statistical analysis. Fisher exact and χ2 tests were used to compare qualitative variables and the Mann-Whitney test was used to compare quantitative data. P = .05 was considered statistically significant.

      Results

       Cross-Sectional Analysis of Lesions Within the Fibrotic Spectrum

       Demographics and Clinical Characteristics

      Forty-four eyes of 39 patients (15 men and 24 women) with a mean (±SD) age of 83 ± 6 years were included in the cross-sectional analysis. nAMD was diagnosed on average 68 ± 31 months before inclusion (range 5-116 months). The mean BCVA at last follow-up was of 42 ± 27 ETDRS letters (Snellen equivalent 20/160). All patients underwent anti-VEGF treatment (ranibizumab and/or aflibercept) with a mean of 20 (±13) intravitreal injections from diagnosis of nAMD to study inclusion. Study eyes did not undergo antiangiogenic treatment on average the last 10 ± 17 months before inclusion.
      In our cross-sectional series, lesions within the fibrotic spectrum were caused by type 1 CNV in 13 of 44 eyes (30%), type 2 CNV in 9 of 44 eyes (20%), mixed type 1 and 2 CNV in 16 of 44 eyes (36%), and type 3 neovascularization in 3 of 44 eyes (7%). The type of initial CNV could not be determined for 3 of 44 eyes. Fibrotic lesions were located in the subfoveal region in 41 of 44 eyes (93%), while the lesions were located in the extrafoveal region in only 3 of 44 eyes (7%).
      In 5 of 39 patients from the cross-sectional series, both eyes exhibited fibrotic lesions and were therefore included according to our criteria, accounting for 10 of 44 eyes in the series (23%). Concerning the fellow eyes of the remaining 34 of 39 patients, they displayed fibrotic lesions in 10 eyes, which did not meet the inclusion criteria (such as history of no antiangiogenic treatment for ≥3 months, absence of exudative signs on spectral-domain OCT, or a lesion area within the borders of the 30° × 30° field of view scan raster). In the remaining 24 of 39 patients, the fellow eye failed to show a fibrotic scar and exhibited type 1 CNV in 8 cases, type 3 NV in 1 case, GA in 9 cases, GA plus CNV in 2 cases, and intermediate AMD in 4 cases.

       Spectral-Domain OCT Features of Lesions Within the Fibrotic Spectrum

      We analyzed a total of 836 spectral-domain OCT slices (44 eyes × 19 slices per eye) through all fibrotic lesions in our cross-sectional cohort and the detailed features studied are shown in Figure 1.
      Fibrosis was located above the RPE in 8 of 44 eyes (40/836 slices) and underneath the RPE in 43 of 44 eyes (430/836 slices). The RPE was not visible on spectral-domain OCT in 6 of 44 eyes (30/836 slices) and was intact in 38 eyes (470/836 slices). On spectral-domain OCT, the borders of the fibrotic lesion were morphologically well defined or fairly defined in 38 of 44 eyes (well defined in 154/836 slices, fairly defined in 316/836), or poorly defined in 6 of 44 eyes (30/836 slices).
      Abnormalities such as hyperreflective lamellae, hyporeflective lamellae, or hyporeflective spaces coexisted in 20 of 44 eyes within lesions belonging to the fibrotic spectrum (120/836 slices; Figure 2). Moreover, the presence of a hyporeflective/slightly hyperreflective band within the width of the fibrotic scar was detected on 22 of 44 eyes (100/836 slices).
      Figure thumbnail gr2
      Figure 2Type A lesions within the fibrotic spectrum, 3 eyes issued from the cross-sectional analysis. (Top) Subtype A1 lesion within the fibrotic spectrum in the left eye of an 81-year-old patient. Multicolor imaging reveals a yellow-greyish lesion in the macular area. Horizontal slice of spectral-domain optical coherence tomography (OCT) guided by infrared imaging (IR) shows a hypereflective, subretinal pigment epithelium (sub-RPE) lesion (white arrow) with intralesional abnormalities: hyporeflective lamellae (yellow arrowhead) and hyporeflective spaces (asterisks). (Middle) Subtype A2 lesion within the fibrotic spectrum in the right eye of an 83-year-old patient. Multicolor imaging reveals a yellowish-green lesion in the macular area. Horizontal slice of spectral-domain (OCT) guided by IR shows a hypereflective, sub-RPE lesion (white arrows) with a hyporeflective band (red dotted arrows) within the entire width of the lesion. (Bottom) Subgroup A3 lesion within the fibrotic spectrum. Color fundus photography reveals a yellowish lesion in the macular area. Horizontal slice of spectral-domain OCT guided by IR shows a compact, homogeneous hypereflective lesion situated beneath the RPE (white arrows) without any intralesional abnormalities.
      The presence of perilesional atrophy, surrounding fibrosis, was detected on 24 of 44 eyes. Outer retinal tubulations were present in 29 of 44 eyes. Complete loss of RPE surrounding the lesion was present in 6 of 44 eyes. Loss of the ellipsoid zone was noted in 19 of 44 eyes.
      The mean central macular thickness was 299 ± 72 μm in our cohort. The mean greatest fibrotic lesion diameter averaged 3128 ± 1051 μm, the mean maximal subfoveal lesion height was 177 ± 127 μm, and the mean maximal lesion height was 277 ± 119 μm.
      Given the aforementioned features, multiple types of fibrotic lesions emerged, based on the location with respect to the RPE, RPE integrity (ie, RPE band intact and identified or not), and the morphologic aspect on spectral-domain OCT.
      First, type A lesions within the fibrotic spectrum were defined by the sub-RPE location on spectral-domain OCT. The RPE in this group was mainly intact and preserved. Lesions included in this group were well defined on spectral-domain OCT. Type A lesions were a common feature in this series and were present in 43 of 44 eyes (98%) and 430/836 slices. Among this group of lesions, we distinguished different patterns based on the sub-RPE reflectivity of the vascularized PED.
      • Subtype A1 (20/44 eyes) was characterized by a heterogeneous hyperreflective lesion with the presence of intralesional abnormalities, as follows: hyperreflective lamellae, hyporeflective lamellae, and hyporeflective spaces. This subtype was present on average in 6 slices per eye, thus on a mean of 120 of 836 slices (Figure 2, Top).
      • Subtype A2 (22/44 eyes) was characterized by the presence of a hyporeflective/slightly reflective band within the entire width of the lesion. This subtype was present in average on 5 slices per eye, thus on a mean of 100 of 836 slices (Figure 2, Middle).
      • Subtype A3 (42/44 eyes) was characterized by the presence of compact, homogeneous hyperreflective lesion situated beneath the RPE, without intralesional irregularities. This subtype was present in average on 5 slices per eye, thus on a mean of 210 of 836 slices (Figure 2, Bottom).
      Second, type B lesions within the fibrotic spectrum were defined by the subretinal localization on spectral-domain OCT. The RPE in these cases was located within the lesion, separating it into 2 parts (subretinal and sub-RPE). Fibrotic lesions included in this type were consistently hyperreflective and well defined. Type B lesions were present in 8 of 44 eyes (19%) and in 40 of 836 slices (Figure 3).
      Figure thumbnail gr3
      Figure 3Type B lesion within the fibrotic spectrum, 2 eyes issued from the transverse analysis. Each line represents the multimodal imaging of a different case. (Top left) Color fundus photography revealing a well-defined yellowish lesion in the macular area. (Top middle and top right) Infrared imaging (middle) guiding a horizontal slice of spectral-domain optical coherence tomography (OCT). Spectral-domain OCT (right) guided by infrared imaging shows a subretinal compact, homogeneous, hypereflective lesion. Note that the retinal pigment epithelium (RPE) (blue arrows) in this case was localized within the lesion, separating it in 2 parts: subretinal (red dotted arrows) and sub-RPE fibrosis (yellow dotted arrows). (Bottom) A second case. (Bottom left) Multicolor imaging, revealing a well-defined, bright-yellow lesion in the macular area. (Bottom middle and bottom right) Infrared imaging guiding a horizontal slice of spectral-domain OCT. Spectral-domain OCT guided by infrared imaging shows a subretinal compact, homogeneous, hypereflective lesion. Note that the RPE (blue arrows) in this case was localized within the lesion, separating it in 2 parts: subretinal (red dotted arrows) and sub-RPE fibrosis (yellow dotted arrows).
      Finally, type C lesions within the fibrotic spectrum were defined by the subretinal localization on spectral-domain OCT, with no discernable RPE and an ill-defined pattern. However, RPE was visible on the edges of the fibrotic lesion. This subtype illustrated a heterogeneous, protuberant lesion. Type C fibrotic lesions were present in 6 of 44 eyes (14%) and in 30 of 836 slices (Figure 4).
      Figure thumbnail gr4
      Figure 4Type C lesion within the fibrotic spectrum. (Top left) Multicolor imaging reveals a yellowish-grey lesion in the macular area. (Top right) Infrared and correscponding scan line. (Bottom) Horizontal slice of spectral-domain optical coherence tomography (OCT) reveals a hypereflective, heterogeneous, prominent, ill-defined, subretinal lesion with no discernable retinal pigment epithelium overlaying the lesions. Note that the retinal pigment epithelium was visible on the edges of the lesion (white arrows).
      The characteristics of each type and subtype of lesions within the fibrotic spectrum are summarized in Table 1. Figure 5 summarizes the different types of lesions described above. These morphologic features and the respective correlations with BCVA, number of intravitreal injections, presence of degenerative signs, and spectral-domain OCT measurements are summarized in Table 2.
      Table 1Cross Sectional Series: Description of Lesions Within the Fibrotic Spectrum According to Spectral-Domain Optical Coherence Tomography Morphologic Features
      Spectral-Domain OCT Morphologic Features
      Lesions Within the Fibrotic SpectrumLocationHyperreflective LesionRPE Totally/Partially PreservedIntralesional Abnormalities (Hyporeflective Spaces, Hyperreflective Lamellae)Hyporeflective Band Through the LesionShape
      Type A
       Subtype A1Sub-RPE+±+Well defined
       Subtype A2Sub-RPE+±+Well defined
       Subtype A3Sub-RPE+±Well defined
      Type BSubretinal+±±Well defined
      Type CSubretinal+±Complex shape
      OCT = optical coherence tomography; RPE = retinal pigment epithelium.
      Figure thumbnail gr5
      Figure 5Illustrative drawing with spectral-domain optical coherence tomography examples of types of lesions within the fibrotic spectrum. Subtypes A1, A2, and A3 correspond to subretinal pigment epithelium (sub-RPE) lesions with (A1 and A2) or without (A3) intralesional abnormalities, while type B corresponds to subretinal fibrosis. A subretinal protuberant lesion with no visible overlaying RPE, however, characterizes type C. BM = Bruch membrane; CC = choriocapillaris.
      Table 2Cross-Sectional Series: Correlations Between Types of Lesions Within the Fibrosis Spectrum and Quantitative Criteria
      Type AType BType C
      Subtype A1Subtype A2Subtype A3
      Eyes (nb)20/4422/4442/448/446/44
      BCVA (ETDRS letters), mean (P)39.7 (.30)
      Quantitative analysis performed using the Student t test.
      47.9 (.14)
      Quantitative analysis performed using the Student t test.
      42.6 (.70)
      Quantitative analysis performed using the Student t test.
      34.6 (.32)
      Quantitative analysis performed using the Student t test.
      34.2 (.46)
      Quantitative analysis performed using the Student t test.
      IVT mean, n (P)23 (.24)
      Quantitative analysis performed using the Student t test.
      17 (.04)
      Quantitative analysis performed using the Student t test.
      21 (.27)
      Quantitative analysis performed using the Student t test.
      15 (.38)
      Quantitative analysis performed using the Student t test.
      15 (.47)
      Quantitative analysis performed using the Student t test.
      Degenerative signs, eyes (P)14/20 (.19)
      Qualitative analysis performed using the χ2 nonparametric test.
      13/22 (1)
      Qualitative analysis performed using the χ2 nonparametric test.
      23 (.6)
      Qualitative analysis performed using the χ2 nonparametric test.
      3 (.31)
      Qualitative analysis performed using the χ2 nonparametric test.
      3 (.69)
      Qualitative analysis performed using the χ2 nonparametric test.
      Eyes with CNV, n (P)
       Type 16/20 (.79)
      Qualitative analysis performed using the χ2 nonparametric test.
      8/22 (1)
      Qualitative analysis performed using the χ2 nonparametric test.
      16/22 (.89)
      Qualitative analysis performed using the χ2 nonparametric test.
      1/8 (.41)
      Qualitative analysis performed using the χ2 nonparametric test.
      1/6 (.97)
      Qualitative analysis performed using the χ2 nonparametric test.
       Type 24/20 (.5)
      Qualitative analysis performed using the χ2 nonparametric test.
      2/22 (.66)
      Qualitative analysis performed using the χ2 nonparametric test.
      6/22 (.63)
      Qualitative analysis performed using the χ2 nonparametric test.
      1/8 (.69)
      Qualitative analysis performed using the χ2 nonparametric test.
      1/6 (.58)
      Qualitative analysis performed using the χ2 nonparametric test.
       Mixed 1 and 26/20 (.63)
      Qualitative analysis performed using the χ2 nonparametric test.
      8/22 (1)
      Qualitative analysis performed using the χ2 nonparametric test.
      14/22 (.25)
      Qualitative analysis performed using the χ2 nonparametric test.
      4/8 (.71)
      Qualitative analysis performed using the χ2 nonparametric test.
      2/6 (.97)
      Qualitative analysis performed using the χ2 nonparametric test.
       Type 31/20 (.87)
      Qualitative analysis performed using the χ2 nonparametric test.
      2/22 (1)
      Qualitative analysis performed using the χ2 nonparametric test.
      3/22 (.29)
      Qualitative analysis performed using the χ2 nonparametric test.
      2/8 (.16)
      Qualitative analysis performed using the χ2 nonparametric test.
      0/6 (.97)
      Qualitative analysis performed using the χ2 nonparametric test.
      Spectral-domain OCT measurements, μm, mean (P)
       CMT307 (.17)
      Quantitative analysis performed using the Student t test.
      306.9 (.16)
      Quantitative analysis performed using the Student t test.
      298.2 (.53)
      Quantitative analysis performed using the Student t test.
      291.5 (.87)
      Quantitative analysis performed using the Student t test.
      315.5 (.63)
      Quantitative analysis performed using the Student t test.
       GLD4025 (.05)
      Quantitative analysis performed using the Student t test.
      3566.5 (.8)
      Quantitative analysis performed using the Student t test.
      3681 (.08)
      Quantitative analysis performed using the Student t test.
      3695.3 (1)
      Quantitative analysis performed using the Student t test.
      4039.8 (.33)
      Quantitative analysis performed using the Student t test.
       MSFH226 (.23)
      Quantitative analysis performed using the Student t test.
      227 (.2)
      Quantitative analysis performed using the Student t test.
      195.3 (.25)
      Quantitative analysis performed using the Student t test.
      225.5 (.14)
      Quantitative analysis performed using the Student t test.
      271.2 (.62)
      Quantitative analysis performed using the Student t test.
       MLH317 (.1)
      Quantitative analysis performed using the Student t test.
      286 (.78)
      Quantitative analysis performed using the Student t test.
      277.2 (.57)
      Quantitative analysis performed using the Student t test.
      312 (.08)
      Quantitative analysis performed using the Student t test.
      383.5 (.08)
      Quantitative analysis performed using the Student t test.
      BCVA = best-corrected visual acuity; CMT = central macular thickness; CNV = choroidal neovascularization; ETDRS = Early Treatment of Diabetic Retinopathy Study; GLD = greatest lesion diameter; IVT = intravitreal injection; MLH = maximal lesion height; MSFH = maximal subfoveal lesion height; OCT = optical coherence tomography.
      a Quantitative analysis performed using the Student t test.
      b Qualitative analysis performed using the χ2 nonparametric test.

       Longitudinal Analysis of Fibrotic Lesions

       Demographics and Clinical Characteristics

      Fifty-four eyes, distinct from the cross-sectional series, adhered to the inclusion criteria. Seven eyes were excluded because of previous laser photocoagulation or PDT. Forty-seven eyes of 47 patients were included in the final longitudinal analysis (17 men and 30 women). nAMD was diagnosed on average 89 ± 35 months before the last visit in the longitudinal analysis cohort. Mean BCVA at last follow-up was of 35 ± 20 ETDRS letters (Snellen equivalent 20/200). All patients underwent anti-VEGF treatment (ranibizumab and/or aflibercept) with a mean of 37 ± 16 intravitreal injections from the time of diagnosis of nAMD to study inclusion. All patients received an as needed regimen. As for the fellow eye of the 47 patients included in the longitudinal series, 11 eyes exhibited intermediate AMD at the initial visit, 14 showed type 1 CNV, 1 displayed type 2 CNV, 2 exhibited type 3 NV, 6 showed GA, 5 displayed GA complicated by CNV, and 7 exhibited fibrosis (Table 3).
      Table 3Longitudinal Series: Distribution of Types of Choroidal Neovascularization Within the Longitudinal Series at Baseline (Neovascular Age-related Macular Degeneration Diagnosis) and at the Time of the Angiofibrotic Switch, According to the Initial Lesions Within the Fibrotic Spectrum Type
      Baseline CNV type (N = 47)CNV TypePercentageEyes, n
      Type 1 CNV55%26
      Type 2 CNV26%12
      Type 3 NV11%5
      Mixed type 1 and 2 CNV6%3
      Aneurysmal type 1 CNV2%1
      Time between baseline CNV diagnosis and angiofibrotic switch, months, mean ± SD20 ± 19
      Fellow eye statusIntermediate AMD, 11/47; type 1 CNV, 14/47; type 2 CNV, 1/47; type 3 NV, 2/47; GA, 6/47; GA plus CNV, 5/47; fibrosis, 7/47
      Angiofibrotic switch (N = 47)Baseline type of CNVFirst fibrosis type
      Type AType BType C
      Type 1 (n = 26)2312
      Type 2 (n =12)372
      Type 3 NV (n = 5)32
      Mixed type 1 and 2 CNV (n = 3)111
      Aneurysmal type 1 CNV (n = 1)1
      AMD = age-related macular degeneration; CNV = choroidal neovascularization; GA = geographic atrophy; SD = standard deviation.

       Spectral-Domain OCT Features of Fibrotic Lesions During Follow-up

      We retrospectively analyzed a total of 4181 spectral-domain OCT slices passing through the center of the fibrotic lesion at each visit for the 47 included eyes. The previously described types of lesions within the fibrotic spectrum were used to define the progression pathway of each patient.
      At baseline nAMD diagnosis, type 1 CNV was present in 26 of 47 eyes (55%), type 2 CNV in 12 of 47 eyes (26%), mixed type 1 and 2 CNV in 3 of 47 eyes (6%), type 3 neovascularization in 5 of 47 eyes (11%), and aneurysmal type 1 in 1 eye (2%). Lesions within the fibrotic spectrum developed after 20 ± 19 months from the baseline diagnosis of CNV. Moreover, macular hemorrhage was observed during follow-up in 13 of 47 eyes (28%). The presence of macular hemorrhage was not statistically associated with a baseline type of CNV (P = 1, Fisher exact test).
      RPE tear occurred during follow-up in 10 of 47 eyes (21%). RPE tear was not statistically associated with a particular type of baseline CNV (P = .313, Fisher exact test). During follow-up, SHRM was noted in 14 of 47 eyes (Figure 6, Figure 7). This SHRM situated above the RPE was statistically associated with type 1 CNV at presentation (P = .007, Fisher exact test) and preceded the transition from type A to type B lesions within the fibrotic spectrum. RPE erosions (Figure 6, Figure 7, Figure 8, Figure 9) were detected in 28 of 47 eyes during follow-up and was not associated with a type of initial CNV in particular (P = .815, Fisher exact test). Perilesional macular atrophy was detected in 13 of 47 eyes on spectral-domain OCT (Figure 9).
      Figure thumbnail gr6
      Figure 6Long-term follow-up of the right eye of a 71-year-old patient, evolving from type 2 choroidal neovascularization (CNV), to type A, type B, and type C fibrosis. (Top panels) Multimodal imaging at baseline diagnosis of neovascular age-related macular degeneration. Multicolor imaging, early fluorescein angiography (FA), and early indocyanine green angiography (ICGA) show a well demarcated lesion, hyperfluorescent on early frames, suggestive for type 2 CNV. Spectral-domain optical coherence tomography at baseline reveals a hypereflective lesion above the retinal pigment epithelium, consistent with the angiographic findings of type 2 CNV. At 1 year of follow-up, the lesion had regressed under the RPE on spectral-domain OCT. Although subretinal fluid was still present, the content of the pigment epithelial detachment seemed rather hypereflective and heterogeneous. This aspect is suggestive of type A fibrosis. At 3 years of follow-up, we noted that fibrosis was located above the RPE (asterisk), as well as underneath the RPE, with a heterogeneous multilayered aspect. This is suggestive of type B fibrosis. On the magnification corresponding to type B fibrosis, note the discontinuous RPE erosions (white arrows). At 5 years from baseline, spectral-domain OCT revealed a hypereflective, protuberant lesion situated in the subretinal space. RPE is not visible within the lesion (white arrowhead for the sudden interruption of the RPE on 1 edge of the lesion), which is typical of type C fibrosis. (Bottom left) Multimodal imaging at last follow-up revealed, on multicolor imaging, a grey-green lesion. (Bottom middle) FA revealed staining with no late leakage from the lesion. (Bottom right) Early indocyanine green angiography shows that the lesion is slightly and heterogeneously hyperfluorescent. Therefore, in this patient the pathway of progression from the initial type 2 CNV to the late-stage fibrotic lesions could be summarized as A → B → C.
      Figure thumbnail gr7
      Figure 7Long-term follow-up of the right eye of an 82-year-old patient evolving from type 1 choroidal neovascularization (CNV) to type A, type B, and type C fibrosis. (Top panels) Multimodal imaging at baseline diagnosis of neovascular age-related macular degeneration. Early and late frames of fluorescein angiography (FA, upper left panels) and early and late indocyanine green angiography (ICGA, right panels) show a hyperfluorescent, ill-defined lesion on early frames, generating a late hyperfluorescent plaque on ICGA, suggestive of type 1 CNV. Spectral-domain optical coherence tomography (OCT) at baseline reveals a moderately hypereflective pigment epithelial detachment (PED), accompanied by subretinal fluid, consistent with the angiographic findings of type 1 CNV. At 1 year of follow-up we noted that the content of the PED seemed multilayered, which is suggestive of type A fibrosis. At 2 years of follow-up from baseline, we noted that while type A fibrosis is still present, a well-defined subretinal hypereflective material (SHRM) was also present (white arrowhead). At 4 years from baseline, fibrosis is located both located above the retinal pigment epithelium (RPE) as a homogenous, hypereflective lesion, as well as underneath the RPE, with a heterogeneous multilayered aspect. This is suggestive of type B fibrosis. Note the chronic subretinal fluid. On the magnification corresponding to type B fibrosis, note the discontinuous RPE erosions (white arrows). At 7 years from baseline, spectral-domain OCT revealed a hypereflective, prominent lesion situated in the subretinal space. RPE is not visible within the lesion, suggesting a progression to type C fibrosis. In this patient, the pathway of progression from the initial type 1 CNV to the late-stage fibrotic lesions could be summarized as A → B → C.
      Figure thumbnail gr8
      Figure 8Long-term follow-up of the right eye of an 83-year-old patient evolving from type 2 choroidal neovascularization (CNV) to type B and type C fibrosis. (Top panels) Multimodal imaging at baseline diagnosis of neovascular age-related macular degeneration. Early fluorescein angiography (FA, top left panel) reveals a hyperfluorescent subfoveal lesion, as well as a subretinal hemorrhage generating masking. Late indocyanine green angiography (ICGA, top right panel) shows a hyperfluorescent late plaque on ICGA. Spectral-domain optical coherence tomography (OCT) at baseline reveals a hypereflective lesion above the RPE and a moderately hypereflective pigment epithelial detachment (PED), accompanied by subretinal fluid, consistent with mixed type 1 and 2 (minimally classic) neovascularization presenting with subretinal hemorrhage. After the monthly loading dose, we noted that a hypereflective, homogenous fibrotic scar was located above the RPE, while the initial small PED had a heterogeneous multilayered aspect. Note the discontinuity of the RPE underneath the fibrotic scar (white arrow). This is suggestive of type B fibrosis. At 2 years of follow-up, spectral-domain OCT revealed a hypereflective, prominent lesion situated in the subretinal space. Note that RPE is not visible within the lesion, suggesting progression from type B to type C fibrosis. In this patient, the pathway of progression from the initial mixed type 1 and 2 CNV to the late-stage fibrotic lesions could be summarized as B → C.
      Figure thumbnail gr9
      Figure 9Long-term follow-up of the left eye of a 90-year-old patient evolving from type 1 choroidal neovascularization (CNV) to a fibroatrophic lesion. (Top panels) Multimodal imaging at baseline diagnosis of neovascular age-related macular degeneration. Early fluorescein angiography (FA, top left panel) reveals an ill-defined hyperfluorescent lesion. Late FA (top middle panel) reveals late leakage and hyperfluorescent pinpoints. Late indocyanine green angiography (ICGA, top right panel) shows a hyperfluorescent late plaque. Spectral-domain optical coherence tomography at baseline revealed a moderately hypereflective pigment epithelial detachment (PED), accompanied by abundant subretinal fluid, consistent with type 1 neovascularization. The patient underwent the monthly loading dose and an as needed regimen afterward. At 1 year of follow-up, the initial PED seemed larger, containing hypereflective and hyporeflective lamellae, consistent with type A fibrosis. Throughout time, the aspect of the PED remained suggestive for type A fibrosis, with a progressive disappearance of the hyporeflective lamellae and a homogenous hypereflective aspect at 3 years of follow-up. Note that at 7 years of follow-up, multicolor imaging (lower left panel) revealed the presence of a yellow-grey macular lesion with concave borders (white arrowhead), as well as of well-defined patches of atrophy. The atrophy is also visible on both infrared imaging (lower middle panel) and fundus autofluorescence imaging (lower right panel). Therefore, in this patient, the pathway of progression from the initial type 1 CNV to fibrotic lesions falls into the fibroatrophic pathway (A → FAL). In all cases, perilesional atrophy “blocked” fibrosis extension in patients presenting with a degree of perilesional atrophy at baseline.

       Progression Pathways of Fibrotic Lesions

      During follow-up, evidence of the angiofibrotic switch (defined previously as the transition from the angiogenic phase to the fibrotic phase)
      • Kuiper E.J.
      • de Smet M.D.
      • van Meurs J.C.
      • et al.
      Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
      was noted after a mean of 20 ± 19 months from the baseline diagnosis of CNV to 1 of the 3 fibrotic lesions:
      • Type A lesions in 30 of 47 eyes. Among the 30 type A lesions, 13 remained stationary throughout follow-up and progressed to FAL (Figure 9). The remaining 17 of 30 eyes progressed from type A toward type B and finally to type C lesions (Figure 6, Figure 7). As shown in Table 4, focal RPE erosion of the RPE associated with SHRM was observed as an early stage during the progression between type A and type B stages, whereas multiple RPE erosions were observed during the progression between type B and type C stages (FGL).
        Table 4Longitudinal Series: Progression Pathways Toward Fibrotic Lesions for Each Type of Baseline Choroidal Neovascularization
        Progression PathwayType 1 CNVType 2 CNVType 3 NVMixed Type 1 and 2 CNVAneurysmal Type 1SHRM, n = 14Macular Hemorrhage, n =13RPE Tear, n = 10RPE Erosions, n = 28Perilesional Macular Atrophy, n = 13
        A → B → C13211A → B
        P < .001, Fisher exact test for the association between A → B progression pathway and occurrence of macular hemorrhage.
        A → B
        P = .058, Fisher exact test for the association between progression pathways and macular hemorrhage.
        B → C
        P = .058, Fisher exact test for the association between progression pathways and macular hemorrhage.
        A → BB → C
        P < .001, Fisher exact test for the association between B → C progression pathway and occurrence of RPE crumbling.
        B → C39221B → C
        P = .058, Fisher exact test for the association between progression pathways and macular hemorrhage.
        B → CB → C
        P < .001, Fisher exact test for the association between B → C progression pathway and occurrence of RPE crumbling.
        A → FAL1012A → A plus GA
        P < .001, Fisher exact test for the association between A → A plus GA progression pathway and occurrence of perilesional atrophy.
        FAL = fibroatrophic lesion; GA = geographic atrophy; RPE = retinal pigment epithelium; SHRM = subretinal hyperreflective material.
        a P < .001, Fisher exact test for the association between A → B progression pathway and occurrence of macular hemorrhage.
        b P = .058, Fisher exact test for the association between progression pathways and macular hemorrhage.
        c P < .001, Fisher exact test for the association between B → C progression pathway and occurrence of RPE crumbling.
        d P < .001, Fisher exact test for the association between A → A plus GA progression pathway and occurrence of perilesional atrophy.
      • Type B lesions in 10 of 47 eyes. In these 10 eyes, the original NV lesions directly progressed to a type B fibrotic scar without evidence of transition through stage A. All eyes (10/10) following this progression pathway evolved toward type C, with progressive RPE erosions and disappearance in all eyes (Figure 8).
      • Type C lesions in 7 of 47 eyes. In these 7 eyes, the original NV lesion directly evolved to a type C fibrotic scar without evidence of transition through stages A or B. These eyes presented with a protuberant fibrotic lesion with no remaining visible RPE layer after the initial monthly loading dose (Figure 8).
      Therefore, 3 progression pathways to fibrotic scars were notable after evaluation of all the spectral-domain OCT scans available from baseline nAMD diagnosis to final follow-up, using the aforementioned scale (A, B, and C) of fibrotic lesions.
      • The first pathway, for which we coined the term “A → B → C,” was characterized by a progression from type A to type B and then type C. This progression pathway appeared in 17 of 47 eyes from our longitudinal cohort and is shone in Figure 6, Figure 7. It is characterized by the initial presence of a multilayered vascularized PED (type A), secondary mostly to type 1 CNV (13/17). The appearance of SHRM above the RPE during follow-up preceded in these cases the development of subretinal fibrosis (type B). All “type A fibrotic lesions” included the A3 subtype defined above. A focal erosion of the RPE was observed as an early stage of progression, at the onset of SHRM (Figure 7). The final step in this progression pathway was loss of the RPE band as identified by spectral-domain OCT, which marked the development of the type C fibrotic lesion. Besides the original type 1 CNV lesion, we observed 3 of 17 type 2 or mixed CNV lesions and 1 of 17 type 3 NV lesions following the same pattern. The latter 4 cases progressed to type A fibrosis and followed the A → B → C pathway.
      • The second pathway, for which we coined the term “B→ C,” was characterized by a direct progression, under anti-VEGF therapy, from the baseline type of CNV to type B and/or C fibrotic lesions. A total of 17 of 47 eyes from our longitudinal series followed this progression pathway. This pathway is shown in Figure 8. It is noteworthy that 7 of 17 eyes transitioned to type C lesions without any evidence of a type B lesion during follow-up. Direct rapid progression to a type C fibrotic lesion was the result of a large subretinal hemorrhage (4/7) or RPE tear (2/7). We cannot exclude the possibility that a type B fibrotic lesion was transient and missed during this progression. The original CNV lesion following this pathway mainly consisted of baseline type 2 or mixed type 1/type 2 CNV (11/17). Besides type 2 CNV, we also observed 3 of 17 type 1 CNVs, 2 of 17 type 3 NVs, and 1 of 17 aneurysmal type 1 CNV follow this second fibrotic pathway. Among these 6 eyes, RPE tears occurred before the progression to type B fibrosis in 4 of 6 eyes (1 type 3 NV and 2 type 1 CNV) and subretinal hemorrhage occurred in 4 of 6 eyes before progression to type B fibrosis (1 case of aneurysmal type 1 CNV, 2 cases of type 1 CNV, and 1 case of type 3 NV).
      • The third pathway, described as “A → FAL,” was associated with perilesional atrophy at the initial visit. The 13 of 47 eyes following this pathway showed persistent type A fibrotic lesions, with the fibrosis confined under the RPE, evidenced by a shadow effect on spectral-domain OCT. This pathway is shown in Figure 9. No large macular hemorrhage was observed during the follow-up of these eyes in this pattern of progression.
      The presence of SHRM during follow-up was significantly associated with CNV progression from type A to type B and then to type C lesions (P < .001, Fisher exact test). SHRM appeared in all cases in the interval between type A and type B lesions. Macular hemorrhage was not associated with any of the 3 progression pathways despite being observed in 13 of 47 cases (4 A → B → C, 4 B → C, and 5 C direct) (P = .058 for the occurrence of macular hemorrhage between stage A → B and B → C). However, macular hemorrhage was absent during follow-up for eyes with FAL (type A).
      RPE erosions were significantly associated with type C fibrotic lesions (P < .001, Fisher exact test). RPE tear occurred in 8 of 10 cases in the interval between type A and B lesions, while in 2 of 10 cases RPE tear occurred in the interval between type B and type C lesions. The presence of macular atrophy was significantly associated with type A lesions (P < .001).

      Discussion

      In this retrospective study, we described the cross-sectional spectral-domain OCT morphologic features of lesions within the fibrotic spectrum caused by nAMD. In a second analysis, we performed a retrospective long-term longitudinal cohort study to elucidate the progression from the original neovascular lesion to the fibrotic scar. We analyzed a total of 5017 spectral-domain OCT slices in the cross-sectional and longitudinal analyses (836 and 4181, respectively).
      The cross-sectional analysis distinguished 3 types of lesions within the fibrotic spectrum. Type A corresponded to well-defined sub-RPE lesions, with or without intralesional abnormalities (43/44, 98% of eyes). Type B corresponded to well-defined hyperreflective lesions situated in the subretinal and sub-RPE space (8/44, 18% of eyes) with an intact RPE band. Type C consisted of prominent, elevated fibrotic lesions, with a complex pattern and RPE atrophy and loss of an identifiable RPE band (6/44, 14% of eyes). Interestingly, these phenotypes often coexisted in the same eye.
      Roger and associates
      • Rogers A.H.
      • Martidis A.
      • Greenberg P.B.
      • Puliafito C.A.
      Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
      proposed, in 2002, a 5-stage classification, based on a highspeed fiberoptic OCT scanner device coupled to a standard slit-lamp biomicroscope, that monitored response in eyes with nAMD treated by PDT.
      • Rogers A.H.
      • Martidis A.
      • Greenberg P.B.
      • Puliafito C.A.
      Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
      The presence of fibrosis was noted in stages IIIa and IIIb based on the severity of subretinal fibrosis and fluid. In stage V with subretinal fluid resolution, a fibrotic lesion merged with the RPE and was associated with overall thinning of the retina.
      • Rogers A.H.
      • Martidis A.
      • Greenberg P.B.
      • Puliafito C.A.
      Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
      With the advent of anti-VEGF therapy, PDT with verteporfin is no longer used for nAMD, being reserved for cases of aneurysmal type 1 neovascularization in combination with anti-VEGF.
      • Koh A.
      • Lee W.K.
      • Chen L.J.
      • et al.
      EVEREST study: efficacy and safety of verteporfin photodynamic therapy in combination with ranibizumab or alone versus ranibizumab monotherapy in patients with symptomatic macular polypoidal choroidal vasculopathy.
      In patients with nAMD, as shown in the outcomes of the ANCHOR study
      • Brown D.M.
      • Kaiser P.K.
      • Michels M.
      • et al.
      ANCHOR Study Group
      Ranibizumab versus verteporfin for neovascular age-related macular degeneration.
      and later confirmed by follow-up studies,
      • Brown D.M.
      • Michels M.
      • Kaiser P.K.
      • et al.
      ANCHOR Study Group
      Ranibizumab versus verteporfin photodynamic therapy for neovascular age-related macular degeneration: two-year results of the ANCHOR study.
      while in the ranibizumab groups (0.3- and 0.5-mg) mean visual acuity increased by 8.5 letters and 11.3 letters, respectively, in the verteporfin group there was a decrease of 9.5 letters in mean BCVA (P < .001).
      • Brown D.M.
      • Kaiser P.K.
      • Michels M.
      • et al.
      ANCHOR Study Group
      Ranibizumab versus verteporfin for neovascular age-related macular degeneration.
      Therefore, a switch in treatment paradigms has occurred, and nearly all patients with nAMD receive currently antiangiogenic therapy.
      In the current literature, the usual annotation of lesions within the fibrotic spectrum varies without consensus, but the most widely used terminology is that of “subretinal fibrosis.”
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Rogers A.H.
      • Martidis A.
      • Greenberg P.B.
      • Puliafito C.A.
      Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
      ,
      • Shah V.P.
      • Shah S.A.
      • Mrejen S.
      • Freund K.B.
      Subretinal hyperreflective exudation associated with neovascular age-related macular degeneration.
      Subretinal fibrosis typically corresponds to type B or type C lesions observed in our cross-sectional series, based on spectral-domain OCT. However, in our study the incidence of these types was rather low: 8 of 44 eyes for type B and 6 of 44 eyes for type C. Therefore, the phenotypic variability of lesions within the fibrotic spectrum, as relates to their location and/or the integrity of the RPE, is much greater than suggested in the previous published literature. Furthermore, while a careful analysis of multiple sections was performed for each patient, no correlation between clinical factors, such as BCVA and number of injections, vs type of fibrosis was observed.
      In addition, different patterns of fibrosis coexisted in the same eye. One explanation could be the fact that foveal involvement was not considered in the A/B/C classification. In addition, even though the greatest linear diameter of the lesions was most commonly associated with the lesion subtype A1 (P = .05), there were no significant correlations between fibrosis type and other quantitative criteria, such as number of intravitreal injections or automatic and manual measurements on spectral-domain OCT.
      In the cross-sectional analysis, we observed 3 subtypes of lesions in the type A fibrotic lesion group, which we labeled as A1, A2, and A3. These subtypes included intralesional abnormalities (A1: 20/44, 46% of eyes), a hyporeflective band across the lesion (A2: 22/44, 50% of eyes), or a grey hypereflective band under the RPE (A3: 42/44, 96% of eyes). We hypothesize that the A1 lesion subtype corresponds to a multilayered pigment epithelium detachment and may correspond to a preliminary stage in the development of the classically compact, hyperreflective lesions described in the literature (Mukkamala SK, et al. IOVS 2012;53:ARVO E-Abstract 2643).
      • Mrejen S.
      • Sarraf D.
      • Mukkamala S.K.
      • Freund K.B.
      Multimodal imaging of pigment epithelial detachment: a guide to evaluation.
      ,
      • Tan A.C.S.
      • Fleckenstein M.
      • Schmitz-Valckenberg S.
      • Holz F.G.
      Clinical application of multicolor imaging technology.
      ,
      • Rahimy E.
      • Freund K.B.
      • Larsen M.
      • et al.
      Multilayered pigment epithelial detachment in neovascular age-related macular degeneration.
      • Mukai R.
      • Sato T.
      • Kishi S.
      A hyporeflective space between hyperreflective materials in pigment epithelial detachment and Bruch’s membrane in neovascular age-related macular degeneration.
      • Querques G.
      • Costanzo E.
      • Miere A.
      • Capuano V.
      • Souied E.H.
      Choroidal caverns: a novel optical coherence tomography finding in geographic atrophy.
      Subtype classification of type A fibrosis establishes a continuum between fibrovascular pigment epithelial detachment and the fibrotic lesion. This is confirmed by our longitudinal analysis, where type A3 indeed precedes B and C forms of fibrosis.
      An important question was therefore to delineate the progression pathways from the initial neovascular lesion to the fibrotic scar observed on spectral-domain OCT, by means of a longitudinal analysis on a well-defined cohort of patients with long-standing nAMD and close spectral-domain OCT follow-up. The results of our longitudinal retrospective analysis suggested that CNV and the ultimate development of fibrosis might be seen as a continuum, progressing from the original CNV through the 2 fibrotic lesions types (sub-RPE type A, subretinal/sub-RPE type B) and finally reaching, after years of follow-up under anti-VEGF treatment, type C fibrotic lesions. In the A → B → C pathway, the presence of SHRM during follow-up was significantly associated with CNV progression from type A to type B and then to type C lesions (P < .001, Fisher exact test). SHRM appeared in all cases in the interval between type A and type B lesions. Our hypothesis would be that a focal erosion of RPE overlying a type A fibrotic lesion may subsequently cause the fibrotic component to emerge from the sub-RPE to the subretinal space, manifested as a SHRM lesion, with subsequent conversion to type B (subretinal) fibrosis, followed by a progressive erosion of the RPE and enlargement of the subretinal fibrotic component in this A → B → C pathway. In the B → C pathway, 11 of 17 eyes had a type 2 CNV or mixed type 1/type 2 CNV as the baseline type of neovascularization. Concerning the A → FAL pathway, baseline lesions in this case were mainly caused by type 1 CNV (10/13 eyes). Our results are consistent with the CATT study analysis at 2 years' follow-up, stating that baseline CNV type on FA predicted scar formation, with type 1 CNV being less likely to evolve to atrophic macular scar compared with type 2 CNV.
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      The 2 final morphologies of the fibrotic process in either pathway were either FAL or FGL.
      It is notable that hemorrhage, SHRM, or an RPE tear consistently preceded the fibrotic switch from type 1 or type 3 NV to FGL (type B and/or type C lesions). Figure 10 shows the main progression pathways from the original CNV to FAL or FGL scarring.
      Figure thumbnail gr10
      Figure 10Illustrative drawing of all patterns of progression, from choroidal neovascularization (CNV) to final fibrotic lesions. The 3 types of NV—type 1 CNV, type 2 CNV, and type 3 NV—are shown on the left. On the right are the final stages of fibrotic lesions: fibroglial lesion (FGL) on the top and fibroatrophic lesion (FAL) on the bottom. The 3 large blue arrows illustrate the 3 main pathways of progression to final fibrotic lesions, from top to bottom: (1) type 2 CNV evolving into type B (subretinal) fibrosis and then FGL (type C); (2) type 1 CNV evolving into type B (subretinal) fibrosis after RPE erosion and subretinal hyperreflective material; and (3) type 1 CNV evolving into an FAL. Thin arrows illustrate some peculiar cases, such as evolvement of any type of neovascularization directly to type B (subretinal) fibrosis and type C fibrosis (FGL) when RPE tears or hemorrhage occur. The shift from type 2 CNV or type 3 NV to a type A fibrosis is also illustrated by thin arrows. For clarity of the figure, aneurysmal type 1 neovascularization is not shown but is integrated in type 1 CNV.
      These pathways of progression within our longitudinal cohort support the cross-sectional analysis and validate a continuum or graduate progression of stages in the final development of fibrosis. In our longitudinal cohort, the angiofibrotic switch occurred 20 months from baseline examination. This is also consistent with current literature, stating that in ≤50% of eyes scars develop after 2 years of anti-VEGF treatment.
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      In another recent study, subretinal fibrosis appeared after 60 months of anti-VEGF therapy in 68.5% of eyes and was a predictor of worse visual outcome.
      • Pedrosa A.C.
      • Sousa T.
      • Pinheiro-Costa J.
      • et al.
      Treatment of neovascular age-related macular degeneration with anti-VEGF agents: predictive factors of long-term visual outcomes.
      Other studies have focused on the functional outcomes of nAMD eyes undergoing anti-VEGF treatment, but were limited by small sample size and short follow-up.
      • Rosenfeld P.J.
      • Brown D.M.
      • Heier J.S.
      • et al.
      Ranibizumab for neovascular age-related macular degeneration.
      • Heier J.S.
      • Boyer D.
      • Nguyen Q.D.
      • et al.
      The 1-year results of CLEAR-IT 2, a phase 2 study of vascular endothelial growth factor trap-eye dosed as-needed after 12-week fixed dosing.
      • Kaiser P.K.
      • Brown D.M.
      • Zhang K.
      • et al.
      Ranibizumab for predominantly classic neovascular age-related macular degeneration: subgroup analysis of first-year ANCHOR results.
      • Unver Y.
      • Yavuz G.
      • Bekiroglu N.
      • Presti P.
      • Li W.
      • Sinclair S.
      Relationships between clinical measures of visual function and anatomic changes associated with bevacizumab treatment for choroidal neovascularization in age-related macular degeneration.
      Daniel and associates
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      described the baseline risk factors of scars (both fibrotic and nonfibrotic) after anti-VEGF treatment with ranibizumab or bevacizumab from the CATT study. In their study, fibrotic scars developed at 1 year of anti-VEGF treatment in 32% of eyes and the incidence of fibrosis was 56% at 5 years of follow-up. Multivariate analysis also identified baseline characteristics that predicted scar formation, such as classic CNV, subretinal fluid, and SHRM.
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      In the CATT study, type 2 CNV was associated with 4.5-fold risk of fibrotic scar formation compared with type 1 CNV. Stevens and associates
      • Stevens T.S.
      • Bressler N.M.
      • Maguire M.G.
      • et al.
      Occult choroidal neovascularization in age-related macular degeneration. A natural history study.
      similarly reported that the presence of classic CNV increases the risk of development of fibrosis in late AMD. In our cross-sectional study, 25 of 44 eyes (57%) exhibited type 2 CNV or mixed type 1 and type 2 CNV at initial onset of nAMD. In our longitudinal analysis, 15 of 47 eyes (32%) presented as a type 2 CNV or mixed type 1 and type 2 CNV. It is clear that type 2 CNV is an important risk factor for fibrotic scar development but it is noteworthy that type 1 CNV was the original lesion in 10 of 13 eyes (77%) of the FAL and 16 of 34 eyes (47%) of the FGL in our study.
      Histologically, it is expected that type 2 CNV may evolve into subretinal fibrotic lesions (type B). However, Dolz-Marco and associates
      • Dolz-Marco R.
      • Phasukkijwatana N.
      • Sarraf D.
      • Freund K.B.
      Regression of type 2 neovascularization into a type 1 pattern after intravitreal anti-vascular endothelial growth factor therapy for neovascular age-related macular degeneration.
      and Coscas and associates
      • Coscas F.
      • Querques G.
      • Forte R.
      • Terrada C.
      • Coscas G.
      • Souied E.H.
      Combined fluorescein angiography and spectral-domain optical coherence tomography imaging of classic choroidal neovascularization secondary to age-related macular degeneration before and after intravitreal ranibizumab injections.
      have shown that type 2 CNV lesions treated with intravitreal antiangiogenic therapy may progress to a type 1 CNV fibrovascular pattern. Similarly, in our longitudinal series, we observed two type 2 CNV cases and 1 mixed CNV case evolving into subtype A3, and then following the A → B → C pathway.
      In the CATT study, a large submacular hemorrhage was associated with a more than 2-fold risk of fibrotic scarring.
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      This has been confirmed by other studies.
      • Scupola A.
      • Coscas G.
      • Soubrane G.
      • Balestrazzi E.
      Natural history of macular subretinal hemorrhage in age-related macular degeneration.
      ,
      • Avery R.L.
      • Fekrat S.
      • Hawkins B.S.
      • Bressler N.M.
      Natural history of subfoveal subretinal hemorrhage in age-related macular degeneration.
      However, Hwang and associates
      • Hwang J.C.
      • Del Priore L.V.
      • Freund K.B.
      • Chang S.
      • Iranmanesh R.
      Development of subretinal fibrosis after anti-VEGF treatment in neovascular age-related macular degeneration.
      showed that subfoveal fibrosis may be identified in eyes without significant subfoveal hemorrhage after anti-VEGF treatment. This is consistent with our findings, in which macular hemorrhage was noted in 13 of 47 (28%) eyes during follow-up. However, macular hemorrhage was not associated with a specific subtype of CNV. It is noteworthy that no hemorrhage was observed in the progression pattern type 1 CNV to FAL. On the other hand, the presence of hemorrhage was found in 4 of 6 cases of non–type 2 CNV that underwent a B → C pathway.
      In our longitudinal analysis, an RPE tear occurred in 10 of 47 (21%) eyes during the follow-up from the CNV lesion to fibrosis. Such an association was not previously described in previous studies on the progression of fibrotic lesions.
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      • Bloch S.B.
      • Lund-Andersen H.
      • Sander B.
      • Larsen M.
      Subfoveal fibrosis in eyes with neovascular age-related macular degeneration treated with intravitreal ranibizumab.
      Sarraf and associates
      • Sarraf D.
      • Joseph A.
      • Rahimy E.
      Retinal pigment epithelial tears in the era of intravitreal pharmacotherapy: risk factors, pathogenesis, prognosis and treatment (an American Ophthalmological Society thesis).
      noted that severe fibrotic scarring may be reduced with continued anti-VEGF therapy after RPE tear development.
      One of the risk factors for scarring identified by the CATT study was the presence of SHRM.
      • Heier J.S.
      • Boyer D.
      • Nguyen Q.D.
      • et al.
      The 1-year results of CLEAR-IT 2, a phase 2 study of vascular endothelial growth factor trap-eye dosed as-needed after 12-week fixed dosing.
      ,
      • Willoughby A.S.
      • Ying G.S.
      • Toth C.A.
      • et al.
      Subretinal hyperreflective material in the comparison of age-related macular degeneration treatments trials.
      In our longitudinal cohort, SHRM appeared in 14 of 47 (almost 30%) eyes and was associated in a statistically significant manner with eyes presenting with type 1 CNV at baseline (P = .007). Moreover, in all cases (14/47), SHRM preceded the progression to type B subretinal fibrotic lesions (P < .001). This is significant in the context of multiple studies showing that SHRM persistence despite anti-VEGF therapy was a risk factor for fibrosis development and had a negative impact on BCVA.
      • Willoughby A.S.
      • Ying G.S.
      • Toth C.A.
      • et al.
      Subretinal hyperreflective material in the comparison of age-related macular degeneration treatments trials.
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      • Casalino G.
      • Bandello F.
      • Chakravarthy U.
      Changes in neovascular lesion hyperreflectivity after anti-VEGF treatment in age-related macular degeneration: an integrated multimodal imaging analysis.
      Willoughby and associates
      • Willoughby A.S.
      • Ying G.S.
      • Toth C.A.
      • et al.
      Subretinal hyperreflective material in the comparison of age-related macular degeneration treatments trials.
      also found that SHRM was present at week 52 in 69.3% of eyes with scars in the CATT cohort. In addition, Pokroy and associates
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      and Casalino and associates
      • Casalino G.
      • Bandello F.
      • Chakravarthy U.
      Changes in neovascular lesion hyperreflectivity after anti-VEGF treatment in age-related macular degeneration: an integrated multimodal imaging analysis.
      showed that well-defined SHRM borders on spectral-domain OCT appear to represent fibrotic tissue or mature neovascular lesions. Pokroy and associates
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      and Casalino and associates
      • Casalino G.
      • Bandello F.
      • Chakravarthy U.
      Changes in neovascular lesion hyperreflectivity after anti-VEGF treatment in age-related macular degeneration: an integrated multimodal imaging analysis.
      suggested that SHRM persistence is consistent with subretinal fibrosis development.
      SHRM may be attributed to a heterogeneous group of lesions, including gray exudative fluid, hemorrhage, vitelliform material, or type 2 CNV.
      • Shah V.P.
      • Shah S.A.
      • Mrejen S.
      • Freund K.B.
      Subretinal hyperreflective exudation associated with neovascular age-related macular degeneration.
      ,
      • Ores R.
      • Puche N.
      • Querques G.
      • et al.
      Gray hyper-reflective subretinal exudative lesions in exudative age-related macular degeneration.
      This hyperreflective subretinal material, lying above the RPE,
      • Mehta H.
      • Tufail A.
      • Daien V.
      • et al.
      Real-world outcomes in patients with neovascular age-related macular degeneration treated with intravitreal vascular endothelial growth factor inhibitors.
      ,
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      may consist of a myriad of elements, from neovascular tissue to fibrin, blood, and lipid.
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      ,
      • Giani A.
      • Luiselli C.
      • Esmaili D.D.
      • et al.
      Spectral-domain optical coherence tomography as an indicator of fluorescein angiography leakage from choroidal neovascularization.
      OCTA can distinguish the vascular and avascular components of SHRM, as demonstrated by Dansingani and associates
      • Dansingani K.K.
      • Tan A.C.S.
      • Gilani F.
      • et al.
      Subretinal hyperreflective material imaged with optical coherence tomography angiography.
      and Kawashima and associates.
      • Kawashima Y.
      • Hata M.
      • Oishi A.
      • et al.
      Association of vascular versus avascular subretinal hyperreflective material with aflibercept response in age-related macular degeneration.
      Subretinal fibrosis may be the consequence of either the natural healing process or different therapies. We propose that an A3 lesion may progress because of RPE erosion, leading to eruption of the fibrotic component into the subretinal space, forming a SHRM aspect, which evolves into type B fibrotic lesion. SHRM development may also take place after anti-VEGF treatment. Continuous treatment over long periods of time may resolve or reduce fluid but can generate an increase in the fibrotic component of SHRM, rendering anti-VEGF therapy less efficacious.
      • Willoughby A.S.
      • Ying G.S.
      • Toth C.A.
      • et al.
      Subretinal hyperreflective material in the comparison of age-related macular degeneration treatments trials.
      ,
      • Pokroy R.
      • Mimouni M.
      • Barayev E.
      • et al.
      Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
      Anti-VEGF molecules induce the blockage of pathologic vessel growth through inhibition of the migration and proliferation of endothelial cells. This skews the angiofibrotic switch toward connective tissue mediators leading to connective tissue deposition, anti-VEGF resistance, and consequent scar formation. This resistance to VEGF-neutralizing molecules has long been accounted for in the oncologic literature.
      • Yang S.
      • Zhao J.
      • Sun X.
      Resistance to anti-VEGF therapy in neovascular age-related macular degeneration: a comprehensive review.
      Pericytes that supply VEGF and other factors to the proliferating endothelial cells of CNV are involved in this process. Pericyte recruitment, survival, and maturation are modulated by the platelet-derived growth factor. Pericytes may contribute to the development of fibrosis either directly by producing collagen or indirectly by differentiating into myofibroblasts and α-smooth muscle actin–expressing and collagen-producing cells that are responsible for the expansion of fibrotic tissue.
      • Harrell C.R.
      • Simovic Markovic B.
      • Fellabaum C.
      • et al.
      Molecular mechanisms underlying therapeutic potential of pericytes.
      Other studies have emphasized the role of (prolonged) anti-VEGF treatment in promoting the transition from angiogenesis to fibrosis.
      • Kuiper E.J.
      • de Smet M.D.
      • van Meurs J.C.
      • et al.
      Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
      ,
      • Kuiper E.J.
      • Van Nieuwenhoven F.A.
      • de Smet M.D.
      • et al.
      The angio-fibrotic switch of VEGF and CTGF in proliferative diabetic retinopathy.
      Connective tissue growth factor is a profibrotic factor whose vitreous levels correlate strongly with the degree of fibrosis in multiple vitreoretinal disorders.
      • Keane P.A.
      • Patel P.J.
      • Liakopoulos S.
      • Heussen F.M.
      • Sadda S.R.
      • Tufail A.
      Evaluation of age-related macular degeneration with optical coherence tomography.
      ,
      • Kuiper E.J.
      • de Smet M.D.
      • van Meurs J.C.
      • et al.
      Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
      Moreover, the ratio of connective tissue growth factor and VEGF seems to be the strongest predictor of fibrosis in eyes with proliferative diabetic retinopathy.
      • Kuiper E.J.
      • de Smet M.D.
      • van Meurs J.C.
      • et al.
      Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
      Several animal models have been applied to study the molecular complexities of subretinal fibrosis in mice and have confirmed the role of macrophage rich peritoneal exudate cells in the myofibrotic changes of RPE cells in vitro.
      • Kimura K.
      • Orita T.
      • Liu Y.
      • et al.
      Attenuation of EMT in RPE cells and subretinal fibrosis by an RAR-γ agonist.
      ,
      • Jo Y.-J.
      • Sonoda K.-H.
      • Oshima Y.
      • et al.
      Establishment of a new animal model of focal subretinal fibrosis that resembles disciform lesion in advanced age-related macular degeneration.
      Platelet-activating factor and the inhibition of platelet-activating factor receptor seem to suppress induced subretinal fibrosis in another animal model.
      • He Y.-G.
      • Wang H.
      • Zhao B.
      • et al.
      Functional analysis of platelet-activating factor in the retinal pigment epithelial cells and choroidal endothelial cells.
      Interleukin-6 and its receptor were also proven to be involved in the development of subretinal fibrosis.
      • Bastiaans J.
      • van Meurs J.C.
      • Mulder V.C.
      • et al.
      The role of thrombin in proliferative vitreoretinopathy.
      All in all, the pathogenic sequence of subretinal fibrosis, even if partially understood, seems to consist of leucocyte induced exudation by a highly permeable neovascular lesion, which initiates the inflammation process, stimulating glial proliferation and ultimately generating subretinal fibrosis.
      • Kimura K.
      • Orita T.
      • Liu Y.
      • et al.
      Attenuation of EMT in RPE cells and subretinal fibrosis by an RAR-γ agonist.
      ,
      • Jo Y.-J.
      • Sonoda K.-H.
      • Oshima Y.
      • et al.
      Establishment of a new animal model of focal subretinal fibrosis that resembles disciform lesion in advanced age-related macular degeneration.
      ,
      • Bastiaans J.
      • van Meurs J.C.
      • Mulder V.C.
      • et al.
      The role of thrombin in proliferative vitreoretinopathy.
      ,
      • Cleary P.E.
      • Ryan S.J.
      Histology of wound, vitreous, and retina in experimental posterior penetrating eye injury in the rhesus monkey.
      The prevention of fibrosis in patients with nAMD is still a challenging goal. Anti–platelet-derived growth factor therapies to target pericytes have met with failure despite initial excitement
      • Jaffe G.J.
      • Ciulla T.A.
      • Ciardella A.P.
      • et al.
      Dual antagonism of PDGF and VEGF in neovascular age-related macular degeneration: a phase IIb, multicenter, randomized controlled trial.
      and an efficacious treatment to prevent the formation of fibrosis is still lacking. Nevertheless, many research groups work on the identification of angiogenic factors involved in wound healing and fibrosis that could lead to emerging therapies. To date, no interventional study has demonstrated the efficacy of any anti–platelet-derived growth factor drug. The criteria selection of eyes for such studies is of major importance to demonstrate efficacy of these drugs. It is possible that development of fibrosis may be an acceptable process without adverse consequences when limited to the sub-RPE space, as with a multilayered PED (ie, type A lesions). However, when fibrosis invades the subretinal space, photoreceptor loss and visual decline may ensue. From our analysis, the occurrence of RPE erosion associated with SHRM may be considered an early stage of the progression to a fibroglial scar. Our hypothesis is that antifibrotic drugs should be considered particularly at this stage, only when fibrosis begins to invade the subretinal space, based on spectral-domain OCT analysis.

       Limitations

      Limitations of our study include the relatively small sample size and its heterogeneity, with no differentiation of patients according to the anti-VEGF treatment agent or duration of treatment. Moreover, we excluded large fibrotic scars that extended beyond the field of view. Several patients were lost to follow-up. Boulanger and associates
      • Boulanger-Scemama E.
      • Querques G.
      • About F.
      • et al.
      Ranibizumab for exudative age-related macular degeneration: a five-year study of adherence to follow-up in a real-life setting.
      found that the dropout rate of patients with nAMD after 5 years of follow-up was 57% and that age and BCVA at baseline and distance from home to hospital were independently associated with long-term drop out (115/201 patients). Stringent inclusion and exclusion criteria for our longitudinal cohort, including a follow up of ≥5 years and strict standards with regard to image quality and the size and localization of the fibrotic lesion within the 30° × 30° field of view can explain our population size.
      Our cross-sectional series included only 44 eyes, but a total of 836 (19 × 44) spectral-domain OCT slices were analyzed. Both eyes were included in 5 patients of the cross-sectional analysis, which may overrepresent a particular individual's response to nAMD as it pertains to fibrosis. Our longitudinal series included only 47 eyes, but our long-term follow-up analysis included a review of a total of 4181spectral-domain OCT slices, which allowed for a detailed morphologic analysis of the process from CNV to final fibrosis. The relatively small numbers of each subtype in the cross-sectional series and the imbalance of baseline lesion types in the longitudinal series (mostly type 1 CNV) is another limitation of our study. However, the predominance of type 1 CNV leading to fibrotic lesions may be considered an important outcome of this study. We demonstrated that macular fibrosis is not only an outcome of type 2 CNV but that fibrosis also occurs frequently as a consequence of type 1 CNV.
      Assessment of the presence or absence of RPE may be limited by spectral-domain OCT technology, especially because fibrosis may have the same reflectivity as the RPE on spectral-domain OCT. Given that our study did not use polarization-sensitive OCT, which may have been able to distinguish the RPE from the fibrotic tissue, we assumed that the lack of visualization of the RPE in group C fibrotic lesions was caused by an altered and/or absent RPE. Last but not least, given the reverse follow-up methodology used for the longitudinal series, we cannot state how type 1 or type 2 CNV progress over time (reported in the CATT study by Daniel and associates
      • Daniel E.
      • Toth C.A.
      • Grunwald J.E.
      • et al.
      Risk of scar in the comparison of age-related macular degeneration treatments trials.
      ,
      • Daniel E.
      • Pan W.
      • Ying G.S.
      • et al.
      Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
      ). Nonetheless, from our reverse follow-up analysis we can determine that FAL mainly originated from type 1 CNV, whereas subretinal fibro glial scar originated either from type 1, type 2 CNV, or type 3 NV. The strengths of the present study were the long-term follow-up of patients with fibrotic scars, allowing insights into the expanded spectrum of fibrotic lesions in nAMD.
      The progression from CNV to fibrosis is more multifaceted than the simple scheme involving type 2 CNV progressing to an FGL. In this study, besides the well-known progression from type 2 CNV to an FGL, we also demonstrated the steps of progression from type 1 fibrovascular CNV to FGL or FAL. This study has shown that any erosion or breach of RPE may lead to an invasion of the fibrotic component under the neurosensory retina, progressing to an FGL lesion. Conversely, some type 1 lesions associated with fibrovascular PED did not invade the subretinal space, and these lesions alternatively progressed to flat FAL. We believe that a description of these different phenotypes and pathways using spectral-domain OCT can provide a better understanding of the multifaceted pathogenic process of scarring leading to fibrosis, contribute to a future international classification of fibrosis, and contribute to the development of potential therapies to prevent fibrosis, vision loss, and blindness.
      This spectral-domain OCT analysis identified 3 main pathways to macular fibrosis leading to 2 types of advanced fibrotic lesions: FALs (absence of proliferation under the subretinal space) or FGLs (fibroglial proliferation in the subretinal space) after RPE erosion. A focal RPE erosion associated with SHRM may represent an early sign of progression from fibrovascular PED to subretinal fibrosis. Our study showed that RPE erosion plus SHRM may constitute a prognostic biomarker for the occurrence of fibrosis, may indicate the need for more aggressive treatment, and may represent a specific target for future antifibrotic drugs. We hope that our description contributes to a better understanding of the stages of the fibrotic process. Understanding of the angiofibrotic switch and detection of the early stages of fibrosis will be of major importance to potentially target a population at greatest risk of scarring and develop a therapeutic approach for the prevention of fibrotic scars. With new therapies promising a brighter future for the management of nAMD, spectral-domain OCT analysis may help determine the precise treatment targets and predict different therapeutic responses.

      CRediT authorship contribution statement

      Eric H. Souied: Conceptualization, Methodology, Writing - original draft. Manar Addou-Regnard: Data curation. Avi Ohayon: Visualization, Investigation. Oudy Semoun: Visualization, Investigation. Giuseppe Querques: Supervision. Rocio Blanco-Garavito: Visualization, Investigation. Roxane Bunod: Visualization. Camille Jung: Software, Methodology. Anne Sikorav: Investigation. Alexandra Miere: Writing - review & editing, Project administration.
      All authors have completed and submitted the ICMJE form for disclosure of potential conflicts of interest. Funding/Support: The authors indicate no financial support or financial conflict of interest. Conceptualization (E.H.S.); Methodology (E.H.S.); Writing - original draft preparation (E.H.S.); Data curation (M.A-R.); Visualization (A.O., O.S., R.B-G., R.B.); Investigation (A.O., O.S., R.B-G., A.S.); Supervision (G.Q.); Software (C.J.); Methodology (C.J.); Writing - reviewing and editing (A.M.); Project administration (A.M.). All authors attest that they meet the current ICMJE criteria for authorship.

      Supplemental Data

      References

        • Bressler N.M.
        Age-related macular degeneration is the leading cause of blindness.
        JAMA. 2004; 291: 1900-1901
        • Wong W.L.
        • Su X.
        • Li X.
        • et al.
        Global prevalence of age-related macular degeneration and disease burden projection for 2020 and 2040: a systematic review and meta-analysis.
        Lancet Glob Health. 2014; 2: e106-e116
        • Zayit-Soudry S.
        • Moroz I.
        • Loewenstein A.
        Retinal pigment epithelial detachment.
        Surv Ophthalmol. 2007; 52: 227-243
        • Mrejen S.
        • Sarraf D.
        • Mukkamala S.K.
        • Freund K.B.
        Multimodal imaging of pigment epithelial detachment: a guide to evaluation.
        Retina. 2013; 33: 1735-1762
        • Schmidt-Erfurth U.
        • Chong V.
        • Loewenstein A.
        • et al.
        Guidelines for the management of neovascular age-related macular degeneration by the European Society of Retina Specialists (EURETINA).
        Br J Ophthalmol. 2014; 98: 1144-1167
        • Cohen S.Y.
        • Oubraham H.
        • Uzzan J.
        • Dubois L.
        • Tadayoni R.
        Causes of unsuccessful ranibizumab treatment in exudative age-related macular degeneration in clinical settings.
        Retina. 2012; 32: 1480-1485
        • Rosenfeld P.J.
        • Shapiro H.
        • Tuomi L.
        • et al.
        Characteristics of patients losing vision after 2 years of monthly dosing in the phase III ranibizumab clinical trials.
        Ophthalmology. 2011; 118: 523-530
        • Gemmy Cheung C.M.
        • Li X.
        • Cheng C.Y.
        • et al.
        Prevalence and risk factors for age-related macular degeneration in Indians: a comparative study in Singapore and India.
        Am J Ophthalmol. 2013; 155 (773.e1–773.e3): 764-773
        • Kawasaki R.
        • Yasuda M.
        • Song S.J.
        • et al.
        The prevalence of age-related macular degeneration in Asians: a systematic review and meta-analysis.
        Ophthalmology. 2010; 117: 921-927
        • Keane P.A.
        • Patel P.J.
        • Liakopoulos S.
        • Heussen F.M.
        • Sadda S.R.
        • Tufail A.
        Evaluation of age-related macular degeneration with optical coherence tomography.
        Surv Ophthalmol. 2012; 57: 389-414
        • Daniel E.
        • Toth C.A.
        • Grunwald J.E.
        • et al.
        Risk of scar in the comparison of age-related macular degeneration treatments trials.
        Ophthalmology. 2014; 121: 656-666
        • Daniel E.
        • Pan W.
        • Ying G.S.
        • et al.
        Comparison of age-related macular degeneration treatment trials. Development and course of scars in the comparison of age-related macular degeneration treatments trials.
        Ophthalmology. 2018; 125: 1037-1046
        • Bloch S.B.
        • Lund-Andersen H.
        • Sander B.
        • Larsen M.
        Subfoveal fibrosis in eyes with neovascular age-related macular degeneration treated with intravitreal ranibizumab.
        Am J Ophthalmol. 2013; 156: 116-124.e1
        • Kuiper E.J.
        • de Smet M.D.
        • van Meurs J.C.
        • et al.
        Association of connective tissue growth factor with fibrosis in vitreoretinal disorders in the human eye.
        Arch Ophthalmol. 2006; 124: 1457-1462
        • Kuiper E.J.
        • Van Nieuwenhoven F.A.
        • de Smet M.D.
        • et al.
        The angio-fibrotic switch of VEGF and CTGF in proliferative diabetic retinopathy.
        PLoS One. 2008; 3: e2675
        • Bloom S.M.
        • Singal I.P.
        The outer bruch membrane layer: a previously undescribed spectral-domain optical coherence tomography finding.
        Retina. 2011; 31: 316-323
        • Miere A.
        • Semoun O.
        • Cohen S.Y.
        • et al.
        Optical coherence tomography angiography features of subretinal fibrosis in age-related macular degeneration.
        Retina. 2015; 35: 2275-2284
        • Roberts P.
        • Sugita M.
        • Deák G.
        • et al.
        Automated identification and quantification of subretinal fibrosis in neovascular age-related macular degeneration using polarization-sensitive OCT.
        Invest Ophthalmol Vis Sci. 2016; 57: 1699-1705
        • Rogers A.H.
        • Martidis A.
        • Greenberg P.B.
        • Puliafito C.A.
        Optical coherence tomography findings following photodynamic therapy of choroidal neovascularization.
        Am J Ophthalmol. 2002; 134: 566-576
        • Chong N.H.V.
        • Keonin J.
        • Luthert P.J.
        • et al.
        Decreased thickness and integrity of the macular elastic layer of Bruch’s membrane correspond to the distribution of lesions associated with age-related macular degeneration.
        Am J Pathol. 2005; 166: 241-251
        • de Boer J.F.
        • Cense B.
        • Park B.H.
        • Pierce M.C.
        • Tearney G.J.
        • Bouma B.E.
        Improved signal-to-noise ratio in spectral-domain compared with time-domain optical coherence tomography.
        Opt Lett. 2003; 28: 2067-2069
        • Toth L.A.
        • Stevenson M.
        • Chakravarthy U.
        Anti-vascular endothelial growth factor therapy for neovascular age-related macular degeneration: outcomes in eyes with poor initial vision.
        Retina. 2015; 35: 1957-1963
        • Keane P.A.
        • Liakopoulos S.
        • Ongchin S.C.
        • et al.
        Quantitative subanalysis of optical coherence tomography after treatment with ranibizumab for neovascular age-related macular degeneration.
        Invest Ophthalmol Vis Sci. 2008; 49: 3115-3120
        • Coscas F.
        • Coscas G.
        • Souied E.
        • Tick S.
        • Soubrane G.
        Optical coherence tomography identification of occult choroidal neovascularization in age-related macular degeneration.
        Am J Ophthalmol. 2007; 144: 592-599
        • Bird A.C.
        • Bressler N.M.
        • Bressler S.B.
        • et al.
        An international classification and grading system for age-related maculopathy and age-related macular degeneration. The International ARM Epidemiological Study Group.
        Surv Ophthalmol. 1995; 39: 367-374
        • Kaiser P.K.
        • Blodi B.A.
        • Shapiro H.
        • Acharya N.R.
        • MARINA Study Group
        Angiographic and optical coherence tomographic results of the MARINA study of ranibizumab in neovascular age-related macular degeneration.
        Ophthalmology. 2007; 114: 1868-1875
        • Tan A.C.S.
        • Fleckenstein M.
        • Schmitz-Valckenberg S.
        • Holz F.G.
        Clinical application of multicolor imaging technology.
        Ophthalmologica. 2016; 236: 8-18
        • Mehta H.
        • Tufail A.
        • Daien V.
        • et al.
        Real-world outcomes in patients with neovascular age-related macular degeneration treated with intravitreal vascular endothelial growth factor inhibitors.
        Prog Retin Eye Res. 2018; 65: 127-146
        • Su D.
        • Lin S.
        • Phasukkijwatana N.
        • et al.
        An updated staging system of type 3 neovascularization using spectral domain optical coherence tomography.
        Retina. 2016; 36: S40-S49
        • Dansingani K.K.
        • Gal-Or O.
        • Sadda S.R.
        • Yannuzzi L.A.
        • Freund K.B.
        Understanding aneurysmal type 1 neovascularization (polypoidal choroidal vasculopathy): a lesson in the taxonomy of “expanded spectra” - a review.
        Clin Exp Ophthalmol. 2018; 46: 189-200
        • Koh A.
        • Lee W.K.
        • Chen L.J.
        • et al.
        EVEREST study: efficacy and safety of verteporfin photodynamic therapy in combination with ranibizumab or alone versus ranibizumab monotherapy in patients with symptomatic macular polypoidal choroidal vasculopathy.
        Retina. 2012; 32: 1453-1464
        • Brown D.M.
        • Kaiser P.K.
        • Michels M.
        • et al.
        • ANCHOR Study Group
        Ranibizumab versus verteporfin for neovascular age-related macular degeneration.
        N Engl J Med. 2006; 355: 1432-1444
        • Brown D.M.
        • Michels M.
        • Kaiser P.K.
        • 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
        • Shah V.P.
        • Shah S.A.
        • Mrejen S.
        • Freund K.B.
        Subretinal hyperreflective exudation associated with neovascular age-related macular degeneration.
        Retina. 2014; 34: 1281-1288
        • Rahimy E.
        • Freund K.B.
        • Larsen M.
        • et al.
        Multilayered pigment epithelial detachment in neovascular age-related macular degeneration.
        Retina. 2014; 34: 1289-1295
        • Mukai R.
        • Sato T.
        • Kishi S.
        A hyporeflective space between hyperreflective materials in pigment epithelial detachment and Bruch’s membrane in neovascular age-related macular degeneration.
        BMC Ophthalmol. 2014; 14: 159
        • Querques G.
        • Costanzo E.
        • Miere A.
        • Capuano V.
        • Souied E.H.
        Choroidal caverns: a novel optical coherence tomography finding in geographic atrophy.
        Invest Ophthalmol Vis Sci. 2016; 57: 2578-2582
        • Pedrosa A.C.
        • Sousa T.
        • Pinheiro-Costa J.
        • et al.
        Treatment of neovascular age-related macular degeneration with anti-VEGF agents: predictive factors of long-term visual outcomes.
        J Ophthalmol. 2017; 2017: 4263017
        • Rosenfeld P.J.
        • Brown D.M.
        • Heier J.S.
        • et al.
        Ranibizumab for neovascular age-related macular degeneration.
        N Engl J Med. 2006; 355: 1419-1431
        • Heier J.S.
        • Boyer D.
        • Nguyen Q.D.
        • et al.
        The 1-year results of CLEAR-IT 2, a phase 2 study of vascular endothelial growth factor trap-eye dosed as-needed after 12-week fixed dosing.
        Ophthalmology. 2011; 118: 1098-1106
        • Kaiser P.K.
        • Brown D.M.
        • Zhang K.
        • et al.
        Ranibizumab for predominantly classic neovascular age-related macular degeneration: subgroup analysis of first-year ANCHOR results.
        Am J Ophthalmol. 2007; 144: 850-857
        • Unver Y.
        • Yavuz G.
        • Bekiroglu N.
        • Presti P.
        • Li W.
        • Sinclair S.
        Relationships between clinical measures of visual function and anatomic changes associated with bevacizumab treatment for choroidal neovascularization in age-related macular degeneration.
        Eye (Lond). 2008; 23: 453-460
        • Stevens T.S.
        • Bressler N.M.
        • Maguire M.G.
        • et al.
        Occult choroidal neovascularization in age-related macular degeneration. A natural history study.
        Arch Ophthalmol. 1997; 115: 345-350
        • Dolz-Marco R.
        • Phasukkijwatana N.
        • Sarraf D.
        • Freund K.B.
        Regression of type 2 neovascularization into a type 1 pattern after intravitreal anti-vascular endothelial growth factor therapy for neovascular age-related macular degeneration.
        Retina. 2017; 37: 222-233
        • Coscas F.
        • Querques G.
        • Forte R.
        • Terrada C.
        • Coscas G.
        • Souied E.H.
        Combined fluorescein angiography and spectral-domain optical coherence tomography imaging of classic choroidal neovascularization secondary to age-related macular degeneration before and after intravitreal ranibizumab injections.
        Retina. 2012; 32: 1069-1076
        • Scupola A.
        • Coscas G.
        • Soubrane G.
        • Balestrazzi E.
        Natural history of macular subretinal hemorrhage in age-related macular degeneration.
        Ophthalmologica. 1999; 213: 97-102
        • Avery R.L.
        • Fekrat S.
        • Hawkins B.S.
        • Bressler N.M.
        Natural history of subfoveal subretinal hemorrhage in age-related macular degeneration.
        Retina. 1996; 16: 183-189
        • Hwang J.C.
        • Del Priore L.V.
        • Freund K.B.
        • Chang S.
        • Iranmanesh R.
        Development of subretinal fibrosis after anti-VEGF treatment in neovascular age-related macular degeneration.
        Ophthalmic Surg Lasers Imaging. 2011; 42: 6-11
        • Sarraf D.
        • Joseph A.
        • Rahimy E.
        Retinal pigment epithelial tears in the era of intravitreal pharmacotherapy: risk factors, pathogenesis, prognosis and treatment (an American Ophthalmological Society thesis).
        Trans Am Ophthalmol Soc. 2014; 112: 142-159
        • Willoughby A.S.
        • Ying G.S.
        • Toth C.A.
        • et al.
        Subretinal hyperreflective material in the comparison of age-related macular degeneration treatments trials.
        Ophthalmology. 2015; 122: 1846-1853.e5
        • Pokroy R.
        • Mimouni M.
        • Barayev E.
        • et al.
        Prognostic value of subretinal hyperreflective material in neovascular age-related macular degeneration treated with bevacizumab.
        Retina. 2018; 38: 1485-1491
        • Casalino G.
        • Bandello F.
        • Chakravarthy U.
        Changes in neovascular lesion hyperreflectivity after anti-VEGF treatment in age-related macular degeneration: an integrated multimodal imaging analysis.
        Invest Ophthalmol Vis Sci. 2016; 57: OCT288-OCT298
        • Ores R.
        • Puche N.
        • Querques G.
        • et al.
        Gray hyper-reflective subretinal exudative lesions in exudative age-related macular degeneration.
        Am J Ophthalmol. 2014; 158: 354-361
        • Giani A.
        • Luiselli C.
        • Esmaili D.D.
        • et al.
        Spectral-domain optical coherence tomography as an indicator of fluorescein angiography leakage from choroidal neovascularization.
        Invest Ophthalmol Vis Sci. 2011; 52: 5579-5586
        • Dansingani K.K.
        • Tan A.C.S.
        • Gilani F.
        • et al.
        Subretinal hyperreflective material imaged with optical coherence tomography angiography.
        Am J Ophthalmol. 2016; 169: 235-248
        • Kawashima Y.
        • Hata M.
        • Oishi A.
        • et al.
        Association of vascular versus avascular subretinal hyperreflective material with aflibercept response in age-related macular degeneration.
        Am J Ophthalmol. 2017; 181: 61-70
        • Yang S.
        • Zhao J.
        • Sun X.
        Resistance to anti-VEGF therapy in neovascular age-related macular degeneration: a comprehensive review.
        Drug Des Devel Ther. 2016; 10: 1857-1867
        • Harrell C.R.
        • Simovic Markovic B.
        • Fellabaum C.
        • et al.
        Molecular mechanisms underlying therapeutic potential of pericytes.
        J Biomed Sci. 2018; 25: 21
        • Kimura K.
        • Orita T.
        • Liu Y.
        • et al.
        Attenuation of EMT in RPE cells and subretinal fibrosis by an RAR-γ agonist.
        J Mol Med. 2015; 93: 749-758
        • Jo Y.-J.
        • Sonoda K.-H.
        • Oshima Y.
        • et al.
        Establishment of a new animal model of focal subretinal fibrosis that resembles disciform lesion in advanced age-related macular degeneration.
        Invest Ophthalmol Vis Sci. 2011; 52: 6089-6095
        • He Y.-G.
        • Wang H.
        • Zhao B.
        • et al.
        Functional analysis of platelet-activating factor in the retinal pigment epithelial cells and choroidal endothelial cells.
        Curr Eye Res. 2009; 34: 957-965
        • Bastiaans J.
        • van Meurs J.C.
        • Mulder V.C.
        • et al.
        The role of thrombin in proliferative vitreoretinopathy.
        Invest Ophthalmol Vis Sci. 2014; 55: 4659-4666
        • Cleary P.E.
        • Ryan S.J.
        Histology of wound, vitreous, and retina in experimental posterior penetrating eye injury in the rhesus monkey.
        Am J Ophthalmol. 1979; 88: 221-231
        • Jaffe G.J.
        • Ciulla T.A.
        • Ciardella A.P.
        • et al.
        Dual antagonism of PDGF and VEGF in neovascular age-related macular degeneration: a phase IIb, multicenter, randomized controlled trial.
        Ophthalmology. 2017; 124: 224-234
        • Boulanger-Scemama E.
        • Querques G.
        • About F.
        • et al.
        Ranibizumab for exudative age-related macular degeneration: a five-year study of adherence to follow-up in a real-life setting.
        J Fr Ophtalmol. 2015; 38: 620-627