This study assesses a sensitive and specific threshold of total area of retinal nonperfusion that is required for any retinal neovascularization to develop on ultra-widefield angiography.
Key Points
Question
What are the amount and distribution of retinal nonperfusion associated with the development of proliferative diabetic retinopathy?
Findings
In a post hoc analysis, a threshold of 118.3 disc areas of retinal nonperfusion was identified to have a specificity of 84.9% for proliferative features, whereas peripheral nonperfusion appears to be larger by at least 48.3 disc areas in eyes with proliferative diabetic retinopathy, with eyes with neovascularization of the optic disc having the largest total area of retinal nonperfusion.
Meaning
These results suggest that eyes with nonproliferative diabetic retinopathy to a threshold of 118.3 disc areas of retinal nonperfusion are at an increased risk of proliferative diabetic retinopathy.
Abstract
Importance
Threshold of retinal nonperfusion for the development of proliferative diabetic retinopathy (PDR) is unclear.
Objectives
To identify a threshold of retinal nonperfusion for the presence of retinal neovascularization and the distribution and area of retinal nonperfusion in eyes with severe nonproliferative diabetic retinopathy (NPDR), PDR, neovascularization of the optic disc (NVD), and retinal neovascularization elsewhere (NVE).
Design, Setting, and Participants
This cross-sectional image analysis study was performed between September 24, 2018, and October 24, 2018, at a multicenter national study in the United Kingdom. Baseline images were obtained from 2 completed randomized clinical trials (Ranibizumab for Diabetic Macular Edema Panretinal Photocoagulation [RDP] study and Clinical Efficacy of Intravitreal Aflibercept vs Panretinal Photocoagulation for Best Corrected Visual Acuity in Patients With Proliferative Diabetic Retinopathy at 52 Weeks [CLARITY] study). The RDP study recruited eyes with severe NPDR between April 1, 2014, and December 31, 2015, and the CLARITY study recruited eyes with PDR between August 22, 2014, and November 20, 2015. Ultra-widefield angiography images of eyes with no prior panretinal photocoagulation treatment were included.
Main Outcomes and Measures
The total area of retinal nonperfusion, the area of posterior pole retinal nonperfusion, and the area of peripheral retinal nonperfusion were measured.
Results
A total of 92 patients (92 eyes) were included in the study: 59 in the PDR group (mean [SD] age, 42 [15] years; 20 female [33.9%]) and 33 in the NPDR group (mean [SD] age, 63 [10] years; 3 female [9.1%]). Forty eyes had NVE and 19 had NVD with or without NVE. We identified a retinal nonperfusion threshold of 118.3 disc areas (DA) with a specificity of 84.9% (95% CI, 68.1% to 94.9%) for PDR. The median area of retinal nonperfusion was 67.8 DA (95% CI, 44.2 to 107.3 DA) in the NPDR eyes and 147.9 DA (95% CI, 127.4 to 173.5 DA) for eyes with proliferative changes, with a difference of 69.0 DA (95% CI, 42.2 to 97.7 DA; P < .001). No difference was found in the median area of posterior nonperfusion between NPDR and PDR, with a difference of 0 DA (95% CI, −6.7 to 5.2 DA; P = .56). As for peripheral nonperfusion, NPDR eyes measured 64.1 DA and PDR eyes measured 130.6 DA, with a difference of 70.8 DA (95% CI, 48.4 to 94.9 DA; P < .001). Eyes with NVD had the largest total area of retinal nonperfusion, with a difference of 65.1 DA (95% CI, 28.6 to 95.8 DA; P < .001) compared with eyes with only NVE.
Conclusions and Relevance
These findings suggest eyes with at least 107.3 DA of nonperfusion are at risk of proliferative disease, and eyes with NVD have the largest area of retinal nonperfusion.
Introduction
Proliferative diabetic retinopathy (PDR) is a complication of diabetes that can result in severe sight impairment if not identified early and treated optimally. The prevalence and incidence of this condition vary depending on the population studied, length of follow-up, baseline retinopathy, and several other factors.1,2 With the increasing prevalence of diabetes globally, the increasing number of people developing PDR in the next decade is concerning.1
The current risk estimates of progression from nonproliferative diabetic retinopathy (NPDR) to PDR are based on the baseline grading of NPDR on color fundus photographs and systemic risk factors.2 Although these are important factors, more accurate determination of PDR is required to personalize new treatment options that are being evaluated to decrease this disease progression. The location of retinal capillary nonperfusion on fluorescein angiography (FA) is related to the presence of retinal neovascularization in eyes with PDR.3 However, the threshold of nonperfusion required for conversion from NPDR to PDR remains unclear.
On color fundus photographs, PDR may present as neovascularization of the optic disc (NVD) and/or retinal neovascularization elsewhere (NVE). High-risk characteristics of PDR are (1) eyes with neovascularization and preretinal or vitreous hemorrhage and (2) eyes with NVD on or within 1 disc diameter of the optic disc equaling or exceeding one-fourth to one-third of a disc area (DA), even in the absence of preretinal or vitreous hemorrhage, as graded on 7-field color photographs.4 To date, high-risk features of PDR on ultra-widefield angiography have not been systematically defined.5
The amount of retinal nonperfusion on FA may explain why some patients develop only NVE and others develop NVD as the first manifestation of PDR. However, the reasons for this from an angiographic or pathogenic point of view remain unclear. The objective of this study was to identify a sensitive and specific threshold of total area of retinal nonperfusion that is required for any retinal neovascularization to develop on ultra-widefield angiography. In addition, we studied the distribution of retinal nonperfusion by evaluating the total area of nonperfusion, posterior segment nonperfusion, and peripheral nonperfusion in eyes with NPDR, eyes with NVE alone, eyes with both NVD and NVE, and eyes with NVD alone.
Methods
This was a retrospective, post hoc, cross-sectional comparative study of baseline retinal images obtained in 2 randomized clinical trials. This study was performed at the Institute of Ophthalmology, University College London, London, United Kingdom. The baseline images analyzed were from the Ranibizumab for Diabetic Macular Edema Panretinal Photocoagulation (RDP) study taken between April 1, 2014, and December 31, 2015, and the Clinical Efficacy of Intravitreal Aflibercept vs Panretinal Photocoagulation for Best Corrected Visual Acuity in Patients With Proliferative Diabetic Retinopathy at 52 Weeks (CLARITY) study, taken between August 22, 2014, and November 20, 2015.6 This image analysis study was performed between September 24, 2018, and October 24, 2018. Written informed consent was obtained from patients during the original studies. Ethical approval was obtained to use these images without further consent in this study. Institutional review board approval was obtained from the London-Westminster Research Ethics Committee for analysis of anonymized data, and the study was conducted in accordance with the tenets of the Declaration of Helsinki.7
Inclusion and Exclusion Criteria
Images were obtained from the RDP and CLARITY studies.6 In short, the RDP study is a clinical trial that recruited eyes with clinically apparent severe NPDR (Early Treatment Diabetic Retinopathy Study [ETDRS] grade 47-53), whereas the CLARITY study recruited patients with PDR. Both studies excluded eyes that received intravitreal anti–vascular endothelial growth factor (VEGF) treatment within 4 months of screening.
Inclusion criteria for our current study were baseline images of eyes with NPDR or PDR from these clinical trials that were imaged with the Optos ultra-widefield system (Optos Plc). Images were assessed for quality and only included if the images were sharp enough to clearly distinguish areas of perfused retina from nonperfused retina. Therefore, images with ungradable secondary significant media opacity, such as a vitreous hemorrhage or cataract and significant obstruction from the ocular adnexa, were excluded. A single fovea-centered FA image captured between 30 and 90 seconds for each eye was included. Exclusion criteria were previous panretinal photocoagulation and no areas of retinal nonperfusion. Eyes with coexisting retinal pathologic features, such as a visible epiretinal membrane that caused retinal vessel distortion, were excluded.
Image Acquisition
The ultra-widefield angiography images were obtained using an ultra-widefield system (mode 200TX; Optos Plc), observing a standard protocol after intravenous bolus infusion of 5 mL of fluorescein sodium, 20%. The protocol consisted of acquiring images in transit phase, arteriovenous phases, and late frames. A single investigator (L.N.) identified the best macula-centered FA image in the arteriovenous phase from the FA series of each eligible eye.
Image Processing
A correction factor was automatically applied for the flattening of the 3-dimensional image to a 2-dimensional image (Optos V2 Vantage Pro software; Optos Plc). The concentric ring template was then added to each image according to previously described methods.8 In brief, this validated method incorporates a macular ring with a radius of 2.5 disc diameters and 5 additional concentric rings (rings 1-5), each with a 2.5–disc diameter increment in the radius. Each of the 6 rings (rings M and 1-5) are divided into 12 segments. Each segment is graded as ungradable, not perfused, or perfused if 50% or more of the segment is involved.8 In addition to quantifying nonperfusion, the concentric ring method allows documentation of location of nonperfusion. The area of each cell in each concentric ring was corrected based on the enlargement factor identified using 3-dimensional printed model eyes.4,9 The montaged image using a steering protocol to obtain 3 images (superior, central, and inferior) introduces projection distortions that as yet have not been studied or validated because the location of the same area of the retina is distorted differently in different images given the different angular location. Therefore, only the central image was used in this study because the projection distortion in this area has previously been studied, and we have made the appropriate corrections.
Image Analysis
Images were categorized as NPDR and PDR. In the PDR group, eyes with new vessels elsewhere alone were classified as NVE and eyes with new vessels on the disc were classified as NVD. These eyes may have new vessels on the disc alone or new vessels on the disc and elsewhere. The images were graded by a retinal specialist (L.N.) using the previously validated concentric rings, which have an intergrader agreement of 0.910 and a test-retest reliability of 0.975.8 The total area of nonperfusion, the posterior pole nonperfusion area, and the peripheral area of nonperfusion were obtained for each of these categories. The posterior pole is represented by rings M and 1, and the peripheral retina is represented by rings 2 through 4. Ring 5 was excluded from the analysis because only the central ultra-widefield image is used, which has a high number of ungradable segments in ring 5, as reported previously.9
Statistical Analysis
The mean and 95% CI were used to describe data from each category. Continuous data were assessed for normality. Comparisons between 2 groups were performed using the Mann-Whitney test, whereas comparisons of multiple (>2) groups were performed using the Kruskal-Wallis test. P values were 2-tailed, and adjustments were not made for multiple analyses. The Youden index was determined to identify the area of retinal nonperfusion threshold with best sensitivity and specificity in identifying PDR. Area under the receiver operating characteristic was used to present the reliability of retinal nonperfusion as a determinant for PDR. Statistical analyses were performed using MedCalc for Windows, version 15.0 (MedCalc Software).
Results
A total of 92 patients were included in the study: 59 in the PDR group (mean [SD] age, 49 [15] years; 20 female [33.9%]) and 33 in the NPDR group (mean [SD] age, 63 [10] years; 3 female [9.1%]). Median visual acuity was 72 (interquartile range, 67-79) in the NPDR group and 84 (interquartile range, 79-87) in the PDR group. A total of 25 patients (42.4%) in the PDR group had diabetes. From the CLARITY study, baseline images from 232 eyes were reviewed, and 123 eyes did not receive prior PDR treatment. Of these, 76 eyes were imaged with the ultra-widefield system, and 59 fulfilled the inclusion criteria of obtaining an image of sufficient quality to grade. From the RDP study, 51 eyes were reviewed, and 33 fulfilled the inclusion criteria. Ninety-two eyes were included in the analysis: 33 were eyes with NPDR and 59 were eyes with PDR, of which 40 had NVE and 19 had NVD.
The median total area of retinal nonperfusion was 67.8 DA (95% CI, 44.2-107.3 DA) for eyes with NPDR and 147.9 DA (95% CI, 127.4-173.5 DA) for eyes with PDR, with a difference of 69.0 DA (95% CI, 42.2-96.7 DA; P < .001). Comparing the distribution of retinal nonperfusion, the median area of posterior pole nonperfusion, represented by rings M and 1, was 15.6 DA (95% CI, 10.4-21.9 DA) in the NPDR group and 10.4 DA (95% CI, 10.4-20.8 DA) in the PDR group, with a difference of 0.0 DA (95% CI, −6.7 to 5.2; P = .56). For peripheral nonperfusion, represented by rings 2 through 4, the median area in the NPDR group measured 50.1 DA (95% CI, 35.3-77.0 DA), and the median area of peripheral nonperfusion in the PDR group measured 136.4 DA (95% CI, 113.0-153.2 DA), with a difference of 70.8 DA (95% CI, 48.4-94.9; P < .001).
A graph representing the area under the receiver operating characteristic curve using the total area of retinal nonperfusion to identify PDR is presented in Figure 1. Using the Youden index, the associated criterion for total area of retinal nonperfusion was 118.3 DA, with a specificity of 84.9% (95% CI, 68.1%-94.9%) and a sensitivity of 66.1% (95% CI, 52.6%-77.9%) for PDR. The sensitivity and specificity of all the cutoff points generated from the receiver operating characteristic curve are represented in Table 1.
Figure 1. Area Under the Receiver Operating Characteristic (AUROC) Curve Using the Total Area of Retinal Nonperfusion and Peripheral Nonperfusion.
Table 1. Sensitivity and Specificity for Total Area of Retinal Nonperfusion in Identifying Proliferative Diabetic Retinopathy Based on the Area Under the Receiver Operating Characteristic Curve.
| Criterion, DA | Sensitivity (95% CI), % | Specificity (95% CI), % |
|---|---|---|
| ≥12.99 | 100 (93.9-100.0) | 0 (0.0-10.6) |
| >12.99 | 100 (93.9-100.0) | 3.01 (0.08-15.8) |
| >22.46 | 98.3 (90.9-100.0) | 6.1 (0.7-20.2) |
| >25.04 | 98.3 (90.9-100.0) | 12.1 (3.4-28.2) |
| >27.67 | 96.6 (88.3-99.6) | 12.1 (3.4-28.2) |
| >31.93 | 96.6 (88.3-99.6) | 21.2 (9.0-38.9) |
| >32.82 | 94.9 (85.9-98.9) | 21.2 (9.0-38.9) |
| >43.17 | 94.9 (85.9-98.9) | 33.3 (18.0-51.8) |
| >43.97 | 93.2 (83.5-98.1) | 33.3 (18.0-51.8) |
| >48.44 | 93.2 (83.5-98.1) | 39.4 (22.9-57.9) |
| >49.13 | 91.5 (81.3-97.2) | 39.4 (22.9-57.9) |
| >49.22 | 91.5 (81.3-97.2) | 42.4 (25.5-60.8) |
| >49.33 | 89.8 (79.2-96.2) | 42.4 (25.5-60.8) |
| >62.17 | 89.8 (79.2-96.2) | 45.4 (28.1-63.6) |
| >64.69 | 86.4 (75.0-94.0) | 45.4 (28.1-63.6) |
| >73.86 | 86.4 (75.0-94.0) | 57.6 (39.2-74.5) |
| >75.99 | 83.0 (71.0-91.6) | 57.6 (39.2-74.5) |
| >76.11 | 83.0 (71.0-91.6) | 60.6 (42.1-77.1) |
| >94.88 | 76.3 (63.4-86.4) | 60.6 (42.1-77.1) |
| >100.36 | 76.3 (63.4-86.4) | 66.7 (48.2-82.0) |
| >109.81 | 71.2 (57.9-82.2) | 66.7 (48.2-82.0) |
| >112.19 | 71.2 (57.9-82.2) | 69.7 (51.3-84.4) |
| >112.93 | 69.5 (56.1-80.8) | 69.7 (51.3-84.4) |
| >113.19 | 69.5 (56.1-80.8) | 72.7 (54.5-86.7) |
| >114.9 | 67.8 (54.4-79.4) | 72.7 (54.5-86.7) |
| >116.64 | 67.8 (54.4-79.4) | 75.8 (57.7-88.9) |
| >116.66 | 66.1 (52.6-77.9) | 75.8 (57.7-88.9) |
| >118.34 | 66.1 (52.6-77.9) | 84.8 (68.1-94.9) |
| >147.89 | 49.2 (35.9-62.5) | 84.8 (68.1-94.9) |
| >149.66 | 49.2 (35.9-62.5) | 90.9 (75.7-98.1) |
| >162.15 | 44.1 (31.2-57.6) | 90.9 (75.7-98.1) |
| >162.26 | 44.1 (31.2-57.6) | 93.9 (79.8-99.3) |
| >201.28 | 17.0 (8.4-29.0) | 93.9 (79.8-99.3) |
| >205.38 | 17.0 (8.4-29.0) | 100 (89.4-100.0) |
| >265.93 | 0 (0.0-6.1) | 100 (89.4-100.0) |
Abbreviation: DA, disc areas.
Comparing NPDR eyes, eyes with NVE only, and eyes that developed NVD with or without NVE, a difference was found in total area of nonperfusion, posterior pole, and peripheral nonperfusion. The median total area was 127.4 DA (95% CI, 102.5-159.8 DA) for the NVE group and 191.8 DA (95% CI, 152.5-227.4 DA) for the NVD group, and the difference was 65.1 DA (95% CI, 28.6-95.8 DA; P < .001).
Reviewing the distribution of nonperfusion among these groups, the median area of posterior pole nonperfusion was 10.4 DA (95% CI, 1.8-13.9 DA) in the NVE group and 26.1 DA (95% CI, 10.4-33.8 DA) for the NVD group, and the difference was 10.4 DA (95% CI, 5.2-20.8 DA; P = .004). For the periphery, the median area of nonperfusion was 113.0 DA (95% CI, 94.3-140.3 DA) for the NVE group and 157.0 DA (95% CI, 143.2-193.5 DA) for the NVD group, with a difference of 50.9 DA (95% CI, 17.4-76.9 DA; P = .002). The distribution of nonperfusion among the groups is presented in Table 2.
Table 2. Median Total Area, Posterior Pole, and Peripheral Area of Retinal Nonperfusion in Eyes With NPDR, PDR, NVE, and NVD .
| Variable | Median (95% CI) | P Value | Median (95% CI) | P Value | ||||
|---|---|---|---|---|---|---|---|---|
| NPDR (n = 33) | PDR (n = 59) | Difference Between NPDR and PDR | NVE (n = 40) | NVD (n = 19) | Difference Between NVE and NVD | |||
| Median total area of retinal nonperfusion, DA | 67.8 (44.2 to 107.3) | 147.9 (127.4 to 173.5) | 69.0 (42.2 to 96.7) | <.001 | 127.4 (102.5 to 159.8) | 191.8 (152.5 to 227.4) | 65.1 (28.6 to 95.8) | <.001 |
| Median posterior pole retinal nonperfusion, DA | 15.6 (10.4 to 21.9) | 10.4 (10.4 to 20.8) | 0.0 (−6.7 to 5.2) | .56 | 10.4 (1.8 to 13.9) | 26.1 (10.4 to 33.8) | 10.4 (5.2 to 20.8) | .004 |
| Median peripheral retinal nonperfusion, DA | 50.1 (35.3 to 77.0) | 136.4 (113.0 to 153.2) | 70.8 (48.4 to 94.9) | <.001 | 113.0 (94.3 to 140.3) | 157.0 (143.2 to 193.5) | 50.9 (17.4 to 76.9) | .002 |
Abbreviations: DA, disc areas; NPDR, nonproliferative diabetic retinopathy; NVD, neovascularization of the optic disc; NVE, neovascularization elsewhere; PDR, proliferative diabetic retinopathy.
Discussion
This study suggests that the total area of retinal nonperfusion is greater among eyes with PDR than among eyes with severe NPDR. However, the difference was predominantly seen in the periphery, with the area of peripheral nonperfusion being significantly higher in eyes with PDR than with NPDR, whereas posterior pole nonperfusion was not significantly different. Although a larger area of retinal nonperfusion has been reported in eyes with PDR, retinal nonperfusion in the periphery appears to be the determining factor in PDR.10 Silva et al11 have also published reports on predominantly peripheral lesions, introducing a 4.7-fold increased risk of developing PDR during 4 years in eyes with nonproliferative changes. The same group also found that predominantly peripheral lesions were associated with a significantly larger area of retinal nonperfusion.12 Our results identified the significance of peripheral nonperfusion as a factor for neovascularization in PDR compared with NPDR, which corroborates the findings by previous groups.3,12 On a separate note, the study by Silva et al11 did not find a difference between the severe NPDR and PDR groups and suggested a plateau of the area of nonperfusion. However, only 16 eyes had PDR in that cohort, which were divided into eyes with high-risk characteristics (n = 3) and eyes without high-risk characteristics (n = 13). The difference observed between our report and the study by Silva et al11 is likely to be secondary to a larger number of eyes with PDR (n = 59) studied in our cohort, thereby probably producing a more reliable result.
We also identified 118.3 DA as a possible threshold with a good specificity of 84.9% for the identification of proliferative changes. Severe NPDR poses an interesting challenge to practitioners because the progression to PDR is a concern. The ETDRS13 suggested that panretinal photocoagulation should be considered in severe NPDR, but the risks and benefits are roughly balanced. Supported by the use of widefield angiography, eyes close to 118.3 DA or at least 107.3 DA, which is the upper limit of the 95% CI of the area of nonperfusion in the NPDR cohort, are presumably at higher risk than eyes at a much lower total area of nonperfusion despite both being classified clinically as severe NPDR. We acknowledge that this is merely an extrapolation of our results, and risk ratios will need to be studied longitudinally before such a conclusion can be made.
There is a difference in the total area of nonperfusion, posterior nonperfusion, and peripheral nonperfusion between eyes with NVE only and eyes with NVD with or without NVE. As discussed above, peripheral nonperfusion appears to be the key difference in NPDR and PDR. However, when studying eyes with NVE and NVD, both posterior and peripheral areas of nonperfusion were significantly larger in eyes with NVD. Peripheral ischemia is an important factor for proliferative changes, and the presence of posterior nonperfusion is a feature in the development of NVD. This is represented in Figure 2, which displays the area of retinal nonperfusion in eyes with NPDR, NVE, and NVD. This finding does not come as a surprise because neovascularization is believed to occur at the margins of perfusion and nonperfusion. Therefore, new vessels on the disc are likely to occur in marked ischemia that involves the peripheral and posterior retina. This finding differs in isolated NVE, in which the posterior retina is not markedly affected compared with the periphery; therefore, the neovascular process occurs at the margins and, in this case, the retina. The reason some eyes develop NVD with no NVE is an interesting discussion. We studied eyes with NVE and NVD and found that posterior pole nonperfusion is unique to eyes with NVD. Niki et al14 reported several different phenotypes in the distribution of ischemia in diabetic retinopathy, most of which was classified as midperipheral ischemia (61.2%) and central ischemia (26.3%). In short, the involvement of new vessels on the disc alone suggests a larger area of retinal nonperfusion that involves the periphery and posterior pole as opposed to eyes with NVE, in which the predominant area of nonperfusion is in the periphery.
Figure 2. Examples of the Distribution of Nonperfusion in Eyes .
A, Nonproliferative diabetic retinopathy (NPDR). B, Neovascularization elsewhere (NVE) (middle). C, Neovascularization of the optic disc (NVD). The posterior nonperfusion is marked in blue, and peripheral nonperfusion is marked in green.
There are several clinical implications from the findings of this report. Eyes with NVE or NVD or both have always been considered under one umbrella for proliferative disease with high-risk features traditionally based on size and presence of hemorrhage. Although this may still be true, the simple presence of NVD alone with no NVE signifies a much larger total area of retinal nonperfusion and involvement of the posterior pole. These eyes may require more aggressive or more panretinal laser treatment than eyes with NVE alone, even involving the posterior pole outside the arcades, although this was not specifically tested in this study. In addition, in the advent of anti-VEGF therapy for diabetic macular edema, reports15,16 have been published of 2-step improvements in diabetic retinopathy severity with anti-VEGF treatment. Diabetic retinopathy severity is graded purely on clinical appearance, which is likely to represent the masking effect of anti-VEGF because the angiographic correlate of this improvement of diabetic retinopathy severity is inconclusive, with contrasting reports17,18,19 regarding reperfusion with anti-VEGF treatment. Therefore, the importance of identifying a threshold for the development of neovascularization is essential in eyes undergoing treatment because cessation of treatment in eyes with more than 100 DA of retinal nonperfusion requires close observation.
Limitations
We acknowledge the limitations involved in this post hoc image analysis study on 2 separate clinical studies (RDP and CLARITY studies). However, these images were obtained as per study protocol and by certified photographers and may be more reliable than routinely performed FA. Although several factors may contribute to the progression or development of retinal nonperfusion, such as glycemic control, hypertension, and type of diabetes, we do not believe that this will confound the results of this study because we looked purely at a cross-sectional area of retinal nonperfusion and neovascularization. Factors such as systemic oxygenation or anemia may contribute to neovascularization, but these data were not collected, which is a limitation of the study. Another limitation is the cross-sectional design of this retrospective study. A prospective cohort study evaluating the progression of retinal nonperfusion and the development of neovascularization may better aid our understanding of the key factors for neovascularization, especially in NVD, NVE, or both. Moreover, this study demonstrates the need for development of a validated computerized tool to measure the total area and distribution of retinal nonperfusion on widefield angiography or widefield optical coherence tomography angiography to aid better stratification of these patients, especially in clinical trials evaluating new treatment options for this high-risk group.
Conclusion
A total of 118.3 DA of total retinal capillary nonperfusion is a specific threshold for the identification of PDR, and eyes with at least 107.3 DA of retinal nonperfusion are at risk for proliferative disease. Peripheral nonperfusion is key in the development of PDR; however, NVD is determined by posterior pole nonperfusion.
References
- 1.Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006;3(11):e442. doi: 10.1371/journal.pmed.0030442 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Harris Nwanyanwu K, Talwar N, Gardner TW, Wrobel JS, Herman WH, Stein JD. Predicting development of proliferative diabetic retinopathy. Diabetes Care. 2013;36(6):1562-1568. doi: 10.2337/dc12-0790 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Lange J, Hadziahmetovic M, Zhang J, Li W. Region-specific ischemia, neovascularization and macular oedema in treatment-naïve proliferative diabetic retinopathy. Clin Exp Ophthalmol. 2018;46(7):757-766. doi: 10.1111/ceo.13168 [DOI] [PubMed] [Google Scholar]
- 4.Nicholson L, Vazquez-Alfageme C, Clemo M, et al. Quantifying retinal area in ultra-widefield imaging using a 3-dimensional (3-D) printed eye model. Ophthalmol Retina. 2018;2:65-71. doi: 10.1016/j.oret.2017.03.011 [DOI] [PubMed] [Google Scholar]
- 5.The Diabetic Retinopathy Study Research Group Photocoagulation treatment of proliferative diabetic retinopathy: clinical application of Diabetic Retinopathy Study (DRS) findings, DRS Report Number 8. Ophthalmology. 1981;88(7):583-600. doi: 10.1016/S0161-6420(81)34978-1 [DOI] [PubMed] [Google Scholar]
- 6.Sivaprasad S, Prevost AT, Vasconcelos JC, et al. ; CLARITY Study Group . Clinical efficacy of intravitreal aflibercept versus panretinal photocoagulation for best corrected visual acuity in patients with proliferative diabetic retinopathy at 52 weeks (CLARITY): a multicentre, single-blinded, randomised, controlled, phase 2b, non-inferiority trial. Lancet. 2017;389(10085):2193-2203. doi: 10.1016/S0140-6736(17)31193-5 [DOI] [PubMed] [Google Scholar]
- 7.World Medical Association World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects. JAMA. 2013;310(20):2191-2194. doi: 10.1001/jama.2013.281053 [DOI] [PubMed] [Google Scholar]
- 8.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. doi: 10.1016/j.ajo.2015.09.003 [DOI] [PubMed] [Google Scholar]
- 9.Nicholson L, Vazquez-Alfageme C, Patrao NV, et al. Retinal nonperfusion in the posterior pole is associated with increased risk of neovascularization in central retinal vein occlusion. Am J Ophthalmol. 2017;182:118-125. doi: 10.1016/j.ajo.2017.07.015 [DOI] [PubMed] [Google Scholar]
- 10.Shimizu K, Kobayashi Y, Muraoka K. Midperipheral fundus involvement in diabetic retinopathy. Ophthalmology. 1981;88(7):601-612. doi: 10.1016/S0161-6420(81)34983-5 [DOI] [PubMed] [Google Scholar]
- 11.Silva PS, Cavallerano JD, Haddad NMN, 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. doi: 10.1016/j.ophtha.2015.01.008 [DOI] [PubMed] [Google Scholar]
- 12.Silva PS, Dela Cruz AJ, Ledesma MG, et al. Diabetic retinopathy severity and peripheral lesions are associated with nonperfusion on ultrawide field angiography. Ophthalmology. 2015;122(12):2465-2472. doi: 10.1016/j.ophtha.2015.07.034 [DOI] [PubMed] [Google Scholar]
- 13.Early Treatment Diabetic Retinopathy Study Research Group Early photocoagulation for diabetic retinopathy. ETDRS report number 9. Ophthalmology. 1991;98(5)(suppl):766-785. doi: 10.1016/S0161-6420(13)38011-7 [DOI] [PubMed] [Google Scholar]
- 14.Niki T, Muraoka K, Shimizu K. Distribution of capillary nonperfusion in early-stage diabetic retinopathy. Ophthalmology. 1984;91(12):1431-1439. doi: 10.1016/S0161-6420(84)34126-4 [DOI] [PubMed] [Google Scholar]
- 15.Wykoff CC, Eichenbaum DA, Roth DB, Hill L, Fung AE, Haskova Z. Ranibizumab induces regression of diabetic retinopathy in most patients at high risk of progression to proliferative diabetic retinopathy. Ophthalmol Retina. 2018;2(10):997-1009. doi: 10.1016/j.oret.2018.06.005 [DOI] [PubMed] [Google Scholar]
- 16.Ip MS, Zhang J, Ehrlich JS. The clinical importance of changes in diabetic retinopathy severity score. Ophthalmology. 2017;124(5):596-603. doi: 10.1016/j.ophtha.2017.01.003 [DOI] [PubMed] [Google Scholar]
- 17.Campochiaro PA, Wykoff CC, Shapiro H, Rubio RG, Ehrlich JS. Neutralization of vascular endothelial growth factor slows progression of retinal nonperfusion in patients with diabetic macular edema. Ophthalmology. 2014;121(9):1783-1789. doi: 10.1016/j.ophtha.2014.03.021 [DOI] [PubMed] [Google Scholar]
- 18.Nicholson L, Crosby-Nwaobi R, Vasconcelos JC, et al. Mechanistic evaluation of panretinal photocoagulation versus aflibercept in proliferative diabetic retinopathy: CLARITY substudy. Invest Ophthalmol Vis Sci. 2018;59(10):4277-4284. doi: 10.1167/iovs.17-23509 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Mir TA, Kherani S, Hafiz G, et al. Changes in retinal nonperfusion associated with suppression of vascular endothelial growth factor in retinal vein occlusion. Ophthalmology. 2016;123(3):625-34.e1. doi: 10.1016/j.ophtha.2015.10.030 [DOI] [PMC free article] [PubMed] [Google Scholar]