Enhanced Depth Imaging Optical Coherence Tomography in Type 2 Diabetes

Abstract

Purpose.: To investigate the changes in macular choroidal thickness in eyes with various stages of diabetic retinopathy, using enhanced depth imaging optical coherence tomography (EDI OCT).

Methods.: Sixty-three consecutive diabetic patients—who presented without diabetic retinopathy (NDR); with diabetic retinopathy (nonproliferative diabetic retinopathy [NPDR]) and no clinically significant macular edema (CSME−); or with NDPR and clinically significant macular edema (CSME+)—underwent EDI OCT. Twenty-one age- and sex-matched healthy subjects (21 eyes) also underwent EDI OCT.

Results.: A total of 63 eyes of 63 consecutive diabetic patients (26 female [41.2%]; mean age 65 ± 9 years, range 48–83 years) were included in the analysis. Mean best-corrected visual acuity was 0.13 ± 0.25 LogMAR (range 0–1). Mean CMT was 272.5 ± 16.2 μm in 21 NDR eyes, 294.5 ± 23.5 μm in 21 NPDR/CSME− eyes, and 385.6 ± 75.1 μm in 21 NPDR/CSME+ eyes. There was no difference in mean subfoveal choroidal thickness among each diabetic group (238.4 ± 47.9 μm [NDR], 207.0 ± 55.9 μm [NPDR/CSME−], 190.8 ± 48.4 μm [NPDR/CSME+]; P = 0.23). The mean subfoveal choroidal thickness was significantly reduced in each diabetic group compared with the control group (309.8 ± 58.5 μm, P < 0.001).

Conclusions.: In diabetic eyes, there is an overall thinning of the choroid on EDI OCT. A decreased choroidal thickness may lead to tissue hypoxia and consequently increase the level of VEGF, resulting in the breakdown of the blood-retinal barrier and development of macular edema.

Introduction

Diabetic retinopathy is due to the breakdown of retinal vasculature integrity ¹ and hemodynamic abnormalities. ² Studies of Doppler flowmetry indicated that the choroidal blood flow may decrease in the early stage of diabetic retinopathy, and decrease further in the presence of macular edema. ³ Similarly, histologic studies suggested that diabetic retinopathy may cause atrophy of the choriocapillaris endothelium. ⁴ The choroidal vasculature, especially the choriocapillaris, provides oxygen and nutrients to the outer retina and is responsible for maintaining the highly metabolically active photoreceptor cells. ⁵ Since there is no retinal vasculature in the foveal region, impairment of the choriocapillaris may cause severe functional damage to the retinal tissue in the fovea. Therefore, in vivo evaluation of the structural changes in the choroid might be very insightful to determine the pathogenesis of progression of the macular changes in diabetic eyes. 

Spectral-domain optical coherence tomography (SD-OCT) technology improves resolution, compared with time-domain OCT. SD-OCT scans show bands that seem to correspond to the anatomic layers of the human retina, and allow seeing even small retinal details of photoreceptor layer, such as the inner segment/outer segment (IS/OS) junction. Recently, a new approach to improve depth imaging by OCT, termed enhanced depth imaging (EDI) OCT, has been shown to be able to reliably image the full-thickness of the choroid. ⁶ EDI OCT uses SD-OCT equipment (Spectralis SD-OCT; Heidelberg Engineering, Heidelberg, Germany) positioned closer to the eye than ordinary, such that a stable inverted image is produced. The net effect of this practice is that the sensitivity of the imaging in deeper layers of tissue is increased. In this fashion, EDI OCT may represent a useful approach to investigate, in vivo, the choroidal changes in eyes with diabetic retinopathy. 

In this study, using EDI OCT, we investigated the choroidal thickness within the macula of eyes with various stages of diabetic retinopathy and compared the measured values with those of healthy normal subjects. 

Methods

Sixty-three consecutive patients were recruited from the diabetic retinopathy outpatient clinic at the University Eye Clinic of San Raffaele Hospital between September 2011 and November 2011. All patients underwent a complete ophthalmologic examination, including EDI OCT, as part of their routine clinical work-up. Inclusion criteria were age >18 years, diagnosis of diabetes type 2, absence of diabetic retinopathy, corresponding to level 10 on the Early Treatment Diabetic Retinopathy Study (ETDRS) scale, or treatment-naïve nonproliferative diabetic retinopathy corresponding to level 20 to 53 on the ETDRS scale, ⁷ as evaluated by experienced ophthalmologists with fundus biomicroscopy. Exclusion criteria were systemic diseases other than diabetes, refractive error of more than −5 diopters (D) in the study eye, and axial length >25 mm (if available); a history of retinal photocoagulation or any other treatment for diabetic retinopathy; previous ocular surgery; proliferative diabetic retinopathy; signs of any other active retinal disease in the study eye such as AMD; choroidal neovascularization (CNV); or vitreoretinal diseases (i.e., vitreomacular traction syndrome and epiretinal membrane). This study was performed in accordance with the ethical standards set forth in the Declaration of Helsinki. Written informed consent was obtained from each patient enrolled and San Raffaele Hospital of Ophthalmology Ethics Committee approval was obtained for this study. 

The diabetic eyes were graded according to the ETDRS scale, and grouped for pathology: without diabetic retinopathy (NDR); with nonproliferative diabetic retinopathy and no clinically significant macular edema (NPDR/CSME−); and with nonproliferative diabetic retinopathy and clinically significant macular edema (NPDR/CSME+). ^7–9  

Age- and sex-matched healthy subjects (control group) without risk factors for diabetes were also included in the study (only one eye per healthy subject). 

Study Protocol

In all subjects, refractive error, duration of diabetes, and HbA1c was recorded on clinical charts. Best-corrected visual acuity (BCVA) was determined with ETDRS charts. Retinal status was evaluated by fundus biomicroscopy after pupil dilation by two experienced retinal physicians (GQ, RL) and recorded on clinical charts. Automated central macular thickness (CMT) measurements were generated by an SD-OCT device (Heidelberg Engineering) using a 19-horizontal line protocol (6 × 6 mm area), each consisting of 1024 A-scans per line (Spectralis Acquisition and Viewing Modules, version 5.3.2; Heidelberg Engineering). All subjects underwent choroidal imaging and thickness measurements by EDI OCT. EDI OCT scans were performed by two retinal physicians (LQ, DC), experienced at performing scans using the SD-OCT device (Heidelberg Engineering). 

Edi Oct

The method of obtaining EDI OCT images has been reported previously. ⁶ The choroid was imaged by positioning the SD-OCT device (Heidelberg Engineering) close enough to the eye to obtain an inverted image. Two 9-mm high-quality line scans through the fovea (one horizontal and one vertical) were obtained for each eye. The line scans were saved for analysis after 100 frames were averaged, using the automatic averaging and eye tracking features of the proprietary device. The resultant images were viewed and measured with the software included in the device (Heidelberg Eye Explorer version 1.7.0.0; Heidelberg Engineering). The choroid was measured by two experienced retinal physicians (GQ, RL), which were blinded to the diagnosis. The choroidal thickness was measured from the outer portion of the hyperreflective line corresponding to the RPE to the hyporeflective line or margin corresponding to the sclerochoroidal interface. These measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea (nine ETDRS sectors). The values of the measurements were compared for each observer and then averaged for analysis. 

Statistical Analysis

Statistical calculations were performed using statistical software (Statistical Package for Social Sciences version 17.0; SPSS, Inc., Inc, Chicago, IL). The difference between healthy subjects (control group) and each diabetic group was generated by conducting repeated measures ANOVA followed by Dunnet's post-hoc test for each individual axial choroidal thickness over the horizontal and vertical line scans (at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea). The concordance correlation coefficient (Pearson correlation) was calculated for interobserver correlations and considered to be strong if the correlation coefficient was >0.8. ¹⁰ Univariate linear regression analyses were performed to evaluate the relationships between choroidal thickness, age, CMT, diabetes duration, and HbA1c in each group (control group and each diabetic group). ¹⁰ Multiple linear regression was used to evaluate the explanatory variables with regard to the dependent variable, BCVA. 

Correlation between BCVA, CMT, and choroidal thickness at the fovea in each diabetic group was analyzed using the Pearson's correlation coefficient. The chosen level of statistical significance was P < 0.05. 

Results

Patient Demographics and Clinical Characteristics

A total of 63 eyes of 63 consecutive diabetic patients (26 female [41.2%], mean age 65 ± 9 years, range 48–83 years) were included in the analysis (Table 1). Mean BCVA was 0.13 ± 0.25 LogMAR (range 0–1; Table 1). Mean CMT was 272.5 ± 16.2 μm in 21 NDR eyes, 294.5 ± 23.5 μm in 21 NPDR/CSME− eyes, and 385.6 ± 75.1 μm in 21 NPDR/CSME+ eyes (Table 1). Eight out of 21 eyes with NPDR/CSME+ showed at least one sector within the macula (90′ superior, 90′ temporal, 90′ inferior, or 90′ nasal to the fovea) without retinal thickening. 

Table 1. 

 

Age, BCVA, CMT, and Foveal Coroidal Thickness in the Three Groups of Diabetic Patients and Healthy Controls

	NDR	NPDR/CSME−	NPDR/CSME+	Controls
Age	65 ± 9 (y)	61.3 ± 10 (y)	67 ± 10 (y)	65 ± 10 (y)
BCVA	0.16 ± 0.35 (LogMAR)	0.05 ± 0.09 (LogMAR)	0.27 ± 0.27 (LogMAR)	0.05 ± 0.11 (LogMAR)
CMT	272.5 ± 16.2 μm	294.5 ± 23.5 μm	385.6 ± 75.1 μm	297.3 ± 53.4 μm
Subfoveal choroidal thickness	238.4 ± 47.9 μm	207.0 ± 55.9 μm	190.8 ± 48.4 μm	309.8 ± 58.5 μm

 CMT, central macular thickness.

Twenty-one eyes of 21 consecutive healthy subjects (10 female [47.6%]; mean age 65 ± 10 years, range 53–83 years) were also included in the analysis (control group; Table 1). Mean BCVA was 0.05 ± 0.11 LogMAR (range 0–0.4) (Table 1). Mean CMT was 297.3 ± 53.4 (Table 1). 

The mean refractive error was −0.29 ± 1.2 D for diabetic patients, and +0.21 ± 1.5 for healthy subjects (control group; P = 0.1). 

EDI OCT

There was no difference in mean subfoveal choroidal thickness among each diabetic group (238.4 ± 47.9 μm [NDR]; 207.0 ± 55.9 μm [NPDR/CSME−]; 190.8 ± 48.4 μm [NPDR/CSME+]; P = 0.23; Table 2). The mean subfoveal choroidal thickness was significantly reduced in each diabetic group compared with the control group (309.8 ± 58.5 μm, P < 0.001; Table 2, Figs. 1–7). Similarly, the mean choroidal thickness was significantly reduced in each diabetic group at each ETDRS sector (individual axial choroidal thickness over the horizontal and vertical line scans at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea) compared with the control group (Table 3). 

Figure 1. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 61-year-old healthy subject (control group) without risk factors for diabetes. The choroid was measured from the outer portion of the hyperreflective line corresponding to the RPE to the inner surface of the sclera. These measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Table 2. 

 

Choroidal Thickness in Each Extrafoveal Evaluated Point in the Three Groups of Diabetic Patients and Healthy Controls

	NDR	P	NPDR−	P	NPDR+	P	Controls
Nasal (1.5 mm)	205.0 ± 52.9 μm	0.01	169.0 ± 50.8 μm	<0.001	159.5 ± 46.4 μm	<0.001	247.3 ± 63.2 μm
Nasal (3 mm)	113.8 ± 44.7 μm	0.005	106.1 ± 51.8 μm	0.001	91.7 ± 37.4 μm	<0.001	183.7 ± 101.7 μm
Temporal (1.5 mm)	196.3 ± 74.3 μm	<0.001	196.7 ± 58.8 μm	<0.001	167.8 ± 41.3 μm	<0.001	282.6 ± 61.4 μm
Temporal (3 mm)	178.3 ± 58.2 μm	<0.001	179.5 ± 51.4 μm	<0.001	165.1 ± 47.7 μm	<0.001	282.8 ± 55.7 μm
Inferior (1.5 mm)	231.6 ± 62.7 μm	0.01	166.2 ± 41.5 μm	<0.001	173.5 ± 36.4 μm	<0.001	284.6 ± 78.9 μm
Inferior (3 mm)	189.4 ± 56.8 μm	<0.001	159.9 ± 51.4 μm	<0.001	153.8 ± 38.4 μm	<0.001	263.3 ± 74.1 μm
Superior (1.5 mm)	250.2 ± 60.3 μm	0.01	204.5 ± 62.7 μm	<0.001	186.8 ± 51.7 μm	<0.001	298.8 ± 69.5 μm
Superior (3 mm)	239.9 ± 64.0 μm	0.01	197.0 ± 69.0 μm	<0.001	176.2 ± 43.7 μm	<0.001	281.6 ± 67.2 μm

P  , statistical difference with control group.

Table 3. 

 

Coefficient of Variation in Each Evaluated Point in the Three Groups of Diabetic Patients and Healthy Controls

	NDR	NPDR−	NPDR+	Controls
Fovea	0.2011	0.2703	0.2537	0.1889
Nasal (1.5 mm)	0.2579	0.3004	0.2907	0.2554
Nasal (3 mm)	0.3929	0.4885	0.4081	0.5533
Temporal (1.5 mm)	0.3784	0.2991	0.2464	0.2173
Temporal (3 mm)	0.3266	0.2862	0.2887	0.1967
Inferior (1.5 mm)	0.2707	0.2497	0.2096	0.2771
Inferior (3 mm)	0.2996	0.3214	0.2493	0.2816
Superior (1.5 mm)	0.2409	0.3065	0.2767	0.2325
Superior (3 mm)	0.2667	0.3503	0.2479	0.2388

In all 8 eyes with NPDR/CSME+ and at least in one sector within the macula without retinal thickening, the choroid appeared thinned in all sectors with corresponding retinal thickening compared with the sector(s) without corresponding retinal thickening (Fig. 7). 

In the NDR group, no correlation was found between BCVA and CMT (Pearson's rho = 0.16, P = 0.47); BCVA and choroidal thickness at the fovea (Pearson's rho = −0.25, P = 0.26); and between CMT and choroidal thickness in the fovea (Pearson's rho = 0.06, P = 0.8). In the NPDR/CSME− group, no correlation was found between BCVA and CMT (Pearson's rho = −0.07, P = 0.7); BCVA and choroidal thickness at the fovea (Pearson's rho = 0.4, P = 0.08); and between CMT and choroidal thickness in the fovea (Pearson's rho = 0.35, P = 0.11). In the NPDR/CSME+ group, no correlation was found between BCVA and CMT (Pearson's rho = 0.26, P = 0.24); BCVA and choroidal thickness at the fovea (Pearson's rho = 0.38, P = 0.08); and between CMT and choroidal thickness in the fovea (Pearson's rho = 0.17, P = 0.45). 

Coefficients of variation for each individual axial choroidal thickness are shown in Table 3. Coefficients of variation in the control group were lower in almost any considered sector, indicating that choroidal changes occur asymmetrically in the macula of diabetic eyes. 

Computed tomography measurements were significantly correlated between the two observers at each location (Pearson's coefficient r = 0.81, P = 0.01). 

A simple regression analysis of macular choroidal thickness is shown in Table 4. Age, BCVA, and CMT were not associated with foveal choroidal thickness in any diabetic group. Age and BCVA were associated with foveal choroidal thickness in the control group (P = 0.002 and P = 0.01, respectively). 

Table 4. 

 

Simple Linear Regression Analysis of Various Factors for the Subfoveal and Foveal Choroidal Thickness in the Three Groups of Diabetic Patients and Healthy Controls

	Unstandardized Coefficients Estimate (Standard Error)	P Value	Correlation Coefficients
NDR
Age	1.39 (1.25)	0.28	0.06
BCVA	−34.58 (30.04)	0.26	0.06
CMT	0.18 (0.67)	0.78	0.004
Duration of diabetes	0.06 (0.20)	0.78	0.02
HbA1C	12.28 (15.32)	0.09	0.18
NPDR−
Age	−1.10 (1.25)	0.38	0.03
BCVA	222.01 (120.91)	0.08	0.15
CMT	0.83 (0.51)	0.11	0.12
Duration of diabetes	0.07 (0.14)	0.62	0.01
HbA1C	25.76 (13.94)	0.08	0.20
NPDR+
Age	−0.86 (1.12)	0.45	0.02
BCVA	68.09 (37.99)	0.08	0.14
CMT	0.11 (0.14)	0.45	0.02
Duration of diabetes	0.07 (0.12)	0.52	0.02
HbA1C	−4.31 (10.46)	0.69	0.01
Controls
Age	−3.42 (0.98)	0.002	0.39
BCVA	−275.85 (96.06)	0.01	0.30
CMT	0.16 (0.25)	0.50	0.02

We performed a stepwise multiple linear regression analysis to determine the explanatory variables most strongly associated with BCVA from among age, CMT, foveal choroidal thickness, duration of diabetes, and HbA1c (except for the control group). In the group NPDR/CSME−, foveal choroidal thickness was the variable associated most significantly with BCVA (P = 0.03); in the group NPDR/CSME+, duration of diabetes (P = 0.03) was a significant predictor of BCVA; in the control group, age (P = 0.05) was the variable most significantly associated with BCVA (Table 5). 

Table 5. 

 

Multiple Linear Regression to Evaluate the Explanatory Variables with Regard to the Dependent Variable BCVA in the Three Groups of Diabetic Patients and Healthy Controls

	Age		Duration of Diabetes		HbA1c		CMT		Choroidal Thickness
	R 2	P	R 2	P	R 2	P	R 2	P	R 2	P
Controls	0.53	0.05	-	-	-	-	0.07	0.2	0.30	0.5
NDR	0.08	0.7	0.48	0.1	0.10	0.4	0.02	0.4	0.06	0.2
NPDR/CSME−	0.02	0.5	0.02	0.5	0.12	0.2	0.006	0.7	0.15	0.03
NPDR/CSME+	0.009	0.6	0.23	0.03	0.15	0.2	0.07	0.2	0.14	0.08

Discussion

In this study, using EDI OCT, we investigated the changes in choroidal thickness within the macula of eyes with various stages of diabetic retinopathy, by comparing the measured values with those of healthy normal subjects. Overall, we found that there was a reduction of choroidal thickness in each diabetic group (NDR, NPDR/CSME−, and NPDR/CSME+) compared with the control group (age- and sex-matched healthy subjects). According to the lower coefficients of variation in the control group, choroidal changes seem to occur asymmetrically in the macula of diabetic eyes. Previous histologic studies revealed that, in eyes with diabetic retinopathy, there is atrophy ⁴ and dropout of the choriocapillaris. ^11,12 Thus, if choroidal thinning is indicative of a loss of choriocapillaris, our in vivo findings may be in agreement with the histologic studies. ^4,12,13 Moreover, our findings are consistent with circulatory studies showing a decreased pulsatile ocular blood flow in diabetic patients. ^13,14 In fact, the dropout of the choriocapillaris could increase vascular resistance, resulting in decreased blood flow in the choriocapillaris. Previous studies demonstrated that a decreased choroidal blood flow may occur before the clinical manifestations of diabetic retinopathy. ³ Similarly, in the current series, the choroidal thickness value obtained in the NDR group was lower than that in the age-matched control group. The choroidal blood flow value in the foveal region in NPDR/CSME+ eyes has been reported to be lower than that in NPDR/CSME− eyes. ³ Accordingly, in our series, the choroidal thickness value at the fovea decreased further in the presence of macular edema (NPDR/CSME+ versus NPDR/CSME− and NDR). These data give insights to the mechanism of diabetic macular edema. The decreased choroidal thickness at the fovea, probably due to the dropout of the choriocapillaris (and determining increased vascular resistance), may cause retinal hypoxia. In fact, it is the role of the choroidal vasculature, especially the choriocapillaris, to provide nutrients to the RPE and outer retinal layers in the foveal region. ⁵ Due to tissue hypoxia, VEGF expression increases in RPE, pericytes, and microvascular endothelial cells, ¹⁵ and may induce breakdown of the blood-retinal barrier, which is on the basis of diabetic macular edema, in patients with diabetes. ^16–18 In turn, a decreased choroidal thickness at the fovea may cause tissue hypoxia and consequently increase the level of VEGF, resulting in the development of macular edema as a result of breakdown of the blood-retinal barrier. 

Recently Esmaeelpour et al. ¹⁹ mapped the choroidal thickness in patients with diabetes using high speed three-dimensional (3D) OCT imaging at 1060 nm. Overall, they found a central and inferior thinning in diabetic eyes compared with healthy eyes. In our series, we used a standard 870 nm SD-OCT device (Heidelberg Engineering). By positioning the device closer to the eye than ordinary, such that a stable inverted image is produced, and the sensitivity of the imaging in deeper layers of tissue is increased (EDI OCT), we found choroidal thickness values overall similar to those reported by Esmaeelpour et al. obtained using 3D 1060-nm OCT imaging. In the series of Esmaeelpour et al., ¹⁹ the generation of choroidal thickness maps based on manual segmentation revealed a reduced subfoveal thickness in all diabetic groups (214 ± 55 μm for NDR, from 208 ± 49 μm to 205 ± 54 μm for NPDR/CSME−, and 211 ± 76 μm for NPDR/CSME+). Similarly, in the current series, the manual measurements of choroidal thickness (future use of the method would benefit from an automated algorithm for making the measurements), which was performed on the horizontal and vertical axis (in the subfoveal area, and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea), revealed a reduced subfoveal thickness in all diabetic groups (238.4 ± 47.9 μm for NDR, 207.0 ± 55.9 μm for NPDR/CSME−, and 190.8 ± 48.4 μm for NPDR/CSME+). 

Interestingly, differently from Esmaeelpour et al., ¹⁹ we found a further reduction of subfoveal choroidal thickness in the presence of CSME. Moreover, while in the series of Esmaeelpour et al. ¹⁹ there was no apparent decrease directly below the lesions, we found a thinning of the choroid in the sector(s) with corresponding retinal thickening compared with the sector(s) without corresponding retinal thickening. 

Diabetic macular edema has been classically attributed to inner retinal capillary changes. ^20,21 However, some eyes with marked macular capillary changes do not develop diabetic macular edema. This suggests that retinal capillary abnormalities are not the only contributor to diabetic macular edema. Our current findings corroborate the idea that coexisting changes in the RPE, choroid, vitreous, or systemic circulation may be necessary for diabetic macular to occur. 

Our study has several limitations. The series presented here is relatively small. However, one should look at the current series in consideration of the strict inclusion criteria for both diabetic and control groups, as well as the similarity of groups with respect to the meaningful characteristics such as age and refraction (which are known to affect the choroid measurements). Moreover, we had no examples of histopathology to correlate to the changes noted on EDI OCT. Finally, our study cannot answer the question of cause-effect relationship. In other words, we cannot definitively conclude whether the decreased choroidal thickness is primarily involved in the development of diabetic retinopathy and macular edema or results from the loss of cellular components that are secondary to diabetic retinopathy and macular edema. 

In conclusion, we showed that in diabetic eyes, there is an overall thinning of the choroid on EDI OCT. A further reduction of choroidal thickness may be detected in the presence of and colocalizes with diabetic macular edema. These data favor the idea that, in diabetic eyes, decreased choroidal thickness may lead to tissue hypoxia and consequently increase the level of VEGF, resulting in the breakdown of the blood-retinal barrier and the development of macular edema. As proof of concept, future studies will investigate whether anti-VEGF treatment would be associated with changes in choroidal thickness. 

Figure 2. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 51-year-old diabetic patient without diabetic retinopathy (NDR group). The choroid was measured from the outer portion of the hyperreflective line corresponding to the RPE to the inner surface of the sclera. These measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Figure 3. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 56-year-old diabetic patient without diabetic retinopathy (NDR group). The measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Figure 4. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 57-year-old diabetic patient with diabetic retinopathy without macular edema (NPDR/CSME− group). The choroid was measured from the outer portion of the hyperreflective line corresponding to the RPE to the inner surface of the sclera. These measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Figure 5. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 56-year-old diabetic patient with diabetic retinopathy without macular edema (NPDR/CSME− group). The measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Figure 6. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 62-year-old diabetic patient with diabetic retinopathy and macular edema (NPDR/CSME+ group). The choroid was measured from the outer portion of the hyperreflective line corresponding to the RPE to the inner surface of the sclera. These measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea.

Figure 7. 

 

Two representative 9-mm high-quality line scans (100 averaged frames) through the fovea, one horizontal (upper panel) and one vertical (bottom panel), from a 64-year-old diabetic patient with diabetic retinopathy and macular edema (NPDR/CSME+ group). The measurements were made of the subfoveal choroid and at 1500-μm intervals from the fovea to 3 mm nasal, 3 mm temporal, 3 mm superior, and 3 mm inferior from the center of the fovea. Of note, the choroid appears thinned in the sectors with corresponding retinal thickening (temporal and inferior) compared with the sectors without corresponding retinal thickening.

Acknowledgments

We thank Giorgio Alto, MS, and Alessio Buzzotta, MS, for their help in collecting data. 

Footnotes

 The authors have no relevant financial interest in the products or companies described in this article.

Footnotes

 Disclosure: G. Querques, Alimera Sciences (C, S), Bayer Schering Pharma (C), Bausch & Lomb (C); R. Lattanzio, None; L. Querques, None; C. Del Turco, None; R. Forte, None; L. Pierro, None; E.H. Souied, Bausch & Lomb-Chauvin (C), Novartis (C), Alcon, (C); F. Bandello, Alcon (C), Alimera Sciences (C), Bayer (C), Allergan (C), Bausch & Lomb-Chauvin (C), Genetec (C), Novartis (C), Pfizer (C), Thea (C)