REVIEW ARTICLE

Spectral Domain Optical Coherence Tomography Features

and Classification Systems for Diabetic Macular Edema:
A Review
Surabhi Ruia, MBBS,* Sandeep Saxena, MS, FRCSEd, FRCS, FRCOphth, FAICO,*
Chui Ming Gemmy Cheung, MD, FRCOphth,† Jagjit S. Gilhotra, MMed, FRANZCO,‡
and Timothy Y.Y. Lai, MD, MMedSc, FRCS, FRCOphth, FHKAM§
that were used to define clinically significant macular edema
Abstract: Spectral domain optical coherence tomography (SD-OCT) is (CSME) by the Early Treatment Diabetic Retinopathy Study
fast becoming the current standard of care for the detection and assessment (ETDRS) Research Group in 1991. Spectral domain OCT (SD-
of diabetic macular edema. With the application of SD-OCT for imaging of OCT) provides unparalleled high-resolution imaging with precise
retinal microstructure and measurement of retinal thickness, new informa- measurement of retinal thickness with high reproducibility and re-
tion regarding disease characteristics has been gathered, which was unrec- liability. Both qualitative analysis based on the morphology and
ognized previously. Retinal thickness measurements on SD-OCT have also reflectivity profile of the retina along with quantitative analysis
been used for deciding the management and monitoring of the disease. provide objective information for the detection of DME, allowing
Since its development, OCT has enhanced the understanding of retinal accurate follow-up for the progression of disease and evaluation of
anatomical changes in diabetic retinopathy. Several authors have used response to treatment.
SD-OCT to classify diabetic macular edema with the purpose of correlat-
ing the pathophysiology with disease severity. The classification systems
have helped monitor the treatment efficacy and provide prognostic infor- CLASSIFICATION SYSTEMS OF DME
mation on the treatment outcome. The following review article summarizes
these classifications.
Diabetic macular edema has been classified on the basis of
various parameters, including the presence of retinal thickening,
Key Words: diabetic retinopathy, diabetic macular edema, external hard exudates, location with relation to the fovea, morphological
limiting membrane, optical coherence tomography, photoreceptor inner patterns of fluid accumulation, vitreomacular interface changes,
segment ellipsoid zone quantitative analysis of fluid accumulation, and microstructural
(Asia Pac J Ophthalmol 2016;5: 360–367) alteration in the retina. Various classifications have attempted to
establish baseline variance in visual acuity, prognosticate the dis-
ease, rationalize treatment modality, and predict treatment out-
T he prevalence of diabetes mellitus is attaining epidemic pro-
portions worldwide, with the number of people with diabetes
mellitus expected to rise to 592 million by 2035.1 Diabetic reti-
comes. The published definitions of DME have been based on 4
examination methods including fundus biomicroscopy, color fun-
dus photography, FA, and OCT.
nopathy (DR) is associated with structural changes in the retina,
On stereoscopic biomicroscopy, the ETDRS defined CSME
with diabetic macular edema (DME) being the most common ma-
as center and noncenter involving on the basis of the presence of
jor cause of vision loss. Diabetic macular edema is characterized
retinal thickening and hard exudates.6 Diabetic macular edema
by a thickening of the macular region caused by breakdown in
with or without visual impairment should be considered for treat-
the inner and outer blood–retinal barrier (BRB) through loss of
ment when it fulfils the ETDRS criteria for CSME. Laser photoco-
pericytes, microaneurysms, and dilated capillaries. Vascular endo-
agulation for DME mostly stabilizes the patient’s vision.7 Clinical
thelial growth factor (VEGF) is considered the main factor that
trials of intravitreal anti-VEGF pharmacotherapies for the treat-
disrupts BRB function. There are approximately 93 million peo-
ment of visual impairment due to center-involved DME observed
ple with DR and 21 million with DME in the world.2 According
an improvement or restoration of visual acuity in DME.8 A panel
to studies on the natural history of DME, 24% of eyes with DME
of experts recommended that current evidence for anti-VEGF treat-
experienced visual loss of at least 3 lines of vision within 3 years.3
ment for DME should be based on involvement of the center of the
Optical coherence tomography (OCT) is a noninvasive, non-
macula.8 Laser photocoagulation based on ETDRS guidelines re-
contact, transpupillary imaging modality that has become the cur-
mains appropriate for the treatment of DME without center involve-
rent standard of care for the detection of DME. It has superseded the
ment, or for DME with center involvement without vision loss.9
traditional and subjective tools like slit-lamp biomicroscopy, fluo-
The ETDRS Report Number 5 previously graded the propor-
rescein angiography (FA),4,5 and stereo color fundus photography
tion of leakage on FA from microaneurysms for classification of
edema as focal or diffuse.10,11 With the availability of OCT, clas-
From the *Department of Ophthalmology, King George’s Medical University,
sification of DME based on FA is much less commonly used by
Lucknow, India; †Medical Retina Service, Singapore National Eye Centre, clinicians. Clinical trials have used these classifications for com-
Singapore; ‡Department of Ophthalmology, University of Adelaide, Adelaide, paring the treatment modalities of DR. The RESTORE and Dia-
South Australia, Australia; and §Department of Ophthalmology and Visual betic Retinopathy Clinical Research Network (DRCR.net) studies
Sciences, The Chinese University of Hong Kong, Shatin, Hong Kong.
Received for publication December 29, 2015; accepted May 11, 2016.
relating to anti-VEGF monotherapy for DME observed no differ-
The authors have no funding or conflicts of interest to declare. ence in visual outcome in the subgroup of focal and diffuse DME.
Reprints: Sandeep Saxena, MS, FRCSEd, FRCS, FRCOphth, FAICO, Department The DRCR.net does not follow strict FA criteria for classification
of Ophthalmology, King George’s Medical University, Lucknow 226003, into focal or diffuse. Furthermore, the terms focal and diffuse DME
India. E-mail: [email protected].
Copyright © 2016 by Asia Pacific Academy of Ophthalmology
are frequently used without clear definitions, and significant
ISSN: 2162-0989 variations exist between different studies involving different
DOI: 10.1097/APO.0000000000000218 examination methods.12

360 www.apjo.org Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016 SD-OCT Features and Classification Systems for DME

TABLE 1. OCT-Based Morphological Classification of DME by Otani et al15

Type 1 (sponge-like swelling) Thickening of the fovea with homogeneous optical reflectivity
Type 2 (cystoid macular edema) Thickening of the fovea with markedly decreased optical reflectivity in outer layers
Type 3 (serous retinal detachment) Thickening of the fovea with subfoveal fluid accumulation and distinct outer border
of detached neurosensory retina
Type 3A Without vitreofoveal traction
Type 3B With apparent vitreofoveal traction

Several studies have correlated retinal sensitivity with visual 3 patterns of fluid accumulation as follows: sponge-like retinal
acuity and retinal thickness at the fovea in patients with DME.13,14 swelling, cystoid macular edema, and serous retinal detachment.
Fundus-monitored microperimetry determines the retinal sensitiv- Initially, the fluid accumulates in the outer retinal layers, with fur-
ity, and retinal thickness is obtained by OCT. The retinal sensitiv- ther cystoid cavity formation correlating with histopathological
ity within the central 2 and 10 degrees of the macula was observed finding of Muller cell necrosis. Persistent edema engages the inner
to be significantly lower in eyes with DME in comparison with retinal layers, leading to atrophy and appearance similar to retinos-
normal eyes. A negative correlation was found between retinal chisis. Serous retinal detachment was interpreted on OCT as de-
sensitivity and foveal retinal thickness. Retinal sensitivity supple- tachment of the neurosensory retina with the presence of
ments the predictive value of OCT in DME.14 underlying fluid (nonreflecting space) above the hyperreflective
line of the retinal pigment epithelium (RPE). Otani et al15 further
observed a significant correlation between visual acuity and reti-
OCT-BASED MORPHOLOGICAL CLASSIFICATION nal thickness regardless of the different tomographic features. In
SYSTEMS FOR DME 2004, Panozzo et al16 classified the morphological pattern in DME,
Otani et al15 proposed the first time domain OCT (TD-OCT) taking into account the size of the cysts (Table 2).
classification for DME based on retinal morphological changes In 2004, Kang et al correlated OCT and FA findings (focal
in 1999 (Table 1). Accumulation of intraretinal fluid leads to in- leakage, diffuse, diffuse cystoid leakage) in DME and inferred a
creased retinal thickness and reduced reflectivity. They described significant correlation between them. Kang et al classified DME

TABLE 2. Classification of DME by Panozzo et al16 Based on Retinal Thickness, Volume, Morphology of Retina, and Macular Traction

Retinal Thickness
Fixation Point Central Zone Perifoveal and Peripheral Areas Volume
Normal 150 + 20 μm 170 + 20 μm 230 + 20 μm 6.5 mm3 ± 1
Borderline 170–210 μm 190–230 μm 250–290 μm Up to 8.0 mm3
Edema >210 μm >230 μm >290 μm ≥8.0 mm3

Morphology
E1 E2 E3
Simple: compact retinal thickening, Cystoid thickening: retinal thickening Neuroepithelial detachment:
no clinically visible cystoid with cysts isolated or associated with
spaces E2A simple or cystoid retinal
Mild: 2–4 central small cysts each with thickening
horizontal diameter 150–200 μm,
vertical diameter 400 μm
E2B
Intermediate: cysts with petaloid
configuration or central cysts with
horizontal diameter less than
300 μm, vertical diameter less
than 600 μm
E2C
Severe: coalescence of cysts with
appearance similar to retinoschisis

Epiretinal (Macular or Foveal) Traction (Tangential or Anteroposterior)

T0 T1 T2 T3
Absence of epiretinal Presence of a continuous line of Presence of continuous line Anteroposterior traction
hyperreflectivity flat hyperreflectivity adherent of hyperreflectivity with with “gull wings”
to the retina without significant multiple points of adhesion configuration
retinal distortion to the retina with distortion

© 2016 Asia Pacific Academy of Ophthalmology www.apjo.org 361

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Ruia et al Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016

TABLE 3A. OCT-Based Classification of DME by Koleva-Georgieva TABLE 4. SD-OCT–Based Classification of DME by Helmy et al19
and Sivkova18
CME I Cysts less than 30% of macular thickness
Mild Small cysts predominantly in the outer retinal layers CME II Cysts between 30% and 60% of macular thickness
Moderate Cysts mainly located in the outer layers CME III Cysts between 60% and 90% of macular thickness
Severe Cysts mainly located in the outer layers predominantly CME IV Cysts more than 90% of the macular thickness
in the fovea A Cysts without any disruption to the ELM or EZ
B Cysts with ELM disruption
into 4 morphological types. Type 1 was described as thickening of C Cysts with EZ disruption
the fovea with homogeneous optical reflectivity throughout the
D Cysts with disruption of both the ELM and EZ
whole layer of the retina, whereas type 2 referred to thickening
of the fovea with markedly decreased optical reflectivity in the CME indicates cystoid macular edema.
outer retinal layers. Type 3 represented foveolar detachment with-
out traction, and type 4 constituted foveolar detachments with ap-
parent vitreofoveal traction. They observed that the morphological QUANTITATIVE ANALYSIS OF DME ON OCT
type of DME on OCT and the type of leakage on FA correlated
with visual acuity.17 In 2008, Koleva-Georgieva and Sivkova clas- Time domain OCT was widely used in clinical practice and
sified DME in terms of the size of cystoid spaces and evaluated was the principal instrument used in numerous studies conducted
the progression of macular edema (Tables 3A and 3B). They by the DRCR.net until 2011. Retinal thickness was measured as
found a correlation between the size of cystoid spaces, retinal the value in microns of the distance between the OCT layers, the
thickness, and visual acuity.18 Recently, Helmy et al19 classified RPE, and the internal limiting membrane. The center point thick-
patients with cystoid macular edema into 4 groups based on the ra- ness was defined as the average of the thickness values at the in-
tio of the vertical height of the largest macular cyst in relation to tersection of the 6 radial scans of the fast macular thickness
the maximum macular thickness, with the use of OCT (Table 4). protocol. The central subfield was defined as the circular area

TABLE 3B. Classification by Koleva-Georgieva With Retinal Thickness, Morphology, Topography, Presence and Severity of Macular
Traction, and Photoreceptor Inner Segment EZ Integrity

Retinal Thickness
No Macular Edema Early Subclinical Macular Edema Established Macular Edema
Normal macular morphology and No clinically detected retinal thickening on Retinal thickening and evident
thickness not reaching the criteria ophthalmoscopy, OCT measured retinal thickness morphological characteristics
for subclinical DME exceeding normal + 2 SD for central fixation of edema
point and fovea
Retinal Morphology
Simple Noncystoid Macular Edema Cystoid Macular Edema Serous Macular Detachment
Increased retinal thickness, intraretinal (a) Mild: cystoid spaces with horizontal
reduced reflectivity, irregularity of the diameter < 300 μm
layered structure, flattening of the foveal (b) Intermediate: cystoid spaces with horizontal
depression without presence of diameter of 300–599 μm
cystoid spaces (c) Severe: cystoid spaces with horizontal
diameter ≥ 600 μm, or large confluent cavities
with retinoschisis appearance
Retinal Topography
Nonsignificant macular edema CSME, as defined by ETDRS and evaluated on
the OCT retinal topography map
Presence and Severity of Macular Traction (PVD and/or ERM)
No Macular Traction Questionable Macular Traction Definite Macular Traction
Presence of complete PVD (Weiss ring Incomplete PVD with perifoveal or peripapillary Incomplete PVD with perifoveal
detected on ophthalmoscopy), or no adhesion and/or globally adherent ERM adhesion and/or focal ERM with
PVD (no visible posterior hyaloid line), without detectable distortion of retinal surface detectable distortion of retinal
and no ERM contour at the points of adhesion contour at the points of adhesion
Retinal Outer Layer Integrity (EZ and ELM)
EZ and ELM intact EZ and ELM with disrupted integrity
*Koleva-Georgieva DN. Optical coherence tomography findings in diabetic macular edema. In: Ola MS, ed. Diabetic Retinopathy. Vienna, Austria:
InTech; 2012.
ERM indicates epiretinal membrane; PVD, posterior vitreous detachment.

362 www.apjo.org © 2016 Asia Pacific Academy of Ophthalmology

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016 SD-OCT Features and Classification Systems for DME

improved transvitreal oxygenation of the retina and removal of a

TABLE 5. Classification of DME by Kim et al34 growth factor reservoir in the premacular hyaloid have been postu-
lated for improved retinal structure and function after vitrectomy.32,33
Type 1 Diffuse retinal thickening Panozzo et al16 studied the role of traction and defined 3
Type 2 Cystoid macular edema types of epiretinal membranes (Table 2). Kang et al17 and Kim
Type 3 Serous retinal detachment without PHT et al34 (Table 5) also considered the presence of vitreofoveal trac-
Type 4 PHT without TRD tion in their classifications. Koleva-Georgieva and Sivkova27,28
Type 5 PHT with TRD took into account the presence or absence of posterior vitreous de-
tachment, presence of epiretinal membrane, and its effect on the
PHT indicates posterior hyaloid traction; TRD, traction retinal detachment. surface of the retina (Table 3B). Studies have now established a
beneficial role of vitrectomy in improving the anatomical and vi-
1 mm in diameter centered around the center point. The central sual outcomes in DME.30
subfield mean thickness was defined as the mean value of the
128 thickness measurements obtained in this circular area. Abso- PHOTORECEPTOR ELLIPSOID ZONE
lute change in retinal thickness was defined as the difference in the DISRUPTION ANALYSIS
thickness obtained at 2 different times. The absolute change in Spectral domain OCT has enabled the visualization of integ-
thickness divided by baseline thickness provides the relative rity of the outer retinal layers, the photoreceptor external limiting
change in thickness.20 Spectral domain OCT has subsequently membrane (ELM), and inner segment ellipsoid zone. Lu et al35
widely replaced TD-OCT. Spectral domain OCT machines gener- suggested that the inner segment ellipsoid zone (EZ) of the photo-
ate higher retinal thickness values relative to TD-OCT machines. receptors corresponds to the highly reflective band of the IS/OS
The median difference between TD-OCT and SD-OCT is esti- junction seen on OCT. An expert panel introduced the term “zone”
mated to be 43 to 67 μm.21 In the recently published DRCR.net as a consensus nomenclature for anatomic regions like the photore-
Protocol T, central subfield thickness (CST) was used as a major ceptor inner segment EZ, which lacks definitely proven evidence
criterion for determining treatment eligibility. A reference limit for a specific reflective structure on OCT.36 Disruptions of these
of at least 250 μm thickness on Stratus (Carl Zeiss Meditec) was layers have been associated with increased severity of DR.37
one of the key inclusion criteria. The recommended equivalent Several authors have also found a positive correlation of visual
values were 320 μm or greater for men or 305 μm or greater for acuity with the integrity of the EZ and ELM.38–40 The postop-
women on Spectralis (Heidelberg Engineering) and 305 μm or erative status of the photoreceptors has also been correlated to
greater for men or 290 μm or greater for women on Cirrus (Carl the final visual outcome after resolution of normal retinal mor-
Zeiss Meditec).22 In a clinical setting, an additional 10% measure- phology after surgery (Fig. 1).41
ment error should be adjusted for change in instrument before
evaluating the change in retinal thickness for monitoring progres-
sion or efficacy of treatment.23 Center point thickness or CST
measured by OCT has been observed to be modestly correlated
with visual acuity. This suggests the role of additional factors like
glycated hemoglobin, age, and other retinal structural alterations
to be responsible for variation in visual acuity at baseline and
follow-up after focal laser photocoagulation.24 However, baseline
CST values have been shown to correlate with visual acuity out-
comes after anti-VEGF therapy. Central subfield thickness less
than 300 μm had lower gains in vision with ranibizumab in com-
parison with those with thicker baseline CST values.25 The
DRCR.net Protocol I study also documented that patients with
poor baseline visual acuity or CST greater than 400 μm experi-
enced more visual gain.26
Retinal thickness and volume were also taken into consider-
ation by Panozzo et al16 in their classification system (Table 2).
Koleva-Georgieva and Sivkova27,28 later put forward a classifica-
tion taking into account thickness, morphology, and topography
of the retina (Table 3B).

VITREOMACULAR INTERFACE ANALYSIS

Recently, the International Vitreomacular Traction Study Group
developed an OCT-based anatomic classification system for dis-
eases of the vitreomacular interface.29 The group defined vitreo-
macular adhesion as perifoveal vitreous separation with remaining
vitreomacular attachment and undisturbed foveal morphologic
features. Vitreomacular traction is characterized by anomalous
posterior vitreous detachment accompanied by anatomic distor-
tion of the fovea. A number of published studies have postulated
the role of tangential traction exerted by the posterior hyaloid on FIGURE 1. Spectral domain optical coherence tomography showing
the retinal surface in DME.30,31 The presence of vitreoretinal trac- grades of disruption of the ELM and EZ. A, Grade 0: no disruption of
tion in treatment unresponsive DME is an indicator for surgical in- ELM and EZ. B, Grade 1: ELM disruption (white arrow), EZ intact. C,
tervention. In addition to relief of traction, other theories such as Grade 2: both ELM and EZ disrupted (yellow arrow).

© 2016 Asia Pacific Academy of Ophthalmology www.apjo.org 363

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Ruia et al Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016

TABLE 6. SD-OCT–Based Classification of Retinal Photore- TABLE 8. SD-OCT–Based Classification of Retinal Photoreceptor
ceptor Inner Segment EZ Disruption in Diabetes Mellitus by ELM and Inner Segment EZ Disruption in Diabetes Mellitus by
Maheshwary et al42 Jain et al44

Grade 0 Intact EZ Grade 0 No disruption of ELM and EZ

Grade 1 Focal EZ disruption of ≤200 μm in length Grade 1 ELM disruption, intact EZ
Grade 2 EZ disruption of >200 μm in length Grade 2 Both ELM and EZ disruption

Maheshwary et al graded the severity of EZ disruption in At present, clinical decision making about DME manage-
eyes with DME in 2010. Grades from each patient’s horizontal ment is mostly based on simple qualitative evaluation of baseline
and vertical SD-OCT scans were added to yield a global disrup- and follow-up OCT scans and thickness measurements.20 Various
tion scale (Table 6).42 The percentage of disruption along the EZ previously mentioned physician-friendly classification systems
band, measured 500 μm in either direction from the foveal center, provide in vivo retinal structural alterations with regard to patho-
was recorded. They demonstrated that the percentage of disrup- physiology of the disease process. These structural alterations cor-
tion of the photoreceptor EZ band was an important predictor of relate with changes in visual acuity. These classification systems
visual acuity in patients with DME. Sharma et al43 in 2014 gave provide a systematic approach to the diagnosis and management
a simplified, comprehensive, and physician-friendly approach to of DME and are useful for execution and analysis of clinical stud-
grading EZ disruption based on SD-OCT findings (Table 7). ies. The recent focus has been on outer retinal structural alter-
The EZ was studied using horizontal and vertical SD-OCT scans ations and integrity of inner retinal layers.42,44,48,49 Changes in
through the fovea. The study also illustrated a significant decrease these parameters correlate with changes in visual acuity and facil-
in visual acuity correlated with an increase in the grade of disrup- itate treatment decision making.
tion of the EZ in patients with DR, and a marked decrease in visual
acuity was found to be associated with global disruption.43
In 2013, Jain et al44 showed for the first time that ELM dis-
MACULAR ISCHEMIA IN DME
ruption occurred earlier than disruption of the EZ (Table 8, Fig. 1). Macular ischemia has been suggested to be associated with
This was based on the observation that the ELM has tight junc- poor visual outcomes in patients with DME.50 Studies have docu-
tions similar to those between RPE cells. Therefore, the ELM acts mented macular ischemia to have a limiting effect on visual out-
like the third outer BRB and its disruption contributes to fluid comes after intravitreal bevacizumab injections in patients
accumulation in DME.45 The disruption of the EZ is secondary with DME.51
to disrupted ELM. This was observed in experimental mice where Fluorescein angiography has been used to identify macular
a mutation leading to disruption of the ELM was found to cause ischemia. The size of the foveal avascular zone (FAZ) and its out-
shortening of the photoreceptor inner segment.46 A significant line, on FA, correspond to the ischemic state of the macula.52 Sev-
positive correlation was found between logMAR visual acuity eral recent reports suggested that OCT can detect the structural
and grade of disruption.44 Helmy et al19 also took into consider- changes in the retina due to ischemic changes. Byeon et al53
ation the presence of photoreceptor ELM and EZ disruption in assessed foveal ganglion cell layer damage in ischemic diabetic
their classification of DME (Table 4). In another study, Murakami maculopathy and postulated that OCT provides objective results
et al overcame the limitation of variability in the reflectivity levels and might be a good noninvasive substitute for FA. Further studies
of continuous EZ bands seen on SD-OCT. Relatively homoge- documented a significant correlation of ganglion cell layer dam-
neous and continuous ELM bands were described as intact, age with duration of diabetes.54 In patients with severe non-
whereas absent or discontinuous ELM bands were described as proliferative and proliferative DR without DME, no significant
disrupted. Quantification of the status of the transverse length of association of FAZ outline or size was found with retinal volume,
each segment of disrupted EZ and ELM band was performed. total retinal thickness, and thickness of the outer and inner retina
Statistical analysis demonstrated that the total additive transverse and ganglion cell layer. There is a contradictory association be-
length of disrupted EZ and ELM within the fovea correlated tween the FAZ outline and continuity of the ELM and EZ. Certain
more closely with logMAR visual acuity than did foveal thickness studies mention foveal ischemia in DME to cause EZ disruption
measurements.47 resulting in outer retinal layer atrophic changes and subsequent
Classifications based on OCT have proved to be useful in visual loss. Studies correlating choroidal thickness and FAZ en-
monitoring disease progression and in clinical trials for the treat- largement in patients with DME have also documented contradic-
ment of DME. Trials on anti-VEGF have recommended the in- tory results.52
itiation of anti-VEGF therapy for foveal center-involving DME In most of the OCT studies in DR, the effects of macular is-
with visual impairment but laser photocoagulation for noncenter- chemia are masked by coexisting DME. Fluorescein angiography
involving cases. Vitrectomy for DME has been found to have remains the gold standard for evaluating the retinal perfusion sta-
good visual outcome in cases with traction over the retina.30 tus and for detecting macular ischemia in patients with DR. Re-
cently, OCT angiography was found to be useful for quantifying
retinal ischemia in diabetic DR. The parafoveal and perifoveal
TABLE 7. SD-OCT–Based Classification of Retinal Photore- vessel density are significantly reduced and low-perfusion areas
ceptor Inner Segment EZ Disruption in Diabetes Mellitus by are significantly increased in eyes with DR. Areas of low perfu-
Sharma et al43 sion in OCT angiogram corresponded to ischemic areas in FA.55

Grade 0 Intact EZ
HYPERREFLECTIVE SPOTS
Grade 1 Focal disruption (localized, subfoveal EZ disruption)
Grade 2 Global disruption (generalized EZ disruption
Bolz et al described the distribution of hyperreflective spots
throughout the macular cube) throughout all the retinal layers in eyes with DME. They hypoth-
esized that the hyperreflective spots represent subclinical features

364 www.apjo.org © 2016 Asia Pacific Academy of Ophthalmology

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016 SD-OCT Features and Classification Systems for DME

of lipoprotein extravasation that act as precursors of hard exu- treatment modality, whether medical or surgical. With various
dates.56 Uji et al documented that the hyperreflective spots were clinical trials on anti-VEGF pharmacotherapy, SD-OCT has be-
located in the outer retina and were closely associated with dis- come an indispensable tool in clinical practice.
ruption of the ELM and EZ. The hyperreflective spots were also
associated with decreased visual acuity. Further, the degenerated REFERENCES
photoreceptors were suggested to give origin to the hyperreflective 1. Guariguata L, Whiting DR, Hambleton I, et al. Global estimates of diabetes
spots.57 Framme et al reported the presence of hyperreflective prevalence for 2013 and projections for 2035. Diabetes Res Clin Pract.
spots in patients with both focal and diffuse DME. After anti- 2014;103:137–149.
VEGF pharmacotherapy, the hyperreflective spots were found
2. Yau JW, Rogers SL, Kawasaki R, et al. Global prevalence and major risk
to reduce significantly. Therefore, the hyperreflective spots were
factors of diabetic retinopathy. Diabetes Care. 2012;35:556–564.
hypothesized to represent clinical markers of inflammatory re-
sponse.58 Recent studies, including experimental studies, have 3. Mohamed Q, Gillies MC, Wong TY. Management of diabetic retinopathy:
documented the location of hyperreflective spots in the inner ret- a systematic review. JAMA. 2007;298:902–916.
inal layers. These were found at an early stage of DR with absence 4. Ozdek SC, Erdinc MA, Gurelik G, et al. Optical coherence
of macular edema or hard exudates. These hyperreflective spots tomographic assessment of diabetic macular edema: comparison with
were even present in the eyes of patients with diabetes without fluorescein angiographic and clinical findings. Ophthalmologica. 2005;
retinopathy. The hyperreflective spots are hypothesized to repre- 219:86–92.
sent the activation of retinal microglia located in the inner retinal 5. Brown JC, Solomon SD, Bressler SB, et al. Detection of diabetic foveal
layers. Microglial cells on activation undergo significant changes edema: contact lens biomicroscopy compared with optical coherence
in shape and size and form microglial aggregates. Further, the tomography. Arch Ophthalmol. 2004;122:330–335.
number of hyperreflective spots increases with the clinical pro- 6. Photocoagulation treatment for diabetic macular edema. Early Treatment
gression of DR and shows an inner to outer retina migration.59 Diabetic Retinopathy Study report number 1. Early Treatment Diabetic
Retinopathy Study Research Group. Arch Ophthalmol. 1985;103:
BIOMARKERS FOR DME 1796–1806.
Biomarkers are a subcategory of medical signs that objec- 7. American Academy of Ophthalmology Retina Panel. Preferred Practice
tively indicate the state of health and wellbeing of an individual. Pattern Guidelines: Diabetic Retinopathy. San Francisco, CA: American
These can be anatomical, biochemical, and molecular param- Academy of Ophthalmology; 2015.
eters or imaging features. In clinical practice, they are useful in 8. Bandello F, Cunha-Vaz J, Chong NV, et al. New approaches for the
refinement of diagnosis, measuring disease progression, or pre- treatment of diabetic macular oedema: recommendations by an expert
dicting and monitoring the effects of therapeutic interventions. panel. Eye (Lond). 2012;26:485–493.
The availability of biomarkers of DR progression offers new per- 9. Treatment techniques and clinical guidelines for photocoagulation of
spectives for understanding DR.60 Cunha-Vaz et al60 identified diabetic macular edema. Early Treatment Diabetic Retinopathy Study
microaneurysm as a biomarker of development of CSME, with report number 2. Early Treatment Diabetic Retinopathy Study Research
subclinical macular edema identified by OCT and multifocal elec- Group. Ophthalmology. 1987;94:761–774.
troretinogram as organ-specific biomarkers of DR. Nunes et al 10. Kinyoun J, Barton F, Fisher M, et al. Detection of diabetic macular edema.
identified 3 different phenotypes of diabetes on the basis of the Ophthalmoscopy versus photography—Early Treatment Diabetic
formation and disappearance of microaneurysm, presence of cap- Retinopathy Study report number 5. The ETDRS Research Group.
illary closure, and alteration of the BRB with associated retinal Ophthalmology. 1989;96:746–750.
edema. These different phenotypes were associated with different 11. Focal photocoagulation treatment of diabetic macular edema. Relationship
risks for progression of DR and development of CSME.61 Re- of treatment effects to fluorescein angiographic and other retinal
cently, OCT-based macular thickness parameters, cube average characteristics at baseline: ETDRS report no. 19. Early Treatment Diabetic
thickness and CST, were observed to be potential imaging bio- Retinopathy Study Research Group. Arch Ophthalmol. 1995;113:
markers for DME.62 They serve as indicators of progression of 1144–1155.
disease within the grade of retinopathy.63 Sharma et al49 docu- 12. Browning DJ, Altaweel MM, Bressler NM, et al. Diabetic macular edema:
mented the role of nitric oxide and oxidative stress markers like what is focal and what is diffuse? Am J Ophthalmol. 2008;146:649–655.
lipid peroxide as indicators of in vivo structural alteration in the
13. Vujosevic S, Midena E, Pilotto E. Diabetic macular edema: correlation
retinal photoreceptor EZ. Recently, Saxena et al64 documented
between microperimetry and optical coherence tomography findings.
the serum levels of urea and creatinine as surrogate markers for Invest Ophthalmol Vis Sci. 2006;47:3044–3051.
disruption of the retinal photoreceptor ELM and EZ observed on
SD-OCT in DR. 14. Okada K, Yamamoto S, Mizunoya S. Correlation of retinal sensitivity
measured with fundus-related microperimetry to visual acuity and retinal
thickness in eyes with diabetic macular edema. Eye (Lond). 2006;20:
CONCLUSIONS 805–809.
To summarize, SD-OCT is an established imaging modality 15. Otani T, Kishi S, Maruyama Y. Patterns of diabetic macular edema with
that has made a significant impact in the diagnostic evaluation optical coherence tomography. Am J Ophthalmol. 1999;127:688–693.
of patients with DME. It helps in understanding the anatomy
16. Panozzo G, Parolini B, Gusson E, et al. Diabetic macular edema: an
and pathophysiology of DME. However, currently there is no in- OCT-based classification. Semin Ophthalmol. 2004;19:13–20.
ternationally accepted, standardized SD-OCT–based classifica-
tion for DME. This probably reflects the complexity of the 17. Kang SW, Park CY, Ham DI. The correlation between fluorescein
changes in DME and the wide range of findings and clinical angiographic and optical coherence tomographic features in clinically
significant diabetic macular edema. Am J Ophthalmol. 2004;137:313–322.
implications. However, careful study of certain SD-OCT–based
changes may demonstrate the sequence of events with the pro- 18. Koleva-Georgieva DN, Sivkova NP. Types of diabetic macular edema
gression of disease and can guide decision making regarding the assessed by optical coherence tomography. Folia Med. 2008;50:30–38.
initiation, continuation, or interruption of therapy. Understanding 19. Helmy YM, Atta Allah HR. Optical coherence tomography classification
disease with the assistance of OCT has helped in choosing the of diabetic cystoid macular edema. Clin Ophthalmol. 2013;7:1731–1737.

© 2016 Asia Pacific Academy of Ophthalmology www.apjo.org 365

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Ruia et al Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016

20. Browning DJ, Glassman AR, Aiello LP, et al. Optical coherence 40. Shin HJ, Lee SH, Chung H, et al. Association between photoreceptor
tomography measurements and analysis methods in optical coherence integrity and visual outcome in diabetic macular edema. Graefes Arch Clin
tomography studies of diabetic macular edema. Ophthalmology. 2008;115: Exp Ophthalmol. 2012;250:61–70.
1366–1371.
41. Yanyali A, Bozkurt KT, Macin A, et al. Quantitative assessment of
21. The Diabetic Retinopathy Clinical Research Network/Writing Committee, photoreceptor layer in eyes with resolved edema after pars plana vitrectomy
Bressler SB, Edwards AR, et al. Reproducibility of spectral domain optical with internal limiting membrane removal for diabetic macular edema.
coherence tomography retinal thickness measurements and conversion to Ophthalmologica. 2011;226:57–63.
equivalent time domain metrics in diabetic macular edema. JAMA
42. Maheshwary AS, Oster SF, Yuson RM, et al. The association between
Ophthalmol. 2014;132:1113–1122.
percent disruption of the photoreceptor inner segment/outer segment
22. Wells JA, Glassman AR, Ayala AR, et al. Aflibercept, bevacizumab, or and visual acuity in diabetic macular edema. Am J Ophthalmol.
ranibizumab for diabetic macular edema. N Engl J Med. 2015;372:1193. 2010;150:63–67.
23. Diabetic Retinopathy Clinical Research Network. Reproducibility of 43. Sharma SR, Saxena S, Mishra N, et al. The association of grades of
macular thickness and volume using Zeiss optical coherence tomography in photoreceptor inner segment-ellipsoid band disruption with severity of
patients with diabetic macular edema. Ophthalmology. 2007;114: retinopathy in type 2 diabetes mellitus. J Case Rep Stud. 2014;2:205.
1520–1525.
44. Jain A, Saxena S, Khanna VK, et al. Status of serum VEGF and ICAM-1
24. Diabetic Retinopathy Clinical Research Network. Relationship between and its association with external limiting membrane and inner
optical coherence tomography–measured central retinal thickness and segment–outer segment junction disruption in type 2 diabetes mellitus.
visual acuity in diabetic macular edema. Ophthalmology. 2007;114: Mol Vis. 2013;19:1760–1768.
525–536.
45. Omri S, Omri B, Savoldelli M, et al. The outer limiting membrane (OLM)
25. Mitchell P, Bandello F, Schmidt-Erfurth U, et al. The RESTORE study: revisited: clinical implications. Clin Ophthalmol. 2010;4:183–195.
ranibizumab monotherapy or combined with laser versus laser monotherapy
for diabetic macular edema. Ophthalmology. 2011;118:615–625. 46. Mehalow AK, Kameya S, Smith RS, et al. CRB1 is essential for external
limiting membrane integrity and photoreceptor morphogenesis in the
26. Hansen LL. Visual acuity response to ranibizumab stratified by baseline
mammalian retina. Hum Mol Genet. 2003;12:2179–2189.
characteristics in diabetic macular edema (RESOLVE study). Eur J
Ophthalmol. 2010;3:634. 47. Murakami T, Nishijima K, Akagi T, et al. Optical coherence tomographic
reflectivity of photoreceptors beneath cystoid spaces in diabetic macular
27. Koleva-Georgieva D, Sivkova N. Assessment of serous macular
edema. Invest Ophthalmol Vis Sci. 2012;53:1506–1511.
detachment in eyes with diabetic macular edema by use of spectral-domain
optical coherence tomography. Graefes Arch Clin Exp Ophthalmol. 2009; 48. Sun JK, Lin MM, Lammer J, et al. Disorganization of the retinal inner
247:1461–1469. layers as a predictor of visual acuity in eyes with center-involved diabetic
macular edema. JAMA Ophthalmol. 2014;132:1309–1316.
28. Koleva-Georgieva DN, Sivkova NP. Optical coherence tomography for the
detection of early macular edema in diabetic patients with retinopathy. Folia 49. Sharma S, Saxena S, Srivastav K, et al. Nitric oxide and oxidative stress is
Med. 2010;52:40. associated with severity of diabetic retinopathy and retinal structural
29. Duker JS, Kaiser PK, Binder S, et al. The International Vitreomacular alterations. Clin Experiment Ophthalmol. 2015;43:429–436.
Traction Study Group classification of vitreomacular adhesion, traction, 50. Sim DA, Keane PA, Zarranz-Ventura J, et al. The effects of macular
and macular hole. Ophthalmology. 2013;120:2611–2619. ischemia on visual acuity in diabetic retinopathy. Invest Ophthalmol Vis Sci.
30. Laidlaw DA. Vitrectomy for diabetic macular oedema. Eye (Lond). 2008; 2013;54:2353–2360.
22:1337–1341. 51. Chung EJ, Roh MI, Kwon OW, et al. Effects of macular ischemia on the
31. Kaiser PK, Riemann CD, Sears JE, et al. Macular traction detachment and outcome of intravitreal bevacizumab therapy for diabetic macular edema.
diabetic macular edema associated with posterior hyaloidal traction. Am J Retina. 2008;28:957–963.
Ophthalmol. 2001;131:44–49. 52. Dmuchowska DA, Krasnicki P, Mariak Z. Can optical coherence
32. Kishi S, Demaria C, Shimizu K. Vitreous cortex remnants at the fovea after tomography replace fluorescein angiography in detection of ischemic
spontaneous vitreous detachment. Int Ophthalmol. 1986;9:253–260. diabetic maculopathy? Graefes Arch Clin Exp Ophthalmol. 2014;252:
731–738.
33. Stefansson E. Ocular oxygenation and the treatment of diabetic retinopathy.
Surv Ophthalmol. 2006;51:364–380. 53. Byeon SH, Chu YK, Lee H, et al. Foveal ganglion cell layer damage in
34. Kim BY, Smith SD, Kaiser PK. Optical coherence tomographic patterns ischemic diabetic maculopathy: correlation of optical coherence
of diabetic macular edema. Am J Ophthalmol. 2006;142:405–412. tomographic and anatomic changes. Ophthalmology. 2009;116:
1949–1959.
35. Lu RW, Curcio CA, Zhang Y, et al. Investigation of the hyper-reflective
inner/outer segment band in optical coherence tomography of living frog 54. VanDijk HW, Verbraak FD, Kok PH, et al. Decreased retinal ganglion cell
retina. J Biomed Opt. 2012;17:060504. layer thickness in patients with type 1 diabetes. Invest Ophthalmol Vis Sci.
2010;51:3660.
36. Staurenghi G, Sadda S, Chakravarthy U, et al. Proposed lexicon for
anatomic landmarks in normal posterior segment spectral-domain optical 55. Jia Y, Gao SS, Hwang TS, et al. Automated quantification of macular
coherence tomography. Ophthalmology. 2014;121:1572–1578. ischemia using optical coherence tomography angiography in diabetic
retinopathy. Invest Ophthalmol Vis Sci. 2015;56:3343.
37. Saxena S, Srivastav K, Akduman L. Spectral domain optical coherence
tomography based alterations in macular thickness and inner segment 56. Bolz M, Schmidt-Erfurth U, Deak G, et al. Optical coherence tomographic
ellipsoid are associated with severity of diabetic retinopathy. Int J hyperreflective foci: a morphologic sign of lipid extravasation in diabetic
Ophthalmol Clin Res. 2015;2:007. macular edema. Ophthalmology. 2009;116:914–920.
38. Otani T, Yamaguchi Y, Kishi S. Correlation between visual acuity and 57. Uji A, Murakami T, Nishijima K, et al. Association between hyperreflective
foveal microstructural changes in diabetic macular edema. Retina. 2010;30: foci in the outer retina, status of photoreceptor layer, and visual acuity in
774–780. diabetic macular edema. Am J Ophthalmol. 2012;153:710–717.
39. Murakami T, Nishijima K, Sakamoto A, et al. Association of 58. Framme C, Schweizer P, Imesch M, et al. Behavior of SD-OCT–detected
pathomorphology, photoreceptor status, and retinal thickness with visual hyperreflective foci in the retina of anti-VEGF–treated patients with
acuity in diabetic retinopathy. Am J Ophthalmol. 2011;151:310–317. diabetic macular edema. Invest Ophthalmol Vis Sci. 2012;53:5814–5318.

366 www.apjo.org © 2016 Asia Pacific Academy of Ophthalmology

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.
Asia-Pacific Journal of Ophthalmology • Volume 5, Number 5, September/October 2016 SD-OCT Features and Classification Systems for DME

59. Vujosevic S, Bini S, Midena G, et al. Hyperreflective intraretinal 62. Ruia S, Saxena S, Prasad S, et al. Correlation of biomarkers thiobarbituric acid
spots in diabetics without and with nonproliferative diabetic reactive substance, nitric oxide and central subfield and cube average thickness
retinopathy: an in vivo study using spectral domain OCT. in diabetic retinopathy: a cross-sectional study. Int J Retin Vitr. 2016;2:1.
J Diabetes Res. 2013;2013:5. 63. Ruia S, Saxena S. Targeted screening of macular edema by spectral domain
60. Cunha-Vaz J, Ribeiro L, Lobo C. Phenotypes and biomarkers of diabetic optical coherence tomography for progression of diabetic retinopathy.
retinopathy. Prog Retin Eye Res. 2014;41:90–111. Indian J Ocular Biol. 2016;1:102.
61. Nunes S, Ribeiro L, Lobo C, et al. Three different phenotypes of mild 64. Saxena S, Ruia S, Prasad S, et al. Increased serum levels of urea and
nonproliferative diabetic retinopathy with different risks for development of creatinine are surrogate markers for disruption of retinal photoreceptor
clinically significant macular edema. Invest Ophthalmol Vis Sci. 2013;54: external limiting membrane and inner segment ellipsoid zone in type 2
4595–4604. diabetes mellitus. Retina. 2016.

Everything you see or hear or experience in any way at all is specific to you. You create a universe by perceiving it,
so everything in the universe you perceive is specific to you.
— Douglas Adams

© 2016 Asia Pacific Academy of Ophthalmology www.apjo.org 367

Copyright © 2016 Asia Pacific Academy of Ophthalmology. Unauthorized reproduction of this article is prohibited.