- Email: [email protected]
OCT Biomarkers as Functional Outcome Predictors in Diabetic Macular Edema Treated with Dexamethasone Implant
Optical Coherence Tomography Biomarkers as Functional Outcome Predictors in Diabetic Macular Edema Treated with Dexamethasone Implant Dinah Zur, MD,1,z Matias Iglicki, MD,2,z Catharina Busch, MD,3 Alessandro Invernizzi, MD,4 Miriana Mariussi, MD,5 Anat Loewenstein, MD,1,6 for the International Retina Group Purpose: Identification and characterization of patients with diabetic macular edema (DME) are important for individualizing treatment and optimizing outcome. We investigated optical coherence tomography (OCT) biomarkers for DME treated by intravitreal dexamethasone (DEX) implant. Design: Multicenter, retrospective, observational cohort study. Participants: A total of 299 eyes from 284 patients treated with DEX implant for DME (naïve, n ¼ 209; refractory, n ¼ 90). Baseline best-corrected visual acuity (BCVA) was between 0.3 and 1.0 on a logarithm of minimum angle of resolution visual chart. Methods: The OCT scans previous to DEX implants were evaluated for submacular fluid, size and location of cystoid changes, inner segment-outer segment (IS-OS) continuity, quantity and location of hyperreflective foci (HRF), vitreomacular interface abnormalities, and epiretinal membrane. The BCVA and central macular thickness were recorded at baseline and at 1, 2, and 4 months after treatment with DEX implants. Correlations between OCT measures and visual outcome were analyzed using the generalized estimating equations procedure. Main Outcome Measures: The correlation between spectral-domain (SD) OCT measures at baseline and BCVA response (mean change from baseline; categorized improvement [<5, 5e9, or 10; Early Treatment Diabetic Retinopathy Study letters] in BCVA) after treatment with a DEX implant. Results: The presence of subretinal fluid (odds ratio [OR], 1.98; 95% confidence interval [CI], 1.23e3.20; P ¼ 0.01), absence of HRF (OR, 3.66; 95% CI, 1.40e9.62; P ¼ 0.01), and integrity of the IS-OS layer (OR, 2.09; 95% CI, 1.30e3.37; P ¼ 0.003) were all predictive of better visual outcome after treatment with DEX implants. Although eyes with naïve DME gained more vision than refractory eyes (P < 0.001), the predictive value of OCT findings did not differ according to this classification. Conclusions: Spectral-domain OCT is useful in identifying various imaging findings in DME. Among eyes with DME, those with submacular fluid, no HRF, and a continuous IS-OS layer responded better to DEX implants than those without these features. These findings call for further study of combinations of OCT and metabolic biomarkers. Ophthalmology 2017;-:1e9 ª 2017 by the American Academy of Ophthalmology Supplemental material available at www.aaojournal.org.
Vision loss associated with diabetic retinopathy is most commonly caused by diabetic macular edema (DME),1 which affects approximately 7% of all diabetic patients.2 Antivascular endothelial growth factor (VEGF) injections are generally considered suitable first-line therapy for center-involved DME and are effective in improving visual acuity (VA), with 10% to 40% of patients achieving significant improvement in best-corrected visual acuity (BCVA) after 1 year of treatment.3e6 Still, a considerable proportion of patients do not respond satisfactorily to antiVEGF agents, even with intensive treatment over the first year.7e9 A recent post hoc analysis of the Diabetic Retinopathy Clinical Research Network’s (DRCR.net) Protocol I study revealed that approximately 40% of eyes gain ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
<5 letters after 3 months and approximately 50% of these maintain poorer long-term visual outcomes after 3 years than eyes with favorable early response.10 Thus, early identification and characterization of patients with DME are critical, as well as the provision of individualized treatment regarding optimal functional outcome and disease management. The efficacy of dexamethasone (DEX) intravitreal implant 0.7 mg (Ozurdex, Allergan, Inc., Irvine, CA) in DME has been shown by improvement in VA and decrease in retinal thickness,11,12 even in patients with DME that is refractory to anti-VEGF treatments.13e15 Corticosteroids and anti-VEGF treatments target different pathways in the pathogenesis of DME.16
http://dx.doi.org/10.1016/j.ophtha.2017.08.031 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 Anatomic measures on spectral-domain (SD) optical coherence tomography (OCT), such as precise evaluation of individual layers, quantification of retinal thickness and macular volume, and qualitative assessment of fluid distribution and existence of hyperreflective foci (HRF),17 could predict treatment success or failure to various therapies. Baseline OCT measures have been investigated regarding their predictive value in patients with DME treated by antiVEGF therapy.18e25 We hypothesized that distinct structural changes identifiable on SD OCT could help predict treatment responses to DEX implants, distinct from findings in eyes treated with anti-VEGF treatments. The purpose of this study was to investigate whether characteristics identified on SD OCT may serve as biomarkers and predict treatment response to DEX implants in patients with DME.
Methods Institutional review board approval was obtained through the individual institutional review boards at the participating institutes for a retrospective consecutive chart review. This international multicenter study included 14 centers (Supplementary Material, available at www.aaojournal.org). The research adhered to the tenets of the Declaration of Helsinki. Patient records from January 1, 2011, to August 1, 2016, were reviewed for cases of DME treated by intravitreal DEX implant.
Study Participants To be included in the analysis, patients had to fulfill the following criteria: (1) age 18 years; (2) type 1 or 2 diabetes mellitus; (3) DME (both naïve and refractory) causing visual loss, with study eye BCVA measuring 0.3 to 1.0 logarithm of the minimum angle of resolution; macular edema defined clinically and by retinal thickness of >250 mm in the central subfield; and intraretinal or subretinal fluid (SRF) seen on SD OCT; (4) treatment with DEX implant. For patients who received bilateral treatment with DEX, both eyes were included. Refractory DME was defined as worsening of BCVA by 2 Early Treatment Diabetic Retinopathy Study lines or reduction of less than 10% of retinal thickness on SD OCT measured 1 month after at least 3 anti-VEGF injections that were given at monthly intervals.10 Only first treatments with DEX implants were considered for the study. Exclusion criteria were (1) another concomitant ocular disease that causes macular edema (i.e., neovascular age-related macular degeneration or choroidal neovascularization due to other reasons, retinal vein occlusion, uveitis, and recent intraocular surgery possibly causing postsurgical macular edema); (2) another ocular condition that compromises VA, except for the presence of cataract; and (3) previous treatment with intraocular corticosteroids within the 6 months before treatment with the DEX implant. Patient charts were reviewed for demographic data, hemoglobin A1C (HbA1C) values, type of retinopathy (nonproliferative or proliferative), previous treatments for DME, and BCVA before the DEX implant and at 1, 2 and 4 months after the injection.
Optical Coherence Tomography Analysis The OCT scans were obtained using SD OCT: Heidelberg Spectralis, Heidelberg, Germany; Optovue Avanti, Fremont, California; Topcon 3D OCT-2000, Tokyo, Japan; and Cirrus, Zeiss, Oberkochen, Germany. Quantitative assessment of DME included central macular thickness (CMT) that was calculated automatically by the instrument and recorded at baseline and at 1, 2, and 4 months after the DEX
2
implant. Qualitative and quantitative evaluations of SD OCT images performed at baseline assessed the presence of several morphologic features (Fig 1), including (1) SRF; (2) cystoid changes in the outer nuclear layer (ONL) and maximal cyst size in the ONL (small <100 mm, large 100e200 mm, 3 ¼ giant >200 mm); (3) presence of cystoid changes in the inner nuclear layer (INL); (4) continuity of the inner segment-outer segment (IS-OS) layer (completely continuous, partly disrupted, completely disrupted); (5) presence of HRF, as well as quantity (few 2e10, moderate 11e20, many >21) and location (between the internal limiting membrane and the INL; between the outer plexiform layer and external limiting membrane; in all retinal layers); (6) status of the vitreomacular interface (detached, vitreomacular adhesion, vitreomacular traction); and (7) presence of an epiretinal membrane. The listed features were evaluated on 3 horizontal OCT scans: 1 b-scan encompassing the fovea, 2 b-scans respectively 500 mm superior and 500 mm inferior to the fovea. Grading of the OCT images was performed by 3 experienced retina specialists (DZ: 109 cases, MI: 100 cases, AI: 90 cases) who were blinded to the functional and anatomic results. Interrater reliability was calculated among 90 cases that were graded by all graders. The CMT was recorded at baseline and at 1, 2, and 4 months after the DEX implant.
Statistical Analysis Interrater reliability was tested as absolute agreement with a 2-way mixed interclass correlation model. To control for the correlated nature of our data, we used a generalized estimating equation (GEE) procedure. Differences in functional treatment response and OCT baseline measures between naïve and refractory patients were calculated by a GEE model and included baseline BCVA as a covariate. Differences in anatomic outcomes were calculated and include baseline CMT as a covariate. The GEE models for outcome measures (BCVA response at 2 and 4 months) were run by testing the following predictors at baseline: (1) BCVA; (2) naïve or refractory DME; (3) the presence of SRF; (4) the presence of ONL cysts; (5) ONL cyst size; (6) presence of INL cysts; (7) IS/OS continuity; (8) HRF presence; (9) HRF quantity; (10) HRF location; (11) vitreoretinal interface; and (12) presence of an epiretinal membrane. Predictors were included stepwise in the GEE model and kept in the model if the P value was less than 0.10. The final GEE model was used to calculate the odds ratios (ORs) and their 95% confidence intervals (CIs), with a change of 5 letters in baseline BCVA considered a standard unit of change. An association of HbA1c levels with baseline and response parameters was examined using a linear GEE model. For BCVA and CMT response after 4 months, baseline levels were included as covariates. The last observation carried forward method was used to impute missing data. Values are presented as mean standard deviation. Statistical analysis was performed with SPSS Statistics 22 (IBM, Armonk, NY).
Results Study Participants A total of 299 eyes from 284 patients were included in the analysis. Demographic data are shown in Table 1. In 209 eyes (70%), DME was naïve; 90 cases (30%) were refractory to previous anti-VEGF injections. The mean number of previous anti-VEGF injections was 7.78.8. A total of 104 eyes (35%) were previously treated by macular laser. A total of 169 eyes (57%) were phakic, and 130 eyes (43%) were pseudophakic. HbA1c levels were available for 180 patients; the mean value was 8.4%2.7%.
Zur et al
OCT Biomarkers for DME with DEX Implant
Figure 1. Optical coherence tomography (OCT) measures. A, Serous macular detachment (SMD) with cystoid diabetic macular edema (DME), showing retinal elevation with an optically clear space between the sensory retina and the retinal pigment epithelium (*). B, Grading of intraretinal cysts: Cystoid DME with a giant outer nuclear layer (ONL) cyst (*), as well as cysts in the inner nuclear layer (INL) (arrow). C, Diabetic macular edema associated with focal vitreomacular traction. DeF, Grading of hyperreflective foci (HRF). D, Few (<10) HRF between the external limiting membrane and the outer plexiform layer. E, Diabetic macular edema with a moderate amount (10e20) of HRF located between the INL and the internal limiting membrane. F, Diabetic macular edema with many (>20) HRF scattered in all retinal layers. GeI, Grading of the inner segment-outer segment (IS-OS) integrity. G, The IS-OS layer is regular and continuous. H, Partial irregularity and lack of continuity of the IS-OS layer (between arrows). I, Complete discontinuity of the ISOS layer (between arrows). ELM ¼ external limiting membrane; ILM ¼ internal limiting membrane; OPL ¼ outer plexiform layer.
Anatomic and Best-Corrected Visual Acuity Baseline Characteristics The OCT baseline characteristics are shown in Table 2. Mean BCVA was decreased at baseline for naïve compared with refractory patients (0.680.32 vs. 0.580.27 [logarithm of the
Table 1. Descriptive Statistics: Demographic Data Age (yrs), mean SD Gender male, n (%) Nonproliferative DR, n (%) Proliferative DR, n (%) Naïve DME, n (%) Refractory DME, n (%) No. of previous anti-VEGF injections, mean SD Previous macular photocoagulation, n (%) Lens status* Phakic, n (%) Pseudophakic, n (%) HbA1c (%), mean SDy
6414 171 (57) 151 (51) 148 (49) 209 (70) 90 (30) 7.78.8 104 (35) 169 (57) 122 (410.8) 8.42.7
DME ¼ diabetic macular edema; DR ¼ diabetes retinopathy; HbA1c ¼ hemoglobin A1c; SD ¼ standard deviation; VEGF ¼ vascular endothelial growth factor. *Missing data for 8 patients (2.7%). y Available for 180 patients.
minimum angle of resolution], P ¼ 0.006). However, CMT did not differ significantly between the groups (508141 vs. 535142, P ¼ 0.13). Furthermore, naïve patients were more likely to present with SRF (OR, 3.57; 95% CI, 2.04e6.26; P < 0.001), smaller ONL cysts (OR, 1.88; 95% CI, 1.25e2.82; P ¼ 0.002), no INL cysts (OR, 3.75; 95% CI, 2.19e6.44; P < 0.001), and no HRF (OR, 7.72; 95% CI, 1.81e32.85; P ¼ 0.01). Interrater reliability among the graders was 0.89 (95% CI, 0.84e0.93).
Functional and Anatomic Outcome At 4 months, 130 eyes (43.5%) showed a 10-letter improvement in BCVA, 148 eyes (49.5%) ranged within a 9-letter change from baseline, and 21 eyes (7%) lost 10 letters. The odds of gaining 10 letters after 2 and 4 months was increased when baseline BCVA was low (2 months: OR, 1.39; 95% CI, 1.25e1.56; P < 0.001; 4 months: OR, 1.15; 95% CI, 1.06e1.25; P ¼ 0.001). Eyes refractory to previous anti-VEGF injections were more likely to gain <10 letters or even lose BCVA after 4 months compared with naïve eyes (OR, 2.07; 95% CI, 1.21e3.54; P ¼ 0.01). At 4 months, CMT decreased by 20% in 204 eyes (68%) and by <20% in 95 eyes. Eyes refractory to previous anti-VEGF injections were half as likely to respond to DEX implants, as assessed by CMT (OR, 2.58; 95% CI, 1.49e4.50; P ¼ 0.001).
3
Ophthalmology Volume -, Number -, Month 2017 Table 2. Optical Coherence Tomography Baseline Measures on Spectral-Domain Optical Coherence Tomography Naïve, n (%) SRF Total, 135 (45.2) ONL cyst, n (%) Total, 249 (83.3) ONL cyst size Small Large Giant INL cysts Total, 144 (48.2) IS-OS continuity Completely continuous Partially continuous Completely disrupted HRF fociepresence Total, 271 (91) HRF foci - quantity Absent Few (2e10) Moderate (11e20) Many (>20) HRF foci - location ILM-INL, 19/247 OPL-ONL, 58/247 All layers, 170/247 Vitreomacular interface PVD VMA VMT
P Value
OR
95% CI
112 (53.4)
Refractory, n (%) 23 (25.6)
<0.001
3.57
2.04e6.26
160 (76.6)
89 (98.9)
<0.001
0.03
0.004e0.21
26 36 98 80
(16.3) (22.5) (61.3) (38.3)
6 16 67 63
(6.7) (18.0) (75.3) (70.8)
0.002
1.88
1.25e2.82
<0.001
0.27
0.16e0.46
73/203 94/203 36/203 183
(36.0) (46.3) (17.7) (87.6)
27 57 6 88
(30.0) (63.3) (6.7) (97.8)
0.95 0.01
1.01 0.32
0.70e1.45 0.13e0.75
26 71 64 48
(12.4) (34.0) (30.6) (23.0)
2 37 27 24
(2.2) (41.1) (30.0) (26.7)
0.02
1.37
1.06e1.78
9/158 (5.7) 37/158 (23.4) 112/158 (70.9)
10/88 (11.4) 21/88 (23.9) 58/88 (64.8)
0.25
0.79
0.53e1.18
183 (87.6) 19 (9.1) 7 (3.3)
67 (74.4) 21 (23.3) 2 (2.2)
0.09
1.57
0.94e2.66
CI ¼ confidence interval; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; PVD ¼ posterior vitreous detachment; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMT ¼ vitreomacular traction.
Optical Coherence Tomography Predictors for Treatment Response The presence of SRF at baseline was a predictor of good functional treatment response after 4 months (SRF: OR, 1.98; 95% CI, 1.23e3.20; P ¼ 0.01) (Fig 1). Moreover, eyes that did not present HRF at baseline were more likely to be good responders at 4 months (absent vs. present HRF: OR, 3.66; 95% CI, 1.40e9.62; P ¼ 0.01). Eyes with continuous IS-OS layers had better treatment response after 2 months (IS/OS continuity: OR, 2.09; 95% CI, 1.30e3.37; P ¼ 0.003). Figure 1 shows OCT biomarkers that were predictive of good BCVA response after 2 and 4 months. The predictive values of all the OCT measures examined are shown in Tables 3 and 4 for responses after 2 and 4 months, respectively. In a subanalysis, we investigated eyes that became good responders only after 4 months and not 2 months. Eyes with SRF at baseline were more likely to gain 10 letters after 4 months, even if they showed a suboptimal response (9 letter change) after 2 months (OR, 3.83; 95% CI, 1.47e10; P ¼ 0.01). Likewise, the absence of HRF at baseline predicted increased odds to gain 10 letters after 4 months, when the response was suboptimal (9 letter change) after 2 months (OR, 5.33; 95% CI, 1.81e15.72; P ¼ 0.002). All OCT biomarkers that were predictive of good treatment response are shown in Figure 2. Eyes in the refractory group were more likely to have a suboptimal response after 2 months than were naïve cases (OR, 3.03; 95% CI, 1.30e7.09; P ¼ 0.01).
4
Higher HbA1c levels were associated with poorer baseline BCVA (P ¼ 0.01), higher baseline CMT (P ¼ 0.03), and a greater decrease in CMT after 4 months (P ¼ 0.007). However, HbA1c levels did not correlate with visual outcome after DEX implant (P ¼ 0.84).
Discussion We identified the presence of SRF, IS/OS continuity, absent HRF, and an attached vitreoretinal interface as biomarkers that predict better visual outcome after DEX implants in eyes with DME. Other studies reported that retinal thickness may serve as a measurement variable in association with treatment outcome in DME.18e26 However, rather than addressing the complexity of structural changes, their approach considered a compound of various retinal and subretinal compartments. In contrast, the systematic analysis used in the current study qualitatively and quantitatively analyzed SD OCT characteristics in DME, while accounting for subtle changes that could provide important information. Although serous macular detachment (SMD) may be difficult to diagnose by slit-lamp biomicroscopy or angiography, SD OCT enables detection of even very subtle SMD. With the use of SD OCT, SMD incidence in DME
Zur et al
OCT Biomarkers for DME with DEX Implant
Table 3. Baseline Predictors of Best-Corrected Visual Acuity Outcome at 2 Months
Baseline Measure Baseline BCVA (in logMAR SD)y Naïve patients Refractory patients Present SRF Absent SRF Present ONL cysts Absent ONL cysts ONL cyst size Small Large Giant Present INL cysts Absent INL cysts IS-OS continuity Completely continuous Partly disrupted Completely disrupted Present HRF Absent HRF HRF quantity Few Moderate Many HRF location ILM-INL OPL-ONL All layers VMI Detached VMA VMT Present ERM Absent ERM
BCVA Gain ‡10 Letters (Total [ 135), n (%)
BCVA Gain <9 Letters, BCVA Loss <9 Letters (Total [ 146), n (%)
BCVA Loss ‡10 Letters (Total [ 21), n (%)
P Value
OR (95% CI)*
0.730.34
0.590.27
0.500.22
<0.001
0.74 (0.67e0.81)
99/209 (47) 36/90 (40) 63/134 (47) 72/165 (44) 116/249 (47) 19/50 (38)
97/209 (46) 46/90 (51) 65/134 (49) 78/165 (47) 114/249 (46) 29/50 (58)
13/209 (6) 8/90 (9) 6/134 (5) 15/165 (9) 19/249 (8) 2/50 (4)
0.62
0.88 (0.54e1.45)
0.43
0.83 (0.53e1.31)
0.92
0.97 (0.54e1.75)
0.90
1.02 (0.73e1.44)
14/32 (44) 22/52 (42) 80/165 (49) 60/143 (42) 73/154 (47)
17/32 (53) 25/52 (48) 72/165 (44) 72/143 (50) 71/154 (46)
1/32 (3) 5/52 (10) 13/165 (8) 11/143 (8) 10/154 (7)
0.58
1.14 (0.72e1.79)
0.01
0.58 (0.38e0.89)
48/100 (48) 69/151 (46) 15/42 (36) 122/271 (45) 13/28 (46)
45/100 (45) 69/151 (46) 26/42 (62) 129/271 (48) 14/28 (50)
7/100 (7) 13/151 (9) 1/42 (2) 20/271 (7) 1/28 (4)
0.66
1.21 (0.52e2.80)
0.16
0.80 (0.59e1.09)
49/108 (45) 43/91 (47) 30/72 (42)
51/108 (51) 41/91 (45) 37/72 (51)
8/108 (7) 7/91 (7) 5/72 (7) 0.004
0.57 (0.38e0.83)
10/19 (53) 30/58 (52) 69/169 (41)
9/19 (47) 23/58 (40) 86/169 (51)
0/19 (0) 5/58 (9) 14/169 (8) 0.02
2.14 (1.15e3.97)
110/250 (44) 18/40 (45) 7/9 (78) 38/87 (44) 97/212 (46)
124/250 (50) 19/40 (48) 0/9 (0) 40/87 (46) 103/212 (49)
16/250 (6) 3/40 (8) 2/9 (22) 9/87 (10) 12/212 (6)
0.66
1.13 (0.66e1.95)
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; ERM ¼ epiretinal membrane; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; logMAR ¼ logarithm of the minimum angle of resolution; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; SD ¼ standard deviation; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMI ¼ vitreomacular interface; VMT ¼ vitreomacular traction. *Odds ratio (OR) for a patient presenting with gain of 10 letters in best corrected visual acuity (BCVA) at 2 months when baseline measure is increased by 1 value. y For every additional 5 letters (0.1 logMAR scale) of baseline BCVA a patient was less likely to present with gain of 10 letters in BCVA at 2 months.
was shown to be much higher than previously supposed.27 However, the pathogenesis of submacular fluid in DME remains poorly defined. Apparently, higher levels of vascular leakage from the macular vasculature play a major role in its development and persistence, in addition to vitreomacular traction.28,29 According to previous reports, the presence of SMD does not seem to correlate with VA in DME.30 The predictive value of SMD at baseline for treatment response to anti-VEGF agents in DME is controversial. Although some studies reported significant improvement in VA when SRF was present at baseline,20,21 others found no difference or even an association with worse functional results.22e24 In a post hoc subanalysis of the RISE and RIDE studies, the presence of submacular fluid predicted excellent visual outcomes in patients treated with ranibizumab.20 However, in sham-treated patients,
submacular fluid was associated with poor visual results and a 4-fold risk of significant vision loss. These findings suggest that persistent SMD may have deleterious effects on visual function. In our study, we did not follow morphologic parameters after DEX treatment and therefore have no information regarding possible cases of persistent submacular fluid or its impact on VA. One small retrospective study investigated the efficacy of DEX implants in the treatment of DME resistant to ranibizumab treatment, according to different morphologic patterns.31 Although no differences between the groups were found at the primary end point after 6 months, a subgroup analysis in the first 4 months revealed better VA outcomes for eyes that presented with SRF at baseline. Our results support this finding and are congruent with recent reports of significantly different aqueous cytokine levels associated with different
5
Ophthalmology Volume -, Number -, Month 2017 Table 4. Baseline Predictors of Best-Corrected Visual Acuity Outcome at 4 Months
Baseline Measure Baseline BCVA (in logMAR SD)y Naïve patients Refractory patients Present SRF Absent SRF Present ONL cysts Absent ONL cysts ONL cyst size Small Large Giant Present INL cysts Absent INL cysts IS-OS continuity Completely continuous Partly disrupted Completely disrupted Present HRF Absent HRF HRF quantity Few Moderate Many HRF location ILM-INL OPL-ONL All layers VMI Detached VMA VMT Present ERM Absent ERM
BCVA Gain ‡10 Letters (Total [ 135), n (%)
BCVA Gain <9 Letters, BCVA loss <9 Letters (Total [ 146), n (%)
BCVA Loss ‡10 Letters (Total [ 21), n (%)
P Value
OR (95% CI)*
0.710.34
0.590.28
0.620.29
0.001
0.88 (0.82e0.95)
103/209 (49) 27/90 (30) 71/134 (53) 59/165 (36) 104/249 (42) 26/50 (52)
92/209 (44) 56/90 (62) 55/134 (41) 93/165 (56) 125/249 (50) 23/50 (46)
14/209 (7) 7/90 (8) 8/134 (6) 13/165 (8) 20/249 (8) 1/50 (2)
0.13
0.68 (0.41e1.12)
0.03
0.60 (0.38e0.96)
0.54
1.24 (0.62e2.48)
0.20
1.26 (0.89e1.78)
12/32 (38) 18/52 (35) 74/165 (45) 57/143 (40) 72/154 (47)
17/32 (53) 29/52 (56) 79/165 (48) 77/143 (54) 70/154 (46)
3/32 (9) 5/52 (10) 12/165 (7) 9/143 (6) 12/154 (8)
0.86
1.04 (0.65e1.68)
0.11
0.75 (0.53e1.07)
49/100 (49) 62/151 (41) 17/42 (41) 110/271 (41) 20/28 (71)
46/100 (46) 74/151 (49) 24/42 (57) 142/271 (52) 6/28 (21)
5/100 (5) 15/151 (10) 1/42 (2) 19/271 (7) 1/28 (7)
0.01
3.43 (1.32e8.95)
0.42
0.88 (0.65e1.20)
42/108 (39) 40/91 (44) 30/72 (39)
57/108 (53) 47/91 (52) 37/72 (53)
9/108 (8) 4/91 (4) 5/72 (8) 0.04
0.64 (0.42e0.98)
10/19 (53) 24/58 (41) 58/169 (34)
8/19 (42) 30/58 (52) 97/169 (57)
1/19 (5) 4/58 (7) 14/169 (8) 0.60
1.18 (0.63e2.20)
109/250 (44) 16/40 (40) 5/9 (56) 34/87 (39) 96/212 (45)
125/250 (50) 23/40 (58) 0/9 (0) 43/87 (49) 105/212 (50)
16/250 (6) 1/40 (2) 4/9 (44) 10/87 (12) 11/212 (5)
0.55
1.18 (0.70e1.99)
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; ERM ¼ epiretinal membrane; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; logMAR ¼ logarithm of the minimum angle of resolution; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; SD ¼ standard deviation; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMI ¼ vitreomacular interface; VMT ¼ vitreomacular traction. *Odd ratio (OR) for a patient presenting with gain of 10 letters in best corrected visual acuity (BCVA) at 4 months when baseline measure is increased by 1 value. y For every additional 5 letters (0.1 logMAR scale) of baseline BCVA a patient was less likely to present with gain of 10 letters in BCVA at 4 months.
morphologic DME patterns.32 Interleukin-6 levels were significantly higher in eyes with SMD than in eyes with solely diffuse or cystoid macular thickening; the latter implies active inflammation. This might explain the better response to DEX implant as an anti-inflammatory drug in eyes with SMD. In DME, cystoid spaces can be related to specific intraretinal layers on SD OCT. The inner and outer plexiform layers were reported to present physical resistance to fluid movements.33 Changes in DME are generally not confined solely to 1 layer, but rather involve several layers simultaneously.34 In our study, 83.3% of all patients presented with ONL cysts; of them, the majority had giant ONL cysts (>200 mm), which mostly occur at a relatively late stage of the disease. Previous studies reported that large ONL cysts have a negative impact on macular
6
function (measured by microperimetry) and predict worse VA outcome after anti-VEGF treatment.35,36 Cystoid changes beneath the outer plexiform layer have been shown to present risk to photoreceptor cells and to impair IS-OS integrity because of the major and irreversible impact on central visual functions.37 In our study, neither the size nor the location of intraretinal cysts altered functional outcome. Apparently, more than the absolute cyst size and volume, the remaining tissue between cysts in the central macula seems to be crucial for good VA.26 We did not analyze this parameter, and further investigation might provide more information about its impact. We did not find the presence of posterior vitreous detachment to be associated with visual outcome after DEX implants. Immunohistologic examinations of the internal limiting excised from eyes with DME indicate proliferative
Zur et al
OCT Biomarkers for DME with DEX Implant
Figure 2. Biomarkers of optical coherence tomography (OCT). Forest plot showing baseline predictors of a 10-letter gain in best-corrected visual acuity (BCVA) at 2 (-) and 4 (:) months in patients with diabetic macular edema (DME) treated with dexamethasone (DEX) implants. Subretinal fluid (SRF): Eyes presenting with SRF were more likely to gain BCVA of 10 letters at 4 months. The inner segment-outer segment (IS/OS) continuity: For every increase in an IS/OS continuity value (completely continuous, partly disrupted, completely disrupted), a patient was less likely to gain BCVA of 10 letters at 2 months. Location of hyperreflective foci (HRF): For every increase in an HRF location value (internal limiting membrane [INL], outer plexiform layer [ONL], all layers), a patient was less likely to gain BCVA of 10 letters at 2 and 4 months. Vitreoretinal interface (VR): For every increase in vitreoretinal interface values (detached, vitreomacular adhesion, vitreomacular traction), a patient was more likely to gain BCVA of 10 letters at 4 months. Error bars presenting logarithmic transformation of 95% confidence intervals (CIs). ILM ¼ internal limiting membrane; OPL ¼ outer plexiform layer.
changes at the vitreomacular interface. Still, the question remains as to whether these are the cause or consequence of DME. Diabetic macular edema associated with posterior hyaloid traction is associated with worse vision.38 Our results support the findings of a subanalysis of the Ranibizumab for Edema of the Macula in Diabetes trial, which showed that the presence of vitreomacular adhesion itself does not preclude worse functional outcome in eyes with DME treated with ranibizumab.39 Generally, cases of vitreomacular traction are excluded from major clinical trials, even though DME is associated with this condition in 28%.40 We did not find any information in the literature regarding the impact of vitreomacular interface configuration on treatment outcome after DEX implant in DME. Hyperreflective foci correlates with lipoprotein extravasation after breakdown of the inner blooderetinal barrier in the initial stages of the development of intraretinal hard exudates.41 The predictive value of HRF on visual outcome after anti-VEGF treatment in DME is unclear.42,43 In the current study, the majority of eyes presented with HRF at baseline (i.e., 91%). We found better visual outcome when HRF was absent at baseline. Poor diabetes control was associated with worse baseline VA. However, there was no association between HbA1C values and improvement in VA. Likewise, improvement in VA, anatomic reduction of macular edema, and improvement in diabetic retinopathy severity score with ranibizumab treatment seem to be independent of baseline HbA1c.44 According to the VIVID and VISTA trials, the benefit of intravitreal aflibercept injections in patients with DME
was less dependent on their presenting glycemic status, compared with laser treatmet.45 To our best knowledge, there are no data available regarding the impact of glycemic control on treatment outcome after intravitreal DEX treatment. In our study, we did not investigate the predictive value of other baseline factors, such as phakic status, baseline acuity, retinal thickness, and previous treatments before the DEX implant. Future prospective investigations of these baseline factors are required to confirm the predictive role of the OCT biomarkers we have identified. Although OCT images are readily available for physicians and provide a magnitude of detailed information, structural changes presumably reflect only part of the complex pathophysiologic processes occurring in DME. A recent study found no association between DME severity on OCT and systemic metabolic and inflammatory factors.46 Intraocular rather than systemic production of proinflammatory cytokines seems to affect the pathogenesis of both diabetic retinopathy and DME.47 Fluorescein angiographies were not considered; therefore, information regarding macular perfusion was not available. Macular ischemia might explain cases of nonresponse to DEX treatment. Recently, baseline choroidal thickness48 and disorganization of the inner retinal layers49 have been reported as predictive markers for visual outcome after DME treatment with anti-VEGF agents. We did not include those measures in the present study. This series demonstrates that certain biomarkers can help predict the response to DEX implant in DME. Future
7
Ophthalmology Volume -, Number -, Month 2017 prospective investigations should consider imaging and systemic biomarkers, and compare anti-VEGF drugs with DEX implants in a head-to-head trial aimed to tailor and optimize patient care.
References 1. Kempen JH, O’Colmain BJ, Leske MC, et al. The prevalence of diabetic retinopathy among adults in the United States. Arch Ophthalmol. 2004;122:552-563. 2. Yau JWY, Rogers SL, Kawasaki R, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556-564. 3. Massin P, Bandello F, Garweg JG, et al. Safety and efficacy of ranibizumab in diabetic macular edema (RESOLVE Study): a 12-month, randomized, controlled, double-masked, multicenter phase II study. Diabetes Care. 2010;33:2399-2405. 4. Mitchell P, Bandello F, Schmidt-Erfurth U, et al. The RESTORE study: ranibizumab monotherapy or combined with laser versus laser monotherapy for diabetic macular edema. Ophthalmology. 2011;118:615-625. 5. Michaelides M, Kaines A, Hamilton RD, et al. A prospective randomized trial of intravitreal bevacizumab or laser therapy in the management of diabetic macular edema (BOLT study) 12month data: report 2. Ophthalmology. 2010;117:1078-1086.e2. 6. Do DV, Nguyen QD, Boyer D, et al. One-year outcomes of the da Vinci Study of VEGF Trap-Eye in eyes with diabetic macular edema. Ophthalmology. 2012;119:1658-1665. 7. Rajendram R, Fraser-Bell S, Kaines A, et al. A 2-year prospective randomized controlled trial of intravitreal bevacizumab or laser therapy (BOLT) in the management of diabetic macular edema: 24-month data: report 3. Arch Ophthalmol. 2012;130:972-979. 8. Elman MJ, Bressler NM, Qin H, et al. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2011;118:609-614. 9. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012;119:789-801. 10. Gonzalez VH, Campbell J, Holekamp NM, et al. Early and long-term responses to anti-vascular endothelial growth factor therapy in diabetic macular edema: analysis of protocol I data. Am J Ophthalmol. 2016;172:72-79. 11. Paper O, Guigou S, Pommier S, et al. Efficacy and safety of intravitreal dexamethasone implant in patients with diabetic macular edema. Ophthalmologica. 2015;233:169-175. 12. Boyer DS, Yoon YH, Belfort R, et al. Three-year, randomized, sham-controlled trial of dexamethasone intravitreal implant in patients with diabetic macular edema. Ophthalmology. 2014;121:1904-1914. 13. Alshahrani ST, Dolz-Marco R, Gallego-Pinazo R, et al. Intravitreal dexamethasone implant for the treatment of refractory macular edema in retina vascular diseases: results of the KKESH International Collaborative Retina Study Group. Retina. 2016;36:131-136. 14. Dutra Medeiros M, Postorino M, Navarro R, et al. Dexamethasone intravitreal implant for treatment of patients with persistent diabetic macular edema. Ophthalmologica. 2014;231:141-146. 15. Pacella F, Romano MR, Turchetti P, et al. An eighteen-month follow-up study on the effects of intravitreal dexamethasone implant in diabetic macular edema refractory to anti-VEGF therapy. Int J Ophthalmol. 2016;9:1427-1432.
8
16. Romero-Aroca P. Targeting the pathophysiology of diabetic macular edema. Diabetes Care. 2010;33:2484-2485. 17. Murakami T, Yoshimura N. Structural changes in individual retinal layers in diabetic macular edema. J Diabetes Res. 2013;2013:920713. 18. Browning DJ, Glassman AR, Aiello LP, et al. Relationship between optical coherence tomography-measured central retinal thickness and visual acuity in diabetic macular edema. Ophthalmology. 2007;114:525-536. 19. Bressler SB, Qin H, Beck RW, et al. Factors associated with changes in visual acuity and central subfield thickness at 1 year after treatment for diabetic macular edema with ranibizumab. Arch Ophthalmol. 2012;130:1153-1161. 20. Sophie R, Lu N, Campochiaro PA. Predictors of functional and anatomic outcomes in patients with diabetic macular edema treated with ranibizumab. Ophthalmology. 2015;122: 1395-1401. 21. Fickweiler W, Hooymans JMM, Los LI, et al. Predictive value of optical coherence tomographic features in the Bevacizumab and Ranibizumab in Patients with Diabetic Macular Edema (BRDME) Study. Retina. 2017. http://dx.doi.org/10.1097/IAE. 0000000000001626 [Epub ahead of print]. 22. Shimura M, Yasuda K, Yasuda M, Nakazawa T. Visual outcome after intravitreal bevacizumab depends on the optical coherence tomographic patterns of patients with diffuse diabetic macular edema. Retina. 2013;33:740-747. 23. Giocanti-aur A, Hrarat L, Qu LM, et al. Functional and anatomical outcomes in patients with serous retinal detachment in diabetic macular edema treated with ranibizumab. Invest Ophthalmol Vis Sci. 2017:797-800. 24. Seo KH, Yu S-Y, Kim M, Kwak HW. Visual and morphologic outcomes of intravitreal ranibizumab for diabetic macular edema based on optical coherence tomography patterns. Retina. 2016;36:588-595. 25. Rayess N, Rahimy E, Ying GS, et al. Baseline choroidal thickness as a predictor for response to anti-vascular endothelial growth factor therapy in diabetic macular edema. Am J Ophthalmol. 2015;159:85-91. 26. Pelosini L, Hull CC, Boyce JF, et al. Optical coherence tomography may be used to predict visual acuity in patients with macular edema. Invest Ophthalmol Vis Sci. 2011;52:2741-2748. 27. Ozdemir H, Karacorlu M, Karacorlu S. Serous macular detachment in diabetic cystoid macular oedema. Acta Ophthalmol Scand. 2005;83:63-66. 28. Kang SW, Park CY, Ham D-I. The correlation between fluorescein angiographic and optical coherence tomographic features in clinically significant diabetic macular edema. Am J Ophthalmol. 2004;137:313-322. 29. Kaiser PK, Riemann CD, Sears JE, Lewis H. Macular traction detachment and diabetic macular edema associated with posterior hyaloidal traction. Am J Ophthalmol. 2001;131:44-49. 30. Alasil T, Keane PA, Updike JF, Dustin L. Relationship between optical coherence tomography retinal parameters and visual acuity in diabetic macular edema. Ophthalmology. 2010;117:2379-2386. 31. Kaldirim H, Yazgan S, Atalay K, et al. Intravitreal dexamethasone implantation in patients with different morphological diabetic macular edema having insufficient response to ranibizumab. Retina. 2017. http://dx.doi.org/10.1097/IAE. 0000000000001648 [Epub ahead of print]. 32. Bandyopadhyay S. Study of aqueous cytokines in patients with different patterns of diabetic macular edema based on optical coherence tomography. Int Ophthalmol. 2017. http://dx.doi. org/10.1007/s10792-017-0453-2 [Epub ahead of print]. 33. Antcliff RJ, Hussain AA, Marshall J. Hydraulic conductivity of fixed retinal tissue after sequential excimer laser ablation:
Zur et al
34. 35. 36.
37.
38. 39.
40.
41.
OCT Biomarkers for DME with DEX Implant
barriers limiting fluid distribution and implications for cystoid macular edema. Arch Ophthalmol. 2001;119:539-544. Bolz M, Ritter M, Schneider M, et al. A systematic correlation of angiography and high-resolution optical coherence tomography in diabetic macular edema. Ophthalmology. 2009;116:66-72. Deák GG, Bolz M, Ritter M, et al. A systematic correlation between morphology and functional alterations in diabetic macular edema. Invest Ophthalmol Vis Sci. 2010;51:6710-6714. Reznicek L, Cserhati S, Seidensticker F, et al. Functional and morphological changes in diabetic macular edema over the course of anti-vascular endothelial growth factor treatment. Acta Ophthalmol. 2013;91:529-536. Murakami T, Nishijima K, Akagi T, et al. Optical coherence tomographic reflectivity of photoreceptors beneath cystoid spaces in diabetic macular edema. Investig Ophthalmol Vis Sci. 2012;53:1506-1511. Agarwal D, Gelman R, Prospero Ponce C, et al. The vitreomacular interface in diabetic retinopathy. J Ophthalmol. 2015;2015:392983. Sadiq MA, Soliman MK, Sarwar S, et al. Effect of vitreomacular adhesion on treatment outcomes in the Ranibizumab for Edema of the Macula in Diabetes (READ-3) Study. Ophthalmology. 2016;123:324-329. Jackson TL, Nicod E, Ms C, et al. Vitreous attachment in agerelated macular degeneration, diabetic macular edema, and retinal vein occlusion: a systematic review and metaanalysis. Retina. 2013;33:1099-1108. Bolz M, Schmidt-Erfurth U, Deak G, et al. Optical coherence tomographic hyperreflective foci. a morphologic sign of lipid extravasation in diabetic macular edema. Ophthalmology. 2009;116:914-920.
42. Framme C, Schweizer P, Imesch M, et al. Behavior of SDOCT - detected hyperreflective foci in the retina of antiVEGFetreated patients with diabetic macular edema. Invest Ophthalmol Vis Sci. 2017;53:5814-5818. 43. Kang J, Chung H, Kim HC. Correlation of optical coherence foci with visual acuity outcomes in different patterns of diabetic macular edema. Retina. 2016;369:1630-1639. 44. Bansal AS, Khurana RN, Wieland MR, et al. Influence of glycosylated hemoglobin on the efficacy of ranibizumab for diabetic macular edema a post hoc analysis of the RIDE/RISE Trials. Ophthalmology. 2015;122:1573-1579. 45. Singh R, Wykoff CC, Brown DM, et al. Outcomes of diabetic macular edema patients by baseline hemoglobin A1c: analyses from VISTA and VIVID. Ophthalmol Retina. 2017 [Epub ahead of print]. 46. Figueras-roca M, Molins B, Sala-puigdollers A, et al. Peripheral blood metabolic and inflammatory factors as biomarkers to ocular findings in diabetic macular edema. PLoS One. 2017;12:e0173865. 47. Vujosevic S, Simó R. Local and systemic inflammatory biomarkers of diabetic retinopathy: an integrative approach. Invest Ophthalmol Vis Sci. 2017;58:68-75. 48. Rayess N, Rahimy E, Ying G, et al. Baseline choroidal thickness as a predictor for response to anti-vascular endothelial growth factor therapy in diabetic macular edema. Am J Ophthalmol. 2015;159:85-91.e3. 49. Sun JK, Lin MM, Lammer J, et al. Disorganization of the retinal inner layers as a predictor of visual acuity in eyes with center-involved diabetic macular edema. JAMA Ophthalmol. 2014;132:1309-1316.
Footnotes and Financial Disclosures Originally received: June 25, 2017. Final revision: August 21, 2017. Accepted: August 22, 2017. Available online: ---.
Author Contributions:
Manuscript no. 2017-1461.
1
Division of Ophthalmology, Tel Aviv Sourasky Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. 2 University of Buenos Aires, Argentina. 3
Department of Ophthalmology, University of Leipzig, Germany.
Eye Clinic - Department of Biomedical and Clinical Science “L. Sacco,” Luigi Sacco Hospital, University of Milan, Milan, Italy.
4
5
Favaloro University Hospital, Buenos Aires, Argentina.
6
Incumbent, Sydney A. Fox Chair in Ophthalmology, Tel Aviv University, Tel Aviv, Israel. Presented: at the Annual Meeting of the American Society of Retinal Specialists, Boston, Massachusetts, August 11e15, 2017.
z
Both Dinah Zur, MD, and Matias Iglicki, MD were equal contributors.
Financial Disclosure(s): The author(s) have made the following disclosure(s): A.L.: Consultant Allergan. The International Retina Group is an independent research branch of the International Retina Panel, which receives travel support by Allergan.
Conception and design: Zur, Iglicki, Busch, Lowenstein Data collection: Zur, Iglicki, Busch, Lowenstein, Invernizzi, Cebecci, Chhablani Analysis and interpretation: Zur, Iglicki, Busch, Lowenstein, Invernizzi Obtained funding: Not applicable Overall responsibility: Zur, Iglicki, Busch, Lowenstein, Invernizzi, Mariussi Abbreviations and Acronyms: BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; CMT ¼ central macular thickness; DEX ¼ dexamethasone; DME ¼ diabetic macular edema; GEE ¼ generalized estimating equation; HbA1C ¼ hemoglobin A1C; HRF ¼ hyperreflective foci; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; OCT ¼ optical coherence tomography; ONL ¼ outer nuclear layer; OR ¼ odds ratio; SD ¼ spectral domain; SMD ¼ serous macular detachment; SRF ¼ subretinal fluid; VA ¼ visual acuity; VEGF ¼ vascular endothelial growth factor. Correspondence: Dinah Zur, MD, Division of Ophthalmology, Tel Aviv Sourasky Medical Center, Weizman 6, Tel Aviv 64239. E-mail: [email protected].
Members of the International Retina Group are listed in the Supplementary Material (available online at www.aaojournal.org).
9
Vision loss associated with diabetic retinopathy is most commonly caused by diabetic macular edema (DME),1 which affects approximately 7% of all diabetic patients.2 Antivascular endothelial growth factor (VEGF) injections are generally considered suitable first-line therapy for center-involved DME and are effective in improving visual acuity (VA), with 10% to 40% of patients achieving significant improvement in best-corrected visual acuity (BCVA) after 1 year of treatment.3e6 Still, a considerable proportion of patients do not respond satisfactorily to antiVEGF agents, even with intensive treatment over the first year.7e9 A recent post hoc analysis of the Diabetic Retinopathy Clinical Research Network’s (DRCR.net) Protocol I study revealed that approximately 40% of eyes gain ª 2017 by the American Academy of Ophthalmology Published by Elsevier Inc.
<5 letters after 3 months and approximately 50% of these maintain poorer long-term visual outcomes after 3 years than eyes with favorable early response.10 Thus, early identification and characterization of patients with DME are critical, as well as the provision of individualized treatment regarding optimal functional outcome and disease management. The efficacy of dexamethasone (DEX) intravitreal implant 0.7 mg (Ozurdex, Allergan, Inc., Irvine, CA) in DME has been shown by improvement in VA and decrease in retinal thickness,11,12 even in patients with DME that is refractory to anti-VEGF treatments.13e15 Corticosteroids and anti-VEGF treatments target different pathways in the pathogenesis of DME.16
http://dx.doi.org/10.1016/j.ophtha.2017.08.031 ISSN 0161-6420/17
1
Ophthalmology Volume -, Number -, Month 2017 Anatomic measures on spectral-domain (SD) optical coherence tomography (OCT), such as precise evaluation of individual layers, quantification of retinal thickness and macular volume, and qualitative assessment of fluid distribution and existence of hyperreflective foci (HRF),17 could predict treatment success or failure to various therapies. Baseline OCT measures have been investigated regarding their predictive value in patients with DME treated by antiVEGF therapy.18e25 We hypothesized that distinct structural changes identifiable on SD OCT could help predict treatment responses to DEX implants, distinct from findings in eyes treated with anti-VEGF treatments. The purpose of this study was to investigate whether characteristics identified on SD OCT may serve as biomarkers and predict treatment response to DEX implants in patients with DME.
Methods Institutional review board approval was obtained through the individual institutional review boards at the participating institutes for a retrospective consecutive chart review. This international multicenter study included 14 centers (Supplementary Material, available at www.aaojournal.org). The research adhered to the tenets of the Declaration of Helsinki. Patient records from January 1, 2011, to August 1, 2016, were reviewed for cases of DME treated by intravitreal DEX implant.
Study Participants To be included in the analysis, patients had to fulfill the following criteria: (1) age 18 years; (2) type 1 or 2 diabetes mellitus; (3) DME (both naïve and refractory) causing visual loss, with study eye BCVA measuring 0.3 to 1.0 logarithm of the minimum angle of resolution; macular edema defined clinically and by retinal thickness of >250 mm in the central subfield; and intraretinal or subretinal fluid (SRF) seen on SD OCT; (4) treatment with DEX implant. For patients who received bilateral treatment with DEX, both eyes were included. Refractory DME was defined as worsening of BCVA by 2 Early Treatment Diabetic Retinopathy Study lines or reduction of less than 10% of retinal thickness on SD OCT measured 1 month after at least 3 anti-VEGF injections that were given at monthly intervals.10 Only first treatments with DEX implants were considered for the study. Exclusion criteria were (1) another concomitant ocular disease that causes macular edema (i.e., neovascular age-related macular degeneration or choroidal neovascularization due to other reasons, retinal vein occlusion, uveitis, and recent intraocular surgery possibly causing postsurgical macular edema); (2) another ocular condition that compromises VA, except for the presence of cataract; and (3) previous treatment with intraocular corticosteroids within the 6 months before treatment with the DEX implant. Patient charts were reviewed for demographic data, hemoglobin A1C (HbA1C) values, type of retinopathy (nonproliferative or proliferative), previous treatments for DME, and BCVA before the DEX implant and at 1, 2 and 4 months after the injection.
Optical Coherence Tomography Analysis The OCT scans were obtained using SD OCT: Heidelberg Spectralis, Heidelberg, Germany; Optovue Avanti, Fremont, California; Topcon 3D OCT-2000, Tokyo, Japan; and Cirrus, Zeiss, Oberkochen, Germany. Quantitative assessment of DME included central macular thickness (CMT) that was calculated automatically by the instrument and recorded at baseline and at 1, 2, and 4 months after the DEX
2
implant. Qualitative and quantitative evaluations of SD OCT images performed at baseline assessed the presence of several morphologic features (Fig 1), including (1) SRF; (2) cystoid changes in the outer nuclear layer (ONL) and maximal cyst size in the ONL (small <100 mm, large 100e200 mm, 3 ¼ giant >200 mm); (3) presence of cystoid changes in the inner nuclear layer (INL); (4) continuity of the inner segment-outer segment (IS-OS) layer (completely continuous, partly disrupted, completely disrupted); (5) presence of HRF, as well as quantity (few 2e10, moderate 11e20, many >21) and location (between the internal limiting membrane and the INL; between the outer plexiform layer and external limiting membrane; in all retinal layers); (6) status of the vitreomacular interface (detached, vitreomacular adhesion, vitreomacular traction); and (7) presence of an epiretinal membrane. The listed features were evaluated on 3 horizontal OCT scans: 1 b-scan encompassing the fovea, 2 b-scans respectively 500 mm superior and 500 mm inferior to the fovea. Grading of the OCT images was performed by 3 experienced retina specialists (DZ: 109 cases, MI: 100 cases, AI: 90 cases) who were blinded to the functional and anatomic results. Interrater reliability was calculated among 90 cases that were graded by all graders. The CMT was recorded at baseline and at 1, 2, and 4 months after the DEX implant.
Statistical Analysis Interrater reliability was tested as absolute agreement with a 2-way mixed interclass correlation model. To control for the correlated nature of our data, we used a generalized estimating equation (GEE) procedure. Differences in functional treatment response and OCT baseline measures between naïve and refractory patients were calculated by a GEE model and included baseline BCVA as a covariate. Differences in anatomic outcomes were calculated and include baseline CMT as a covariate. The GEE models for outcome measures (BCVA response at 2 and 4 months) were run by testing the following predictors at baseline: (1) BCVA; (2) naïve or refractory DME; (3) the presence of SRF; (4) the presence of ONL cysts; (5) ONL cyst size; (6) presence of INL cysts; (7) IS/OS continuity; (8) HRF presence; (9) HRF quantity; (10) HRF location; (11) vitreoretinal interface; and (12) presence of an epiretinal membrane. Predictors were included stepwise in the GEE model and kept in the model if the P value was less than 0.10. The final GEE model was used to calculate the odds ratios (ORs) and their 95% confidence intervals (CIs), with a change of 5 letters in baseline BCVA considered a standard unit of change. An association of HbA1c levels with baseline and response parameters was examined using a linear GEE model. For BCVA and CMT response after 4 months, baseline levels were included as covariates. The last observation carried forward method was used to impute missing data. Values are presented as mean standard deviation. Statistical analysis was performed with SPSS Statistics 22 (IBM, Armonk, NY).
Results Study Participants A total of 299 eyes from 284 patients were included in the analysis. Demographic data are shown in Table 1. In 209 eyes (70%), DME was naïve; 90 cases (30%) were refractory to previous anti-VEGF injections. The mean number of previous anti-VEGF injections was 7.78.8. A total of 104 eyes (35%) were previously treated by macular laser. A total of 169 eyes (57%) were phakic, and 130 eyes (43%) were pseudophakic. HbA1c levels were available for 180 patients; the mean value was 8.4%2.7%.
Zur et al
OCT Biomarkers for DME with DEX Implant
Figure 1. Optical coherence tomography (OCT) measures. A, Serous macular detachment (SMD) with cystoid diabetic macular edema (DME), showing retinal elevation with an optically clear space between the sensory retina and the retinal pigment epithelium (*). B, Grading of intraretinal cysts: Cystoid DME with a giant outer nuclear layer (ONL) cyst (*), as well as cysts in the inner nuclear layer (INL) (arrow). C, Diabetic macular edema associated with focal vitreomacular traction. DeF, Grading of hyperreflective foci (HRF). D, Few (<10) HRF between the external limiting membrane and the outer plexiform layer. E, Diabetic macular edema with a moderate amount (10e20) of HRF located between the INL and the internal limiting membrane. F, Diabetic macular edema with many (>20) HRF scattered in all retinal layers. GeI, Grading of the inner segment-outer segment (IS-OS) integrity. G, The IS-OS layer is regular and continuous. H, Partial irregularity and lack of continuity of the IS-OS layer (between arrows). I, Complete discontinuity of the ISOS layer (between arrows). ELM ¼ external limiting membrane; ILM ¼ internal limiting membrane; OPL ¼ outer plexiform layer.
Anatomic and Best-Corrected Visual Acuity Baseline Characteristics The OCT baseline characteristics are shown in Table 2. Mean BCVA was decreased at baseline for naïve compared with refractory patients (0.680.32 vs. 0.580.27 [logarithm of the
Table 1. Descriptive Statistics: Demographic Data Age (yrs), mean SD Gender male, n (%) Nonproliferative DR, n (%) Proliferative DR, n (%) Naïve DME, n (%) Refractory DME, n (%) No. of previous anti-VEGF injections, mean SD Previous macular photocoagulation, n (%) Lens status* Phakic, n (%) Pseudophakic, n (%) HbA1c (%), mean SDy
6414 171 (57) 151 (51) 148 (49) 209 (70) 90 (30) 7.78.8 104 (35) 169 (57) 122 (410.8) 8.42.7
DME ¼ diabetic macular edema; DR ¼ diabetes retinopathy; HbA1c ¼ hemoglobin A1c; SD ¼ standard deviation; VEGF ¼ vascular endothelial growth factor. *Missing data for 8 patients (2.7%). y Available for 180 patients.
minimum angle of resolution], P ¼ 0.006). However, CMT did not differ significantly between the groups (508141 vs. 535142, P ¼ 0.13). Furthermore, naïve patients were more likely to present with SRF (OR, 3.57; 95% CI, 2.04e6.26; P < 0.001), smaller ONL cysts (OR, 1.88; 95% CI, 1.25e2.82; P ¼ 0.002), no INL cysts (OR, 3.75; 95% CI, 2.19e6.44; P < 0.001), and no HRF (OR, 7.72; 95% CI, 1.81e32.85; P ¼ 0.01). Interrater reliability among the graders was 0.89 (95% CI, 0.84e0.93).
Functional and Anatomic Outcome At 4 months, 130 eyes (43.5%) showed a 10-letter improvement in BCVA, 148 eyes (49.5%) ranged within a 9-letter change from baseline, and 21 eyes (7%) lost 10 letters. The odds of gaining 10 letters after 2 and 4 months was increased when baseline BCVA was low (2 months: OR, 1.39; 95% CI, 1.25e1.56; P < 0.001; 4 months: OR, 1.15; 95% CI, 1.06e1.25; P ¼ 0.001). Eyes refractory to previous anti-VEGF injections were more likely to gain <10 letters or even lose BCVA after 4 months compared with naïve eyes (OR, 2.07; 95% CI, 1.21e3.54; P ¼ 0.01). At 4 months, CMT decreased by 20% in 204 eyes (68%) and by <20% in 95 eyes. Eyes refractory to previous anti-VEGF injections were half as likely to respond to DEX implants, as assessed by CMT (OR, 2.58; 95% CI, 1.49e4.50; P ¼ 0.001).
3
Ophthalmology Volume -, Number -, Month 2017 Table 2. Optical Coherence Tomography Baseline Measures on Spectral-Domain Optical Coherence Tomography Naïve, n (%) SRF Total, 135 (45.2) ONL cyst, n (%) Total, 249 (83.3) ONL cyst size Small Large Giant INL cysts Total, 144 (48.2) IS-OS continuity Completely continuous Partially continuous Completely disrupted HRF fociepresence Total, 271 (91) HRF foci - quantity Absent Few (2e10) Moderate (11e20) Many (>20) HRF foci - location ILM-INL, 19/247 OPL-ONL, 58/247 All layers, 170/247 Vitreomacular interface PVD VMA VMT
P Value
OR
95% CI
112 (53.4)
Refractory, n (%) 23 (25.6)
<0.001
3.57
2.04e6.26
160 (76.6)
89 (98.9)
<0.001
0.03
0.004e0.21
26 36 98 80
(16.3) (22.5) (61.3) (38.3)
6 16 67 63
(6.7) (18.0) (75.3) (70.8)
0.002
1.88
1.25e2.82
<0.001
0.27
0.16e0.46
73/203 94/203 36/203 183
(36.0) (46.3) (17.7) (87.6)
27 57 6 88
(30.0) (63.3) (6.7) (97.8)
0.95 0.01
1.01 0.32
0.70e1.45 0.13e0.75
26 71 64 48
(12.4) (34.0) (30.6) (23.0)
2 37 27 24
(2.2) (41.1) (30.0) (26.7)
0.02
1.37
1.06e1.78
9/158 (5.7) 37/158 (23.4) 112/158 (70.9)
10/88 (11.4) 21/88 (23.9) 58/88 (64.8)
0.25
0.79
0.53e1.18
183 (87.6) 19 (9.1) 7 (3.3)
67 (74.4) 21 (23.3) 2 (2.2)
0.09
1.57
0.94e2.66
CI ¼ confidence interval; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; PVD ¼ posterior vitreous detachment; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMT ¼ vitreomacular traction.
Optical Coherence Tomography Predictors for Treatment Response The presence of SRF at baseline was a predictor of good functional treatment response after 4 months (SRF: OR, 1.98; 95% CI, 1.23e3.20; P ¼ 0.01) (Fig 1). Moreover, eyes that did not present HRF at baseline were more likely to be good responders at 4 months (absent vs. present HRF: OR, 3.66; 95% CI, 1.40e9.62; P ¼ 0.01). Eyes with continuous IS-OS layers had better treatment response after 2 months (IS/OS continuity: OR, 2.09; 95% CI, 1.30e3.37; P ¼ 0.003). Figure 1 shows OCT biomarkers that were predictive of good BCVA response after 2 and 4 months. The predictive values of all the OCT measures examined are shown in Tables 3 and 4 for responses after 2 and 4 months, respectively. In a subanalysis, we investigated eyes that became good responders only after 4 months and not 2 months. Eyes with SRF at baseline were more likely to gain 10 letters after 4 months, even if they showed a suboptimal response (9 letter change) after 2 months (OR, 3.83; 95% CI, 1.47e10; P ¼ 0.01). Likewise, the absence of HRF at baseline predicted increased odds to gain 10 letters after 4 months, when the response was suboptimal (9 letter change) after 2 months (OR, 5.33; 95% CI, 1.81e15.72; P ¼ 0.002). All OCT biomarkers that were predictive of good treatment response are shown in Figure 2. Eyes in the refractory group were more likely to have a suboptimal response after 2 months than were naïve cases (OR, 3.03; 95% CI, 1.30e7.09; P ¼ 0.01).
4
Higher HbA1c levels were associated with poorer baseline BCVA (P ¼ 0.01), higher baseline CMT (P ¼ 0.03), and a greater decrease in CMT after 4 months (P ¼ 0.007). However, HbA1c levels did not correlate with visual outcome after DEX implant (P ¼ 0.84).
Discussion We identified the presence of SRF, IS/OS continuity, absent HRF, and an attached vitreoretinal interface as biomarkers that predict better visual outcome after DEX implants in eyes with DME. Other studies reported that retinal thickness may serve as a measurement variable in association with treatment outcome in DME.18e26 However, rather than addressing the complexity of structural changes, their approach considered a compound of various retinal and subretinal compartments. In contrast, the systematic analysis used in the current study qualitatively and quantitatively analyzed SD OCT characteristics in DME, while accounting for subtle changes that could provide important information. Although serous macular detachment (SMD) may be difficult to diagnose by slit-lamp biomicroscopy or angiography, SD OCT enables detection of even very subtle SMD. With the use of SD OCT, SMD incidence in DME
Zur et al
OCT Biomarkers for DME with DEX Implant
Table 3. Baseline Predictors of Best-Corrected Visual Acuity Outcome at 2 Months
Baseline Measure Baseline BCVA (in logMAR SD)y Naïve patients Refractory patients Present SRF Absent SRF Present ONL cysts Absent ONL cysts ONL cyst size Small Large Giant Present INL cysts Absent INL cysts IS-OS continuity Completely continuous Partly disrupted Completely disrupted Present HRF Absent HRF HRF quantity Few Moderate Many HRF location ILM-INL OPL-ONL All layers VMI Detached VMA VMT Present ERM Absent ERM
BCVA Gain ‡10 Letters (Total [ 135), n (%)
BCVA Gain <9 Letters, BCVA Loss <9 Letters (Total [ 146), n (%)
BCVA Loss ‡10 Letters (Total [ 21), n (%)
P Value
OR (95% CI)*
0.730.34
0.590.27
0.500.22
<0.001
0.74 (0.67e0.81)
99/209 (47) 36/90 (40) 63/134 (47) 72/165 (44) 116/249 (47) 19/50 (38)
97/209 (46) 46/90 (51) 65/134 (49) 78/165 (47) 114/249 (46) 29/50 (58)
13/209 (6) 8/90 (9) 6/134 (5) 15/165 (9) 19/249 (8) 2/50 (4)
0.62
0.88 (0.54e1.45)
0.43
0.83 (0.53e1.31)
0.92
0.97 (0.54e1.75)
0.90
1.02 (0.73e1.44)
14/32 (44) 22/52 (42) 80/165 (49) 60/143 (42) 73/154 (47)
17/32 (53) 25/52 (48) 72/165 (44) 72/143 (50) 71/154 (46)
1/32 (3) 5/52 (10) 13/165 (8) 11/143 (8) 10/154 (7)
0.58
1.14 (0.72e1.79)
0.01
0.58 (0.38e0.89)
48/100 (48) 69/151 (46) 15/42 (36) 122/271 (45) 13/28 (46)
45/100 (45) 69/151 (46) 26/42 (62) 129/271 (48) 14/28 (50)
7/100 (7) 13/151 (9) 1/42 (2) 20/271 (7) 1/28 (4)
0.66
1.21 (0.52e2.80)
0.16
0.80 (0.59e1.09)
49/108 (45) 43/91 (47) 30/72 (42)
51/108 (51) 41/91 (45) 37/72 (51)
8/108 (7) 7/91 (7) 5/72 (7) 0.004
0.57 (0.38e0.83)
10/19 (53) 30/58 (52) 69/169 (41)
9/19 (47) 23/58 (40) 86/169 (51)
0/19 (0) 5/58 (9) 14/169 (8) 0.02
2.14 (1.15e3.97)
110/250 (44) 18/40 (45) 7/9 (78) 38/87 (44) 97/212 (46)
124/250 (50) 19/40 (48) 0/9 (0) 40/87 (46) 103/212 (49)
16/250 (6) 3/40 (8) 2/9 (22) 9/87 (10) 12/212 (6)
0.66
1.13 (0.66e1.95)
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; ERM ¼ epiretinal membrane; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; logMAR ¼ logarithm of the minimum angle of resolution; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; SD ¼ standard deviation; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMI ¼ vitreomacular interface; VMT ¼ vitreomacular traction. *Odds ratio (OR) for a patient presenting with gain of 10 letters in best corrected visual acuity (BCVA) at 2 months when baseline measure is increased by 1 value. y For every additional 5 letters (0.1 logMAR scale) of baseline BCVA a patient was less likely to present with gain of 10 letters in BCVA at 2 months.
was shown to be much higher than previously supposed.27 However, the pathogenesis of submacular fluid in DME remains poorly defined. Apparently, higher levels of vascular leakage from the macular vasculature play a major role in its development and persistence, in addition to vitreomacular traction.28,29 According to previous reports, the presence of SMD does not seem to correlate with VA in DME.30 The predictive value of SMD at baseline for treatment response to anti-VEGF agents in DME is controversial. Although some studies reported significant improvement in VA when SRF was present at baseline,20,21 others found no difference or even an association with worse functional results.22e24 In a post hoc subanalysis of the RISE and RIDE studies, the presence of submacular fluid predicted excellent visual outcomes in patients treated with ranibizumab.20 However, in sham-treated patients,
submacular fluid was associated with poor visual results and a 4-fold risk of significant vision loss. These findings suggest that persistent SMD may have deleterious effects on visual function. In our study, we did not follow morphologic parameters after DEX treatment and therefore have no information regarding possible cases of persistent submacular fluid or its impact on VA. One small retrospective study investigated the efficacy of DEX implants in the treatment of DME resistant to ranibizumab treatment, according to different morphologic patterns.31 Although no differences between the groups were found at the primary end point after 6 months, a subgroup analysis in the first 4 months revealed better VA outcomes for eyes that presented with SRF at baseline. Our results support this finding and are congruent with recent reports of significantly different aqueous cytokine levels associated with different
5
Ophthalmology Volume -, Number -, Month 2017 Table 4. Baseline Predictors of Best-Corrected Visual Acuity Outcome at 4 Months
Baseline Measure Baseline BCVA (in logMAR SD)y Naïve patients Refractory patients Present SRF Absent SRF Present ONL cysts Absent ONL cysts ONL cyst size Small Large Giant Present INL cysts Absent INL cysts IS-OS continuity Completely continuous Partly disrupted Completely disrupted Present HRF Absent HRF HRF quantity Few Moderate Many HRF location ILM-INL OPL-ONL All layers VMI Detached VMA VMT Present ERM Absent ERM
BCVA Gain ‡10 Letters (Total [ 135), n (%)
BCVA Gain <9 Letters, BCVA loss <9 Letters (Total [ 146), n (%)
BCVA Loss ‡10 Letters (Total [ 21), n (%)
P Value
OR (95% CI)*
0.710.34
0.590.28
0.620.29
0.001
0.88 (0.82e0.95)
103/209 (49) 27/90 (30) 71/134 (53) 59/165 (36) 104/249 (42) 26/50 (52)
92/209 (44) 56/90 (62) 55/134 (41) 93/165 (56) 125/249 (50) 23/50 (46)
14/209 (7) 7/90 (8) 8/134 (6) 13/165 (8) 20/249 (8) 1/50 (2)
0.13
0.68 (0.41e1.12)
0.03
0.60 (0.38e0.96)
0.54
1.24 (0.62e2.48)
0.20
1.26 (0.89e1.78)
12/32 (38) 18/52 (35) 74/165 (45) 57/143 (40) 72/154 (47)
17/32 (53) 29/52 (56) 79/165 (48) 77/143 (54) 70/154 (46)
3/32 (9) 5/52 (10) 12/165 (7) 9/143 (6) 12/154 (8)
0.86
1.04 (0.65e1.68)
0.11
0.75 (0.53e1.07)
49/100 (49) 62/151 (41) 17/42 (41) 110/271 (41) 20/28 (71)
46/100 (46) 74/151 (49) 24/42 (57) 142/271 (52) 6/28 (21)
5/100 (5) 15/151 (10) 1/42 (2) 19/271 (7) 1/28 (7)
0.01
3.43 (1.32e8.95)
0.42
0.88 (0.65e1.20)
42/108 (39) 40/91 (44) 30/72 (39)
57/108 (53) 47/91 (52) 37/72 (53)
9/108 (8) 4/91 (4) 5/72 (8) 0.04
0.64 (0.42e0.98)
10/19 (53) 24/58 (41) 58/169 (34)
8/19 (42) 30/58 (52) 97/169 (57)
1/19 (5) 4/58 (7) 14/169 (8) 0.60
1.18 (0.63e2.20)
109/250 (44) 16/40 (40) 5/9 (56) 34/87 (39) 96/212 (45)
125/250 (50) 23/40 (58) 0/9 (0) 43/87 (49) 105/212 (50)
16/250 (6) 1/40 (2) 4/9 (44) 10/87 (12) 11/212 (5)
0.55
1.18 (0.70e1.99)
BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; ERM ¼ epiretinal membrane; HRF ¼ hyperreflective foci; ILM ¼ internal limiting membrane; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; logMAR ¼ logarithm of the minimum angle of resolution; ONL ¼ outer nuclear layer; OPL ¼ outer plexiform layer; OR ¼ odds ratio; SD ¼ standard deviation; SRF ¼ subretinal fluid; VMA ¼ vitreomacular adhesion; VMI ¼ vitreomacular interface; VMT ¼ vitreomacular traction. *Odd ratio (OR) for a patient presenting with gain of 10 letters in best corrected visual acuity (BCVA) at 4 months when baseline measure is increased by 1 value. y For every additional 5 letters (0.1 logMAR scale) of baseline BCVA a patient was less likely to present with gain of 10 letters in BCVA at 4 months.
morphologic DME patterns.32 Interleukin-6 levels were significantly higher in eyes with SMD than in eyes with solely diffuse or cystoid macular thickening; the latter implies active inflammation. This might explain the better response to DEX implant as an anti-inflammatory drug in eyes with SMD. In DME, cystoid spaces can be related to specific intraretinal layers on SD OCT. The inner and outer plexiform layers were reported to present physical resistance to fluid movements.33 Changes in DME are generally not confined solely to 1 layer, but rather involve several layers simultaneously.34 In our study, 83.3% of all patients presented with ONL cysts; of them, the majority had giant ONL cysts (>200 mm), which mostly occur at a relatively late stage of the disease. Previous studies reported that large ONL cysts have a negative impact on macular
6
function (measured by microperimetry) and predict worse VA outcome after anti-VEGF treatment.35,36 Cystoid changes beneath the outer plexiform layer have been shown to present risk to photoreceptor cells and to impair IS-OS integrity because of the major and irreversible impact on central visual functions.37 In our study, neither the size nor the location of intraretinal cysts altered functional outcome. Apparently, more than the absolute cyst size and volume, the remaining tissue between cysts in the central macula seems to be crucial for good VA.26 We did not analyze this parameter, and further investigation might provide more information about its impact. We did not find the presence of posterior vitreous detachment to be associated with visual outcome after DEX implants. Immunohistologic examinations of the internal limiting excised from eyes with DME indicate proliferative
Zur et al
OCT Biomarkers for DME with DEX Implant
Figure 2. Biomarkers of optical coherence tomography (OCT). Forest plot showing baseline predictors of a 10-letter gain in best-corrected visual acuity (BCVA) at 2 (-) and 4 (:) months in patients with diabetic macular edema (DME) treated with dexamethasone (DEX) implants. Subretinal fluid (SRF): Eyes presenting with SRF were more likely to gain BCVA of 10 letters at 4 months. The inner segment-outer segment (IS/OS) continuity: For every increase in an IS/OS continuity value (completely continuous, partly disrupted, completely disrupted), a patient was less likely to gain BCVA of 10 letters at 2 months. Location of hyperreflective foci (HRF): For every increase in an HRF location value (internal limiting membrane [INL], outer plexiform layer [ONL], all layers), a patient was less likely to gain BCVA of 10 letters at 2 and 4 months. Vitreoretinal interface (VR): For every increase in vitreoretinal interface values (detached, vitreomacular adhesion, vitreomacular traction), a patient was more likely to gain BCVA of 10 letters at 4 months. Error bars presenting logarithmic transformation of 95% confidence intervals (CIs). ILM ¼ internal limiting membrane; OPL ¼ outer plexiform layer.
changes at the vitreomacular interface. Still, the question remains as to whether these are the cause or consequence of DME. Diabetic macular edema associated with posterior hyaloid traction is associated with worse vision.38 Our results support the findings of a subanalysis of the Ranibizumab for Edema of the Macula in Diabetes trial, which showed that the presence of vitreomacular adhesion itself does not preclude worse functional outcome in eyes with DME treated with ranibizumab.39 Generally, cases of vitreomacular traction are excluded from major clinical trials, even though DME is associated with this condition in 28%.40 We did not find any information in the literature regarding the impact of vitreomacular interface configuration on treatment outcome after DEX implant in DME. Hyperreflective foci correlates with lipoprotein extravasation after breakdown of the inner blooderetinal barrier in the initial stages of the development of intraretinal hard exudates.41 The predictive value of HRF on visual outcome after anti-VEGF treatment in DME is unclear.42,43 In the current study, the majority of eyes presented with HRF at baseline (i.e., 91%). We found better visual outcome when HRF was absent at baseline. Poor diabetes control was associated with worse baseline VA. However, there was no association between HbA1C values and improvement in VA. Likewise, improvement in VA, anatomic reduction of macular edema, and improvement in diabetic retinopathy severity score with ranibizumab treatment seem to be independent of baseline HbA1c.44 According to the VIVID and VISTA trials, the benefit of intravitreal aflibercept injections in patients with DME
was less dependent on their presenting glycemic status, compared with laser treatmet.45 To our best knowledge, there are no data available regarding the impact of glycemic control on treatment outcome after intravitreal DEX treatment. In our study, we did not investigate the predictive value of other baseline factors, such as phakic status, baseline acuity, retinal thickness, and previous treatments before the DEX implant. Future prospective investigations of these baseline factors are required to confirm the predictive role of the OCT biomarkers we have identified. Although OCT images are readily available for physicians and provide a magnitude of detailed information, structural changes presumably reflect only part of the complex pathophysiologic processes occurring in DME. A recent study found no association between DME severity on OCT and systemic metabolic and inflammatory factors.46 Intraocular rather than systemic production of proinflammatory cytokines seems to affect the pathogenesis of both diabetic retinopathy and DME.47 Fluorescein angiographies were not considered; therefore, information regarding macular perfusion was not available. Macular ischemia might explain cases of nonresponse to DEX treatment. Recently, baseline choroidal thickness48 and disorganization of the inner retinal layers49 have been reported as predictive markers for visual outcome after DME treatment with anti-VEGF agents. We did not include those measures in the present study. This series demonstrates that certain biomarkers can help predict the response to DEX implant in DME. Future
7
Ophthalmology Volume -, Number -, Month 2017 prospective investigations should consider imaging and systemic biomarkers, and compare anti-VEGF drugs with DEX implants in a head-to-head trial aimed to tailor and optimize patient care.
References 1. Kempen JH, O’Colmain BJ, Leske MC, et al. The prevalence of diabetic retinopathy among adults in the United States. Arch Ophthalmol. 2004;122:552-563. 2. Yau JWY, Rogers SL, Kawasaki R, et al. Global prevalence and major risk factors of diabetic retinopathy. Diabetes Care. 2012;35:556-564. 3. Massin P, Bandello F, Garweg JG, et al. Safety and efficacy of ranibizumab in diabetic macular edema (RESOLVE Study): a 12-month, randomized, controlled, double-masked, multicenter phase II study. Diabetes Care. 2010;33:2399-2405. 4. Mitchell P, Bandello F, Schmidt-Erfurth U, et al. The RESTORE study: ranibizumab monotherapy or combined with laser versus laser monotherapy for diabetic macular edema. Ophthalmology. 2011;118:615-625. 5. Michaelides M, Kaines A, Hamilton RD, et al. A prospective randomized trial of intravitreal bevacizumab or laser therapy in the management of diabetic macular edema (BOLT study) 12month data: report 2. Ophthalmology. 2010;117:1078-1086.e2. 6. Do DV, Nguyen QD, Boyer D, et al. One-year outcomes of the da Vinci Study of VEGF Trap-Eye in eyes with diabetic macular edema. Ophthalmology. 2012;119:1658-1665. 7. Rajendram R, Fraser-Bell S, Kaines A, et al. A 2-year prospective randomized controlled trial of intravitreal bevacizumab or laser therapy (BOLT) in the management of diabetic macular edema: 24-month data: report 3. Arch Ophthalmol. 2012;130:972-979. 8. Elman MJ, Bressler NM, Qin H, et al. Expanded 2-year follow-up of ranibizumab plus prompt or deferred laser or triamcinolone plus prompt laser for diabetic macular edema. Ophthalmology. 2011;118:609-614. 9. Nguyen QD, Brown DM, Marcus DM, et al. Ranibizumab for diabetic macular edema: results from 2 phase III randomized trials: RISE and RIDE. Ophthalmology. 2012;119:789-801. 10. Gonzalez VH, Campbell J, Holekamp NM, et al. Early and long-term responses to anti-vascular endothelial growth factor therapy in diabetic macular edema: analysis of protocol I data. Am J Ophthalmol. 2016;172:72-79. 11. Paper O, Guigou S, Pommier S, et al. Efficacy and safety of intravitreal dexamethasone implant in patients with diabetic macular edema. Ophthalmologica. 2015;233:169-175. 12. Boyer DS, Yoon YH, Belfort R, et al. Three-year, randomized, sham-controlled trial of dexamethasone intravitreal implant in patients with diabetic macular edema. Ophthalmology. 2014;121:1904-1914. 13. Alshahrani ST, Dolz-Marco R, Gallego-Pinazo R, et al. Intravitreal dexamethasone implant for the treatment of refractory macular edema in retina vascular diseases: results of the KKESH International Collaborative Retina Study Group. Retina. 2016;36:131-136. 14. Dutra Medeiros M, Postorino M, Navarro R, et al. Dexamethasone intravitreal implant for treatment of patients with persistent diabetic macular edema. Ophthalmologica. 2014;231:141-146. 15. Pacella F, Romano MR, Turchetti P, et al. An eighteen-month follow-up study on the effects of intravitreal dexamethasone implant in diabetic macular edema refractory to anti-VEGF therapy. Int J Ophthalmol. 2016;9:1427-1432.
8
16. Romero-Aroca P. Targeting the pathophysiology of diabetic macular edema. Diabetes Care. 2010;33:2484-2485. 17. Murakami T, Yoshimura N. Structural changes in individual retinal layers in diabetic macular edema. J Diabetes Res. 2013;2013:920713. 18. Browning DJ, Glassman AR, Aiello LP, et al. Relationship between optical coherence tomography-measured central retinal thickness and visual acuity in diabetic macular edema. Ophthalmology. 2007;114:525-536. 19. Bressler SB, Qin H, Beck RW, et al. Factors associated with changes in visual acuity and central subfield thickness at 1 year after treatment for diabetic macular edema with ranibizumab. Arch Ophthalmol. 2012;130:1153-1161. 20. Sophie R, Lu N, Campochiaro PA. Predictors of functional and anatomic outcomes in patients with diabetic macular edema treated with ranibizumab. Ophthalmology. 2015;122: 1395-1401. 21. Fickweiler W, Hooymans JMM, Los LI, et al. Predictive value of optical coherence tomographic features in the Bevacizumab and Ranibizumab in Patients with Diabetic Macular Edema (BRDME) Study. Retina. 2017. http://dx.doi.org/10.1097/IAE. 0000000000001626 [Epub ahead of print]. 22. Shimura M, Yasuda K, Yasuda M, Nakazawa T. Visual outcome after intravitreal bevacizumab depends on the optical coherence tomographic patterns of patients with diffuse diabetic macular edema. Retina. 2013;33:740-747. 23. Giocanti-aur A, Hrarat L, Qu LM, et al. Functional and anatomical outcomes in patients with serous retinal detachment in diabetic macular edema treated with ranibizumab. Invest Ophthalmol Vis Sci. 2017:797-800. 24. Seo KH, Yu S-Y, Kim M, Kwak HW. Visual and morphologic outcomes of intravitreal ranibizumab for diabetic macular edema based on optical coherence tomography patterns. Retina. 2016;36:588-595. 25. Rayess N, Rahimy E, Ying GS, et al. Baseline choroidal thickness as a predictor for response to anti-vascular endothelial growth factor therapy in diabetic macular edema. Am J Ophthalmol. 2015;159:85-91. 26. Pelosini L, Hull CC, Boyce JF, et al. Optical coherence tomography may be used to predict visual acuity in patients with macular edema. Invest Ophthalmol Vis Sci. 2011;52:2741-2748. 27. Ozdemir H, Karacorlu M, Karacorlu S. Serous macular detachment in diabetic cystoid macular oedema. Acta Ophthalmol Scand. 2005;83:63-66. 28. Kang SW, Park CY, Ham D-I. The correlation between fluorescein angiographic and optical coherence tomographic features in clinically significant diabetic macular edema. Am J Ophthalmol. 2004;137:313-322. 29. Kaiser PK, Riemann CD, Sears JE, Lewis H. Macular traction detachment and diabetic macular edema associated with posterior hyaloidal traction. Am J Ophthalmol. 2001;131:44-49. 30. Alasil T, Keane PA, Updike JF, Dustin L. Relationship between optical coherence tomography retinal parameters and visual acuity in diabetic macular edema. Ophthalmology. 2010;117:2379-2386. 31. Kaldirim H, Yazgan S, Atalay K, et al. Intravitreal dexamethasone implantation in patients with different morphological diabetic macular edema having insufficient response to ranibizumab. Retina. 2017. http://dx.doi.org/10.1097/IAE. 0000000000001648 [Epub ahead of print]. 32. Bandyopadhyay S. Study of aqueous cytokines in patients with different patterns of diabetic macular edema based on optical coherence tomography. Int Ophthalmol. 2017. http://dx.doi. org/10.1007/s10792-017-0453-2 [Epub ahead of print]. 33. Antcliff RJ, Hussain AA, Marshall J. Hydraulic conductivity of fixed retinal tissue after sequential excimer laser ablation:
Zur et al
34. 35. 36.
37.
38. 39.
40.
41.
OCT Biomarkers for DME with DEX Implant
barriers limiting fluid distribution and implications for cystoid macular edema. Arch Ophthalmol. 2001;119:539-544. Bolz M, Ritter M, Schneider M, et al. A systematic correlation of angiography and high-resolution optical coherence tomography in diabetic macular edema. Ophthalmology. 2009;116:66-72. Deák GG, Bolz M, Ritter M, et al. A systematic correlation between morphology and functional alterations in diabetic macular edema. Invest Ophthalmol Vis Sci. 2010;51:6710-6714. Reznicek L, Cserhati S, Seidensticker F, et al. Functional and morphological changes in diabetic macular edema over the course of anti-vascular endothelial growth factor treatment. Acta Ophthalmol. 2013;91:529-536. Murakami T, Nishijima K, Akagi T, et al. Optical coherence tomographic reflectivity of photoreceptors beneath cystoid spaces in diabetic macular edema. Investig Ophthalmol Vis Sci. 2012;53:1506-1511. Agarwal D, Gelman R, Prospero Ponce C, et al. The vitreomacular interface in diabetic retinopathy. J Ophthalmol. 2015;2015:392983. Sadiq MA, Soliman MK, Sarwar S, et al. Effect of vitreomacular adhesion on treatment outcomes in the Ranibizumab for Edema of the Macula in Diabetes (READ-3) Study. Ophthalmology. 2016;123:324-329. Jackson TL, Nicod E, Ms C, et al. Vitreous attachment in agerelated macular degeneration, diabetic macular edema, and retinal vein occlusion: a systematic review and metaanalysis. Retina. 2013;33:1099-1108. Bolz M, Schmidt-Erfurth U, Deak G, et al. Optical coherence tomographic hyperreflective foci. a morphologic sign of lipid extravasation in diabetic macular edema. Ophthalmology. 2009;116:914-920.
42. Framme C, Schweizer P, Imesch M, et al. Behavior of SDOCT - detected hyperreflective foci in the retina of antiVEGFetreated patients with diabetic macular edema. Invest Ophthalmol Vis Sci. 2017;53:5814-5818. 43. Kang J, Chung H, Kim HC. Correlation of optical coherence foci with visual acuity outcomes in different patterns of diabetic macular edema. Retina. 2016;369:1630-1639. 44. Bansal AS, Khurana RN, Wieland MR, et al. Influence of glycosylated hemoglobin on the efficacy of ranibizumab for diabetic macular edema a post hoc analysis of the RIDE/RISE Trials. Ophthalmology. 2015;122:1573-1579. 45. Singh R, Wykoff CC, Brown DM, et al. Outcomes of diabetic macular edema patients by baseline hemoglobin A1c: analyses from VISTA and VIVID. Ophthalmol Retina. 2017 [Epub ahead of print]. 46. Figueras-roca M, Molins B, Sala-puigdollers A, et al. Peripheral blood metabolic and inflammatory factors as biomarkers to ocular findings in diabetic macular edema. PLoS One. 2017;12:e0173865. 47. Vujosevic S, Simó R. Local and systemic inflammatory biomarkers of diabetic retinopathy: an integrative approach. Invest Ophthalmol Vis Sci. 2017;58:68-75. 48. Rayess N, Rahimy E, Ying G, et al. Baseline choroidal thickness as a predictor for response to anti-vascular endothelial growth factor therapy in diabetic macular edema. Am J Ophthalmol. 2015;159:85-91.e3. 49. Sun JK, Lin MM, Lammer J, et al. Disorganization of the retinal inner layers as a predictor of visual acuity in eyes with center-involved diabetic macular edema. JAMA Ophthalmol. 2014;132:1309-1316.
Footnotes and Financial Disclosures Originally received: June 25, 2017. Final revision: August 21, 2017. Accepted: August 22, 2017. Available online: ---.
Author Contributions:
Manuscript no. 2017-1461.
1
Division of Ophthalmology, Tel Aviv Sourasky Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel. 2 University of Buenos Aires, Argentina. 3
Department of Ophthalmology, University of Leipzig, Germany.
Eye Clinic - Department of Biomedical and Clinical Science “L. Sacco,” Luigi Sacco Hospital, University of Milan, Milan, Italy.
4
5
Favaloro University Hospital, Buenos Aires, Argentina.
6
Incumbent, Sydney A. Fox Chair in Ophthalmology, Tel Aviv University, Tel Aviv, Israel. Presented: at the Annual Meeting of the American Society of Retinal Specialists, Boston, Massachusetts, August 11e15, 2017.
z
Both Dinah Zur, MD, and Matias Iglicki, MD were equal contributors.
Financial Disclosure(s): The author(s) have made the following disclosure(s): A.L.: Consultant Allergan. The International Retina Group is an independent research branch of the International Retina Panel, which receives travel support by Allergan.
Conception and design: Zur, Iglicki, Busch, Lowenstein Data collection: Zur, Iglicki, Busch, Lowenstein, Invernizzi, Cebecci, Chhablani Analysis and interpretation: Zur, Iglicki, Busch, Lowenstein, Invernizzi Obtained funding: Not applicable Overall responsibility: Zur, Iglicki, Busch, Lowenstein, Invernizzi, Mariussi Abbreviations and Acronyms: BCVA ¼ best-corrected visual acuity; CI ¼ confidence interval; CMT ¼ central macular thickness; DEX ¼ dexamethasone; DME ¼ diabetic macular edema; GEE ¼ generalized estimating equation; HbA1C ¼ hemoglobin A1C; HRF ¼ hyperreflective foci; INL ¼ inner nuclear layer; IS-OS ¼ inner segment-outer segment; OCT ¼ optical coherence tomography; ONL ¼ outer nuclear layer; OR ¼ odds ratio; SD ¼ spectral domain; SMD ¼ serous macular detachment; SRF ¼ subretinal fluid; VA ¼ visual acuity; VEGF ¼ vascular endothelial growth factor. Correspondence: Dinah Zur, MD, Division of Ophthalmology, Tel Aviv Sourasky Medical Center, Weizman 6, Tel Aviv 64239. E-mail: [email protected].
Members of the International Retina Group are listed in the Supplementary Material (available online at www.aaojournal.org).
9