Journal français d’ophtalmologie (2018) 41, 733—738

Disponible en ligne sur

ScienceDirect
www.sciencedirect.com

ORIGINAL ARTICLE

Comparison of the effect of ranibizumab

and dexamethasone implant on serous
retinal detachment in diabetic macular
edema夽
Comparaison de l’effet du ranibizumab et de l’implant de dexamethasone sur
le décollement séreux rétinien dans l’œdème maculaire diabétique

A. Demircan ∗, A. Ozkaya , Z. Alkin , B. Kemer ,

C. Yesilkaya , G. Demir

University of Health Sciences, Beyoglu Eye Training and Research Hospital, Bereketzade Cami
Sok. No. 2 Beyoglu, Istanbul, Turkey

Received 7 February 2018; accepted 15 March 2018

Available online 10 September 2018

KEYWORDS Summary
Diabetic macular Purpose. — To compare the efficacy of intravitreal ranibizumab (IVR) and intravitreal dexam-
edema; ethasone implant (IDI) on neurosensory retinal detachment (SRD) associated with diabetic
Subfoveal macular edema (DME) in the early treatment period.
neurosensory retinal Methods. — This was a retrospective, interventional, case-control study. After three monthly
detachment; loading doses of IVR or an initial IDI injection, the changes in best-corrected visual acuity (BCVA),
Dexamethasone; central macular thickness (CMT) on OCT, and presence and height of SRD were evaluated.
Ranibizumab Results. — The IVR and IDI groups consisted of 101 and 35 eyes, respectively. The mean
changes in CMT in the IVR and IDI groups were 204.4 ± 176.6 and 311.4 ± 163, respectively
(P < 0.001). The mean changes in SRD height in the IVR and IDI groups were 133.6 ± 92.1 and
168.6 ± 103.9 ␮m, respectively. The decrease in SRD height was significantly greater in the IDI
group than in the IVR group (P = 0.002). The SRD resolved completely in 72.2% and 71.4% of the
patients in the IVR and IDI groups, respectively (P = 0.9).

夽 Involved in design and conduct of the study (AD, AO, ZA, BK, CY, GD); preparation and review of the study (AD, AO, ZA, GD); data

collection (AD, BK, CY); and statistical analysis (AO).

∗ Corresponding author.

E-mail address: alidemircanctf@yahoo.com (A. Demircan).

https://doi.org/10.1016/j.jfo.2018.03.004
0181-5512/© 2018 Elsevier Masson SAS. All rights reserved.
734 A. Demircan et al.

Conclusion. — The mean reduction in CMT and SRD height was greater in the IDI group than in
the IVR group. There was a negative correlation between baseline best-corrected visual acuity
(BCVA) and SRD height and also between BCVA and CMT.
© 2018 Elsevier Masson SAS. All rights reserved.

MOTS CLÉS Résumé

Œdème maculaire Objectif. — Comparer l’efficacité du ranibizumab en intravitréen (IVR) et de l’implant intrav-
diabétique ; itréen de dexamethasone (IDI) dans le décollement de rétine neurosensorielle (DSR) en
Décollement rétinéen association avec l’œdème maculaire diabétique (OMD) dans les phases précoces du traitement.
neurosensoriel Méthodes. — Nous avons effectué une étude cas-témoins, rétrospective et interventionnelle.
sous-fovéolaire ; Les patients ont bénéficié soit de trois injections mensuelles de IVR ou l’injection initiale
Dexamethasone ; d’IDI, nous avons évalué les changements de la meilleure acuité visuelle corrigée (MAVC) et
Ranibizumab de l’épaisseur maculaire central à l’OCT (EMC), en même temps, la présence et la hauteur du
DSR.
Résultats. — Au total, 101 yeux ont reçu IVR, alors que 35 yeux ont été traités avec IDI.
Les changements moyens de l’EMC dans les deux groupes IVR et IDI étaient 204,4 ± 176,6 et
311,4 ± 163, respectivement (P < 0,001). Les changements moyens de la hauteur du DSR pour
chaque groupe IVR et IDI étaient 133,6 ± 92,1 et 168,6 ± 103,9 ␮m, respectivement. La diminu-
tion de la hauteur du DSR était significativement plus grande chez les sujets ayant reçu IID
(P = 0,002). Le DSR s’est complètement résolu chez 72,2 % des sujets du groupe IVR et 71,4 %
des sujets du groupe IDI (P = 0,9).
Conclusion. — La réduction moyenne de l’EMC et du DSR était plus grande chez les sujets du
groupe IDI comparé à ceux des IVR. Il existe une corrélation négative entre la MAVC de base et
la hauteur du DSR, ainsi qu’entre la MAVC et l’EMC.
© 2018 Elsevier Masson SAS. Tous droits réservés.

Introduction Some noninvasive biomarkers of retinal inflammation in

Diabetic macular edema (DME) is the main cause of
visual loss associated with diabetic retinopathy (DR). DME
affects around 20% of patients with diabetic retinopathy 10
years after diagnosis [1]. Although hyperglycemia has been
accepted as the main pathological factor contributing to
DME development, the precise mechanism is still unclear.
However, hyperglycemia causes DME by activating four main
glucose metabolic pathways:
• diacylglycerol/protein kinase C;
• advanced glycation end products/receptor for advanced
glycation end products;
• polyol (sorbitol) pathway, and;
• hexosamine pathway [2].
Activation of these pathways can lead to increased oxida-
tive stress, inflammation, and vascular dysfunction [3].
Oxidative stress and inflammation leads to upregulation of
growth factors and cytokines such as vascular endothelial
growth factor (VEGF), angiopoietins, tumor necrosis fac-
tor, interleukins (ILs), and matrix metalloproteinases, which
contribute to the breakdown of the blood-retinal barrier and
consequent DME development [3]. It is currently postulated
that DR is both a vascular and neuroinflammatory disease
[4,5].
DR and DME were recently reported using optical coherence
tomography (OCT) findings [4,6—8]. One of these biomarkers
was subfoveal neurosensory retinal detachment (SRD) [6,8].
In this study, we aimed to compare the effects of intrav-
itreal ranibizumab (IVR) and an intravitreal dexamethasone
implant (IDI) in the treatment of DME-associated SRD.
Methods
In this retrospective, interventional case-control study,
we reviewed the records of treatment-naïve patients
with center-involved DME-associated subfoveal neurosen-
sory retinal detachment (SRD) who were treated with
either one intravitreal implant of 0.7 mg dexamethasone
(Ozurdex, Allergan, Inc., Irvine, California, USA) or with
three monthly intravitreal injections of 0.5 mg ranibizumab
(Lucentis, Genentech, San Francisco, USA). The patients
were evaluated one month after one intravitreal dexam-
ethasone implant or three ranibizumab injections. Written
informed consent was obtained from all patients before the
treatment. The study adhered to the tenets of the Declara-
tion of Helsinki.
To be included in the study, each patient was required to
meet several criteria:
Ranibizumab versus ozurdex in SRD
• age ≥18 years and;
• patients having newly diagnosed DME-associated SRD who
had undergone three IVR injections or one IDI injection.
Patients were excluded from the study if they had a
retinal disease other than DR (such as age-related macular
degeneration, retinal vein occlusion), a vitreoretinal surface
disorder, and/or history of prior vitrectomy.
Data collected from the patient’s records included age,
sex, best-corrected visual acuity (BCVA), central macular
thickness (CMT), SRD height, and intraocular pressure (IOP)
at baseline and the first follow-up visit one month after the
initial three IVR injections or one IDI injection.
All patients underwent a standardized examination
including BCVA measurements using the Early Treatment
Diabetic Retinopathy Study chart at 4 m, slit-lamp biomi-
croscopy, IOP measurements with applanation tonometry,
and biomicroscopic fundus examination. Fundus photog-
raphy, fluorescein angiography (FA) (HRA-2; Heidelberg
Engineering, Heidelberg, Germany), and OCT imaging
(Spectralis; Heidelberg Engineering, Heidelberg, Germany)
were performed before treatment. All examinations were
repeated one month after three IVR injections or one IDI,
except for FA. FA was repeated according to the physi-
cian’s discretion. OCT was used for the detection of macular
edema and (CMT) measurements. CMT, defined as the mean
thickness of the neurosensory retina in a central 1 mm
diameter area, was computed using OCT mapping software
generated by the device. Diabetic macular edema was diag-
nosed using FA and OCT, and patients with a CRT of >300
microns were considered to have DME. SRD height was mea-
sured manually with the built in caliper function of the OCT
device.
Injection method
All injections were performed under sterile conditions after
topical anesthesia and 10% povidone-iodine scrub (Beta-
dine; Purdue Pharma, Stamford, CT, USA) was used on the
lids and lashes, and 5% povidone-iodine was then admin-
istered on the conjunctival sac. A sustained release IDI
®
(Ozurdex 0.7 mg, Allergan inc., Irvine, CA) was injected
through the pars plana at 3.5 to 4 mm to the limbus with
a customized, single-use, 22-gauge applicator, or alterna-
tively, ranibizumab (Lucentis; Novartis, Basel, Switzerland)
was injected with a 30-gauge needle through the pars plana
at 3.5 to 4 mm posterior to the limbus. Patients were
then instructed to consult the hospital if they experienced
decreased vision, eye pain, and/or any new symptoms.
The primary outcome measures in this study included
changes in BCVA and OCT with respect to CMT and presence
and height of SRD at baseline and at the first visit one month
after the injection.
Statistical analysis
Means, standard deviations, medians, minimum and maxi-
mum values, frequencies, and ratios were used to describe
the data set. The distribution of data was measured using the
Kolmogorov-Smirnov test. The Wilcoxon test was used in the
analysis of dependent quantitative data. The Mann-Whitney
U test was used in the analysis of independent quantitative
735
data. The chi2 test was used to analyze qualitative inde-
pendent data, and Spearman’s correlation analysis was used
to assess relationships between the variables. SPSS program
version 22.0 was used for the analyses.
Results
A total of 136 eyes of (136 patients) were included the study.
The IVR group consisted of 101 eyes (69.7%), and the IDI
group consisted of 35 eyes (24.1%). Baseline characteristics
and clinical data of the patients and statistical analyses are
summarized in Table 1.
The mean BCVA values of the groups were not significantly
different before the injections (P = 0.064). BCVA significan-
tly increased after the injections in both groups (P = 0.03
and P = 0.001 for IVR and IDI groups, respectively). The mean
BCVA changes before and after the injections were not sta-
tistically significant between the groups (P = 0.146).
The mean CMTs before the injections were 643.9 ± 148.6
and 605.3 ± 140.8 ␮m in the IDI and IVR groups, respectively
(P = 0.225). CMT significantly decreased after the injections
in both groups. The mean changes were 311.4 ± 163 and
204.4 ± 176.6 ␮m in the IDI and IVR groups, respectively
(P < 0.001 for both groups). The decrease in CMT was sig-
nificantly greater in the IDI group than in the IVR group
(P = 0.002) (Fig. 1).
The mean SRD height before the injections was
212.5 ± 157 and 166.5 ± 101.4 ␮m in the IDI and IVR
groups, respectively; there were no statistically significant
differences between the groups (P = 0.059). SRD height sig-
nificantly decreased after the injections in both groups.
The mean change was 168.6 ± 103.9 and 133.6 ± 92.1 ␮m
in the IDI and IVR groups, respectively (P < 0.001 for both
groups). The decrease in SRD height was significantly greater
in the IDI group than in the IVR group (P = 0.002). SRD com-
pletely resolved in 73% and 71.4% of the patients in the
IVR and IDI groups, respectively (P = 0.9). There was a nega-
tive correlation between baseline BCVA with reinjection SRD
height (r = 0.296, P < 0.001) and reinjection CMT (r = 0.292,
P < 0.001) in the entire study population (Fig. 2).
No injection-related endophthalmitis was noted after IVR
and IDI injections. Only four patients in the IDI group showed
a transient IOP rise of >10 mmHg.
Discussion
SRD is accumulation of the fluid in the subretinal space.
The posterior border of the detached retina is clearly deter-
mined as a hyperreflective line on the upper edge of the fluid
in optical coherent tomography [9]. In spite of high SRD fre-
quency in DME, the exact importance and pathophysiology
of SRD has not yet been determined [10,11]. Leakage from
retinal or choroidal circulation into the subretinal space
exceeding its drainage capacity is thought to be the main
mechanism [12]. Disruption of the outer blood-retinal bar-
rier (BRB) may be an important factor in the pathogenesis
of DME and SRD. The diabetes-induced outer BRB dysfunc-
tion has been observed in humans and animals [13,14]. Also
Xu H and Le Y demonstrated substantial leakages of macro-
molecules and fluid from the outer BRB to subretinal space
736 A. Demircan et al.

Table 1 Comparison of demographic and clinical data of the groups.

IVR Group IDI Group P

Mean ± S.D./n-% Mean ± S.D./n-%

Age 58.0 ± 8.9 60.3 ± 9.0 0.279m
Gender
Female 51 50.5% 16 45.7% 0.626x2
Male 50 49.5% 19 54.3%
Visual acuity, (LogMar)
Before treatment 0.8 ± 0.4 1.0 ± 0.6 0.064m
After treatment 0.7 ± 0.4 0.8 ± 0.6 0.201m
Difference 0.1 ± 0.3 0.2 ± 0.3 0.146m
Intra group p 0.003w 0.001w
CMT, ()
Before treatment 605.3 ± 140.8 643.9 ± 148.6 0.225m
After treatment 409.9 ± 144.9 332.5 ± 142.0 0.006m
Difference 204.5 ± 176.6 311.4 ± 163.0 0.002m
Intra group p 0.000w 0.000w
SRD Height group p
Before treatment 166.5 ± 101.4 212.5 ± 157.0 0.059m
After treatment 32.9 ± 68.7 43,9 ± 77.5 0.271m
Difference 133.6 ± 92.1 133.6 ± 103.9 0.020m
Intra group p 0.000w 0.000w
m: Mann-whitney u test; X2 : Chi2 test; w : Wilcoxon test; SD: standard deviation; IVR: intravitreal ranibizumab; IDI: intravitreal dex-
amethasone implant; CMT: central macular thickness; SRD: subfoveal neurosensorial retinal detachment. Characters in bold indicates
statistically significant values.

Figure 2. Correlation between visual acuity and subfoveal serous

retinal detachment height.
Figure 1. The mean central macular thickness changes.

Normally, the external limiting membrane is permeable

in diabetic animal retina. They stated that the outer BRB to fluid and albumin [16]. However, when the inner blood-
breakdown resulted with leakage of the molecules and influx retinal barrier is disrupted, the excess fluid might reach the
of osmolytes and this caused accumulation of fluid in the subretinal space in large amounts, and the retinal pigment
subretinal space and serous retinal detachment [15]. epithelium may fail to remove the fluid, resulting in SRD
Ranibizumab versus ozurdex in SRD
[17]. Integrity of the external limiting membrane (ELM) is
also an important factor for protection against SRD in DME.
There are no anatomic barriers to fluid movement in the
retina, and only the ELM has a barrier function with narrow
channels against cells and large molecules [16]. When the
ELM is impaired, it allows fluid to move into the subretinal
space along with lipids, proteins, and inflammatory cells.
These changes may lead to the development of SRD [6]. This
hypothesis was supported by Ota et al. [18]. ELM disconti-
nuity was frequently observed in eyes with SRD, and most
of these eyes showed hyperreflective dots on OCT images.
Although some authors described these spots as a precursor
of hard exudates, Vujosevic et al. defined them as activated
microglial cells [8,18,19]. Aloisi hypothesized that disrup-
tion of the ELM in eyes with DR could be accompanied by cell
damage, and these damaged cells and debris were strong
attractors of scavenger cells to the retina [20]. Omri et al.
demonstrated that there was a blockage of the normal cell
trafficking process between the retina and choroid in dia-
betic rats due to the decrease in retinal pigment epithelium
(RPE) pores, which resulted in subretinal accumulation of
activated microglia/macrophages [21]. These cells could be
the source of IL-6 [20]. Sonoda et al. demonstrated a corre-
lation between SRD and higher intravitreal levels of IL-6 [6].
Recently, Vujosevic et al. presented intraretinal hyperreflec-
tive retinal spots, subfoveal neuroretinal detachment, and
increased foveal autofluorescence as retinal inflammatory
biomarkers in DR [8].
In the current study, we evaluated the effects of IDI
and ranibizumab (as anti-inflammatory and anti-vascular
endothelial growth factor agents, respectively) on SRD-
associated DME. In the literature, there is some controversy
about the most appropriate SRD treatment. Kim et al.
reported that intravitreal bevacizumab injections were
more effective in diffuse retinal thickness-type than in cys-
toid macular edema or serous retinal detachment (SRD) DME
types [22].
Seo et al. reported that ranibizumab was effective in
the treatment of different DME types (diffuse, cystoid, and
serous), but poorer visual acuity outcomes, closely related
to ELM and ellipsoid zone integrity, occurred more fre-
quently in the SRD-type [23]. Shimura et al. showed that
the effectiveness of intravitreal bevacizumab (IVB) in redu-
cing macular edema was weakest in the SRD-type [24].
Ercalık et al. compared intravitreal ranibizumab (IVR) com-
bined with posterior sub-tenon injections of triamcinolone
acetonide with IVR alone in SRD treatment. The authors
reported that SRD disappeared with a higher rate with the
combined therapy at the 1-month follow-up, but this rate
was not significant at the 3-month follow-up [25]. Liu et al.
compared intravitreal triamcinolone acetonide (IVTA) with
IVB in the treatment of different types of DME. They demon-
strated that IVTA treatment was more favorable in SRD-type
DME with respect to functional and morphological improve-
ments [26].
No studies have included large sample sizes (such as ours
did) to compare dexamethasone with ranibizumab in the
treatment of SRD-associated DME. Vujosevic et al. analyzed
and compared changes in retinal inflammatory biomarkers
(such as SRD) after treatment with intravitreal dexam-
ethasone and IVR. Seventeen of 49 patients had SRD in
this study (seven in the dexamethasone group, 10 in the
737
ranibizumab group). The researchers reported no significant
differences in SRD resolution between the two treatment
groups although there was a trend toward a higher response
in the dexamethasone group (86% versus 50%). Although they
documented a significant decrease in CMT and resolution of
SRD after both steroid and anti-VEGF treatments, they noted
that CMT decrease was greater after the steroid treatment
[8]. The findings of our study correspond to those of Vujo-
sevic et al., but different from theirs, we focused only on
SRD, and our sample size was larger. Our study included 136
eyes of 136 patients; all patients had SRD.
Although morphological and functional improvements
were observed after both IVR and IDI injections, the
decreases in CMT and SRD height were significantly greater
in the IDI group when compared with the IVR group (Fig. 1).
This superiority of IDI over IVR might be related to the role
of inflammation in the formation of SRD. This relation was
described in previous studies [6,16]. Vujosevic et al. also
proposed that SRD was an inflammatory biomarker in DME
[8]. SRD completely resolved in 73% and 71.4% of patients in
the IVR and IDI groups, respectively (P = 0.34).
BCVA significantly increased after the injections in both
groups. The mean BCVA changes before and after the injec-
tions were not statistically significant between the groups.
There was a negative correlation between baseline BCVA,
SRD height, and CMT (Fig. 2).
No data are currently available on long-term follow-up
results after steroid and anti-VEGF treatment of SRD in DME.
We compared the early results of a single injection of IDI
with a loading dose (three injections) of ranibizumab as the
first-line treatment. When starting anti-VEGF treatment, a
loading dose should at least be performed before evaluating
any stabilizing effects. Therefore, it is reasonable that these
dosing regimens may be clinically comparable [8]. This study
was not a treatment regimen study as it focused only on SRD
as one of the inflammation-associated biomarkers in DME
[8].
This study included limitations such as the lack of patient
and treatment selection criteria due to retrospective design
and the non-classification of patients according to duration
and severity of diabetes, unequal contribution between the
two treatment groups with respect to patient numbers, and
short follow-up periods. The duration of SRD might be a con-
tributing factor to the treatment response; however, we had
no data about this secondary to the retrospective nature of
our study.
In conclusion, we compared two intravitreal agents (IDI
versus IVR) in SRD-associated DME and found more morpho-
logical improvement in the IDI group rather than in the IVR
group. The mean CMT and SRD height reduction was greater
in the IDI group versus the IVR group. In the light of this
study IDI might be preferred in SRD associated with DME.
Prospective, randomized studies and long-term follow-up
periods are needed to compare the differences in specific
morphological and functional responses to the two types of
treatment.
Funding
No financial support was received for this submission.
738
Acknowledgments
The authors thank Dr. Cem Kesim for French and English
editing of the manuscript.
The paper has not been presented in a meeting or pub-
lished elsewhere.
Disclosure of interest
The authors declare that they have no competing interest.
References
[1] Early Treatment Diabetic Retinopathy Study Research Group:
photocoagulation for diabetic macular edema: ETDRS report
No. 4. Int Ophtahlmol Clin 1987;27:265—72.
[2] Zhang X, Zeng H, Bao S, Wang N, Gillies MC. Diabetic macular
edema: new concepts in patho-physiology and treatment. Cell
Biosci 2014;4:4—27.
[3] Das A, McGuire PG, Rangasamy. Diabetic macular edema:
pathophysiology and novel therapeutic targets. Ophthalmology
2015;122:1375—94.
[4] Madeira MH, Boia R, Santos PF, Ambrosio AF, Santiago
AR. Contribution of microglia-mediated neuroinflamma-
tion to retinal degenerative diseases. Mediators Inflamm
2015;2015:673090.
[5] Tang J, Kern TS. Inflammation in diabetic retinopathy. Prog
Retin Eye Res 2011;30:343—58.
[6] Sonoda S, Sakamoto T, Yamashita T, et al. Retinal morphologic
changes and concentrations of cytokines in eyes with diabetic
macular edema. Retina 2014;34:741—8.
[7] Vujosevic S, Casciano M, Pilotto E, et al. Diabetic macular
edema: fundus autofluorescence and functional correlations.
Invest Ophthalmol Vis Sci 2011;52:442—8.
[8] Vujosevic S, Torresin T, Bini S, et al. Imaging retinal inflam-
matory biomarkers after intravitreal steroid and anti-VEGF
treatment in diabetic macular oedema. Acta Ophthalmol
2017;95:464—71.
[9] Otani T, Kishi S, Maruyama Y. Patterns of diabetic macular
edema with optical coherence tomography. Am J Ophthalmol
1999;127:688—93.
[10] Gaucher D, Sebah C, Erginay A, et al. Optical coherence
tomography features during the evolution of serous retinal
detachment in patients with diabetic macular edema. Am J
Ophthalmol 2008;145:289—96.
[11] Ozdemir H, Karacorlu M, Karacorlu S. Serous macular detach-
ment in diabetic cystoid macular oedema. Acta Ophthalmol
Scand 2005;83:63—6.
A. Demircan et al.
[12] Weinberg D, Jampol LM, Schatz H, Brady KD. Exudative reti-
nal detachment following central and hemicentral retinal vein
occlusions. Arch Ophthalmol 1990;108:271—5.
[13] Vinores SA, Gadegbeku C, Campochiaro PA, Green WR. Immuno-
histochemical localization of blood-retinal barrier breakdown
in human diabetics. Am J Pathol 1989;134:231—5.
[14] Vinores SA, Derevjanik NL, Ozaki H, Okamoto N, Campochiaro
PA. Cellular mechanisms of blood-retinal barrier dysfunction in
macular edema. Doc Ophthalmol 1999;97:217—28.
[15] Xu HZ, Le YZ. Significance of outer blood-retina barrier break-
down in diabetes and ischemia. Invest Ophthalmol Vis Sci
2011;52:2160—4.
[16] Marmor MF. Mechanisms of fluid accumulation in retinal edema.
Doc Ophthalmol 1999;97:239—49.
[17] Kang SW, Park CY, Ham D. The correlation between fluorescein
angiographic and optical coherence tomographic features in
clinically significant diabetic macular oedema. Am J Ophthal-
mol 2004;137:313—22.
[18] Ota M, Nishijima K, Sakamoto A, et al. Optical coherence
tomographic evaluation of foveal hard exudates in patients
with diabetic maculopathy accompanying macular detach-
ment. Ophthalmology 2010;117:1996—2002.
[19] Vujosevic S, Bini S, Midena G, et al. Hyperreflective
intraretinal spots in diabetics without and with non-
proliferative diabetic retinopathy: an in vivo study
using spectral domain OCT. J Diabetes Res 2013:491835,
http://dx.doi.org/10.1155/2013/491835 [Epub 2013 Dec 9].
[20] Aloisi F. Immune function of microglia. Glia 2001;36:165—79.
[21] Omri S, Behar-Cohen F, de Kozak Y, et al.
Microglia/macrophages migrate through retinal epithe-
lium barrier by a transcellular route in diabetic retinopathy:
role of PKC in the Goto Kakizaki rat model. Am J Pathol
2011;179:942—53.
[22] Kim M, Lee P, Kim Y, Yu SY, Kwak HW. Effect of intravitreal beva-
cizumab based on optical coherence tomography patterns of
diabetic macular edema. Ophthalmologica 2011;226:138—44.
[23] Seo KH, Yu SY, Kim M, Kwak HW. Visual and morphologic
outcomes of intravitreal ranibizumab for diabetic macular
edema based on optical coherence tomography patterns.
Retina 2016;36:588—95.
[24] Shimura M, Yasuda K, Yasuda M, Nakazawa T. Visual outcome
after intravitreal bevacizumab depends on the optical coher-
ence tomographic patterns of patients with diffuse diabetic
macular edema. Retina 2013;33:740—7.
[25] Ercalik NY, Yenerel NM, Imamoglu S, et al. Combined intravit-
real ranibizumab and sub-tenon injection of triamcinolone for
the treatment of diabetic macular edema with retinal detach-
ment. J Ocul Pharmacol Ther 2016;32:225—9.
[26] Liu Q, Hu Y, Yu H, et al. Comparison of intravitreal triam-
cinolone acetonide versus intravitreal bevacizumab as the
primary treatment of clinically significant macular edema.
Retina 2015;35:272—9.