Download Visual Acuity Is Correlated with the Area of the Foveal Avascular

Visual Acuity Is Correlated with the
Area of the Foveal Avascular Zone
in Diabetic Retinopathy and Retinal
Vein Occlusion
Chandrakumar Balaratnasingam, MD, PhD,1,2,3,4 Maiko Inoue, MD,1,2 Seungjun Ahn, MS,5
Jesse McCann, MD, PhD,1,2,3 Elona Dhrami-Gavazi, MD,1,2,3 Lawrence A. Yannuzzi, MD,1,2
K. Bailey Freund, MD1,2,3
Purpose: To determine if the area of the foveal avascular zone (FAZ) is correlated with visual acuity (VA) in
diabetic retinopathy (DR) and retinal vein occlusion (RVO).
Design: Cross-sectional study.
Participants: Ninety-five eyes of 66 subjects with DR (65 eyes), branch retinal vein occlusion (19 eyes), and
central retinal vein occlusion (11 eyes).
Methods: Structural optical coherence tomography (OCT; Spectralis, Heidelberg Engineering) and OCT
angiography (OCTA; Avanti, Optovue RTVue XR) data from a single visit were analyzed. FAZ area, point thickness
of central fovea, central 1-mm subfield thickness, the occurrence of intraretinal cysts, ellipsoid zone disruption,
and disorganization of retinal inner layers (DRIL) length were measured. VA was also recorded. Correlations
between FAZ area and VA were explored using regression models. Main outcome measure was VA.
Results: Mean age was 62.9!13.2 years. There was no difference in demographic and OCT-derived
anatomic measurements between branch retinal vein occlusion and central retinal vein occlusion groups (all
P " 0.058); therefore, data from the 2 groups were pooled together to a single RVO group for further statistical
comparisons. Univariate and multiple regression analysis showed that the area of the FAZ was significantly
correlated with VA in DR and RVO (all P # 0.003). The relationship between FAZ area and VA varied with age
(P ¼ 0.026) such that for a constant FAZ area, an increase in patient age was associated with poorer vision (rise in
logarithm of the minimum angle of resolution visual acuity). Disruption of the ellipsoid zone was significantly
correlated with VA in univariate and multiple regression analysis (both P < 0.001). Occurrence of intraretinal cysts,
DRIL length, and lens status were significantly correlated with VA in the univariate regression analysis (P # 0.018)
but not the multiple regression analysis (P " 0.210). Remaining variables evaluated in this study were not
predictive of VA (all P " 0.225).
Conclusions: The area of the FAZ is significantly correlated with VA in DR and RVO and this relationship is
modulated by patient age. Further study about FAZ area and VA correlations during the natural course of retinal
vascular diseases and following treatment is warranted. Ophthalmology 2016;123:2352-2367 ª 2016 by the
American Academy of Ophthalmology.
Supplemental material is available at www.aaojournal.org.
Diabetic retinopathy (DR) is the leading cause of vision loss
in the working-age population in developed countries.1
According to data from the National Eye Institute, there
were approximately 7.7 million diagnosed cases of DR in
the United States alone in 2010.2 Retinal venous occlusion
(RVO), due to central retinal vein occlusion (CRVO) or
branch retinal vein occlusion (BRVO), is the second most
common retinal vascular disorder leading to significant
vision loss, with an estimated number of global cases in
2010 exceeding 16.4 million.3 The prevalence of DR and
RVO is projected to increase to pandemic portions over
2352
ª 2016 by the American Academy of Ophthalmology
Published by Elsevier Inc.
the next 30 years, as is the socioeconomic burden
associated with these conditions.4 Thus, there is an urgent
need to improve our understanding of the pathophysiologic
mechanisms and anatomic correlates for significant visual
loss due to these disorders.
The foveal avascular zone (FAZ) is a specialized region
of the human retina that approximates the region of highest
cone photoreceptor density and oxygen consumption.5,6
Histologic techniques7e9 and a range of in vivo imaging
modalities10e19 have highlighted the variability in FAZ
topology in normal human eyes. In healthy eyes, the size of
http://dx.doi.org/10.1016/j.ophtha.2016.07.008
ISSN 0161-6420/16
Balaratnasingam et al
the FAZ does not seem to influence visual function,9,17,20
but the relationship between FAZ size and visual acuity
(VA) in retinal vascular diseases remains a matter of
conjecture. Much of our understanding concerning the
relationship between FAZ topology and visual function has
been attained from studies that utilized fluorescein angiography (FA) techniques to visualize the retinal circulation.21,22 Although these studies significantly aided our
understanding of the pathogenic mechanisms leading to
vision loss in retinal vascular diseases, they seldom
accounted for the influence of other anatomic changes, such
as ellipsoid zone (EZ) disruption,23 that may have also
affected visual function.
Volumetric imaging using optical coherence tomography angiography (OCTA) permits rapid, noninvasive
evaluation of the retinal circulation.24e29 Recent work has
shown that OCTA permits reproducible measurement of
FAZ dimensions and provides quantitative vascular information comparable to histologic examination.9,30 The
technique of OCTA thus appears suited for delineating
relationships between the morphometric properties of the
FAZ and VA. In this report, a biomorphometric analysis
incorporating OCTA and structural optical coherence
tomography (OCT) data is used to determine the significant
predictors of VA in DR and RVO. We show that FAZ area
is an independent predictor of VA in DR and RVO. We
also demonstrate that patient age modulates the relationship
between FAZ area and VA in retinal vascular diseases. The
results of this report, therefore, have important clinical
applications.
Methods
This study followed the tenets of the Declaration of Helsinki and
was approved by the Institutional Review Board at North Shore
Long Island Jewish Health System. Data were stored and managed
in compliance with guidelines from the Health Insurance Portability and Accountability Act.
Subjects
Consecutive cases of DR, BRVO, and CRVO seen between
August 2014 and October 2015 by 2 retina specialists (L.A.Y.
and K.B.F.) at Vitreous Retina Macula Consultants of New York
were enrolled in this study. Treatment-naïve and treated cases
were included. The diagnosis of diabetes mellitus was based on
the results of fasting blood samples and glycosylated hemoglobin
(HbA1c). Only patients with clinical signs of diabetic retinopathy, according to the Early Treatment Diabetic Retinopathy
Study (ETDRS) grading criteria,31 were included in this study
(Fig 1). Patients with diabetes mellitus without evidence of
diabetic retinopathy (level 10 scale ETDRS criteria) were not
included in this study. Central retinal vein occlusion was
determined by the presence of retinal vein dilation, retinal
edema, or scattered superficial or deep hemorrhages with or
without the presence of optic disc edema/hyperemia (Fig 1).
Branch retinal vein occlusion was determined by the presence
of the same clinical features as CRVO, but confined to a focal
region in the retina corresponding to a specific arteriovenous
crossing (Fig 1). Long-standing vein occlusion was determined
by the presence of occluded or sheathed retinal veins and/or
vascular anastomoses. With the exception of 4 cases (1 case of
(
Foveal Avascular Zone
BRVO and 3 cases of DR), the clinical evaluation of all subjects
included imaging with FA (Optos 200Tx [Optos, Dunfermline,
Scotland, United Kingdom], Topcon TRC 501x fundus camera
[Topcon Imagenet, Tokyo, Japan], or Heidelberg Spectralis
HRAþOCT [Heidelberg Engineering, Heidelberg, Germany]).
Other inclusion criteria for this study included the following: (1)
clear ocular media; (2) absence of significant refractive error; (3)
absence of significant concurrent ocular diseases. Patients
deemed to have unsatisfactory OCTA images such that FAZ area
could not be reliably measured, as described below, were also
excluded from this study. All subjects underwent slit-lamp biomicroscopy, dilated funduscopic examination, and measurement
of pinhole VA on the day of OCTA imaging. Demographic and
clinical information, including treatment history, lens status
(phakic or pseudophakic), and duration of disease, was obtained
from patient records.
Optical Coherence TomographyeDerived
Measurements of Foveal Anatomy
All patients were imaged with Spectralis spectral-domain OCT (SD
OCT) (Heidelberg Engineering, Heidelberg, Germany) on the day
of OCTA imaging. A raster scan protocol centered at the fovea
(range, 20& '25& e30& '20& ) was used. The following assessments
of foveal structure were attained using the macular volume scan
(Figs 2 and 3):
1. Point thickness of the central fovea (point foveal thickness,
PFT): Determined using the B-scan image of the central
fovea and defined as the distance between the retinal
pigment epithelium and inner limiting membrane (ILM).
Calipers provided by the OCT software were used to
determine this measurement. Point thickness of the central
fovea was measured by 2 independent examiners (M.I. and
J.M.) and the average measurement was used for statistical
analysis.
2. Occurrence of intraretinal cystoid changes: Categorically
graded as being present or absent by 2 independent
observers (M.I. and J.M.). Intraretinal cysts were identified using previously determined OCT criteria32 and
were defined as the occurrence of round or oval
hyporeflective spaces arranged in linear aggregates at
the level of the inner nuclear or outer nuclear/Henle
fiber layers. When the evaluation was inconsistent
between the 2 graders, a third masked reader (C.B.)
made the final arbitration.
3. Integrity of the ellipsoid zone in the central fovea: The
central 3 mm of the fovea, as circumscribed by the ETDRS
grid, was evaluated. Disruption of the EZ was categorically
graded as being present or absent by 2 independent
observers (M.I. and J.M.). An absent grading was denoted
if there was any disruption of the EZ on OCT. When the
evaluation was inconsistent between the 2 graders, a third
masked reader (C.B.) made the final arbitration.
4. Central 1-mm subfield thickness (CST): Recorded from the
retinal thickness ETDRS grid generated by Spectralis software (Heidelberg Engineering, Heidelberg, Germany). The
occurrence of diabetic macular edema, defined as a central
foveal thickness of greater than 275 mm (ETDRS central
subfield),33 was also evaluated in the diabetic cohort.
5. Disorganization of the retinal inner layers (DRIL) length:
Previously reported definitions of DRIL were used.34,35
Disorganization of the retinal inner layers length was
defined as the horizontal extent (in microns) for which any
boundaries between the ganglion celleinner plexiform
layer complex, inner nuclear layer, and outer nuclear layer
2353
Ophthalmology Volume 123, Number 11, November 2016
Figure 1. Clinical and fluorescein angiographic features of typical cases of diabetic retinopathy (A, B), branch retinal vein occlusion (C, D), and central
retinal vein occlusion (E, F) included in this study.
could not be identified. The horizontal extent of DRIL
within each B-scan image within the central 3 mm of the
fovea, circumscribed using the ETDRS grid, was assessed.
Average of measurements was used to derive a global
DRIL measurement for the eye. Global DRIL measurement
was determined by 2 independent examiners (M.I. and
C.B.) and the average measurement was used for quantitative analysis.
2354
6. Foveal pit volume: This measurement was determined by
applying the Cavalieri principle of stereologic analysis36 as
illustrated in Figure 3. The following steps were used to
calculate foveal pit volume:
i. The rim of the foveal pit on each B-scan image was
determined using ImageJ (a publicly available image
processing program developed by Wayne Rasband,
Balaratnasingam et al
(
Foveal Avascular Zone
Figure 2. Optical coherence tomography (OCT)-derived anatomic measurements of the fovea. Color fundus (A), OCT angiography (B), and structural OCT
image of the fovea (C) from an eye with diabetic retinopathy are provided. Foveal avascular zone (FAZ) area (red shade) was determined using the OCT
angiography image and manual tracing techniques. B-scan images of the fovea were used to categorically grade the presence or absence of intraretinal cysts and
ellipsoid zone (EZ) disruption. Point foveal thickness (PFT) was measured as the distance between the retinal pigment epithelium and inner limiting membrane.
Disorganization of the retinal inner layers length was also measured. Automated measurement of the 1-mm central subfield thickness (213 mm in this case),
calculated using Heidelberg Spectralis software (top right corner of the color image), was also used in the regression analysis. The area imaged by OCT angiography (green inset) and the region from where the single B-scan image was acquired (white line) are illustrated in the color photograph.
National Institutes of Health, Bethesda, MD; http://
rsb.info.nih.gov/ij/index.html). Specifically, the first
point away from the foveal center (in nasal and temporal directions) where the slope of the ILM was
measured to be zero was defined as the rim. The angle
tool on ImageJ was used to determine this
measurement.
ii. The area confined by the ILM and a straight line connecting the 2 points denoting the foveal pit rim on the
B-scan image, as determined in step (i), was then
calculated using manual planimetry.
iii. The volume of the foveal pit between 2 adjacent OCT
slices (herein called a segment) was then determined
using this equation:
Vol ¼ d
!
"
Ax þ Axþ1
2
where: Vol ¼ volume of foveal pit within segment bounded by
OCT slice number x and xþ1
d ¼ distance between adjacent slices in mm
A ¼ area between the ILM and a straight line connecting
the edges of the foveal pit rim in mm2
x ¼ OCT slice number
iv. Foveal pit volume was calculated by summing the volumes
of individual segments. The number of segments in the
OCT volume was (n)1), where n ¼ total number of slices
that spanned the foveal pit.
Measurement of the Area of the Foveal
Avascular Zone
The area of the FAZ was measured using OCTA. The instrument
used was based on the Optovue RTVue XR Avanti (Optovue, Inc,
Fremont, CA) to obtain split-spectrum amplitude-decorrelation
angiography images. This instrument has an A-scan rate of 70 000
scans per second, using a light source centered on 840 nm and a
bandwidth of 45 nm. Each OCTA volume contains 304'304
A-scans with 2 consecutive B-scans captured at each fixed position
before proceeding to the next sampling location. The scan area was
2355
Ophthalmology Volume 123, Number 11, November 2016
Figure 3. Calculation of foveal pit volume using optical coherence tomography (OCT). Color fundus (A) and OCT angiography image (B) of a patient
with a superior branch retinal vein occlusion are presented. Three contiguous OCT slices illustrate the application of the Cavalieri principle in stereology for
calculating foveal pit volume. Areas of the central fovea from which B-scan images were acquired are presented on the reflectance image (C; white lines).
The rim of the foveal pit was defined as the first point away from the central fovea in the B-scan image where the slope of the inner limiting membrane
(ILM) was zero. Manual tracing of the ILM and a straight line connecting the 2 points denoting the foveal rim was used to calculate the cross-sectional area
of the foveal pit (pink shade) in each B-scan image. The volume of a segment, defined as the volume between 2 contiguous OCT slices, was then determined
by averaging the area of the foveal pit in the 2 slices and multiplying it by the distance (d) separating the slices. Summation of volumetric measurements from
each segment allowed calculation of the entire volume of the foveal pit. The area of the foveal avascular zone as seen on OCT angiography (red shade) is
clearly discernible as is the contour of the foveal pit upon 3-dimensional visualization of the structural OCT (D) data. The area of OCT angiography imaging
is denoted by a green inset on the color image.
2'2 mm or 3'3 mm centered on the fovea. Each OCTA volume is
acquired in approximately 3 seconds and 2 orthogonal OCTA
volumes were acquired to perform motion correction to minimize
motion artifacts arising from microsaccades and fixation changes.
Angiography information displayed is the average of the decorrelation values when viewed perpendicularly through the thickness
being evaluated. If the image processed with motion-correction
software demonstrated artifact involving the FAZ in the form of
2356
doubling of vascular structures or sideways shearing, then the case
was excluded from analysis. Only OCTA images with signal
strength above 50 were included in this study.
Image analysis to calculate FAZ area was performed using
Adobe Photoshop CC (Adobe Systems, Inc., San Jose, CA) and
our previously reported technique.37 Images from screen capture
program integrated in the AngioView software, representing the
full thickness of the retina (inner and outer capillary networks
Balaratnasingam et al
combined), were used for quantitative analysis. Gain and contrast
were adjusted if necessary, using the brightness/contrast function,
to allow clear delineation of the FAZ. The magic wand tool was
then used to manually demarcate the boundaries of the FAZ after
the innermost capillaries in the fovea were identified. We used a
tolerance of 15 and repeatedly used the magic wand tool to select
the FAZ. If the selection spilled outside of the FAZ, the excess
selection was trimmed using the lasso tool. If there were
regions in the FAZ where pixel value exceeded the selected
pixel value plus tolerance, then they were not selected by the
magic wand tool. In these instances, the lasso tool was also
used to manually select pixel outliers. The area measurement
function was then used to determine the area of the FAZ in
pixels2 and this value was converted to micrometers squared
(mm2) using scale conversion. The FAZ area was determined
by 2 independent examiners (M.I. and J.M.) and the average
measurement was used for quantitative analysis.
Statistical Analysis
Data were summarized with descriptive statistics. Univariate
regression models estimated using generalized estimating equations (GEE) were fit using FAZ area, foveal pit volume, age, lens
status (phakic or pseudophakic), PFT, CST, disease (DR or RVO),
sex (male or female), occurrence of intraretinal cysts, occurrence of
EZ disruption, DRIL length, or previous treatment (no treatment,
laser photocoagulation, intravitreal antievascular endothelial
growth factor [VEGF] therapy, vitrectomy, or combination therapy) as a single predictor and VA as the outcome. Treatment was a
categorical variable with 5 levels; thus, we created 4 dummy variables (laser photocoagulation, intravitreal anti-VEGF injection,
combination therapy, and vitrectomy) to be directly included in the
regression model, and also contain the same information as the
single categorical variable. The reference group for treatment was
the cohort of subjects that had not received any previous treatment.
Generalized estimating equations were used to account for correlations of observations from the same subject, as some patients had
bilateral imaging. Interactions between FAZ area and age/CST/
disease (DR or RVO)/integrity of the EZ/occurrence of intraretinal
cysts/DRIL length/foveal pit volume were then explored using
regression analysis (Appendix, available at www.aaojournal.org).
Results from univariate regression models and also models
assessing interactions were then used to create a final
multivariate model in which VA was the outcome. A summary
of all the models explored in this study is provided in the
Appendix. Visual acuity measurements were converted to
logarithm of the minimum angle of resolution (logMAR) units
before analysis. A P value less than or equal to 0.05 was
considered significant. Statistical analysis was performed using
R 3.1.2 and Stata 12. Descriptive results in this manuscript are
provided as mean ! standard deviation.
Results
General
A total of 95 eyes from 66 subjects (36 male and 30 female) were
analyzed. The mean age of the cohort was 62.9!13.2 years. Mean VA
was 0.25!0.24 logMAR (mean Snellen acuity of 20/35). A total of 53
right eyes and 42 left eyes were evaluated. Twenty-six eyes were
pseudophakic. The cohort comprised 65 eyes with DR and 30 eyes
with RVO (19 with BRVO and 11 with CRVO). The diabetic cohort
comprised 13 eyes with mild DR, 22 eyes with moderate DR, 6 eyes
with severe DR, and 24 eyes with proliferative DR according to the
(
Foveal Avascular Zone
ETDRS scale. Twenty-five eyes in the entire cohort had not received
any previous treatment, 25 eyes had previously received intravitreal
anti-VEGF therapy, 17 eyes had received laser photocoagulation, 3
eyes had undergone vitrectomy, and 25 eyes had received a combination of the above nonsurgical treatments. The condition of diabetic
macular edema, defined as a central subfield thickness greater than
275 mm,33 was present in 38 of 65 diabetic eyes (58.4%). Mean signal
strength of OCTA images for the cohort was 65.2!5.1.
Summary statistics for FAZ area, lens status, VA, duration of
disease, PFT, CST, frequency of intraretinal cysts, foveal pit volume, DRIL length, and frequency of EZ disruption for the entire
cohort are provided in Table 1. There was no difference in
demographic and OCT-derived anatomic measurements (all
P " 0.058) between BRVO and CRVO groups (Table S1,
available at www.aaojournal.org); therefore, data from the 2
groups were pooled together to a single RVO group for further
statistical comparisons. Interobserver disagreement on the status
of the ellipsoid zone or intraretinal cysts required a tie-breaker
examination in 3 and 4 instances, respectively.
Mean age (P ¼ 0.822), VA (P ¼ 0.974), PFT (P ¼ 0.050), CST
(P ¼ 0.056), DRIL length (P ¼ 0.252), foveal pit volume (P ¼ 0.060),
frequency of pseudophakia (P ¼ 0.274), and frequency of EZ
disruption (P ¼ 0.473) was not significantly different between the DR
and RVO groups. Duration of disease was significantly greater in the
DR group compared with the RVO group (91.8 vs. 34.9 months;
P < 0.001). The frequency of treatment variables was significantly
greater in the DR group than in the RVO group (P ¼ 0.042). Mean
FAZ area was significantly greater in the DR group (P ¼ 0.019). A
greater proportion of eyes in the RVO demonstrated intraretinal cysts
on OCT (46.7% vs. 24.6%; P ¼ 0.033).
Mean DRIL length for the cohort was 235.2!506.4 mm. There
was a moderate positive correlation between FAZ area and DRIL
length (Pearson correlation coefficient ¼ 0.3961) that was statistically significant (P < 0.001; Fig 4).
It was possible to calculate foveal pit volume in only 73 of 95
eyes. In the remaining eyes, retinal thickening and intraretinal
cystic changes distorted the shape of the foveal pit such that it was
not possible to reliably identify its margins (Fig 5). Mean foveal pit
volume for the eyes that were measured was 0.14!0.11 mm3. A
moderate positive correlation between FAZ area and foveal pit
volume (P ¼ 0.001) was identified, with a Pearson correlation
coefficient of 0.3744 (Fig 6).
Association of Foveal Avascular Zone Area and
Other Variables with Visual Acuity
Results from univariate GEE analyses are summarized in Table 2.
Foveal avascular zone area (P < 0.001), the occurrence of EZ
disruption (P < 0.001), the occurrence of intraretinal
cysts (P ¼ 0.018), lens status (P ¼ 0.002), and DRIL length
(P < 0.001) were significantly associated with VA. Age
(P ¼ 0.600), PFT (P ¼ 0.283), CST (P ¼ 0.226), foveal pit
volume (P ¼ 0.237), disease (DR or RVO; P ¼ 0.489), female
sex (P ¼ 0.225), and history of previous treatment were not
associated with VA (all P " 0.072).
Point foveal thickness and CST were highly correlated
(r ¼ 0.893) and similarly associated with VA. Therefore, CST was
used as a predictor in subsequent multiple regression models and
PFT measurements were excluded from further analyses.
2357
Ophthalmology Volume 123, Number 11, November 2016
Table 1. Summary of Clinical and Optical Coherence TomographyeDerived Anatomic Measurements for Entire Cohort
Quantiles
Age (yrs)
VA (logMAR)
Duration of disease (mos)
Frequency of pseudophakia (%)
FAZ area (mm2)
Point foveal thickness (mm)
Central 1-mm subfield thickness (mm)
DRIL length (mm)
Foveal pit volume (mm3)
Frequency of intraretinal cysts (%)
Frequency of ellipsoid zone disruption (%)
Mean
SD
Minimum
First Quantile
Median
Third Quantile
Maximum
62.90
0.25
78.00
27.40
0.55
256.80
311.70
235.20
0.14
31.60
28.40
13.20
0.24
101.90
d
0.32
91.00
84.40
506.40
0.11
d
d
27.00
0.00
0.00
d
0.12
103.00
149.00
0.00
0.01
d
d
57.00
0.10
4.50
d
0.34
204.50
258.00
d
0.07
d
d
65.00
0.18
30.00
d
0.47
230.00
300.00
d
0.12
d
d
71.00
0.30
134.00
d
0.66
281.00
346.00
d
0.17
d
d
93.00
1.30
480.00
d
1.96
642.00
638.00
2305.70
0.70
d
d
DRIL ¼ disorganization of the retinal inner layers; FAZ ¼ foveal avascular zone; logMAR ¼ logarithm of the minimum angle of resolution; VA ¼ visual
acuity.
Interactions Between Foveal Avascular Zone
Area and Other Variables
The results of multivariate GEE analyses where interactions
between FAZ area and another predictor were explored using VA
as the outcome are summarized in Table 3. The only interaction
that was observed to be significant was the interaction between
FAZ area and age (P ¼ 0.026). Interactions between FAZ
area and CST (P ¼ 0.615), FAZ area and disease (DR or RVO;
P ¼ 0.598), FAZ area and the occurrence of intraretinal cysts
(P ¼ 0.453), FAZ area and the integrity of the EZ (P ¼ 0.655),
FAZ area and DRIL length (P ¼ 0.245), and FAZ area and
foveal pit volume (P ¼ 0.810) were not significant.
Predictive Model of Visual Acuity Using Foveal
Avascular Zone Area and Other Variables
Because FAZ area, the occurrence of intraretinal cysts, the occurrence
of EZ disruption, DRIL length, and lens status were found to be significant predictors of VA in univariate GEE analyses, they were used in
the final predictive model. Age was also used in the final predictive
model because the interaction between age and FAZ area was found to
be significant in multivariate GEE analyses. We also wanted to
determine if the type of retinal vascular disease (DR or RVO) would be
a significant predictor of VA, so this variable was also included in the
final model. The final model was therefore defined as follows:
Outcome: VA. Predictors: FAZ area, age, disease (DR or
RVO), occurrence of intraretinal cysts, occurrence of EZ disruption, lens status (phakic or pseudophakic), DRIL length, and an
interaction between FAZ area and age.
The results of this model are summarized in Table 4. FAZ area (P ¼
0.003), age (P < 0.001), EZ disruption (P < 0.001), and the interaction
between FAZ area and age (P < 0.001) were significant predictors of
VA. Disease (DR or RVO), lens status (phakic or pseudophakic),
DRIL length, and the occurrence of intraretinal cysts were not found
to be significant predictors of VA using this model (all P " 0.210).
Figure 7B is a graphical illustration of the final regression
model demonstrating how the relationship between FAZ area and
mean VA varies with age. Data from subjects above the mean
age of the cohort (62.9 years) are denoted in red and those
below the mean age are denoted in blue. Statistical techniques
previously described by Cohen et al38 were used to generate
2358
simple slopes for different age groups using the predictive
model. The predicted slope of the relationship between FAZ area
and VA for a population 1 standard deviation above the mean
age of the cohort (dashed line), the mean age of the cohort
(continuous line), and 1 standard deviation below the mean age
of the cohort (dotted line) are provided. The slopes of the 3 lines
are different, thereby providing evidence that the relationship
between FAZ area and VA varies with age.
Case Illustrations
A comparison of VA from 2 patients with DR of similar age and
OCT-derived anatomic measurements (aside from FAZ area) are
presented in Figure 8.
Case A is the right eye of a 70-year-old female patient that was
diagnosed with type 2 diabetes mellitus 26 years ago. Intraretinal
cysts were seen on SD OCT, PFT was 239 mm, the integrity of
the EZ was judged to be intact, and DRIL was present. The area of
the FAZ was measured as 0.44 mm2 and the VA was recorded
as 20/40.
Case B is the left eye of a 71-year-old male patient that was
diagnosed with type 2 diabetes mellitus 29 years previously.
Intraretinal cysts were seen on SD OCT, PFT was 254 mm, the EZ
was intact, and DRIL was present. The area of the FAZ was
measured as 0.76 mm2 and the VA was recorded as 20/100.
Discussion
We utilized OCT-derived anatomic measurements and
regression models to determine if FAZ area is a significant
predictor of VA in DR and RVO. Although robust statistical
evaluations demonstrated consistent findings across multiple
analyses, we emphasize that the strength of our conclusions
is modulated by the limited sample size of our cohort. The
major findings of this study are as follows: (1) FAZ area is
significantly correlated with VA in diabetic retinopathy and
retinal vein occlusion; (2) age modulates the relationship
between FAZ area and VA such that for a constant FAZ area
a relatively older group of patients will have poorer vision
(i.e., a higher mean logMAR visual acuity).
Balaratnasingam et al
(
Foveal Avascular Zone
Figure 4. Relationship between foveal avascular zone (FAZ) area and disorganization of retinal inner layers (DRIL) length. FAZ area was significantly and
positively correlated to DRIL length, as illustrated by the clinical images of 2 different patients (A, B). The area imaged by optical coherence tomography
angiography is denoted by the green inset on each fluorescein angiogram image and the area of the FAZ in each eye is highlighted in red shade. DRIL length
was defined as the horizontal extent (in microns) for which any boundaries between the ganglion celleinner plexiform layer complex, inner nuclear layer,
and outer nuclear layer could not be identified.
The FAZ is a specialized region of the human macula
that contains the highest density of cone photoreceptors.5
With respect to metabolic activity, oxygen consumption
within the macula, per gram of tissue, is also greater than
most other organs in the human body.39 Cells within the
FAZ are principally nourished by the choroid and during
physiologic conditions the choroidal circulation is able to
meet the metabolic demands of the FAZ. Therefore, it is
2359
Ophthalmology Volume 123, Number 11, November 2016
Figure 5. Distortion of foveal pit morphology due to retinal structural changes. Color (A), fluorescein angiography (B), and optical coherence tomography
(OCT) angiography (C) images of the macula of a patient with diabetic retinopathy are presented. Marked thickening of the central fovea is evident on the
thickness map generated by Spectralis software (Heidelberg Engineering, Heidelberg, Germany; D) and the foveal pit is not discernible upon 3-dimensional
visualization of the structural OCT (E) data. B-scan images (IeIII) of the central fovea demonstrate disruption of the foveal pit due to intraretinal cystic
changes. Sites from which B-scan images were acquired are denoted on the color image (white lines) and the area studied using OCT angiography is denoted
on the fluorescein angiogram (yellow inset). It was not possible to calculate foveal pit volume in these eyes, but the area of the foveal avascular zone (red
shade) was readily measurable.
not surprising that a correlation between FAZ size and VA
in the normal human eye has not been reported. Samara
et al17 quantified the area of the superficial FAZ in 70
normal eyes using OCTA, and although there was
significant variation in FAZ area (range, 0.071e0.527
mm2), the VA of every eye in their cohort was recorded
as 20/20.17 Similarly, Laatikainen and Larinkari20
measured the diameter of the FAZ in 167 healthy eyes
using FA. Despite great variation in FAZ diameter (range,
2360
0.25e0.85 mm; approximated from graphical data
presented in their manuscript), the VA of all subjects in
their study was recorded as 20/25 or better.
Yu et al showed that the inner retina was rendered relatively anoxic following experimental retinal arterial occlusion,40 thereby providing evidence that vasculogenic insults
to the fovea perturb the delicate balance between oxygen
supply and consumption within the inner retina. Because
the choroid is unable to sufficiently oxygenate the inner
Balaratnasingam et al
Figure 6. Relationship between foveal avascular zone (FAZ) area and
foveal pit volume in diabetic retinopathy and retinal vein occlusion. A
scatterplot demonstrates a significant relationship, with a Pearson correlation coefficient of 0.3744.
retina following retinal vascular injury, it is plausible that
FAZ size, and therefore the area of the macula devoid of
a retinal blood supply, would correlate to the degree of
visual dysfunction in DR and RVO. However, previous
reports concerning the relationship between FAZ size and
VA in retinal vascular diseases have reached discordant
conclusions. Arend et al21 compared FAZ size between
diabetic patients with decreased VA (20/50 or worse) and
(
Foveal Avascular Zone
unaffected VA (median VA of 20/25) and showed that the
FAZ was enlarged by 73% in eyes with decreased VA. In
their study, scanning laser video-fluorescein angiograms
were used to quantify FAZ topology and fundus photography was used to grade the occurrence of macular edema.
Remky et al22 evaluated FAZ size, using digitized FA
images acquired with a scanning laser ophthalmoscope, in
24 eyes with acute CRVO and compared it to
measurements from 98 healthy volunteers. Although FAZ
size was increased in the CRVO group (0.32!0.17 mm2)
compared with the control group (0.22!0.07 mm2), it was
not found to be correlated with VA. Parodi et al41 also
used FA techniques to compare FAZ area between 20
patients with BRVO and 41 control subjects. Mean FAZ
area was shown to be greater in eyes with BRVO
(0.56!0.34 mm2) compared with controls (0.26!0.07
mm2). Their study also showed that VA impairment due
to BRVO was correlated with FAZ enlargement. Most of
these studies only examined the association between FAZ
size and VA using FA and omitted the influence of other
disease-induced structural changes, such as EZ disruption,23 DRIL length,35 and intraretinal cystic changes,42 in
their statistical models. This may be one explanation for
the discrepancies in findings between previous FA-based
studies.
Optical coherence tomography angiography is a relatively new imaging modality that utilizes flow properties
within a defined volume of tissue to visualize vascular
structures and therefore obviates dye administration.43 Over
the past year, there have been a considerable number of
studies that have reported the spectrum of retinal vascular
changes due to DR, BRVO, and CRVO using
OCTA.24,44e47 With regard to quantitative evaluation,
recent evidence also suggests that OCTA is a reliable
technique for measuring FAZ area that compares favorably
Table 2. Results of Simple Regression Models (Generalized Estimating Equations) Using Visual Acuity as the Outcome
2
FAZ area (mm )
Age (yrs)
Lens status
Point foveal thickness (mm)
Central 1-mm subfield thickness (mm)
Disease (DR or RVO)
Female
Intraretinal cysts
Ellipsoid zone disruption
DRIL length (mm)
Foveal pit volume (mm3)
Treatment
No treatment (Reference Group)
Laser photocoagulation
Anti-VEGF injection
Combination therapy*
Vitrectomy
Coefficient
Standard Error
P Value
95% CI
0.34
0.001
0.1735
0.0004
0.0006
0.04
0.07
0.14
0.34
0.00018
0.2641
0.07
0.002
0.0566
0.0003
0.0005
0.06
0.05
0.06
0.06
0.000046
0.2232
<0.001
0.600
0.002
0.283
0.226
0.489
0.225
0.018
<0.001
<0.001
0.237
(0.19, 0.48)
()0.003, 0.005)
(0.06263, 0.2844)
()0.0003, 0.001)
()0.0004, 0.002)
()0.081, 0.169)
()0.04, 0.17)
(0.02, 0.25)
(0.22, 0.45)
(0.00009, 0.00027)
()0.1732, 0.7015)
0.007
0.101
0.127
0.003
0.051
0.06
0.071
0.032
0.897
0.091
0.072
0.918
()0.09, 0.11)
()0.02, 0.22)
()0.01, 0.27)
()0.059, 0.066)
CI ¼ confidence interval; DR ¼ diabetic retinopathy; DRIL ¼ disorganization of the retinal inner layers; FAZ ¼ foveal avascular zone; RVO ¼ retinal vein
occlusion; VEGF ¼ vascular endothelial growth factor.
*Two or more of the listed treatments.
2361
Ophthalmology Volume 123, Number 11, November 2016
Table 3. Multiple Regression Models (Generalized Estimating Equations) Exploring Interactions Between Foveal Avascular Zone Area
and Other Variables Using Visual Acuity as the Outcome
Interaction Terms
FAZ
FAZ
FAZ
FAZ
FAZ
FAZ
FAZ
area
area
area
area
area
area
area
'
'
'
'
'
'
'
Age
CST
Disease
Integrity of EZ
Intraretinal cysts
DRIL length
Foveal pit volume
Coefficient
Standard Error
P Value
95% CI
0.014
)0.001
)0.081
0.063
0.098
)0.002
0.135
0.006
0.002
0.154
0.14
0.131
0.0001
0.561
0.026
0.615
0.598
0.655
0.453
0.245
0.810
(0.002, 0.026)
()0.004, 0.002)
()0.382, 0.220)
()0.213, 0.338)
()0.159, 0.355)
()0.0004, 0.0001)
()0.965, 1.234)
CI ¼ confidence interval; CST ¼ central 1-mm subfield thickness; DRIL ¼ disorganization of the retinal inner layers; EZ ¼ ellipsoid zone; FAZ ¼ foveal
avascular zone.
to histology.9 Carpineto et al30 quantified FAZ area in 60
eyes using OCTA and demonstrated a high intraobserver
intersession concordance correlation coefficient (mean
0.99) and high interobserver intrasession concordance
correlation coefficient (mean 0.99). The advantages of
OCTA for reliably evaluating FAZ size were exploited in
a number of recent studies that quantified the topologic
features of the FAZ in DR and RVO. Salz et al48 used a
prototype swept-source OCT to delineate the innermost
vessels surrounding the avascular zone in DR, which they
denoted as the foveal nonflow zone, and showed that the
size of the foveal nonflow zone was significantly greater in
eyes with proliferative DR compared with normal eyes and
those with nonproliferative DR. Takase et al49 used OCTA
to show that FAZ area was significantly greater in the DR
group compared with a control group but, similar to the
report by Salz et al,48 correlations between FAZ size and
VA were not performed. Samara et al50 evaluated 17 eyes
with BRVO and compared FAZ area measurements
between the eye with BRVO and the fellow normal eye
from the same subject. They found that FAZ area in
BRVO was significantly enlarged at the level of the deep
capillary network only. They also reported that the area of
the deep FAZ correlated positively with logMAR VA.
The important work by Samara et al,50 however, did not
consider the influence of other disease-induced anatomic
changes, such as EZ disruption and intraretinal cysts, which
may have also influenced VA. Therefore, it was not possible
to determine if VA change was correlated to FAZ size or
was due to other retinal anatomic changes that may have
confounded the analysis. Similarly, a number of other
studies that have reported the correlation between FAZ size
and VA also did not account for the effects of confounding
variables.21,22,33,41,51
In this study, FAZ area together with other OCT-derived
anatomic measurements such as PFT, CST, intraretinal
cysts, DRIL length, and the integrity of the EZ were used to
determine the significant predictors of VA in DR and RVO.
Important clinical variables such as age, sex, and disease
type (DR or RVO) were also used in the analysis. We
attained FAZ measurements using the entire volume scan
and did not stratify capillary networks into superficial and
deep networks to avoid errors associated with image
segmentationdShahlaee et al52 previously showed that
interobserver agreement did not meet the lowest
acceptable grader agreement for isolated measurements of
the deep vascular network using OCTA. We demonstrate
that FAZ area is a significant predictor of VA in DR and
RVO. Univariate and multiple regression analysis using
GEEs revealed that irrespective of disease type (DR or
RVO), FAZ area was significantly correlated with VA,
thereby implicating a degree of overlap in the pathways
Table 4. Results of the Final Multiple Regression Model where the Outcome Was Visual Acuity and the Predictors Included Foveal
Avascular Zone Area, Age, Disease, Occurrence of Ellipsoid Zone Disruption, Occurrence of Intraretinal Cysts, Lens Status, Disorganization of the Retinal Inner Layers, and the Interaction Between Foveal Avascular Zone Area and Age
2
FAZ area (mm )
Age (yrs)
Disease (DR or RVO)
Ellipsoid zone disruption
Intraretinal cysts
Lens status (phakic or pseudophakic)
DRIL length
FAZ area ' Age
Constant
Coefficient
Standard Error
P Value
95% CI
)0.703
)0.007
0.041
0.245
)0.002
0.077
0.000
0.012
0.531
0.239
0.002
0.043
0.057
0.043
0.062
0.000
0.003
0.147
0.003
0.001
0.344
<0.001
0.962
0.210
0.240
<0.001
<0.001
()1.171, )0.235)
()0.011, )0.003)
()0.044, 0.126)
(0.133, 0.357)
()0.087, 0.083)
()0.044, 0.126)
()0.000, )0.001)
(0.006, 0.018)
(0.244, 0.818)
CI ¼ confidence interval; DR ¼ diabetic retinopathy; DRIL ¼ disorganization of the retinal inner layers; FAZ ¼ foveal avascular zone; RVO ¼ retinal vein
occlusion.
2362
Balaratnasingam et al
Figure 7. Correlation between area of the foveal avascular zone (FAZ) and
visual acuity (VA). A, A scatterplot of the entire cohort is provided. B, A
graphical illustration of the final regression model demonstrating how the
relationship between FAZ area and mean VA varies with age is presented.
Blue and red circles represent subjects below and above the mean age of the
cohort (62.9 years), respectively. The predicted slope of the relationship
between FAZ area and VA for a population 1 standard deviation above the
mean age of the cohort (dashed line), the mean age of the cohort
(continuous line), and 1 standard deviation below the mean age of the
cohort (dotted line) are provided. The slopes of the 3 lines are different,
thereby providing evidence that the relationship between FAZ area and
VA varies with age. BCVA ¼ best-corrected visual acuity.
by which retinal vascular diseases lead to visual
dysfunction.
Foveal homeostasis is critically regulated by retinal glia.53
Astrocytes and Muller cells are the predominant glia subtypes
(
Foveal Avascular Zone
in the primate macula.53,54 From an embryological standpoint,
the pattern of astrocyte development closely approximates
retinal vascularization and a number of studies have shown
that glial cells codistribute with blood vessels in the retina.54,55
Therefore, it is expected that eyes with a relatively larger FAZ
also have a lower number of glia in the macula. During
physiologic states, the concentration of glial and vascular elements in the fovea in those with larger FAZs may be sufficient
to maintain visual function. However, following insults that
perturb the structural and biochemical systems that ordinarily
support macular homeostasis,56 these mechanisms may
rapidly decompensate; more so when there is an associated
increase in energy demands by regional neurons. We
postulate that eyes with larger FAZs have an increased
susceptibility to visual dysfunction due to the lower baseline
concentrations of glial and vascular elements. We also
postulate that such eyes are more vulnerable to imbalances
in energy supply/demand relationships during disease states
and may suffer a greater level of injury than an eye with a
smaller FAZ, following an insult of comparable magnitude.
We emphasize, however, that our hypothesis is entirely
speculative and focused work with histologic correlation is
required for validation of these concepts.
The other major finding in this study is that age modulates the relationship between FAZ area and VA in DR and
RVO. For a constant FAZ area, mean logMAR VA was
found to be higher in a relatively older age group. Agerelated decline in VA is known to occur in patients absent
of ocular disease; however, there is little evidence to suggest
that it is due to age-related changes in FAZ topology.57 In
the study by Laatikainen and Larinkari,20 a gradual
enlargement in FAZ size was noted until the fifth decade
of life in healthy eyes after which significant increases
were not detected. Their study concluded that the size of
the FAZ did not affect VA in healthy eyes. Further work
is required to understand the cellular basis for the
interaction between FAZ area and age in retinal vascular
diseases. Again, disease-induced alterations to Muller
cells58 and astrocytes59 may underlie this interaction.
Disorganization of inner retinal layers is an OCT-derived
biomarker that is predictive of baseline VA and VA after
resolution of diabetic macular edema.34,35 Histologic correlations for DRIL have not been performed, but DRIL is
thought to correlate to regions where synaptic connections
of bipolar, horizontal, and amacrine cells within the inner
retina have been disrupted.35 The inner retina is
predominantly nourished by the retinal circulation60,61;
therefore, it is plausible that vasculogenic insults at the level
of the inner retina will result in observable retinal structural
changes that manifest as DRIL on OCT. The data in our
study support this speculation, as we found a significant
positive correlation between FAZ size and DRIL length.
Although our results implicate an important relationship
between FAZ size and DRIL, it will be essential to validate
these findings by defining the temporal relationships
between these 2 variables.
The FAZ is required for foveal development and complete foveal excavation.62 Although causeeeffect
relationships in foveal development are being refined, the
2363
Ophthalmology Volume 123, Number 11, November 2016
Figure 8. Case illustrations. Patient A and patient B are of similar age and demonstrate similar foveal anatomic changes due to diabetic retinopathy, with the
exception of the area of the foveal avascular zone. Patient B has a significantly larger foveal avascular zone area and also a lower visual acuity (VA).
formation of the FAZ before the foveal pit during
embryologic development suggests that the size of the
FAZ regulates the degree of foveal excavation.63 Dubis
et al18 calculated foveal pit volume in 42 subjects using
OCT techniques and demonstrated that it was significantly
correlated with FAZ area. Because foveal pit volume is
intrinsically related to FAZ area, foveal pit volume might
2364
be considered a useful biomarker of visual function in
retinal vascular disease. However, in this study we found
that it was not possible to reliably measure foveal pit
volume in 22 of 95 eyes (23.2%) due to retinal thickening
and intraretinal cystic changes that subsequently distorted
the contour of the foveal pit. Furthermore, foveal pit
volume was not shown to be significantly associated with
Balaratnasingam et al
VA using regression analysis. Collectively, these findings
suggest that FAZ area has greater utility as a biomarker of
visual function than foveal pit volume in DR and RVO.
Arguably, FAZ area is also easier to quantify than foveal
pit volume in the clinical setting using current image
analysis software.
We acknowledge several limitations of this study,
including its retrospective design and the limited number of
subjects. Prospective studies of larger sample size will be
important to reaffirm the findings of this report. Another
limitation of this study is the nonstandardized manner in
which VA measurements were attained. Future studies that
aim to evaluate the relationship between VA and FAZ area
should utilize standardized VA measurement protocols such
as those used in clinical trials.64 Also, because image
artifacts due to poor fixation or media opacities may
preclude clear visualization of the macula, it may not be
possible to measure FAZ area using OCTA in all patients
with retinal vascular disease.65 Additionally, when
capillary flow rates fall below the threshold detected by
OCTA algorithms, the terminal capillary ring may not be
completely visualized in certain instances and FAZ area
measurements may therefore be influenced. We also
emphasize that the etiopathology of retinal vascular
diseases, particularly DR, is complex. Several important
predictors that have been shown to correlate with disease
progression and visual loss, such as age of disease onset,
duration of disease, HbA1c measurements, renal disease,
and hypertension, were not used in the regression
analyses.66e68 Some of these parameters can be difficult
to reliably discern from patient medical records.69,70
Furthermore, The Diabetes Control and Complication
Trial showed that approximately 11% of the variation in
retinopathy risk was associated with HbA1c readings and
duration of diabetes,71 suggesting that 89% of the variation
in risk was due to factors independent of HbA1c.72 Because
OCT is widely used in the clinical management of retinal
vascular diseases, the purpose of this report was to
demonstrate that FAZ area is a useful metric for grading
the functional severity of maculopathy due to DR and
RVO. Automated algorithms and advances in OCTA
software may facilitate rapid acquisition of FAZ area
measurements, thereby enabling clinicians to routinely
integrate morphologic information about the FAZ into
clinical practice.48,51 Because FAZ area is also correlated
to foveal capillary density and foveal intercapillary distances in some conditions,73 by extension, it is possible that
these latter quantitative metrics may also be correlated with
VA. We emphasize that biomorphometric modeling
performed in this study utilized data from a single visit;
therefore, the results of this study cannot be used to
determine the risk of disease progression or predict
response to treatment. The previous study by Sim et al74
showed that the rate of FAZ enlargement ranged between
5% and 10% of baseline FAZ area per year in eyes with
diabetic macular ischemia; therefore, further study about
FAZ area and VA correlations during the natural course
of retinal vascular diseases and following treatment is
warranted.
(
Foveal Avascular Zone
References
1. From the Centers for Disease Control and Prevention. Blindness caused by diabeteseMassachusetts, 1987-1994. JAMA
1996;276:1865–6.
2. Diabetic Retinopathy. Available at: https://nei.nih.gov/eyedata/
diabetic. Accessed May 2, 2016.
3. Rogers S, McIntosh RL, Cheung N, et al. The prevalence of
retinal vein occlusion: pooled data from population studies
from the United States, Europe, Asia, and Australia.
Ophthalmology 2010;117:313–319 e1.
4. Boyle JP, Honeycutt AA, Narayan KM, et al. Projection of
diabetes burden through 2050: impact of changing demography and disease prevalence in the U.S. Diabetes Care
2001;24:1936–40.
5. Jonas JB, Schneider U, Naumann GO. Count and density of
human retinal photoreceptors. Graefes Arch Clin Exp Ophthalmol 1992;230:505–10.
6. Yu DY, Cringle SJ, Su EN. Intraretinal oxygen distribution in
the monkey retina and the response to systemic hyperoxia.
Invest Ophthalmol Vis Sci 2005;46:4728–33.
7. Yu PK, Balaratnasingam C, Cringle SJ, et al. Microstructure
and network organization of the microvasculature in the
human macula. Invest Ophthalmol Vis Sci 2010;51:6735–43.
8. Yu PK, Balaratnasingam C, Morgan WH, et al. The structural
relationship between the microvasculature, neurons, and glia
in the human retina. Invest Ophthalmol Vis Sci 2010;51:
447–58.
9. Mammo Z, Balaratnasingam C, Yu P, et al. Quantitative
noninvasive angiography of the fovea centralis using speckle
variance optical coherence tomography. Invest Ophthalmol
Vis Sci 2015;56:5074–86.
10. Kim DY, Fingler J, Zawadzki RJ, et al. Noninvasive imaging
of the foveal avascular zone with high-speed, phase-variance
optical coherence tomography. Invest Ophthalmol Vis Sci
2012;53:85–92.
11. Zheng Y, Gandhi JS, Stangos AN, et al. Automated segmentation of foveal avascular zone in fundus fluorescein angiography. Invest Ophthalmol Vis Sci 2010;51:3653–9.
12. Popovic Z, Knutsson P, Thaung J, et al. Noninvasive imaging
of human foveal capillary network using dual-conjugate
adaptive optics. Invest Ophthalmol Vis Sci 2011;52:2649–55.
13. Chui TY, VanNasdale DA, Elsner AE, Burns SA. The association between the foveal avascular zone and retinal thickness. Invest Ophthalmol Vis Sci 2014;55:6870–7.
14. Pinhas A, Razeen M, Dubow M, et al. Assessment of perfused
foveal microvascular density and identification of nonperfused
capillaries in healthy and vasculopathic eyes. Invest
Ophthalmol Vis Sci 2014;55:8056–66.
15. Schmoll T, Singh AS, Blatter C, et al. Imaging of the parafoveal capillary network and its integrity analysis using fractal
dimension. Biomed Opt Express 2011;2:1159–68.
16. Kuehlewein L, Tepelus TC, An L, et al. Noninvasive visualization and analysis of the human parafoveal capillary network
using swept source OCT optical microangiography. Invest
Ophthalmol Vis Sci 2015;56:3984–8.
17. Samara WA, Say EA, Khoo CT, et al. Correlation of foveal
avascular zone size with foveal morphology in normal eyes
using optical coherence tomography angiography. Retina
2015;35:2188–95.
18. Dubis AM, Hansen BR, Cooper RF, et al. Relationship
between the foveal avascular zone and foveal pit morphology.
Invest Ophthalmol Vis Sci 2012;53:1628–36.
2365
Ophthalmology Volume 123, Number 11, November 2016
19. Tam J, Martin JA, Roorda A. Noninvasive visualization
and analysis of parafoveal capillaries in humans. Invest
Ophthalmol Vis Sci 2010;51:1691–8.
20. Laatikainen L, Larinkari J. Capillary-free area of the fovea with
advancing age. Invest Ophthalmol Vis Sci 1977;16:1154–7.
21. Arend O, Wolf S, Harris A, Reim M. The relationship of
macular microcirculation to visual acuity in diabetic patients.
Arch Ophthalmol 1995;113:610–4.
22. Remky A, Wolf S, Knabben H, et al. Perifoveal capillary
network in patients with acute central retinal vein occlusion.
Ophthalmology 1997;104:33–7.
23. Maheshwary AS, Oster SF, Yuson RM, et al. The association
between percent disruption of the photoreceptor inner
segment-outer segment junction and visual acuity in diabetic
macular edema. Am J Ophthalmol 2010;150:63–67.e1.
24. Schwartz DM, Fingler J, Kim DY, et al. Phase-variance optical
coherence tomography: a technique for noninvasive angiography. Ophthalmology 2014;121:180–7.
25. Fingler J, Readhead C, Schwartz DM, Fraser SE. Phasecontrast OCT imaging of transverse flows in the mouse
retina and choroid. Invest Ophthalmol Vis Sci 2008;49:
5055–9.
26. Makita S, Jaillon F, Yamanari M, et al. Comprehensive in vivo
micro-vascular imaging of the human eye by dual-beam-scan
Doppler optical coherence angiography. Opt Express
2011;19:1271–83.
27. Jia Y, Bailey ST, Hwang TS, et al. Quantitative optical
coherence tomography angiography of vascular abnormalities
in the living human eye. Proc Natl Acad Sci U S A 2015;112:
E2395–402.
28. Xu J, Han S, Balaratnasingam C, et al. Retinal angiography
with real-time speckle variance optical coherence tomography.
Br J Ophthalmol 2015;99:1315–9.
29. An L, Wang RK. In vivo volumetric imaging of vascular
perfusion within human retina and choroids with optical micro-angiography. Opt Express 2008;16:11438–52.
30. Carpineto P, Mastropasqua R, Marchini G, et al. Reproducibility and repeatability of foveal avascular zone measurements
in healthy subjects by optical coherence tomography angiography. Br J Ophthalmol 2016;100:671–6.
31. Grading diabetic retinopathy from stereoscopic color fundus
photographsean extension of the modified Airlie House
classification. ETDRS report number 10. Early Treatment
Diabetic Retinopathy Study Research Group. Ophthalmology
1991;98(5 Suppl):786–806.
32. Trichonas G, Kaiser PK. Optical coherence tomography imaging
of macular oedema. Br J Ophthalmol 2014;98(Suppl 2):ii24–9.
33. Sim DA, Keane PA, Fung S, et al. Quantitative analysis of
diabetic macular ischemia using optical coherence tomography. Invest Ophthalmol Vis Sci 2014;55:417–23.
34. Radwan SH, Soliman AZ, Tokarev J, et al. Association of
disorganization of retinal inner layers with vision after resolution of center-involved diabetic macular edema. JAMA
Ophthalmol 2015;133:820–5.
35. Sun JK, Lin MM, Lammer J, et al. Disorganization of the
retinal inner layers as a predictor of visual acuity in eyes with
center-involved diabetic macular edema. JAMA Ophthalmol
2014;132:1309–16.
36. Prakash YS, Smithson KG, Sieck GC. Application of the
Cavalieri principle in volume estimation using laser confocal
microscopy. Neuroimage 1994;1:325–33.
37. Balaratnasingam C, Chae B, Remmer MH, et al. The spatial
profile of macular pigments is related to the topological
characteristics of the foveal avascular zone. Invest Ophthalmol
Vis Sci 2015;56:7859–65.
2366
38. Cohen J, Cohen P, West SG, Aiken LS. Applied Multiple
Regression/Correlation Analysis for the Behavioral Sciences.
3rd ed. Mahwah, New Jersey: Lawrence Erlbaum Associates;
2003.
39. Roach ES, Bettermann K, Biller J. Toole’s Cerebrovascular Disorders. 6th ed. New York: Cambridge University Press; 2010:33.
40. Yu DY, Cringle SJ, Yu PK, Su EN. Intraretinal oxygen distribution and consumption during retinal artery occlusion and
graded hyperoxic ventilation in the rat. Invest Ophthalmol Vis
Sci 2007;48:2290–6.
41. Parodi MB, Visintin F, Della Rupe P, Ravalico G. Foveal
avascular zone in macular branch retinal vein occlusion. Int
Ophthalmol 1995;19:25–8.
42. Simader C, Ritter M, Bolz M, et al. Morphologic parameters
relevant for visual outcome during anti-angiogenic therapy of
neovascular age-related macular degeneration. Ophthalmology
2014;121:1237–45.
43. Mariampillai A, Standish BA, Moriyama EH, et al. Speckle
variance detection of microvasculature using swept-source
optical coherence tomography. Opt Lett 2008;33:1530–2.
44. de Carlo TE, Chin AT, Bonini Filho MA, et al. Detection of
microvascular changes in eyes of patients with diabetes but not
clinical diabetic retinopathy using optical coherence tomography angiography. Retina 2015;35:2364–70.
45. Coscas F, Glacet-Bernard A, Miere A, et al. Optical coherence
tomography angiography in retinal vein occlusion: evaluation
of superficial and deep capillary plexa. Am J Ophthalmol
2016;161:160–71. e1e2.
46. Agemy SA, Scripsema NK, Shah CM, et al. Retinal vascular
perfusion density mapping using optical coherence tomography angiography in normals and diabetic retinopathy patients.
Retina 2015;35:2353–63.
47. Kashani AH, Lee SY, Moshfeghi A, et al. Optical coherence
tomography angiography of retinal venous occlusion. Retina
2015;35:2323–31.
48. Salz DA, de Carlo TE, Adhi M, et al. Select features of
diabetic retinopathy on swept-source optical coherence
tomographic angiography compared with fluorescein angiography and normal eyes. JAMA Ophthalmol 2016;134:
644–50.
49. Takase N, Nozaki M, Kato A, et al. Enlargement of foveal
avascular zone in diabetic eyes evaluated by en face optical
coherence tomography angiography. Retina 2015;35:2377–83.
50. Samara WA, Shahlaee A, Sridhar J, et al. Quantitative optical
coherence tomography angiography features and visual function in eyes with branch retinal vein occlusion. Am J Ophthalmol 2016;166:76–83.
51. Hwang TS, Gao SS, Liu L, et al. Automated quantification of
capillary nonperfusion using optical coherence tomography
angiography in diabetic retinopathy. JAMA Ophthalmol
2016;134:367–73.
52. Shahlaee A, Pefkianaki M, Hsu J, Ho AC. Measurement of
foveal avascular zone dimensions and its reliability in healthy
eyes using optical coherence tomography angiography. Am J
Ophthalmol 2016;165:201–2.
53. Bringmann A, Pannicke T, Grosche J, et al. Muller cells in the
healthy and diseased retina. Prog Retin Eye Res 2006;25:
397–424.
54. Distler C, Weigel H, Hoffmann KP. Glia cells of the monkey
retina. I. Astrocytes. J Comp Neurol 1993;333:134–47.
55. Gariano RF, Sage EH, Kaplan HJ, Hendrickson AE.
Development of astrocytes and their relation to blood vessels in fetal monkey retina. Invest Ophthalmol Vis Sci
1996;37:2367–75.
Balaratnasingam et al
56. Cai J, Boulton M. The pathogenesis of diabetic retinopathy:
old concepts and new questions. Eye 2002;16:242–60.
57. Gittings NS, Fozard JL. Age related changes in visual acuity.
Exp Gerontol 1986;21:423–33.
58. Mizutani M, Gerhardinger C, Lorenzi M. Muller cell changes
in human diabetic retinopathy. Diabetes 1998;47:445–9.
59. Ly A, Yee P, Vessey KA, et al. Early inner retinal astrocyte
dysfunction during diabetes and development of hypoxia,
retinal stress, and neuronal functional loss. Invest Ophthalmol
Vis Sci 2011;52:9316–26.
60. Yu DY, Yu PK, Cringle SJ, et al. Functional and morphological characteristics of the retinal and choroidal vasculature.
Prog Retin Eye Res 2014;40:53–93.
61. Yu DY, Cringle SJ. Oxygen distribution and consumption within
the retina in vascularised and avascular retinas and in animal
models of retinal disease. Prog Retin Eye Res 2001;20:175–208.
62. Springer AD, Hendrickson AE. Development of the primate
area of high acuity, 3: temporal relationships between pit
formation, retinal elongation and cone packing. Vis Neurosci
2005;22:171–85.
63. Provis JM, Sandercoe T, Hendrickson AE. Astrocytes and
blood vessels define the foveal rim during primate retinal
development. Invest Ophthalmol Vis Sci 2000;41:2827–36.
64. Dong LM, Marsh MJ, Hawkins BS. Measurement and analysis
of visual acuity in multicenter randomized clinical trials in the
United States: findings from a survey. Ophthalmic Epidemiol
2003;10:149–65.
65. Spaide RF, Fujimoto JG, Waheed NK. Image artifacts in optical coherence tomography angiography. Retina 2015;35:
2163–80.
(
Foveal Avascular Zone
66. Leese G. Longitudinal study examining the risk factors for
proliferative retinopathy and maculopathy in type-I diabetes:
The Royal College of Physicians of Edinburgh Diabetes
Register Group. Eye 2004;18:814–20.
67. Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence and
major risk factors of diabetic retinopathy. Diabetes Care
2012;35:556–64.
68. Moss SE, Klein R, Klein BE. Ten-year incidence of visual loss
in a diabetic population. Ophthalmology 1994;101:1061–70.
69. Banerjee D, Chung S, Wong EC, et al. Underdiagnosis of
hypertension using electronic health records. Am J Hypertens
2012;25:97–102.
70. Staff M, Roberts C, March L. The completeness of electronic
medical record data for patients with Type 2 Diabetes in primary care and its implications for computer modelling of
predicted clinical outcomes. Prim Care Diabetes 2016;10:
352–9.
71. Lachin JM, Genuth S, Nathan DM, et al. Effect of glycemic
exposure on the risk of microvascular complications in the
diabetes control and complications trial)revisited. Diabetes
2008;57:995–1001.
72. Hirsch IB, Brownlee M. Beyond hemoglobin A1c)need for
additional markers of risk for diabetic microvascular complications. JAMA 2010;303:2291–2.
73. Miri S, Shrier EM, Glazman S, et al. The avascular zone and
neuronal remodeling of the fovea in Parkinson disease. Ann
Clin Transl Neurol 2015;2:196–201.
74. Sim DA, Keane PA, Zarranz-Ventura J, et al. Predictive factors for the progression of diabetic macular ischemia. Am J
Ophthalmol 2013;156:684–92.
Footnotes and Financial Disclosures
Originally received: May 17, 2016.
Final revision: July 3, 2016.
Accepted: July 10, 2016.
Available online: August 11, 2016.
Author Contributions:
Conception and design: Balaratnasingam, Inoue, Ahn, Yannuzzi, Freund
Manuscript no. 2016-1056.
1
LuEsther T. Mertz Retinal Research Center, Manhattan Eye, Ear and
Throat Hospital, New York, New York.
2
Vitreous Retina Macula Consultants of New York, New York, New York.
3
Department of Ophthalmology, New York University School of Medicine, New York, New York.
4
Centre for Ophthalmology and Visual Science, Lions Eye Institute, University of Western Australia, Perth, Australia.
5
Department of Biostatistics, Columbia University Medical Center, New
York, New York.
Financial Disclosures:
The authors have made the following disclosures: L.Y.: Honorarium e
Genentech (San Franciscio, CA) for the retina fellow teaching program.
K.B.F.: Consultant e Genentech (San Francisco, CA), Optos (Dunfermline,
UK), Bayer HealthCare (New York), Optovue (Fremont, CA), DigiSight
(San Francisco, CA), and Heidelberg Engineering (Heidelberg, Germany)
(honorarium for each).
Financial Support: LuEsther T. Mertz Retinal Research Center, Manhattan
Eye, Ear and Throat Hospital, New York, New York, The Macula Foundation, Inc., New York, New York, and RANZCO Eye Foundation, New
South Wales, Australia. The funding organizations had no role in the design
or conduct of this research.
Analysis and interpretation: Balaratnasingam, Inoue, Ahn, McCann,
Dhrami-Gavazi, Freund
Data acquisition and/or research execution: Balaratnasingam, Inoue, Ahn,
McCann, Dhrami-Gavazi, Yannuzzi
Obtained funding: Not applicable
Overall responsibility: Balaratnasingam, Inoue, Ahn, Yannuzzi, Freund
Abbreviations and Acronyms:
BRVO ¼ branch retinal vein occlusion; CRVO ¼ central retinal vein
occlusion; CST ¼ central subfield thickness; DR ¼ diabetic retinopathy;
DRIL ¼ disorganization of the retinal inner layers; ETDRS ¼ Early
Treatment Diabetic Retinopathy Study; EZ ¼ ellipsoid zone;
FA ¼ fluorescein angiography; FAZ ¼ foveal avascular zone;
GEE ¼ generalized estimating equations; ILM ¼ inner limiting membrane;
LogMAR ¼ logarithm of the minimum angle of resolution; OCT ¼ optical
coherence tomography; OCTA ¼ optical coherence tomography angiography; PFT ¼ point foveal thickness; RVO ¼ retinal venous occlusion;
SD OCT ¼ spectral-domain optical coherence tomography; VA ¼ visual
acuity; VEGF ¼ vascular endothelial growth factor.
Correspondence:
Chandrakumar Balaratnasingam, MD, PhD, LuEsther T Mertz Retinal
Research Center and Vitreous, Retina, Macula Consultants of New York,
Manhattan Eye, Ear and Throat Hospital, New York, NY 10022. E-mail:
[email protected].
2367