Ophthalmic Surgery, Lasers and Imaging Retina

Clinical Science 

Detection of Choroidal Neovascular Membrane Beneath Pigment Epithelial Detachment Using SD-OCTA

Arthi G. Venkat, MS, MD; Justis P. Ehlers, MD; Peter K. Kaiser, MD; Rishi P. Singh, MD; Andrew P. Schachat, MD; Sunil K. Srivastava, MD; Daniel F. Martin, MD; Aleksandra V. Rachitskaya, MD

Abstract

BACKGROUND AND OBJECTIVE:

To identify choroidal neovascular membrane (CNVM) associated with spectral-domain optical coherence tomography (SD-OCT)-defined pigment epithelial detachment (PED) using SD-OCT angiography (SD-OCTA).

PATIENTS AND METHODS:

Sixty-nine patients with same-day OCT and OCTA imaging were reviewed, and 41 eyes of 29 patients with PEDs were included. OCTs were analyzed for PED type, fluid, and subretinal hyperreflective material (SHRM).

RESULTS:

Twenty-seven eyes (66%) demonstrated CNVM on OCTA beneath all subtypes of PED. Twenty-two eyes (75.9%) with fluid or SHRM demonstrated CNVM on OCTA (P = .036). Fluid corresponded in a statistically significant manner with treatment (P = .0032), whereas SHRM did not (P = .613). OCTA-defined CNVM showed borderline statistically significant correlation to treatment (P = .05). Increased choroidal flow signal seen in 50% of eyes did not demonstrate statistically significant correlation to the presence of fluid on SD-OCT (P = .2798) or treatment decision (P = .678). A subset of 14 untreated eyes with CNVM was analyzed, 21% of which required treatment at subsequent visits.

CONCLUSIONS:

OCTA-defined CNVM was seen in all subtypes of PED in clinically active and inactive disease. The role of OCTA in predicting need for treatment remains to be established.

[Ophthalmic Surg Lasers Imaging Retina. 2019;50:620–626.]

Abstract

BACKGROUND AND OBJECTIVE:

To identify choroidal neovascular membrane (CNVM) associated with spectral-domain optical coherence tomography (SD-OCT)-defined pigment epithelial detachment (PED) using SD-OCT angiography (SD-OCTA).

PATIENTS AND METHODS:

Sixty-nine patients with same-day OCT and OCTA imaging were reviewed, and 41 eyes of 29 patients with PEDs were included. OCTs were analyzed for PED type, fluid, and subretinal hyperreflective material (SHRM).

RESULTS:

Twenty-seven eyes (66%) demonstrated CNVM on OCTA beneath all subtypes of PED. Twenty-two eyes (75.9%) with fluid or SHRM demonstrated CNVM on OCTA (P = .036). Fluid corresponded in a statistically significant manner with treatment (P = .0032), whereas SHRM did not (P = .613). OCTA-defined CNVM showed borderline statistically significant correlation to treatment (P = .05). Increased choroidal flow signal seen in 50% of eyes did not demonstrate statistically significant correlation to the presence of fluid on SD-OCT (P = .2798) or treatment decision (P = .678). A subset of 14 untreated eyes with CNVM was analyzed, 21% of which required treatment at subsequent visits.

CONCLUSIONS:

OCTA-defined CNVM was seen in all subtypes of PED in clinically active and inactive disease. The role of OCTA in predicting need for treatment remains to be established.

[Ophthalmic Surg Lasers Imaging Retina. 2019;50:620–626.]

Introduction

Pigment epithelial detachment (PED) is a pathologic detachment of the retinal pigment epithelium (RPE), most notably found in exudative and non-exudative age-related macular degeneration (AMD). Although the pathophysiology of PED formation is incompletely understood, it is presumed that degenerative changes in Bruch's membrane are related both to PED and choroidal neovascular membrane (CNVM) formation.1 Histologic study of eyes with drusenoid PEDs has demonstrated correlation of RPE pathology to spectral-domain optical coherence tomography (SD-OCT) findings.2 SD-OCT has allowed further characterization of PEDs by morphology as drusenoid, serous or fibrovascular,3 which has allowed further inquiry into the pathogenic implications of each morphologic subtype. Cukras et al. found the natural course of drusenoid PED to correlate with progression to advanced AMD and vision loss.4 The role of vascularized PED in RPE tear formation has been investigated and contraction of CNVM and physical traction of overlying PED has been implicated as a cause of RPE tear formation.5,6

The localization of CNVM beneath PED might therefore provide a useful anatomical correlation in predicting disease course, response to treatment, and potentially RPE tear formation. The current study is a retrospective review of patients with PEDs using OCT angiography (OCTA) to assess the location of CNVM with respect to PED. OCTA is a relatively novel imaging modality that can identify retinal and choroidal vascular changes without the use of injectable dyes. Previous studies that utilized prototype doppler SD-OCT and phase-variance SD-OCT, a modality that uses consecutive B-scans to determine regions of higher vascular flow, demonstrated the ability to identify vascular signals within PEDs in the setting of AMD and other pathophysiologic processes.7,8 With the advent of noninvasive technology that allows the juxtaposition of angiographic evaluation alongside structural B-scan images obtained by SD-OCT, it is possible to co-localize a CNVM lesion with overlying PED. Visualization of CNVM with respect to PED is additionally superior on OCTA as compared to fluorescein angiography (FA), as FA is unable to demonstrate the location of both CNVM and PED.

The purpose of the current study is to examine the role of OCTA as a diagnostic modality in eyes with PED on SD-OCT with respect to the identification of CNVM. The current study seeks to identify morphologic factors on SD-OCT such as fluid and subretinal hyperreflective material (SHRM) and to compare the predictive utility of these factors to CNVM identification on OCTA in diagnosis and treatment decisions.

Patients and Methods

The current study was approved by the institutional review board of the Cleveland Clinic Foundation. This retrospective review analyzed consecutive patients imaged with OCTA (AngioPlex; Zeiss, Oberkochen, Germany) and SD-OCT (Zeiss, Oberkochen, Germany) on the same visit between October 2016 and March 2017. All patients with same-day imaging on OCTA and SD-OCT were reviewed, and the 69 patients who demonstrated PED on SD-OCT were included. Any patient who did not demonstrate PED on SD-OCT or did not receive SD-OCT and OCTA imaging at the same visit was excluded.

SD-OCT images were analyzed for morphological characteristics including PED height, characterization of PED type as drusenoid, serous or fibrovascular, presence or absence of intraretinal fluid (IRF) or subretinal fluid (SRF), and presence or absence of SHRM. OCTA images were analyzed to determine the presence or absence of CNVM on en face images as well as changes in flow signal on B-scan images; an example of such a composite image is shown in Figure 1. The investigator reviewing SD-OCT and OCTA was initially blinded to the patients' charts to prevent bias of interpretation by the original physician. Subsequent to image analysis, patient demographics, diagnoses, visual acuity and clinical examination at the time of imaging, history of anti-vascular endothelial growth factor (VEGF) therapy, therapy at the time of imaging, and subsequent treatment(s) at the following visit were recorded.

Spectral-domain optical coherence tomography angiography en face images of the same eye (3 mm × 3 mm and 6 mm × 6 mm) with corresponding B-scans. These images demonstrate clear en-face choroidal neovascular membrane with corresponding increased flow signal on B-scan images that correlate to the location of the en-face lesion (arrows).

Figure 1.

Spectral-domain optical coherence tomography angiography en face images of the same eye (3 mm × 3 mm and 6 mm × 6 mm) with corresponding B-scans. These images demonstrate clear en-face choroidal neovascular membrane with corresponding increased flow signal on B-scan images that correlate to the location of the en-face lesion (arrows).

Statistical analysis was performed using R (R Core Team 2013, Vienna, Austria) and GraphPad (La Jolla, CA) software. Student's t-test was used for continuous measures, and Chi squared test was used for categorical factors.

Results

Of 69 eyes with OCTA and SD-OCT imaging at the same visit, 41 eyes of 29 patients were included based on the presence of a PED on SD-OCT. Sixty-nine percent of patients were female, and the average age was 79 years ± 10.2 years (range: 29 years to 98 years). Of included patients, 73.2% carried a clinical diagnosis of exudative AMD, with other included diagnoses being nonexudative AMD (9.8%), myopic degeneration (4.9%), and central serous chorioretinopathy (7.3%) (Table 1). Twenty-five eyes (61%) had received previous anti-VEGF therapy. The mean logMAR visual acuity (VA) at the time of imaging was 0.58 (20/76).

Baseline Characteristics

Table 1:

Baseline Characteristics

PEDs were analyzed based on morphology, height, and correlation with CNVM on en face OCTA imaging (Table 2). In all 27 eyes with CNVM on OCTA, the CNVM was located beneath the PED. Among all eyes, the majority of PEDs were drusenoid (75.6%), followed by fibrovascular (17.1%) and serous (7.3%) subtypes. Twenty-seven eyes demonstrated CNVM on en face OCTA and 14 did not. CNVM was identified by OCTA beneath all three subtypes of PEDs. In the 14 eyes without CNVM on OCTA, all PEDs were found to be drusenoid. Average PED height was greater in eyes with CNVM on OCTA (208 μm ± 145 μm) than those without CNVM (140 μm ± 91 μm), but this tendency did not demonstrate statistical significance (P = .622). In 37 eyes (90%), a PED was diagnosed at the initial visit with a Cleveland Clinic retina specialist. Of the four eyes in which PED was diagnosed after the initial visit, three developed PEDs at the same visit at which they were diagnosed with CNVM, and one developed a PED following a switch from bevacizumab (Avastin; Genentech, South San Francisco, CA) to aflibercept (Eylea; Regeneron, Tarrytown, NY) therapy. The average time between the diagnosis of PED and the imaging visit was 925 days ± 872 days.

Pigment Epithelial Detachment Characteristics

Table 2:

Pigment Epithelial Detachment Characteristics

Eyes with and without IRF, SRF, and SHRM were analyzed for the presence of CNVM on OCTA (Table 3). Of 16 eyes with IRF on SD-OCT, 13 (81.3%) had CNVM on en face OCTA. Of 25 eyes that did not have IRF on SD-OCT, 14 (56%) had CNVM on en face OCTA. The 14 eyes with SRF demonstrated CNVM on OCTA 71.4% of the time, and 63% of the 27 eyes without SRF were found to have OCTA defined CNVM. Combining the groups with IRF and SRF, 75.9% of eyes with fluid demonstrated CNVM on OCTA, and 41.7% of eyes without fluid had CNVM on OCTA. SHRM was found in seven eyes, five of which demonstrated CNVM on OCTA. Of the 34 eyes without SHRM, 22 (65%) had OCTA defined CNVM. None of these groups demonstrated a statistically significant correlation of OCTA defined CNVM with these SD-OCT findings; P values are shown in Table 3.

Structural Characteristics of SD-OCT in Eyes With and Without CNVM on En Face OCTA

Table 3:

Structural Characteristics of SD-OCT in Eyes With and Without CNVM on En Face OCTA

Imaging characteristics were also analyzed with respect to treatment decision at the initial visit (Table 4). Age and gender did not correlate in a statistically significant manner to the decision to treat. The presence of fluid demonstrated a statistically significant correlation to treatment decision (P = .0032), and the presence of SHRM did not (P = .613). Seventy-eight percent of eyes that were treated had both fluid on SD-OCT and CNVM on en face OCTA. The correlation of OCTA defined CNVM with treatment decision on the day of imaging demonstrated borderline statistical significance (P = .05). When combined, the presence of fluid and the presence of CNVM on OCTA showed a statistically significant correlation with treatment decision (P = .0007), whereas the presence of increased flow signal on OCTA structural B-scan was not significant with respect to treatment decision (P = .678).

Correlation of Imaging Characteristics With Treatment Decision at Initial Visit

Table 4:

Correlation of Imaging Characteristics With Treatment Decision at Initial Visit

Increased flow signal on OCTA structural B-scan images was also analyzed with respect to the en face choroidal images on OCTA (Table 5). Seventy percent of eyes with CNVM identified on en face OCTA also demonstrated increased flow signal on the B-scan image, and 30% did not. Of the 19 eyes that showed CNVM on en face OCTA and increased flow signal on B-scan, 14 had fluid on SD-OCT and five did not; all eight eyes with en face CNVM without increased flow signal on structural B-scan demonstrated fluid on SD-OCT. There was no statistically significant correlation between SD-OCT fluid and increased flow signal on OCTA structural B-scan in eyes with CNVM on en face OCTA images (P = .2798).

Correlation of En Face CNVM Identification With Flow Signal on OCTA

Table 5:

Correlation of En Face CNVM Identification With Flow Signal on OCTA

Overall, 18 eyes (44%) received treatment at the imaging visit. The mean pre-treatment logMAR VA in this group was 0.59 (20/78) and the mean post-treatment logMAR VA was 0.49 (20/62).

Table 6 details the treatment decision and follow up in the 12 eyes with en face OCTA defined CNVM who were not treated. Nine of the 12 eyes did not require treatment in subsequent visits. One patient in this group failed to return for follow-up. Two eyes required subsequent treatment due to increased fluid on SD-OCT but did not receive OCTA imaging at subsequent visits. Nine eyes also demonstrated increased flow signal on the OCTA B-scan, one of which required treatment at subsequent visits. One of the three eyes that did not demonstrate increased flow signal required multiple treatments at subsequent visits.

Table 7 details treatment decision and follow-up in the 14 eyes that did not demonstrate CNVM on en face OCTA. Seven eyes in this group demonstrated fluid on SD-OCT, and of these, three were treated at the initial visit. All three of these eyes required treatments at subsequent visits. No eyes in this group without fluid were treated initially or at subsequent visits.

Very few eyes were imaged on OCTA at subsequent visits. One patient who was imaged multiple times on OCTA is shown in Figure 2; this patient did not demonstrate fluid on SD-OCT at the initial imaging visit but did have a mature CNVM on OCTA with clearly visible increased flow signal (Figure 2A). The patient was not treated at the imaging visit, but at her following visit developed subsequent fluid on SD-OCT and was treated. Her first repeat image on OCTA was obtained 2 months after treatment and demonstrated involution of her CNVM on en face imaging with persistent increased flow signal on B-scan (Figure 2B). This patient required one treatment in this eye at a subsequent visit and has not required further treatment.

Pre-treatment (A) and post-treatment (B) spectral-domain optical coherence tomography angiography images. These images demonstrate the presence of increased flow signal in both mature and involuted en face choroidal neovascular membrane (CNVM) (arrows), indicating that flow signal abnormalities on B-scan do not correspond to CNVM activity.

Figure 2.

Pre-treatment (A) and post-treatment (B) spectral-domain optical coherence tomography angiography images. These images demonstrate the presence of increased flow signal in both mature and involuted en face choroidal neovascular membrane (CNVM) (arrows), indicating that flow signal abnormalities on B-scan do not correspond to CNVM activity.

Discussion

Current management of patients with PED on SD-OCT is largely dependent upon the clinician's judgement, and the decision to treat with anti-VEGF therapy is dependent upon suspicion for CNVM, which is influenced by the presence or absence of fluid on SD-OCT as a surrogate marker of active CNVM. OCTA is a modality by which CNVM, active or inactive, can be identified. Although several studies have been done to analyze CNVM on OCTA, the current study is the first to analyze the diagnostic and prognostic role of OCTA in eyes with PEDs on SD-OCT.

All morphologic subtypes of PED demonstrated CNVM on OCTA. CNVM was identified on OCTA in eyes with and without fluid or SHRM on SD-OCT. Although fluid on SD-OCT alone and fluid on SD-OCT with CNVM on OCTA showed statistical significance in correlation with the decision to treat, OCTA-defined CNVM alone did not. This finding begs the question of the utility of OCTA in identifying CNVM as it pertains to treatment decision. In these patients, according to current treatment guidelines, OCTA would not have served as an adequate surrogate test for SD-OCT. Moreover, CNVM identified on OCTA may not be actively exudative. The presence of increased flow signal on OCTA in “nonexudative” macular degeneration is well-documented and has led to a discussion regarding new terminology in the classification of ARMD.9

Of interest is the outcome of patients who demonstrated CNVM on OCTA and were not treated. The decision to hold treatment was made by the clinician based on historical factors (ie, the stability of fluid over time, previous response or lack thereof to anti-VEGF), visual prognosis, and in one case, patient preference. Despite the presence of CNVM on OCTA, nine of 12 of these eyes did not require treatment at subsequent visits and demonstrated stability over time. In these eyes, the presence of fluid was also not predictive of future need for treatment, as eight of the 12 eyes demonstrated IRF or SRF on SD-OCT.

The advent of OCTA has incited questions regarding the implications of this new imaging modality on treatment. Multimodal imaging in retinal disease, specifically AMD and similar macular diseases, has conferred significant advantage in diagnosis, but also frequently confounds diagnosis when artifact and misinterpretation create discrepancies between clinical diagnosis and diagnostic imaging modalities. The presence of artifact may account for the patients who did not demonstrate clear CNVM on OCTA but were treated based on the presence of fluid on SD-OCT.

Increased flow signal did not correlate consistently with CNVM activity. Not all eyes with CNVM on en face OCTA demonstrated increased flow signal, and only 50% of treated eyes demonstrated increased flow signal. The example in Figure 2 demonstrates persistent flow signal in a CNVM that has undergone involution post-treatment and did not require anti-VEGF at subsequent visits. The limitation of these observations is that OCTA was not obtained in patients following the initial imaging visit at regular intervals, and the response of flow signal to treatment cannot be interpreted. Low or absent flow signal on B-scan in eyes with en face CNVM on SD-OCTA may be due to limited choroidal penetration; comparisons of SD-OCTA and swept-source OCTA (SS-OCTA) have determined superior ability of SS-OCTA in CNVM assessment due to deeper penetration of light waves.10

Limitations of the study include the inability to determine CNVM incidence on OCTA in eyes with PEDs because of the inherent bias by which practitioners decide to obtain OCTA at our institution. OCTA is often obtained in patients in whom CNVM is predicted; therefore, due to higher pre-test probability that CNVM is present, the incidence of CNVM in the studied patients is inherently higher. Future studies would prospectively analyze all eyes with PED on clinical exam using SD-OCT and OCTA to eliminate bias. A study of treatment-naïve eyes with OCTA obtained at regular intervals throughout treatment would allow more definitive assessment of en face morphology and flow signal as markers of response to therapy. The current study did not compare OCTA to the gold standard imaging study for identification of CNVM, fluorescein angiography (FA). OCTA has demonstrated a high correspondence to FA-identified CNVM in prospective studies,11 and in smaller studies has demonstrated superior imaging of CNVM in cases where FA findings were ambiguous.12 Because FA is no longer routinely obtained in patients with the conditions studied, a comparison between it and OCTA was unable to be drawn in this study. However, previous studies have compared the two imaging modalities and have found OCTA to be superior to FA in the imaging of deep retinal capillary networks.13

The presence of fluid on SD-OCT appears to be the best predictor of need for future treatments in patients with PED. The current study did not demonstrate consistent correlation of OCTA defined CNVM with SD-OCT findings or the need for future treatment. Further prospective studies are warranted to further elucidate the role of OCTA in diagnosis, treatment and prognosis of CNVM.

References

  1. Pauleikhoff D, Löffert D, Spital G, et al. Pigment epithelial detachment in the elderly. Clinical differentiation, natural course and pathogenetic implications. Graefes Arch Clin Exp Ophthalmol. 2002;240(7):533–538. https://doi.org/10.1007/s00417-002-0505-8 PMID: doi:10.1007/s00417-002-0505-8 [CrossRef]12136282
  2. Balaratnasingam C, Messinger JD, Sloan KR, Yannuzzi LA, Freund KB, Curcio CA. Histologic and Optical Coherence Tomographic Correlates in Drusenoid Pigment Epithelium Detachment in Age-Related Macular Degeneration. Ophthalmology. 2017;124(5):644–656. https://doi.org/10.1016/j.ophtha.2016.12.034 PMID: doi:10.1016/j.ophtha.2016.12.034 [CrossRef]28153442
  3. Mrejen S, Sarraf D, Mukkamala SK, Freund KB. Multimodal imaging of pigment epithelial detachment: a guide to evaluation. Retina. 2013;33(9):1735–1762. https://doi.org/10.1097/IAE.0b013e3182993f66 PMID: doi:10.1097/IAE.0b013e3182993f66 [CrossRef]23873168
  4. Cukras C, Agron E, Klein M, et al. Natural history of drusenoid pigment epithelial detachment in age-related macular degeneration: Age-Related Eye Disease Study Report No. 28. Ophthalmology. 2010;117(3):489–499. https://doi.org/10.1016/j.ophtha.2009.12.002 PMID: doi:10.1016/j.ophtha.2009.12.002 [CrossRef]20079925
  5. Lafaut BA, Aisenbrey S, Vanden Broecke C, et al. Clinicopathological correlation of retinal pigment epithelial tears in exudative age related macular degeneration: pretear, tear, and scarred tear. Br J Ophthalmol. 2001;85(4):454–460. https://doi.org/10.1136/bjo.85.4.454 PMID: doi:10.1136/bjo.85.4.454 [CrossRef]11264137
  6. Meyer CH, Toth CA. Retinal pigment epithelial tear with vitreomacular attachment: a novel pathogenic feature. Graefes Arch Clin Exp Ophthalmol. 2001;239(5):325–333. https://doi.org/10.1007/s004170100259 PMID: doi:10.1007/s004170100259 [CrossRef]11482335
  7. McClintic SM, Kim DY, Fingler J, et al. Detection of pigment epithelial detachment vascularization in age-related macular degeneration using phase-variance OCT angiography. Clin Ophthalmol. 2015;9:1299–1305. https://doi.org/10.2147/OPTH.S83701 PMID:26203220
  8. Miura M, Makita S, Iwasaki T, Yasuno Y. Three-dimensional visualization of ocular vascular pathology by optical coherence angiography in vivo. Invest Ophthalmol Vis Sci. 2011;52(5):2689–2695. https://doi.org/10.1167/iovs.10-6282 PMID: doi:10.1167/iovs.10-6282 [CrossRef]21273541
  9. Roisman L, Zhang Q, Wang RK, et al. Optical Coherence Tomography Angiography of Asymptomatic Neovascularization in Intermediate Age-Related Macular Degeneration. Ophthalmology. 2016;123(6):1309–1319. https://doi.org/10.1016/j.ophtha.2016.01.044 PMID: doi:10.1016/j.ophtha.2016.01.044 [CrossRef]26876696
  10. Zhang Q, Chen CL, Chu Z, et al. Automated quantitation of choroidal neovascularization: A comparison study between spectral-domain and swept-source OCT angiograms. Invest Ophthalmol Vis Sci. 2017;58(3):1506–1513. https://doi.org/10.1167/iovs.16-20977 PMID: doi:10.1167/iovs.16-20977 [CrossRef]28273317
  11. Coscas GJ, Lupidi M, Coscas F, Cagini C, Souied EH. OPTICAL COHERENCE TOMOGRAPHY ANGIOGRAPHY VERSUS TRADITIONAL MULTIMODAL IMAGING IN ASSESSING THE ACTIVITY OF EXUDATIVE AGE-RELATED MACULAR DEGENERATION: A New Diagnostic Challenge. Retina. 2015;35(11):2219–2228. https://doi.org/10.1097/IAE.0000000000000766 PMID: doi:10.1097/IAE.0000000000000766 [CrossRef]26398697
  12. Malihi M, Jia Y, Gao SS, et al. Optical coherence tomographic angiography of choroidal neovascularization ill-defined with fluorescein angiography. Br J Ophthalmol. 2017;101(1):45–50. https://doi.org/10.1136/bjophthalmol-2016-309094 PMID: doi:10.1136/bjophthalmol-2016-309094 [CrossRef]
  13. Spaide RF, Klancnik JM Jr, Cooney MJ. Retinal vascular layers imaged by fluorescein angiography and optical coherence tomography angiography. JAMA Ophthalmol. 2015;133(1):45–50. https://doi.org/10.1001/jamaophthalmol.2014.3616 PMID: doi:10.1001/jamaophthalmol.2014.3616 [CrossRef]

Baseline Characteristics

Number of Patients29
  Male9 (31%)
  Female20 (68.9%)

Age (Median ± IQR)82 ± 8

Number of Eyes41

Diagnoses
  Exudative AMD30 (73.2%)
  Non-exudative AMD4 (9.8%)
  Myopic degeneration2 (4.9%)
  Central serous chorioretinopathy3 (7.3%)
  Other2 (4.9%)

Pigment Epithelial Detachment Characteristics

Total41
Fibrovascular7 (17.1%)
Drusenoid31 (75.6%)
Serous3 (7.3%)
CNVM present on en face OCTA27
Fibrovascular7 (25.9%)
Drusenoid17 (62.9%)
Serous3 (0.11%)
CNVM absent on en face OCTA14
Drusenoid14 (100%)
Average PED Height (M ± SD)182 ± 124
With CNVM on en face OCTA (N = 27)208 ± 145
Without CNVM on en face OCTA (N = 14)140 ± 91
P value.622

Structural Characteristics of SD-OCT in Eyes With and Without CNVM on En Face OCTA

Intraretinal Fluid Present16
  With CNVM13 (81.3%)
Intraretinal Fluid Absent25
  With CNVM14 (56%)
P value.096
Subretinal Fluid Present14
  With CNVM10 (71.4%)
Subretinal Fluid Absent27
  With CNVM17 (63%)
P value.588
Any Fluid Present29
  With CNVM22 (75.9%)
Any Fluid Absent12
  With CNVM5 (41.7%)
P value.036
Subretinal Hyperreflective Material Present7
  With CNVM5 (71.4%)
Subretinal Hyperreflective Material Absent34
  With CNVM22 (65%)
P value.733

Correlation of Imaging Characteristics With Treatment Decision at Initial Visit

VariableTotal (N = 41)Treatment (N = 18)No Treatment (N = 23)P Value

Mean Age (M ± SD)79.2 ± 10.272.8 ± 16.279.2 ± 11.9.16

Gender.80††
  Male12 (29.2%)10 (55.5%)3 (13%)
  Female29 (70.7%)8 (44.4%)20 (87%)

With IRF/SRF29 (70.7%)17 (94.4%)12 (52.2%).0032*††

Without IRF/SRF12 (29.3%)1 (5.6%)11 (47.8%)

With SHRM12 (29.3%)6 (33.3%)6 (26.0%).613††

Without SHRM29 (70.7%)12 (66.7%)17 (73.9%)

With CNVM on en-face27 (65.8%)15 (83.3%)12 (52.2%).05**††

Without CNVM on en-face14 (34.1%)3 (16.7%)11 (47.8%)

Fluid on SD-OCT and CNVM on en-face18 (43.9%)14 (77.7%)4 (17.4%).0007*††

No fluid on SD-OCT and no CNVM on en-face7 (17%)07 (30.4%)

Increased flow signal19 (46.3%)9 (50%)10 (43.5%).678††

Normal/absent flow signal22 (53.7%)9 (50%)13 (56.5%)

Correlation of En Face CNVM Identification With Flow Signal on OCTA

VariableCNVM Present on En Face (N = 27)CNVM Absent on En Face (N = 14)P Value
Increased Flow Signal19 (70%)0.001*††
Normal/Absent Flow Signal8 (30%)14 (100%)
Authors

From Cleveland Clinic Foundation, Cole Eye Institute, Cleveland, Ohio.

Paper presented at the 2017 American Society of Retina Specialists and the 2017 American Academy of Ophthalmology Annual Meeting.

Dr. Ehlers has received personal fees from Zeiss and Leica; grants and personal fees from Novartis, Aerpio, Genentech, Regeneron, and Alcon; and grants, personal fees, and other funding from Thrombogenics outside the submitted work. Dr. Kaiser has received non-financial support from Carl Zeiss Meditec and personal fees from Bayer, Regeneron, Novartis, Allergan, and Kanghong outside the submitted work. Dr. Singh has receieved personal fees from Zeiss, Optos, Genentech, Regeneron, and Alcon/Novartis during the conduct of the study. Dr. Schachat has received royalties from Elsevier and an editor stipend from American Academy of Ophthalmology outside the submitted work, is a part-time employee of the State of Ohio, and is a full-time employee of the Cleveland Clinic. Dr. Srivstava has received grants from Allergan, personal fees from Optos, and grants and personal fees from Regeneron, Santen, Sanofi, Eyepoint, and Bausch + Lomb outside the submitted work. Dr. Rachitskaya has received personal fees from Allergan, Zeiss, and Alcon outside the submitted work. The remaining authors report no relevant financial disclosures.

Dr. Singh did not participate in the editorial review of this manuscript.

Address correspondence to Aleksandra V. Rachitskaya, MD, Cleveland Clinic Foundation, Cole Eye Institute, 9500 Euclid Ave, i-30, Cleveland, OH 44195; email: rachita@ccf.org.

Received: August 29, 2018
Accepted: March 11, 2019

10.3928/23258160-20191009-04

Sign up to receive

Journal E-contents