If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanCollege of Medicine, Chang Gung University, Taoyuan, Taiwan
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanCollege of Medicine, Chang Gung University, Taoyuan, Taiwan
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanDepartment of Ophthalmology, Chang Gung Memorial Hospital, Keelung, Taiwan
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanCollege of Medicine, Chang Gung University, Taoyuan, Taiwan
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanCollege of Medicine, Chang Gung University, Taoyuan, Taiwan
Department of Ophthalmology, Chang Gung Memorial Hospital, Linkou Medical Center, Taoyuan, TaiwanCollege of Medicine, Chang Gung University, Taoyuan, Taiwan
To investigate vascular abnormalities, including persistent avascular retina (PAR), in a large cohort of patients with regressed or treated retinopathy of prematurity (ROP) through long-term follow-up until school age.
Design
Retrospective large cohort study.
Participants
We included pediatric patients (<18 years old) with history of nontreated or treated ROP (treated with either photocoagulation or intravitreal injection) regularly followed up until 2020.
Methods
Upon enrollment, we categorized patients into four groups: prematurity, regressed ROP, and IVI and laser treatment ROP groups. All patients had undergone visual acuity examinations, optical coherence tomography, and ultra-widefield fluorescein angiography.
Main Outcome Measures
Percentage of eyes with PAR (≥area of 2 disc diameters from the ora serrata to vascular termini) and with vascular abnormalities in the peripheral and posterior retina.
Results
In total, we included 187 eyes of 95 patients. The prevalence of PAR was 0%, 33.33%, and 31.65% in eyes in the prematurity, regressed ROP, and IVI treatment groups, respectively (p < 0.0001). There was no significant difference in the percentage of PAR eyes when comparing regressed ROP group (33.33%) with the IVI treatment group (31.65%). At least one type of vascular abnormality was noted until school age in all (100%) treated ROP eyes. Although multivariate analysis found a significant association between IVI treatment and PAR (Odds Ratio [OR]: 10.28, 95% CI: 3.29–32.14) until the school age of 6-8 years old, there were no stage 3 eyes in the spontaneously-regressed group, suggesting that stage 3 ROP in the IVI group could also drive the association.
Conclusion
Approximately one-third of ROP eyes with spontaneous regression or IVI treatment still exhibit PAR when the child reaches school age. Several distinct vascular anomalies at the vascular–avascular juncture and within the vascularized retina can persist in these children. The clinical significance of these anomalies and the decision whether to treat them warrant further investigation to optimize their outcomes.
Retinopathy of prematurity (ROP) is a widely known retinal vascular disorder in preterm infants and a leading cause of visual disability or blindness in children. Advances in antenatal care have resulted in an increase in the survival rate of infants with extremely low birth weight. Consequently, the incidences of ROP and treatment-requiring ROP are increasing worldwide.
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
In certain patients with or without treatment, the retina may fail to fully vascularize or may develop vascular abnormalities, thus demonstrating persistent avascular retina (PAR) or anomalous vessel findings at the periphery. Because of the advent of technologies such as ultra-widefield fluorescein angiography (UWFFA) and spectral-domain optical coherence tomography (SD-OCT), persistent vascular abnormalities can be detected more readily and investigated.
The natural history of peripheral nonperfusion and vascular abnormalities, which persist beyond the acute phase of ROP, remains poorly understood. Although patients with history of type I ROP are more likely to exhibit abnormal foveal development and poor vision,
a discrepancy still exists between the structural and functional findings of vascular abnormalities in the long term. For instance, circumferential atypical vessels are detected in patients with spontaneously regressed ROP and favorable vision.
Follow-up to Age 4 Years of Treatment of Type 1 Retinopathy of Prematurity Intravitreal Bevacizumab Injection versus Laser: Fluorescein Angiographic Findings.
Blair MP, Hilkert Rodriguez S, Chan RVP, Shapiro MJ. Re: Lepore et al.: Follow-up to age 4 years of treatment of type 1 retinopathy of prematurity intravitreal bevacizumab injection versus laser: fluorescein angiographic findings (Ophthalmology. 2018;125:218-226). Ophthalmology 2018;125(10):e70-e1.
The significance of such vessel anomalies and whether they persist to school age remain uncertain.
Studies on vascular anomalies evaluated using fluorescein angiography (FA) in patients with treated or nontreated ROP are scant, and they have used a small sample.
Follow-up to Age 4 Years of Treatment of Type 1 Retinopathy of Prematurity Intravitreal Bevacizumab Injection versus Laser: Fluorescein Angiographic Findings.
In addition, few studies have investigated school-age children with ROP through UWFFA to elucidate the prevalence of vascular abnormalities in the long term. This study investigated persistent vascular abnormalities, including PAR, in a large cohort of patients with regressed or treated ROP through long-term follow-up until the patients had reached school age.
PATIENTS AND METHODS
Study Population and Grouping
This study adhered to the tenets of the Declaration of Helsinki and was approved by the institutional review board (IRB) of our main institution (IRB No. 201801566A3). We obtained written informed consent from the parent or guardian of each participant.
This study investigated pediatric patients (<18 years old) with history of nontreated or treated ROP (treated with either photocoagulation or intravitreal injection [IVI]) and regularly followed up between 2015 and 2020 at the medical center (Chang Gung Memorial Hospital, Linkou, Taoyuan, Taiwan). All recruited patients had undergone UWFFA at least once during their clinical follow-up and were free from other ocular diseases. We excluded patients from the study if they had history of cerebral damage or severe congenital defects that prevented their cooperation in the tests conducted, as well as those with history of adverse reactions to fluorescein sodium or with a detached retina upon examination. Both eyes were examined in each patient.
We recruited patients from our clinic during clinical follow-up and classified them into the following four groups: prematurity group, comprising control subjects with prematurity history (gestational age [GA] ≤ 37 weeks) but with no diagnosis of ROP; regressed ROP group, comprising children with regressed ROP who had not received any treatment; laser treatment ROP group, comprising children with history of threshold ROP treated with laser photocoagulation; and IVI treatment ROP group, comprising children with history of threshold ROP treated with IVI of anti-VEGF agents. If a patient had received any photocoagulation treatment after IVI, we categorized them into the laser group for vascular anomaly analysis only because photocoagulation is a permanent retina-scar-inducing ablative treatment that affects the evaluation of PAR.
Examinations of Visual Acuity, Refractive Errors, and Optical Components
Uncorrected visual acuity (VA) and best-corrected VA (BCVA) were measured for all patients. Automatic cycloplegic refraction was measured using an automatic keratorefractometer (KR-8100A, Topcon, Tokyo, Japan), and this was followed by measurement of clinical refraction to refine the outcome. We converted the Snellen VA into the logarithm of the minimum angle of resolution (logMAR) VA prior to statistical analysis. Medical data including GA, birth weight (BW), stage of maximal severity and zone affected in the acute disease and plus diseases were all recorded.
We determined macula status using SD-OCT (Triton, Topcon, Tokyo, Japan). For each eligible patient, both eyes were scanned after pharmaceutical papillary dilation. The examiner was masked, and imaging was performed by a single experienced technician. An OCT scan was repeated if the image was decentered or contained segmentation errors. We considered images with signal strength index < 50 as suboptimal and excluded them.
Ultra-Widefield Fundoscopy, FA Imaging, and Image Analysis
We employed a retinal imaging system (California, Optos, MA, USA) equipped with optomap imaging modalities including indocyanine green angiography (ICG), FA and composite color views comprising red-free and 200° views of the retina (approximately 80% of the entire retina) in a single image.
In accordance with the standardized protocol for FA, a bolus of 10% fluorescein solution at a dose of 0.1 mL/kg was administered through the intravenous (IV) route, and this was followed by administration of an isotonic saline flush. For patients intolerable to venipuncture for the IV route, this study performed an oral fluorescein protocol in accordance with previously published methods with modifications
Comparison between Oral and Intravenous Ultrawide-Field Fluorescein Angiography in the Clinical Follow-up of Children with a History of Retinopathy of Prematurity or Prematurity.
: two ampoules of 10% fluorescein dye (maximum dose of 25 mg/kg) were mixed with approximately 30 mL of orange juice. The mixture was ingested by the patient. We then obtained noncontact, high-resolution UWFFA images for both eyes of each patient. If more than a single series of UWFFA images was obtained, we included only the most recent study in the analysis.
Three masked, experienced retinal specialists (EYCK, YSH, and WWC) were responsible for examining and reviewing the images. If any discrepancies existed in the diagnosis, votes were casted to decide the final diagnosis. We processed all images by using ImageJ software (Image Jdev, Fiji contributors, USA) as the standardized protocol for review.
We determined interrater reliability through calculation of the Cohen kappa coefficient. Images with poor fluorescein quality were excluded if all reviewers agreed.
Definition of PAR
We adopted and modified the definition of PAR in UWFFA images from that in a previous study.
We first determined the vessel termini and ora serrata of the retina in each image acquired from the patients. The area between the vessel termini and ora serrata was measured in the unit of disc diameters (DD) as the nonperfusion area (Fig. 1). We made three separate measurements for all images and calculated the average area. This paper reports the smallest PAR within all images of the same eye to prevent overestimation.
Figure 1Definition of peripheral avascular retina (PAR). Color fundus (A) and fluorescent angiography (B) of the patient. First, the vessel termini (dotted curved white line) and ora serrata (area temporal to curved white line) were identified from ultra-widefield photography and fluorescein angiography images. The corresponding area between the vessel termini and ora serrata in fluorescein angiography was measured in units of disc diameters (DD; dotted yellow circle) and indicated the nonperfusion area. We defined PAR as the nonperfusion area with measurements ≥2.0 DD, which is three standard deviations more than the normal value.
reported that the physiologic avascular retina of children of age < 13 years usually extends ≤1.5 DD temporally or ≤1.0 DD nasally from the ora serrata. Therefore, we defined the PAR area as the nonperfusion area with measurements ≥2.0 DD, which is three standard deviations more than the normal value (Fig. 1).
Due to the ablative nature of laser photocoagulation therapy of the peripheral retina, the definition of PAR did not apply to retinas treated previously with laser in this cohort. This was because avascularity cannot be definitively determined in peripheral areas scarred with photocoagulation. Thus, we excluded laser-treated eyes from only the analysis of PAR.
Features of Vascular Abnormalities in FA Images
In addition to PAR, reviewers evaluated each set of FA images for vascular abnormalities at the vascular–avascular juncture and within the vascularized retina. For each image, if there was disagreement between the two reviewers, a third reviewer (KJC) evaluated the image again under blinded circumstances and determined the outcome.
The types of vascular abnormalities defined at the vascular–avascular juncture included vascular leakage, arteriovenous (AV) shunt formation, tangle formation, and abnormal branching (Fig. 2). This study examined the following features within the vascularized retina: capillary bed loss, abnormal hyperfluorescent lesions, abnormal choroidal flushing, tortuosity of vessels, and loss of the fovea avascular zone (Fig. 3).
Figure 2Series of ultra-widefield fluorescein angiography images from four school-age children with history of retinopathy of prematurity showing various features at the vascular–avascular junction. (A) Large area of persistent avascular retina (PAR) with massive leakage at the junction (white arrows). (B) Arteriolar–venular shunt (yellow arrows) observed at the juncture of vascular and avascular retina. (C) Enlarged picture. Tangle formation (white arrowheads) is evident at the juncture with prominent PAR. (D) Similarly extensive PAR but with abnormal branching (dichotomous-like; dotted yellow circle) of the peripheral vessels.
Figure 3Series of ultra-widefield fluorescein angiography images showing various features of vascular anomalies within the vascularized retina. (A) Prominent vessel tortuosity and multiple capillary loops (white arrows) in a 6-year-old boy treated with laser photocoagulation. (B) Capillary bed loss (yellow arrowhead) observed within the posterior pole. (C) Staining within the fovea and loss of the foveal avascular zone (dotted white circle). Additionally, regions with capillary bed loss (dotted yellow squares) and abnormal hyperfluorescent spots (yellow arrow) are observed. (D) Abnormal choroidal flushing (white arrowheads) can be seen within the vascularized retina in a 4-year-old girl with history of intravitreal bevacizumab injection. In addition, vascular leakage and tangle formation can be seen near the vascular–avascular juncture in this patient.
We performed statistical analyses using SAS software version 9.4 (SAS Institute Inc., Cary, NC, USA). The Kruskal–Wallis test was used to compare continuous variables such as age, GA, and BW between the groups. We employed chi-square analysis (χ2) or the Fisher exact test for comparing categorical outcomes between the four groups. Post hoc analysis was performed using the Bonferroni test and adjustment. We used univariate and multivariate logistic regressions to identify the risk factors for PAR. This study considered a p value of <0.05 to indicate statistical significance.
RESULTS
Patients and Treatment
A total of 187 eyes of 95 patients met our inclusion criteria and completed all examinations. These included 46 eyes in the prematurity group, 9 in the regressed ROP group, 53 in the laser treatment ROP group, and 79 in the IVI treatment ROP group. Male patients were predominant in each group, with comparable proportions among the groups (p = 0.335).
The mean GA and BW of the preterm controls (33.78 ± 3.75 weeks and 1994.96 ± 765.3 g, respectively) were significantly higher than those in the other groups (p < 0.0001). The mean ages at the time points of FA imaging and examination completion were 7.58 ± 1.98, 6.39 ± 1.05, 8.16 ± 1.35, and 6.7 ± 2.02 for the control, regressed ROP, laser treatment, and IVI treatment groups, respectively, with age in the laser treatment group significantly higher than that in the other three groups (p < 0.0001).
The mean spherical (−3.05 ± 3.67 D) and cylindrical (−1.71 ± 1.21 D) refractive errors were significantly higher in the laser treatment group than the other three groups (p < 0.001). The corrected Snellen VA and logMAR VA were similar in the IVI and laser treatment groups via post hoc analysis, but they were lower in these groups than in the preterm control and regressed ROP groups (p < 0.0001).
In the laser and IVI treatment groups, stage 3 (81.13% and 94.94%, respectively) and zone 2 ROP (84.91% and 75.95%, respectively) were the main form of ROP. The demographics and ocular characteristics of the four categorized groups upon their enrollment are shown in Table 1.
Table 1Demographics Data for Infants with ROP in Multiple Treatment Groups within Follow-Up (Total No. of Eyes = 187)
Variables
Prematurity
Regresssed ROP
ROP Treated with Laser
ROP Treated with IVI
P value
46 eyes
9 eyes
53 eyes
79 eyes
Male, n (%)
29 (63.04)
7 (77.78)
40 (75.47)
49 (62.03)
0.335
GA, weeks (mean ± SD)
33.78 ± 3.75
27.08 ± 2.16
25.66 ± 1.69
27.12 ± 2.76
<.0001
Birth weight, grams (mean ± SD)
1994.96 ± 765.3
1002.22 ± 258.06
747.28 ± 115.8
971.44 ± 316.96
<.0001
Age, y/o (mean ± SD)
7.58 ± 1.98
6.39 ± 1.05
8.16 ± 1.35
6.7 ± 2.02
<.0001
Oral FA, n (%)
15 (32.61)
3 (33.33)
12 (22.64)
19 (24.05)
0.63
Spherical power, D (mean ± SD)
0.19 ± 0.51
0.58 ± 0.7
-3.05 ± 3.67
-0.76 ± 2.93
<.0001
Cylindrical power, D (mean ± SD)
-0.77 ± 1.12
-0.78 ± 0.86
-1.71 ± 1.21
-0.88 ± 0.98
<.001
Axis (mean ± SD)
140.2 ± 68.44
106.43 ± 88.77
108.78 ± 80.53
122.72 ± 72.88
0.418
Corrected Snellen VA (mean ± SD)
0.98 ± 0.05
0.94 ± 0.1
0.81 ± 0.23
0.83 ± 0.24
<.0001
Corrected logMAR VA (mean ± SD)
0.01 ± 0.03
0.03 ± 0.05
0.12 ± 0.17
0.11 ± 0.19
<.0001
Stage, n (%)
<.0001
0
46 (100)
1 (11.11)
0 (0)
0 (0)
1
0 (0)
1 (11.11)
0 (0)
3 (3.79)
2
0 (0)
7 (77.78)
6 (11.32)
0 (0)
3
0 (0)
0 (0)
43 (81.13)
75 (94.94)
4
0 (0)
0 (0)
4 (7.55)
1 (1.27)
Zone, n (%)
<.0001
1
0 (0)
0 (0)
8 (15.09)
17 (21.52)
2
0 (0)
5 (55.56)
45 (84.91)
60 (75.95)
3
0 (0)
0 (0)
0 (0)
2 (2.53)
Plus disease, n (%)
0 (0)
0 (0)
42 (79.25)
57 (72.15)
0.196
Abbreviations: GA: Gestational age; IVI: Intravitreal injection; logMAR: Logarithm of the minimum angle of resolution; ROP: Retinopathy of prematurity; VA: Visual acuity.
Prevalence of PAR and Vascular Abnormalities under UWFFA analysis
UWFFA images of 187 eyes were available and reviewed. We evaluated the interrater reliability on the basis of substantial agreement regarding the occurrence of PAR and vascular abnormalities; the Cohen kappa coefficient was 0.95 (95% CI: 0.89–0.99).
We detected PAR in all groups with ROP except the preterm controls, and did not define PAR in the laser treatment group, as stated earlier. Three eyes (33.33%) in the regressed ROP group and 25 eyes (31.65%) in the IVI treatment group met the criteria for PAR. We found significance (p < 0.0001) only in comparisons with the preterm controls and discovered no statistical significance between the two groups with ROP (regressed ROP vs. IVI treatment ROP) in terms of PAR (Table 2).
Table 2Prevalence of Persistent Avascular Retina (PAR) and Vessels Abnormalities under Long-Term Follow-Up
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
Abbreviations: IVI: Intravitreal injection; ROP: Retinopathy of prematurity; N/A: Not applicable
1 Other vascular abnormalities defined as any one of the vascular abnormalities defined in the vascular-avascular juncture or the vascularized retina; excluding PAR
[A, B, C]: Inter-group comparisons using post hoc test; different (or the same) letters represent significant (or insignificant) statistical differences.
All treated ROP eyes exhibited at least one type of vascular abnormality as defined in this study. As many as 53 laser-treated eyes (100.0%) and 79 IVI-treated eyes (100.0%) were observed to have vascular abnormalities, which was significantly higher than the numbers in either the control (10.87%) or regressed ROP (66.67%) group (p < 0.0001). Post hoc analysis revealed that the proportion of regressed ROP eyes with vascular abnormalities was significantly higher than that for the controls (p = 0.001).
Subgroup Analysis of Eyes with Combined Treatment
Twenty-eight eyes received combined treatment (both IVI and laser), and they were categorized as laser-treated eyes in imaging analysis due to the long-term presence of a photocoagulation scar on the retina. Supplementary Table 1 is a subgroup table comprising the characteristics of these eyes.
From the subgroup analysis table, 28 eyes within the cohort had received combined treatment with IVI and laser. Out of these 28 eyes, 24 (85.71%) eyes were of stage 3 ROP and 22 eyes (62.96%) were of zone 2 ROP. Only 6 eyes (37.04%) were of zone 1 ROP, therefore contributing to 6 out of 8 eyes with zone 1 ROP in the laser group. The proportion of eyes which received laser after the initial IVI was 85.71%.
In a scenario whereby non-perfusion area > 2 DD within the photocoagulated retina was considered as analogous to PAR, 12 (42.86%) eyes of combined-treated ROP eyes were comparable to 9 (36.00%) eyes which were treated with laser only (p=0.61).
SD-OCT Analysis of Macula Thickness
The SD-OCT analysis revealed that the average central macula thickness was greatest in the laser-treated eyes followed by the regressed ROP eyes without treatment, IVI-treated eyes, and control eyes (Table 3; mean ± standard deviation thicknesses: 283.55 ± 33.07 μm, 253.56 ± 17.16 μm, and 253.43 ± 24.10 μm, 239.15 ± 20.51 μm, respectively, p < 0.0001). The measurements of retinal thicknesses in the other Early Treatment Diabetic Retinopathy Study areas (3 and 6 mm) exhibited similar trends among the four groups (Table 3).
Table 3Retinal Nerve Fiber Layer Thickness in the Macula Measured Using OCT in accordance to ETDRS Areas
Variables
Prematurity
Regresssed ROP
ROP Treated with Laser
ROP Treated with IVI
P value
46 eyes
9 eyes
53 eyes
79 eyes
OCT macula thickness, um (mean ± SD)
Central
239.15 ± 20.51
253.56 ± 17.16
283.55 ± 33.07
253.43 ± 24.1
<.0001
Temporal 3mm
305.15 ± 11.68
303.78 ± 16.95
310.11 ± 32.47
287.97 ± 20.43
<.0001
Nasal 3mm
316.3 ± 13.06
319.78 ± 17.85
325.09 ± 29.97
304.29 ± 22.79
<.0001
Superior 3mm
317.83 ± 12.27
320.33 ± 18.63
324.17 ± 26.12
301.75 ± 21.91
<.0001
Inferior 3mm
313.04 ± 12.18
312.56 ± 21.7
321.7 ± 46.75
297.79 ± 20.47
<.0001
Temporal 6mm
270.13 ± 11.33
271.44 ± 17.53
280.91 ± 26.12
259.95 ± 41.35
<.0001
Nasal 6mm
302.7 ± 14.59
303.44 ± 22.93
308.77 ± 28.7
287.03 ± 20.13
<.0001
Superior 6mm
286.33 ± 11.94
284.67 ± 19.1
295.55 ± 23.01
270.91 ± 18.03
<.0001
Inferior 6mm
275.46 ± 20.63
276.56 ± 16.54
281.55 ± 26.5
259 ± 19.32
<.0001
Abbreviations: ETDRS: Early treatment diabetic retinopathy study; IVI: Intravitreal injection; OCT: Optical coherence tomography; ROP: Retinopathy of prematurity
We report vascular findings from images of the same groups of eyes as in the PAR analysis. Regarding abnormalities at the vascular–avascular juncture of the eyes, the proportion of tangle formation in the IVI treatment group (51.90%) was higher than that in the regressed ROP (11.11%) and laser treatment (33.96%) groups (Table 4; p < 0.0001). We obtained similar results for the rate of abnormal branching (11.11%, 32.08%, and 53.16% for regressed ROP, laser treatment, and IVI treatment groups, respectively; p < 0.0001).
Table 4Vascular Abnormalities by Fluorescein Angiography at Vascular-Avascular Juncture of Retina
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
Abbreviations: AV: Arterio-venous; IVI: Intravitreal injection; ROP: Retinopathy of prematurity
[A, B, C]: Inter-group comparisons using post hoc test; different (or the same) letters represent significant (or insignificant) statistical differences.
A higher proportion of laser-treated eyes showed vascular leakage and AV shunting (52.83% and 29.11%, respectively) compared with IVI-treated eyes (29.11% and 26.58%, respectively). Post hoc analysis revealed significant differences among these two groups in terms of vascular leakage and AV shunting (p = 0.006 and 0.045, respectively).
We also detected vascular abnormalities in regressed ROP eyes, with two (22.22%) eyes presenting AV shunts, one (11.11%) eye having tangle formation, and one (11.11%) other eye exhibiting abnormal branching. Only one (2.17%) eye in the control group had an identifiable abnormal branching pattern in the peripheral area, but this proportion was still significantly lower than those in the laser and IVI treatment groups (p < 0.01 in post hoc analysis).
Abnormalities within the vascularized retina are detailed in Table 5. The percentage of capillary loops was significantly higher in the IVI treatment group than in the laser treatment and regressed ROP groups (73.42%, 47.17%, and 33.33%, respectively; p < 0.05). A significantly higher proportion of eyes in both the laser and IVI treatment groups had capillary dropouts, hyperfluorescent lesions, choroidal flushing, abnormal fovea vascularization, and vessel tortuosity than in the control and regressed ROP groups (p < 0.05). Post hoc analysis revealed no significant difference in hyperfluorescent lesions, choroidal flushing, abnormal fovea, and vessel tortuosity between the laser and IVI treatment groups.
Table 5Vascular Abnormalities by Fluorescein Angiography within Vascularized Retina
Variables
Prematurity
Regressed ROP
ROP Treated with Laser
ROP Treated with IVI
P value
46 eyes
9 eyes
53 eyes
79 eyes
Vascularized retina
Capillary loops, n (%)
0 (0)[A]
3 (33.33)[B]
25 (47.17)[B]
58 (73.42)[C]
<.0001
Capillary dropouts, n (%)
0 (0)[A]
0 (0)[A,B]
11 (20.75)[B]
24 (30.38)[B]
<.0001
Hyperfluorescent lesions, n (%)
1 (2.17)[A]
0 (0)[A,B]
14 (26.42)[C]
31 (39.24)[C]
<.0001
Abnormal choroidal flushing, n (%)
0 (0)[A]
0 (0)[A,B]
12 (22.64)[C]
29 (36.71)[C]
<.0001
Abnormal fovea vascularization, n (%)
0 (0)[A]
4 (44.44)[B]
33 (62.26)[B]
49 (62.03)[B]
<.0001
Vessel tortousity, n (%)
3 (6.52)[A]
4 (44.44)[B]
31 (58.49)[B]
34 (43.04)[B]
<.0001
Abbreviations: IVI: Intravitreal injection; ROP: Retinopathy of prematurity [A, B, C]: Inter-group comparisons using post hoc test; different (or the same) letters represent significant (or insignificant) statistical differences.
Regardless of treatment, all eyes with ROP history were significantly more likely to have abnormal fovea vascularization and vessel tortuosity than the preterm controls (p < 0.001). Post hoc analysis revealed no significant difference among the three groups (regressed ROP, laser treatment, and IVI treatment groups).
Risk Factor Analysis for Prevalence of PAR
In school-age children with nontreated or IVI-treated ROP, age at the time of examination, IVI treatment, zone 1 ROP and decreased central macula thickness as revealed by SD-OCT were significantly correlated with the prevalence of PAR in univariate logistic regression analysis (Table 6). We discovered no correlations of gender predilection, GA, BW, zone 1, occurrence of other vascular abnormalities, regressed ROP, or best corrected VA in logMAR with the prevalence of PAR.
Table 6Logistic Regression Analysis of Risk Factors for Persistent Avascular Retina (PAR)
We found a significant correlation between PAR and IVI treatment (OR: 10.28, 95% CI: 3.29–32.14) in multivariate logistic regression. We separately constructed two multivariate models with adjustment for central macula thickness and age at the time of examination to exclude any inherent collinearity relationship between these two variables. Both multivariate analyses revealed similar outcomes—that IVI treatment was an independent risk factors for the persistence of PAR until school age. We obtained similar outcomes in another model with adjustment for anti-VEGF agents, including bevacizumab, aflibercept, and ranibizumab (data not shown in the table). The results of multivariate analysis are shown in Table 6.
Although lasered eyes were excluded from the PAR logistic regression analysis, an additional logistic analysis was performed for when considering non-perfusion area > 2 DD within the photocoagulated retina as analogous to PAR (Supplementary Table 2). The additional multivariate analysis revealed that combined treatment (IVI and laser) eyes were not an independent risk factor (OR: 1.24, 95% CI: 0.40-3.84) for persistent non-perfusion area in the school age.
DISCUSSION
Our study found that PAR and other vascular abnormalities persisted to the age range of 6–8 years, suggesting the long-term effect of prematurity or treatment on retinal circulation in these patients. The prevalence of PAR was significantly correlated with IVI treatment in the analysis adjusted for other factors. Vascular abnormalities—including vascular leakage, shunt formation, capillary loops, and other abnormalities—were noted in all laser- or IVI-treated eyes and were shown to persist into school age. Vision, however, appeared to be stable even when some of the patients had persistent vascular activity with active dye leakage. To the best of our knowledge, this is the first study that has examined vascular abnormalities and the incidence of PAR following a standardized definition and has identified the risk factors for PAR until school age among patients with ROP.
In this study, the prevalence of PAR was 33.33% and 31.65% for eyes with regressed ROP and IVI treatment, respectively. There was no significant difference in the proportion of eyes between these two groups, implicating that the initial ROP pathogenesis process could also play a role in PAR. Few studies have investigated the prevalence of persistent PAR and vascular abnormalities in school-age ROP children. Hamad et al.
reported late-onset retinal findings such as retinal holes and tears in the avascular retina of untreated ROP eyes. They noted incomplete vascularization posterior to zone III in 71.6% of eyes in the cohort of patients with ROP with a mean age of 34.5 years. However, no strict definition of PAR was employed in the study.
Through FA, Lepore et al. found peripheral or posterior vascular abnormalities in all (100%) bevacizumab-treated eyes of children of age 4 years.
Follow-up to Age 4 Years of Treatment of Type 1 Retinopathy of Prematurity Intravitreal Bevacizumab Injection versus Laser: Fluorescein Angiographic Findings.
Similarly, a study examined the early sequelae of ROP through FA performed at the mean age of 7.7 years; the results revealed abnormal branching and leakage on the temporal peripheral retinas in spontaneously regressed ROP, but the frequency of these anomalies was not reported.
Until now, the natural course of retinal vascularization in both nontreated and treated ROP has remained uncertain, as has the molecular nature of its impact on retinal function, the farthest extent of retinal vascularization,
Comparison of fluorescein angiographic findings in type 1 and type 2 retinopathy of prematurity with intravitreal bevacizumab monotherapy and spontaneous regression.
This further highlights the need for the long-term investigation and longitudinal follow-up of these children.
The majority of studies have focused on FA images acquired within a brief period beyond the period specified in the standard monitoring guidelines for receiving treatment for ROP (i.e., during infancy).
Fluorescein angiographic observations of peripheral retinal vessel growth in infants after intravitreal injection of bevacizumab as sole therapy for zone I and posterior zone II retinopathy of prematurity.
Using similar criteria as those in our study (area of 2 DD from the ora serrata to vascular termini), Tahijia et al. reported avascular retina up to 24 weeks in 11 of 20 eyes (55%) receiving bevacizumab injection.
Fluorescein angiographic observations of peripheral retinal vessel growth in infants after intravitreal injection of bevacizumab as sole therapy for zone I and posterior zone II retinopathy of prematurity.
However, wide-field photography was not utilized in the study. These differences in the prevalence of PAR reported across various studies could be related to different demographics and ethnicities of the included patients, ages upon observation, and imaging systems used.
Our results demonstrated that IVI treatment with anti-VEGFs is an independent risk factor for PAR until the age of 6–8 years. This finding is consistent with the findings of previous studies showing that anomalies such as large avascular areas were present at 9 months of age and persisted up until 4 years of age.
Follow-up to Age 4 Years of Treatment of Type 1 Retinopathy of Prematurity Intravitreal Bevacizumab Injection versus Laser: Fluorescein Angiographic Findings.
However, owing to real world limitation, there were no stage 3 eyes in the spontaneously-regressed ROP group, while almost 95% eyes in the IVI-treated ROP group were of stage 3. Therefore, it may be that stage 3 ROP, rather than IVI, is the risk factor for PAR. In our regression analysis, we discovered no correlation of PAR with BCVA, central macula thicknesses, or determinants of prematurity, such as low GA and BW. To evaluate if combined treatment of IVI and laser might affect outcomes, a regression sub-analysis was performed in which the non-perfusion areas > 2DD (devoid of both choroidal and retinal perfusion) within the photocoagulated retina of lasered eyes were taken into consideration (Supplementary Table 1). The results revealed that combined therapy of IVI and laser during the acute phase was not an independent risk factor for peripheral non-perfusion area in school age. However, since laser-ablated areas of peripheral retina can preclude standardized evaluation of non-perfusion area, this outcome must be interpreted cautiously.
VEGF is an important angiogenic factor that regulates normal retinal vasculature development during the embryonic stage.
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
The long-term effect of anti-VEGF therapy on retinal vascularization in children with ROP was uncertain until now. The prevalence of PAR is postulated to be related to both ROP and disruption of VEGF action by anti-VEGF therapy during infancy because both the regressed ROP and IVI treatment ROP groups demonstrated PAR, but the prematurity group without ROP did not.
The clinical significance of these abnormal vascular lesions reported in ROP eyes, regardless of treatment, should be considered. Do angiographic anomalies still detected in school age need to be treated? Findings such as vascular leakage, tangle formation, or NV formation could be concerning when discovered during follow-up in children with ROP history. However, whether all these lesions require treatment remains unclear. In our cohort, two eyes each in the IVI and laser treatment groups (12.7% and 18.9%, respectively; Fig. 4) were treated during clinical follow-up, although all were asymptomatic. The main factor affecting whether these lesions should be treated was active vascular activity with significant leakage of NV in the eyes. However, the vision of most of these patients was not affected due to the peripheral location of the lesion.
Figure 4Ultra-widefield photograph of a 7-year-old child with retinopathy of prematurity and previously treated with laser photocoagulation (A) and early phase of fluorescein angiography (B) with leakage in late frames (C). Ultra-widefield photograph of an 8-year-old child with ROP treated with intravitreal anti-VEGF therapy (D) with early (E) and late leakage (F) as well as detectable peripheral avascular retina with neovascular tufts at the juncture. (Reproduced from Tsai et al. with permission.
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
Currently, insufficient evidence exists to support routine treatment of PAR in school-age children with treated or nontreated ROP because laser treatment is not without complications.
Our research raises important questions: are the lesions associated with a risk of reactivation in the future? When is treatment indicated, if ever? In a recent study, Hamad et al. postulated that factors contributing to late-onset complications of untreated ROP include atrophic holes within the peripheral avascular retina; thus, the authors recommended performing UWFFA examinations at regular intervals.
However, the study did not report the frequency of such complications in eyes treated with anti-VEGF and the benefits of prophylactic treatment. In the present study, BCVA in our patients was discovered to be uncorrelated with persistent PAR in multivariate analysis. Additional prospective studies with a standardized protocol and treatment strategy for PAR in ROP eyes might provide guidance on the clinical management of such patients.
The limitations of our study include its retrospective design, small number of enrolled patients, and the single ethnicity of the patient population. During image analysis, the image reviewers could not be masked to the treatment group because lasered-eyes had distinctive ablated peripheral retinae. In addition, because PAR cannot be clearly defined in photocoagulated eyes due to retinal tissue being scarred, a biased impression may have been introduced that PAR was not present in these eyes. In addition, the UWFFA images, which were acquired in school age (6-8 years old), can make evaluation of PAR extent (or area) in area/zone terms at this stage difficult, due to imaging artifacts such as eyelids artifacts and non-cooperativity. The number of eyes with spontaneously regressed ROP in this cohort was small, which might limit generalizability of the results with regards to this particular group. Two eyes with stage 3, zone 2 ROP without plus disease with spontaneous regression were followed-up but ultimately excluded for this study, due to no available UWFFA image. The strengths of this study include its direct comparison of vascular findings among groups under a well-defined UWFFA protocol and assessment of the PAR risk factors up to school age in a long-term cohort.
In conclusion, this study revealed that persistent PAR occurred in one-third (33.3% and 31.65%, respectively) of ROP eyes in school-age patients in spontaneous regression and IVI treatment groups. Several distinct vascular anomalies were detected at the vascular–avascular juncture and within the vascularized retina of ROP eyes, especially in eyes treated with IVI or laser. IVI treatment was identified as a risk factor for PAR in this study population. However, the absence of stage 3 ROP eyes in the regressed ROP group implicated that stage 3 ROP, rather than IVI, could be the risk factor for PAR. The clinical significance and long-term effect of these abnormal vascular patterns require further prospective studies to provide evidence-based guidance on their management.
Tsai AS, Chou HD, Ling XC, et al. Assessment and management of retinopathy of prematurity in the era of anti-vascular endothelial growth factor (VEGF). Prog Retin Eye Res 2021:101018.
Follow-up to Age 4 Years of Treatment of Type 1 Retinopathy of Prematurity Intravitreal Bevacizumab Injection versus Laser: Fluorescein Angiographic Findings.
Blair MP, Hilkert Rodriguez S, Chan RVP, Shapiro MJ. Re: Lepore et al.: Follow-up to age 4 years of treatment of type 1 retinopathy of prematurity intravitreal bevacizumab injection versus laser: fluorescein angiographic findings (Ophthalmology. 2018;125:218-226). Ophthalmology 2018;125(10):e70-e1.
Comparison between Oral and Intravenous Ultrawide-Field Fluorescein Angiography in the Clinical Follow-up of Children with a History of Retinopathy of Prematurity or Prematurity.
Comparison of fluorescein angiographic findings in type 1 and type 2 retinopathy of prematurity with intravitreal bevacizumab monotherapy and spontaneous regression.
Fluorescein angiographic observations of peripheral retinal vessel growth in infants after intravitreal injection of bevacizumab as sole therapy for zone I and posterior zone II retinopathy of prematurity.
This study was supported by Chang Gung Memorial Hospital Research Grants (CMRPG3M0131 and CMRPG3L0151), a Ministry of Science and Technology Research Grant (MOST 109-2314-B-182A-019-MY3). The sponsors had no role in the design or conduct of this research.
The authors also thank the statistical assistance and wish to acknowledge the support of the Maintenance Project of the Center for Big Data Analytics and Statistics at the Chang Gung Memorial Hospital for the study design and monitoring, data analysis, and interpretation. We acknowledge Wallace Academic Editing for assistance in editing this manuscript.
Financial Support
The authors have no proprietary or commercial interest in any materials discussed in this article.
About one-third of retinopathy of prematurity (ROP) eyes with spontaneous regression or with intravitreal injection treatment exhibited persistent avascular retina (PAR) until school age under ultra-wide-field fluorescein angiography.
00013-1&id=gr1.jpg){kind=link}
00013-1&id=gr2.jpg){kind=link}
00013-1&id=gr3.jpg){kind=link}
00013-1&id=gr4.jpg){kind=link}