Ocular imaging is a cornerstone of diagnosis and management of diabetic macular edema (DME). Historically, fluorescein angiography (FA) has played a pivotal role in qualifying vascular leakage and guiding therapy for DME.1 In the past two decades, optical coherence tomography (OCT) has revolutionized diagnosis of DME, as well as led to the identification of prognostic biomarkers such as distribution of edema, disruption to the inner segment ellipsoid (ISe) band and external limiting membrane (ELM), and disorganization of the inner retinal layers (DRIL).2–5
More recently, ultra-widefield (UW) imaging enabled documentation of peripheral lesions, leading to more detailed description of disease severity and prognosis of progression.6,7 When combined with FA, ultra-widefield fluorescein angiography (UWFA) has been shown to identify 3.9-times more nonperfusion, 1.9-times more neovascularization, and 1.1-times more retinal pathology as compared with standard seven-fields photography.8
Peripheral ischemia may be quantified through the calculation of an ischemic index (II).9 Such an index gives a ratio of nonperfused to perfused retina and has been shown to be associated with the prevalence of DME in a retrospective study of 122 eyes (odds ratio [OR] = 3.75; 95% CI, 1.26–11.13; P < .02).10 Furthermore, DME that is recalcitrant to macular photocoagulation was found to be associated with a worse II in a retrospective study of 148 eyes with persistent DME.11 These patients also required a greater number of treatments with macular laser photocoagulation and had a lesser reduction in the central macular thickness (CMT).
It is unclear what prognostic information can be gained from UWFA for patients with DME being treated with anti-vascular endothelial growth factor (VEGF) drugs. Herein we report the prognostic value of UWFA from a prospective clinical trial where patients with persistent DME were switched from bevacizumab (Avastin; Genentech, South San Francisco, CA) to aflibercept (Eylea; Regeneron, Tarrytown, NY).
Patients and Methods
All patients gave informed consent to participate in this prospective, nonrandomized, open-label clinical trial. The study adhered to the Declaration of Helsinki and was registered on the Australian and New Zealand Clinical Trials Registry (ACTRN12614001307695).
Full inclusion and exclusion criteria for the patients enrolled in this trial have previously been reported.12 Briefly, patients were aged 18 years or older and had DME with a central macular thickness greater than 300 μm in the spectral-domain OCT (SD-OCT) (Spectralis; Heidelberg Engineering, Heidelberg, Germany), best-corrected visual acuity (BCVA) between 34 and 85 Early Treatment of Diabetic Retinopathy Study (ETDRS) letters, and at least four previous intravitreal injections of bevacizumab (2.5 mg/0.1 mL) in the 6 months prior to enrollment. Patients then received intravitreal aflibercept (2.0 mg/0.1mL) in one eye only every 4 weeks for a total of five injections, at which point the treatment interval was extended to injections every 8 weeks for a total follow-up of 48 weeks. BCVA and CMT assessments occurred at every 4 weeks, with UWFA occurring at baseline and at 48 weeks.
UWFA images were acquired using the Optos 200TX (Optos Plc, Dunfermline, Scotland). An intravenous bolus of 5 mL of 10% w/v fluorescein was given and images were obtained in the transit phase (up to 45 seconds) arteriovenous phase (1 to 2 minutes) and during recirculation (up to 10 minutes). A single best image from the arteriovenous phase of the study eye was selected for grading.
Calculation of Ischemic Index
The ischemic index was calculated using the concentric rings method previously described.13 Briefly, UWFA images for each patient at baseline and 48 weeks were overlaid with the template of seven concentric rings (Figure 1) as supplied in the supplement to the publication by Nicholson et al.13 Using ImageJ software (NIH, Bethesda, Maryland), the template was resized and repositioned for each image such that the innermost ring was equal in size to the optic disc and the central point of the template was placed at the fovea. Each of these seven rings was divided into 12 equal segments subtending an angle of 30° at the fovea. Each segment was graded as perfused, nonperfused, or not gradable if more than half of the segment consisted of one of the three. Grading was validated by an independent trained external grader (TP). Areas with scatter laser were deemed not gradable and were excluded from the calculation of the index.
Ultra-widefield fluorescein angiography with concentric rings overlay.
The II was calculated from the grading by multiplying each segment by the total disc area represented and then dividing nonperfused retina by the total gradable area. An II of 50% or above was considered to be high as previously described and further analyses were based on this definition.14 The macular ischemic index (MII) was calculated from the 12 sectors comprising the innermost ring of the template.
Optical Coherence Tomography Grading
Images were graded for morphology of DME (intraretinal fluid and/or subretinal fluid [SRF]), presence or absence of disorganization of the inner retinal layers (DRIL) greater than 50%, ISe band disruption, and external limiting membrane (ELM) disruption in a 1-mm area centered around the fovea, as previously described and reported.2,12,15,16
All statistical tests were performed and figures produced using IBM SPSS software (version 22; SPSS Inc., Chicago, IL). Patients were grouped by baseline II greater than or less than or equal to 50%. Data were confirmed to be distributed normally using Shapiro-Wilk tests. Homogeneity of data was confirmed using Levene's test for all independent samples' t-tests. Adjustment was made to the analysis using the Welch-Satterthwaite method for data that were not homogenous. Fisher's exact test was used to analyze categorical data when sample sizes in groups were small. Pearson's correlation coefficient was calculated for correlation analyses. For all analyses, a P value of less than .05 was considered to be statistically significant.
Of the 43 patients recruited, one withdrew consent after the baseline visit, one was withdrawn due to a retinal detachment in the study eye after the second injection, three did not have UWFA performed at baseline, and six did not have UWFA performed at 48-weeks. The characteristics of the 38 patients with complete UWFA data at baseline included for analysis are summarized in Table 1.
Baseline Characteristics of Patients Included in Analysis
BCVA improved by a mean ± standard deviation of 4.0 ± 7.2 letters to 72.2 ± 9.1 letters (P = .002), and CMT reduced by 60 μm ± 111 μm to 356 μm ± 109 μm (P = .002) in the 38 patients with UWFA data at baseline during the 48-week study period.
There was no significant change in mean II (24.9 ± 32.5% to 22.7 ± 29.7%; P = .36) or MII (8.1 ± 23.5% to 7.9 ± 19.7%; P = .88) from baseline to 48-weeks. There was correlation between II and MII at baseline (r = 0.66; P < .001) and at 48-weeks (r = 0.57; P < .001).
Patients with an II greater than 50% (n = 9) at baseline had a poorer baseline visual acuity (VA) (60.1 ± 10.2 letters vs. 70.7 ± 9.0 letters; P = .005; Figure 2A) and a worse MII (6.9 ± 25% vs. 56 ± 52%; P < .001) compared to patients with a lower II (≤ 50%). These patients gained significantly more letters of vision at 48 weeks (8.3 ± 9.3 letters vs. 2.6 ± 5.9 letters; P = .03). At 48 weeks, there was no significant difference in absolute VA in patients with an II greater than 50% compared to those with an index less than or equal to 50% (68.4 ± 6.0 letter vs. 73.3 ± 9.6 letters; P = .16; Figure 2B). There was no significant difference in baseline (69.4 ± 9.3 letters vs. 63.0 ± 13 letters; P = .14) or final (72.6 ± 9.4 letters vs. 70.3 ± 7.9 letters; P = .55) BCVA in patients with the absence or presence of macular ischemia, nor were the BCVA gains in these two groups different (3.2 ± 6.9 letters vs. 7.3 ± 8.1 letters; P = .18).
Boxplot showing baseline (A) and 48-week (B) visual acuities grouped by baseline ischemic index.
There was no difference in CMT at baseline or 48-weeks for patients with an II greater than 50% at baseline compared with those less than or equal to 50% (429 μm ± 61 μm vs. 412 μm ± 101 μm; P = .63 and 395 μm ± 112 μm vs. 343 μm ± 107 μm; P = .22, respectively; Figure 3). There was no difference in change in CMT for patients with an II greater than 50% (−34 μm ± 73 μm vs. −68 μm ± 120 μm; P = .42).
Boxplot showing baseline (A) and 48-week (B) central macular thickness grouped by baseline ischemic index.
There was no significant correlation between CMT at baseline (r = 0.18; P = .29) or change in CMT at 48-weeks (r = 0.16; P = .33) with the MII.
There was no significant correlation between baseline II and HbA1c (r = 0.16; P = .40), duration of diabetes (r = 0.25; P = .13), or number of previous anti-VEGF injections for DME (r = −0.31; P = .06). There was no significant correlation between baseline MII and HbA1c (r = 0.01; P = .96), duration of diabetes (r = 0.21; P = .20), or number of previous anti-VEGF injections for DME (r = −0.23; P = .17).
There was no correlation between II or MII and presence of SRF, disorganization of the inner retinal layers, external limiting membrane, or ISe band disruption (Table 2).
Correlation of OCT Parameters With Baseline Ischemic Index and Macular Ischemic Index
Among patients with persistent DME and significant prior treatment with bevacizumab, a high baseline II greater than 50% was associated with a poorer baseline VA. However, these patients had a similar final VA to those with a low baseline II when therapy was switched to aflibercept.
VA is dependent on the health of the macula, both in the available blood supply as well as the integrity of the various cells involved in phototransduction. In this study, VA was not associated with a thicker CMT but with worse macular ischemia. Recent studies utilizing OCT angiography have identified a negative correlation between macular capillary density and VA.17 OCT biomarkers such as ISe band and ELM disruption, DRIL and presence of SRF are all associated with a poorer VA.2–5 Although ischemia may be postulated to explain these structural abnormalities, presence of these factors did not correlate with macular ischemia in this study.
There are several potential explanations for a greater gain in vision for the patients with a higher baseline II. Firstly, the starting VA was significantly lower, meaning that there was more potential for vision gain. Secondly, there may be a “ceiling effect” to the amount of vision that can be gained in this cohort of patients with persistent DME. Finally, ischemia and hypoxia are strongly implicated in the pathogenesis of DME. Areas of untreated retinal non-perfusion may stimulate the production of mediators such as VEGF-A and placental growth factor that contribute to the formation and persistence of DME. Although there is no control group to compare to in this study, these factors may be more effectively inhibited by aflibercept leading to improved outcomes in these patients.18
It has been hypothesized that scatter photocoagulation to areas of peripheral ischemia may help in the management of DME. Complete resolution of macular edema following panretinal photocoagulation was demonstrated in a case series of 17 eyes with florid proliferative DR and DME.19 Worsening of DME was reduced during a period of 6 months in a clinical trial of 52 patients randomized to a single dose of bevacizumab either with or without targeted photocoagulation.20 Reduced levels of VEGF in the eye following panretinal photocoagulation (PRP) may be responsible for this effect.21 However, PRP is also known to exacerbate macular edema likely through transient increases in inflammatory cytokines and VEGF.22,23
Most recently, monotherapy with ranibizumab (Lucentis; Genentech, South San Francisco, CA) was shown to have similar outcomes to combination therapy with ranibizumab and targeted laser photocoagulation in a 3-year, randomized trial of 40 eyes from 29 patients with DME and significant peripheral ischemia.24 There were no differences in treatment burden or visual or anatomical outcomes. The authors suggested areas of nonperfusion may represent dead rather than stressed tissue and thus do not contribute to increased production of factors driving DME. This is similar to the outcomes presented in the RELATE study, where patients with branch and central retinal vein occlusion who were randomized to intravitreal ranibizumab therapy with targeted scatter laser had similar vision and anatomical outcomes to those treated with ranibizumab monotherapy.25
Contrary to Wessel et al., we did not find any correlation between diabetes control and II, nor did we find a relationship between duration of diabetes and II.10 Furthermore, we did not confirm findings of other studies presenting data suggesting a reversal of areas of nonperfusion with intravitreal injection of anti-VEGF drugs or dexamethasone implant.26,27
The strengths of this study are in the prospective and standardized nature of data collection in a clinical trial setting. The trial participants received the per protocol treatment during the study period and had retinal imaging performed at standardized time points. The methodology for grading ischemia has been previously validated.
There are inherent limitations in the data and analyses performed due to lack of a control arm as well as a relatively small sample size from a single center. Patients included had heterogeneous prior treatments including macular and panretinal photocoagulation. Furthermore, correction for peripheral distortion and introducing validated, reliable computer-based segmentation for the calculation of peripheral ischemia may yield more accurate results in the future.28,29
The exploratory analyses from this study demonstrate that significant retinal ischemia may correlate with poorer VA and a greater capacity for vision improvement in patients with persistent DME switched to aflibercept. Additionally, there appears to be no clear association between degree of peripheral ischemia and severity of macular edema, suggesting that other factors may be involved in macular thickening in DME. Future directions for these findings are to assess the effect of peripheral ischemia as a biomarker of treatment response to anti-VEGF drugs in treatment naïve eyes as well as those treated with other modalities such as corticosteroids that target different pathological pathways.
- Early Treatment Diabetic Retinopathy Study research group. Photocoagulation for diabetic macular edema. Early Treatment Diabetic Retinopathy Study report number 1. Early Treatment Diabetic Retinopathy Study research group. Arch Ophthalmol. 1985;103(12):1796–1806. https://doi.org/10.1001/archopht.1985.01050120030015 PMID: doi:10.1001/archopht.1985.01050120030015 [CrossRef]2866759
- Maheshwary AS, Oster SF, Yuson RM, Cheng L, Mojana F, Freeman WR. The association between percent disruption of the photoreceptor inner segment-outer segment junction and visual acuity in diabetic macular edema. Am J Ophthalmol. 2010;150(1):63–67.e1. https://doi.org/10.1016/j.ajo.2010.01.039 PMID: doi:10.1016/j.ajo.2010.01.039 [CrossRef]20451897
- Murakami T, Nishijima K, Sakamoto A, Ota M, Horii T, Yoshimura N. Association of pathomorphology, photoreceptor status, and retinal thickness with visual acuity in diabetic retinopathy. Am J Ophthalmol. 2011;151(2):310–317. https://doi.org/10.1016/j.ajo.2010.08.022 PMID: doi:10.1016/j.ajo.2010.08.022 [CrossRef]
- Otani T, Yamaguchi Y, Kishi S. Correlation between visual acuity and foveal microstructural changes in diabetic macular edema. Retina. 2010;30(5):774–780. https://doi.org/10.1097/IAE.0b013e3181c2e0d6 PMID: doi:10.1097/IAE.0b013e3181c2e0d6 [CrossRef]
- 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(11):1309–1316. https://doi.org/10.1001/jamaophthalmol.2014.2350 PMID: doi:10.1001/jamaophthalmol.2014.2350 [CrossRef]25058813
- Silva PS, Cavallerano JD, Haddad NM, et al. Peripheral Lesions Identified on Ultrawide Field Imaging Predict Increased Risk of Diabetic Retinopathy Progression over 4 Years. Ophthalmology. 2015;122(5):949–956. https://doi.org/10.1016/j.ophtha.2015.01.008 PMID: doi:10.1016/j.ophtha.2015.01.008 [CrossRef]25704318
- Silva PS, El-Rami H, Barham R, et al. Hemorrhage and/or Microaneurysm Severity and Count in Ultrawide Field Images and Early Treatment Diabetic Retinopathy Study Photography. Ophthalmology. 2017;124(7):970–976. https://doi.org/10.1016/j.ophtha.2017.02.012 PMID: doi:10.1016/j.ophtha.2017.02.012 [CrossRef]28336057
- Wessel MM, Aaker GD, Parlitsis G, Cho M, D'Amico DJ, Kiss S. Ultra-wide-field angiography improves the detection and classification of diabetic retinopathy. Retina. 2012;32(4):785–791. https://doi.org/10.1097/IAE.0b013e3182278b64 PMID: doi:10.1097/IAE.0b013e3182278b64 [CrossRef]
- Tsui I, Kaines A, Havunjian MA, et al. Ischemic index and neovascularization in central retinal vein occlusion. Retina. 2011;31(1):105–110. https://doi.org/10.1097/IAE.0b013e3181e36c6d PMID: doi:10.1097/IAE.0b013e3181e36c6d [CrossRef]
- Wessel MM, Nair N, Aaker GD, Ehrlich JR, D'Amico DJ, Kiss S. Peripheral retinal ischaemia, as evaluated by ultra-widefield fluorescein angiography, is associated with diabetic macular oedema. Br J Ophthalmol. 2012;96(5):694–698. https://doi.org/10.1136/bjophthalmol-2011-300774 PMID: doi:10.1136/bjophthalmol-2011-300774 [CrossRef]22423055
- Patel RD, Messner LV, Teitelbaum B, Michel KA, Hariprasad SM. Characterization of ischemic index using ultra-widefield fluorescein angiography in patients with focal and diffuse recalcitrant diabetic macular edema. Am J Ophthalmol. 2013;155(6):1038–1044.e2. https://doi.org/10.1016/j.ajo.2013.01.007 PMID: doi:10.1016/j.ajo.2013.01.007 [CrossRef]23453693
- Bahrami B, Hong T, Zhu M, Schlub TE, Chang A. Switching therapy from bevacizumab to aflibercept for the management of persistent diabetic macular edema. Graefes Arch Clin Exp Ophthalmol. 2017;255(6):1133–1140. https://doi.org/10.1007/s00417-017-3624-y PMID: doi:10.1007/s00417-017-3624-y [CrossRef]28238195
- Nicholson L, Vazquez-Alfageme C, Ramu J, et al. Validation of concentric rings method as a topographic measure of retinal nonperfusion in ultra-widefield fluorescein angiography. Am J Ophthalmol. 2015;160(6):1217–1225.e2. https://doi.org/10.1016/j.ajo.2015.09.003 PMID: doi:10.1016/j.ajo.2015.09.003 [CrossRef]26384165
- Sim DA, Keane PA, Rajendram R, et al. Patterns of peripheral retinal and central macula ischemia in diabetic retinopathy as evaluated by ultra-widefield fluorescein angiography. Am J Ophthalmol. 2014;158(1):144–153.e1. https://doi.org/10.1016/j.ajo.2014.03.009 PMID: doi:10.1016/j.ajo.2014.03.009 [CrossRef]24709807
- Sun JK, Radwan SH, Soliman AZ, et al. Neural Retinal Disorganization as a Robust Marker of Visual Acuity in Current and Resolved Diabetic Macular Edema. Diabetes. 2015;64(7):2560–2570. https://doi.org/10.2337/db14-0782 PMID: doi:10.2337/db14-0782 [CrossRef]25633419
- Bahrami B, Hong T, Schlub TE, Chang AA. AFLIBERCEPT FOR PERSISTENT DIABETIC MACULAR EDEMA: Forty-Eight-Week Outcomes. Retina. 2019;39(1):61–68. https://doi.org/10.1097/IAE.0000000000002253 PMID: doi:10.1097/IAE.0000000000002253 [CrossRef]
- Samara WA, Shahlaee A, Adam MK, et al. Quantification of Diabetic Macular Ischemia Using Optical Coherence Tomography Angiography and Its Relationship with Visual Acuity. Ophthalmology. 2017;124(2):235–244. https://doi.org/10.1016/j.ophtha.2016.10.008 PMID: doi:10.1016/j.ophtha.2016.10.008 [CrossRef]
- Papadopoulos N, Martin J, Ruan Q, et al. Binding and neutralization of vascular endothelial growth factor (VEGF) and related ligands by VEGF Trap, ranibizumab and bevacizumab. Angiogenesis. 2012;15(2):171–185. https://doi.org/10.1007/s10456-011-9249-6 PMID: doi:10.1007/s10456-011-9249-6 [CrossRef]22302382
- Gaucher D, Fortunato P, LeCleire-Collet A, et al. Spontaneous resolution of macular edema after panretinal photocoagulation in florid proliferative diabetic retinopathy. Retina. 2009;29(9):1282–1288. https://doi.org/10.1097/IAE.0b013e3181a91e9f PMID: doi:10.1097/IAE.0b013e3181a91e9f [CrossRef]19629019
- Takamura Y, Tomomatsu T, Matsumura T, et al. The effect of photocoagulation in ischemic areas to prevent recurrence of diabetic macular edema after intravitreal bevacizumab injection. Invest Ophthalmol Vis Sci. 2014;55(8):4741–4746. https://doi.org/10.1167/iovs.14-14682 PMID: doi:10.1167/iovs.14-14682 [CrossRef]25028357
- Aiello LP, Avery RL, Arrigg PG, et al. Vascular endothelial growth factor in ocular fluid of patients with diabetic retinopathy and other retinal disorders. N Engl J Med. 1994;331(22):1480–1487. https://doi.org/10.1056/NEJM199412013312203 PMID: doi:10.1056/NEJM199412013312203 [CrossRef]7526212
- McDonald HR, Schatz H. Macular edema following panretinal photocoagulation. Retina. 1985;5(1):5–10. https://doi.org/10.1097/00006982-198500510-00002 PMID: doi:10.1097/00006982-198500510-00002 [CrossRef]4001591
- Ito A, Hirano Y, Nozaki M, Ashikari M, Sugitani K, Ogura Y. Short pulse laser induces less inflammatory cytokines in the murine retina after laser photocoagulation. Ophthalmic Res. 2015;53(2):65–73. https://doi.org/10.1159/000366520 PMID: doi:10.1159/000366520 [CrossRef]25572382
- Brown DM, Ou WC, Wong TP, Kim RY, Croft DE, Wykoff CCDAVE Study Group. Targeted Retinal Photocoagulation for Diabetic Macular Edema with Peripheral Retinal Nonperfusion: Three-Year Randomized DAVE Trial. Ophthalmology. 2018;125(5):683–690. https://doi.org/10.1016/j.ophtha.2017.11.026 PMID: doi:10.1016/j.ophtha.2017.11.026 [CrossRef]29336896
- Campochiaro PA, Hafiz G, Mir TA, et al. Scatter Photocoagulation Does Not Reduce Macular Edema or Treatment Burden in Patients with Retinal Vein Occlusion: the RELATE Trial. Ophthalmology. 2015;122(7):1426–1437. https://doi.org/10.1016/j.ophtha.2015.04.006 PMID: doi:10.1016/j.ophtha.2015.04.006 [CrossRef]25972260
- Levin AM, Rusu I, Orlin A, et al. Retinal reperfusion in diabetic retinopathy following treatment with anti-VEGF intravitreal injections. Clin Ophthalmol. 2017;11:193–200. https://doi.org/10.2147/OPTH.S118807 PMID: doi:10.2147/OPTH.S118807 [CrossRef]28176934
- Querques L, Parravano M, Sacconi R, Rabiolo A, Bandello F, Querques G. Ischemic index changes in diabetic retinopathy after intravitreal dexamethasone implant using ultra-widefield fluorescein angiography: a pilot study. Acta Diabetol. 2017;54(8):769–773. https://doi.org/10.1007/s00592-017-1010-1 PMID: doi:10.1007/s00592-017-1010-1 [CrossRef]28577137
- Tanchon C, Srivastava SK, Ehlers JP. Automated quantitative analysis of leakage and ischemia for ultra-widefield angiography in retinal vascular disease. Invest Ophthalmol Vis Sci. 2015;56(7):3067.
- Tan CS, Chew MC, van Hemert J, Singer MA, Bell D, Sadda SR. Measuring the precise area of peripheral retinal non-perfusion using ultra-widefield imaging and its correlation with the ischaemic index. Br J Ophthalmol. 2016;100(2):235–239. https://doi.org/10.1136/bjophthalmol-2015-306652 PMID: doi:10.1136/bjophthalmol-2015-306652 [CrossRef]
Baseline Characteristics of Patients Included in Analysis
|Number of patients||38|
|Age (years), mean ± SD||62.6 ± 9.7|
|Male, n (%)||26 (68)|
|Right eyes, n (%)||18 (47)|
|Duration of diabetes (years), mean ± SD||17.5 ± 11.1|
|HbA1c (%), mean ± SD||7.9 ± 1.7|
|Duration of anti-VEGF treatment (months), mean ± SD||25.0 ± 22.2|
|Total number of anti-VEGF injections, mean ± SD||16.4 ± 11.2|
|Prior treatments in study eye|
| Focal/grid macular photocoagulation, n (%)||15 (39.5)|
| Panretinal photocoagulation, n (%)||15 (39.5)|
| Vitrectomy, n (%)||5 (13.1)|
|Baseline BCVA (letter score), mean ± SD||68.2 ± 10.2|
|Baseline CMT (μm), mean ± SD||416 ± 93|
Correlation of OCT Parameters With Baseline Ischemic Index and Macular Ischemic Index
|Characteristic||External Limiting Membrane Disruption||Presence of Subretinal Fluid||Presence of Inner Segment Ellipsoid Band Disruption||Disorganization ofthe Retinal Inner Layers >50%|
|Baseline ischemic index||r= 0.24, P= .14||r = −0.07, P= .66||r = 0.14, P= .42||r = 0.08, P= .65|
|Baseline macular ischemic index||r = 0.35, P= .06||r = −0.14, P= .42||r = 0.17, P= .30||r = 0.18, P= .28|