Choroidal Measurements in Decision Making for Retinopathy of Prematurity: A Decision Tree Analysis

doi:10.21203/rs.3.rs-2969157/v1

Download PDF

Article

Choroidal Measurements in Decision Making for Retinopathy of Prematurity: A Decision Tree Analysis

https://doi.org/10.21203/rs.3.rs-2969157/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Objective: To compare the choroidal thickness and vascular profile of premature infants with ROP (retinopathy of prematurity) using a handheld SD-OCT device.

Design: A prospective cross-sectional case-control study

Subject: All ROP patients examined under sedation or general anesthesia in Farabi Eye Hospital or Mahdie Hospital from 2020 to 2022.

Methods: We performed horizontal SD-OCT scans through the fovea in 164 eyes of 82 premature infants. Premature infants included 2 groups (infants with ROP requiring treatment [as treatment group] vs. infant without ROP or with ROP not- requiring treatment [as no-treatment group]) Choroidal thicknesses (CT) were measured at 5 points, including the fovea, 250 µm, and 500µm mm nasal and temporal to the fovea. The choroidal vascularity index (CVI) and choroidal stromal index (CSI) were also calculated. The classification and regression tree (CRT) algorithm was used to predict the need for treatment based on all OCT characteristics.

Main outcome: Treatment-require ROP patients.

Results: Mean CT was higher in 500 micrometers nasal to the fovea compared to temporal CT (275.8 ± 64.8 and 257.1 ± 57.07, P value<0.03). No statistically significant difference was found regarding CVI, corrected CVI, and temporal and nasal CT in the treatment group versus the no-treatment group. The foveal CT was significantly lower in ROP patients with the plus disease compared to not-plus ROP (P value=0.03. ANOVA, Bonferroni posthoc test). CT was not significantly different between plus and pre-plus patients (P-value=0.9, ANOVA, Bonferroni posthoc test). No significant relationship was found between the stage of ROP and choroidal thickness (P value>0.05, GEE). The decision tree analysis showed that in infants with ROP, the most important predictor for the need for treatment is CVI.

Conclusion:

This study delineated the possible effectiveness of choroidal measurements as an additive to decision-making for ROP. We also demonstrated that choroidal involution is associated with the presence of pre-plus/plus, not with the stage of ROP. We demonstrated that choroidal measurements are very sensitive but not specific tools for assessing the need for treatment in ROP patients.

Retinopathy of prematurity (ROP) is a major cause of treatable visual loss in the pediatric population(1). The main pathogenesis behind this disease is the exposure of the immature retina to relative hyperoxia, resulting in the reduction of growth factors(2, 3). Although retinal neovascularization was the primary suspect for the visual loss, the revelation of central photoreceptor loss raises a probable role of choroidal vasculature (4–6). Choroidal involution was also demonstrated by Oxygen induced retinopathy (OIR) models in the rat(4). With the advent of handheld Ocular coherence tomography, a growing body of literature evaluates choroidal parameters in ROP (7–9). Choroidal thinning was detectable in patients with a history of treatment for ROP and also spontaneously regressed ROP(10, 11). Therefore, choroidal involution may be responsible for permanent vision loss in these patients. Despite many studies evaluating OCT characteristics in ROP, none of them considered its usage in planning for the treatment of these patients. In this study, we used decision tree analysis to find whether choroidal parameters could be used as a marker to determine the need for treatment.

Subjects

This cross-sectional case-control study was performed in Farabi Eye Hospital, Tehran University of medical science between 2019 and 2021. This study was adhered to the tenets of the Declaration of Helsinki and was approved by the local ethics committee of the Tehran University of Medical Sciences (https://ethics.research.ac.ir/IR.TUMS.FARABIH.REC.1400.065). Informed consent was obtained from all the parents or the legal guardians of the involved infants. We included 156 eyes of 81 premature neonates. All infants were born between 2019 and 2021.

The preterm infants with type 1 ROP received treatment. The patients who received general anesthesia or sedation for treatment were included in this study as treatment group. They were examined using handheld portable SD-OCT (Optovue iVue SD-OCT Wellness report; Optovue Corporation, Fremont, CA) after pupil dilatation and before treatment. Some preterm infants were examined in the neonatal intensive care unit (NICU), and routinely they received sedation for fundus examination with sclera depression for controlling stress and pain in addition to topical anesthesia. The patients in NICU who did not require treatment were chosen as a control group (no-treatment group), and OCT was performed for them.

We have three sedation protocols: 1. In IVB treated group, the sedation protocol was oral sucrose 0.5-1 mg/kg 2minute before IVB injection and repeated 2 minutes later. 2. In NICU-admitted infants (control group or no-treatment group), the sedation protocol was nasal administration of ketamine 2-3mg/kg plus fentanyl 2–3 µg/kg. 3. In the laser-treated group, IV ketamine 1mg/kg with suppository acetaminophen 20mg/kg was used, and if needed, ketamine 1mg/kg was added. In cases that were not sedated with this protocol, sevoflurane 3% with oxygen through a mask for 2–3 minutes was administered.

Intravitreal injections were performed under topical anesthesia and light sedation due to the general condition. The OCT was performed for them. After administration of 10% povidone-iodine for periocular skin and 5% povidone-iodine for ocular surface, half of the adult doses of bevacizumab (0.625 mg/0.025 mL) were injected with a 30-gauge needle 1-1.5 mm behind the limbus into the vitreous cavity. Topical gentamycin or sulfacetamide was given for three days post-injection. In cases that need laser ablation, the indirect laser was performed in avascular areas using a confluent or near confluent pattern with moderate intensity after general anesthesia and full pupillary dilation. Topical gentamycin or sulfacetamide, topical mydrax 0.5%, and topical betamethasone were prescribed for seven days in patients receiving indirect laser.

In order to control imaging artifacts, the examination without any sedation or general anesthesia and images with low quality were excluded. Patients with unstable general conditions or hazy media, such as corneal edema, were also excluded. ROP staging, zone, and presence of plus were recorded. Of 156 eyes, 133 required treatment (treatment group). Twenty two eyes required Bevacizumab injection, and the other 111 eyes received indirect laser photocoagulation.

Image processing:

OCT scans were obtained by a single experienced examiner. Choroidal thickness (CT) was defined as the distance from Bruch’s membrane to the inner part choroidoscleral junction. Choroidal thickness was calculated at the subfoveal, 250 µm, and 500 µm nasal and temporal to the fovea manually using ImageJ software (National Institutes of Health, Bethesda, MD, USA) (Fig. 1). The average of these calculations was considered as foveal choroidal thickness (FCT). Previously described, Agrawal et al. protocol was applied to imaging to calculate luminal area (LA), stromal area (SA), and choroidal vascularity index (CVI)(12). We also calculated the choroidal stromal index (CSI) by subtracting CVI from 1 (CSI = 1-CVI).

All of the choroidal parameters were measured by an experienced co-author (MRMB). To evaluate the observer variability, the measurement was re-analyzed two times for 100 cases, and the average values were used for the analysis. To evaluate the intra-observer reliability of the CVI measurements, we calculated the intraclass correlation coefficient (ICC). The ICC was 0.95 with a 95% confidence interval (CI) of 0.91 to 0.97 (p < 0.001) (excellent reliability).

Statistical analysis:

We described choroidal parameters using mean (standard deviation) and median (interquartile range). We used the generalized estimating equation (GEE) to assess the correlation of the ROP severity and the choroidal thickness with consideration of the possible correlation of measurements in bilateral samples. In the comparison of treatment type and no-treatment group, multiple comparisons were considered by the Sidak method. We used a decision tree analysis to obtain a method for the prediction of the needs for treatment. The classification and regression tree (CRT) approach was used to grow the tree. The maximum depth of the tree was set at five levels, and the minimum number of cases of the parent node and the child node was set to 20 and 5, respectively. In addition, the sensitivity, specificity, and diagnostic accuracy of the model were reported. The inter-grader reliability for each parameter was also calculated using the intraclass correlation coefficient (ICC). All statistical analysis performed by SPSS (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp).

Patient demographics

In the analyses, 156 eyes of 81 children with ROP were included. Of 133 eyes that underwent treatment, 22 received intravitreal injections of Bevacizumab, and 111 were treated with laser photocoagulation. The findings were compared to 23 eyes of ROP children who didn’t require treatment. The demographic features of the study population are presented in Table 1.

Table 1

Demographic data of all ROP patients
Variable	Label	Need to treat		P value	Treatment		P value
Variable	Label	No	Yes	P value	IVB	Laser	P value
Sex	Male	7 (33.3%)	69 (51.1%)	0.115	11 (50%)	58 (51.3%)	0.800
Sex	Female	14 (66.7%)	66 (48.9%)		11 (50%)	55 (48.7%)	0.800
Laterality	OD	10 (47.6%)	68 (50.4%)	0.815	11 (50.0%)	57 (50.4%)	0.969
Laterality	OS	11 (52.4%)	67 (49.6%)		11 (50.0%)	56 (49.6%)	0.969
Zone	1	0 (0.0%)	4 (3.3%)	0.000	4 (20%)	0 (0.0%)	0.000
	2 posterior	0 (0.0%)	7 (5.7%)		7 (35%)	0 (0.0%)
	2	10 (47.6%)	106 (87.8%)		8 (40%)	98 (97.0%)
	3	5 (23.8%)	4 (3.3%)		1 (5%)	3 (3.0%)
	Complete	6 (28.6%)	0 (0.0%)		0 (0.0%)	0 (0.0%)
Stage	0	6 (40.0%)	0 (0.0%)	0.000	0 (0.0%)	0 (0.0%)	0.090
	1	9 (60.0%)	1 (0.8%)		1 (5%)	0 (0.0%)
	2	0 (0.0%)	7 (5.7%)		1 (5%)	6 (5.9%)
	3	0 (0.0%)	105 (86.8%)		20 (90%)	87 (86.1%)
	4	0 (0.0%)	8 (6.5%)		0 (0.0%)	8 (7.9%)
PLUS	0	17 (100.0%)	1 (0.8%)	0.000	1 (5%)	0 (0.0%)	0.010
	Pre Plus	0 (0.0%)	19 (15.4%)		0 (0.0%)	19 (18.8%)
	Plus	0 (0.0%)	101 (83.7%)		19 (95. %)	82 (81.2%)
IVB: Intravitreal Bevacizumab

The Choroidal thickness measurements in the horizontal foveal scan are displayed in Tables 2 and 3. In all the subjects, the highest choroidal thickness was in the subfoveal region (temporal: P value < 0.05 and nasal: P value < 0.05. t-test). The mean CT was higher in 500 micrometers nasal to the fovea in comparison with temporal CT (275.8 ± 64.8 and 257.1 ± 57.07, P value < 0.03, t-test). However, at the 250-micrometer eccentricity, there wasn’t any significant difference between nasal and temporal borders.

Table 2

Comparison of Choroidal parameter in ROP-treated and Not-treated ROP patients
Variable		Need to treat		Diff	95% CI (confidence interval)		P value
Variable		No	Yes	Diff	Lower	Upper	P value
Total Area	Mean ± SD	349 ± 73	438 ± 114	-86.4	-145.4	-27.3	0.005
Total Area	Median (range)	350 (250 to 480)	440 (210 to 810)
Stromal Area	Mean ± SD	282 ± 61	313 ± 72	-31.0	-69.4	7.4	0.112
Stromal Area	Median (range)	284 (179 to 381)	309 (182 to 528)
Luminal Area	Mean ± SD	67 ± 23	124 ± 76	-55.7	-93.8	-17.6	0.005
Luminal Area	Median (range)	71 (32 to 102)	101 (26 to 340)
Choroidal Vascular Index (CVI)	Mean ± SD	81 ± 6	73 ± 11	7.4	1.7	13.1	0.012
Choroidal Vascular Index (CVI)	Median (range)	80 (72 to 91)	75 (48 to 92)
Choroidal Stromal Index(CSI)	Mean ± SD	19 ± 6	27 ± 11	-7.4	-13.1	-1.7	0.012
Choroidal Stromal Index(CSI)	Median (range)	20 (9 to 28)	25 (8 to 52)
Foveal Choroidal Thickness (FCT)	Mean ± SD	254 ± 56	305 ± 83	-43	-85.8	-0.7	0.04
Foveal Choroidal Thickness (FCT)	Median (range)	245 (170 to 390)	300 (130 to 480)
Temporal Choroidal Thickness 500 (TCT 500)	Mean ± SD	238 ± 55	263 ± 57	-24.6	-52.5	3.4	0.084
Temporal Choroidal Thickness 500 (TCT 500)	Median (range)	(147 to 323)	(128 to 433)
Temporal Choroidal Thickness 250 (TCT250)	Mean ± SD	243 ± 49	272 ± 61	-28.5	-57.5	0.5	0.054
Temporal Choroidal Thickness 250 (TCT250)	Median (range)	246 (152 to 309)	274 (135 to 461)
Sub Foveal Choroidal Thickness (SFCT)	Mean ± SD	247 ± 53	284 ± 66	-36.1	-67.7	-4.5	0.026
Sub Foveal Choroidal Thickness (SFCT)	Median (range)	250 (152 to 330)	290 (132 to 462)
Nasal Choroidal Thickness 250 (NCT 250)	Mean ± SD	239 ± 49	284 ± 68	-44.5	-76.8	-12.3	0.007
Nasal Choroidal Thickness 250 (NCT 250)	Median (range)	236 (135 to 314)	289 (108 to 435)
Nasal Choroidal Thickness 500 (NCT500)	Mean ± SD	249 ± 54	282 ± 66	-31.7	-63.2	-0.1	0.049
Nasal Choroidal Thickness 500 (NCT500)	Median (range)	249 (138 to 334)	283 (152 to 487)
SD: standard deviation

Table 3

Comparison of Choroidal parameter in IVB-treated and Laser-treated ROP patients
Variable		Treatment		Diff	95% CI (confidence interval)		P value
Variable		IVB	Laser	Diff	Lower	Upper	P value
Area	Mean ± SD	398 ± 84	444 ± 117	-54.6	-128.0	18.7	0.142
Area	Median (range)	370 (310 to 520)	440 (210 to 810)
Stromal Area	Mean ± SD	265 ± 61	321 ± 71	-52.7	-98.1	-7.3	0.024
Stromal Area	Median (range)	269 (187 to 394)	316 (182 to 528)
Luminal Area	Mean ± SD	133 ± 54	123 ± 79	-2.8	-52.5	46.9	0.910
Luminal Area	Median (range)	126 (45 to 202)	97 (26 to 340)
Choroidal Vascular Index (CVI)	Mean ± SD	67 ± 12	74 ± 11	-4.1	-11.4	3.1	0.259
Choroidal Vascular Index (CVI)	Median (range)	65 (53 to 86)	75 (48 to 92)
Choroidal Stromal Index (CSI)	Mean ± SD	32 ± 12	26 ± 11	4.1	-3.1	11.4	0.259
Choroidal Stromal Index (CSI)	Median (range)	35 (14 to 47)	25 (8 to 52)
Foveal Choroidal Thickness (FCT)	Mean ± SD	274 ± 63	309 ± 86	-38.8	-85.5	7.8	0.1
Foveal Choroidal Thickness (FCT)	Median (range)	248 (200 to 420)	300 (130 to 580)
Temporal Choroidal Thickness 500 (TCT 500)	Mean ± SD	234 ± 43	268 ± 58	-31.2	-64.9	2.6	0.070
Temporal Choroidal Thickness 500 (TCT 500)	Median (range)	(194 to 322)	(128 to 433)
Temporal Choroidal Thickness 250 (TCT 250)	Mean ± SD	245 ± 51	276 ± 62	-30.2	-66.2	5.9	0.100
Temporal Choroidal Thickness 250 (TCT 250)	Median (range)	227 (200 to 370)	280 (135 to 461)
Sub Foveal Choroidal Thickness (SFCT)	Mean ± SD	260 ± 57	288 ± 67	-27.5	-67.1	12.1	0.171
Sub Foveal Choroidal Thickness (SFCT)	Median (range)	242 (200 to 410)	298 (132 to 462)
Nasal Choroidal Thickness 250 (NCT250)	Mean ± SD	258 ± 72	289 ± 68	-34.4	-75.1	6.3	0.096
Nasal Choroidal Thickness 250 (NCT250)	Median (range)	230 (197 to 433)	295 (108 to 435)
Nasal Choroidal Thickness 500 (NCT500)	Mean ± SD	258 ± 91	286 ± 62	-31.8	-71.1	7.5	0.111
Nasal Choroidal Thickness 500 (NCT500)	Median (range)	230 (168 to 487)	290 (152 to 423)
SD: standard deviation, IVB: intravitreal bevacizumab.

Significant differences were found between the treatment and control group only regarding the total area and luminal area after Sidak correction for multiple comparisons (total area: 438 ± 114 and 349 ± 73, luminal area: 124 ± 76 and 67 ± 23 in treatment and control groups retrospectively all P value < 0.0051) (Table 2). No statistically significant difference was found regarding CVI, corrected CVI, and temporal and nasal CT in children with ROP receiving treatment versus the control group. In the treatment group, there was not any significant difference between the laser and IVB groups regarding choroidal parameters after Sidak correction for multiple comparisons.

The foveal CT was significantly lower in ROP patients with the plus disease compared to non-plus ROP (P value = 0.03. ANOVA, Bonferroni posthoc test). CT was not significantly different between plus and pre-plus patients (P-value = 0.9, ANOVA, Bonferroni posthoc test). No significant relationship was found between the stage of ROP and choroidal thickness (P value > 0.05, GEE).

Decision tree analysis

CRT algorithm was used to predict the need for treatment based on all OCT characteristics. The tree diagram (Fig. 2) shows tree construction based on the test sample of 115 cases. There are seven nodes that consist of 4 terminal nodes, and the depth of the tree is 3. The Parent node has 20 absences (17.4%) and 95 presence (82.6%) of indication for treatment. The first discriminator, "CSI," splits the root node into two child nodes: low avascularity (less than or equal to 29.1) (node 1, n = 82) and high (more than 29.1) (terminal node 2, n = 33). The improvement for this classification is 0.028. The next discriminator is the mean choroidal thickness in 250 µm of nasal to the fovea"NCT250", split into low choroidal thickness (less than or equal to 243, terminal node 3, n = 26) and high choroidal thickness (more than 243, node 4, n = 56). The improvement for this classification is 0.020. When the nasal choroidal thickness is more than 243, then the final discriminator is again NCT 250 (0.009 improvement), which produces two terminal nodes, 5 and 6. Percentages in each category and in each joint category are shown in Fig. 2. The predictors: SFCT, CVI, CT 500, and CT 250 didn’t contribute to the classification tree used. The prediction accuracy over the entire sample size was 82.6%, with sensitivity 100% and specificity of 0%. The misclassification rate was 0.174, with standard error of 0.035.

The current study aimed to assess the choroidal profile in treatment–requiring ROP neonates and compare it with ROP neonates not requiring treatment. Our study showed that foveal choroidal thickness was significantly lower in ROP patients with the plus disease compared to non-plus ROP (P value = 0.03. ANOVA, Bonferroni posthoc test). However, CT was not significantly different between plus and pre-plus patients (P-value = 0.9, ANOVA, Bonferroni posthoc test). We found no significant relationship between the different staging of ROP and choroidal thickness. It might be explained partly by that choroidal vasculature especially choriocapillaris might be involved other than choroidal thickness. The assessment of choriocapillaris is very difficult in premature infant.

Retinopathy of prematurity (ROP) is considered a disease of retinal vascularization; however, recent evidence demonstrates that choroidal vasculature abnormality is also important and is responsible for outer retinal dysfunction and visual loss(13). Several clinical studies have demonstrated choroidal thinning associated with ROP in older children and adults using OCT imaging (8–11). Anderson and colleagues reported lower subfoveal choroidal thickness in children and young adults with a history of ROP treatment with laser ablation and/or cryotherapy compared with healthy controls(14). Similar results have been shown by Bowl et al. in a study that compared choroidal thickness in 17 young children with a history of treated ROP or spontaneously regressed ROP. They also showed that reduced choroidal thickness was linked to ROP severity(11). However, few studies in the literature evaluated choroidal thickness in infants with ROP (7, 8, 15). Erol and his colleagues studied subfoveal choroidal thickness in 80 premature infants. They found that the thickness of the choroid decreased with the severity of ROP(8). In another study by Mangalesh et al., it is shown that the presence of pre-plus/plus disease versus without accompanied by thinner choroid(16). In consensus with previous studies(16), we also demonstrated the choroid was thinner in ROP neonates with the plus disease compared to not-plus disease. It has been suggested that thinning of the choroid in ROP neonates may be due to oxidative stress and choroidal vascular loss (4, 7, 17, 18). This choroidal vascular loss has been proven angiographically. In a study by Islam et al., the presence of choroidal hypo fluorescence in the central and or peripheral retina of ROP patients was demonstrated(19).

Interestingly, our results indicated that the choroidal thickness didn’t significantly differ in ROP neonates requiring treatment (tROP) in comparison with no need treatment group (nROP). A previous study by Erol and colleagues showed that choroidal thickness in ROP patients with a grade 2 or 3 was lower than grade 0. However, they showed no significant difference between the choroidal thickness of grade 1 and grade 2 /3 ROP patients(8). These findings agree with our results, as 60% of the untreated group was stage 1 ROP, and 92.7% of the treated group was grade 2/3 ROP. Also, Mangalesh et al. reported thinner choroid was associated with pre-plus or plus disease and lower gestational age and birth weight but not the ROP stage(16).

Consistent with previous studies, we found that choroidal involution is associated with the presence of pre-plus/plus, not with the ROP stage.

According to our data, the choroidal vascularity index (CVI) and choroidal stromal index (CSI) weren’t significantly different in the treatment and untreated groups. CVI is a novel means of choroidal evaluation introduced by Agrawal et al. in 2016(12). It has been shown that this parameter is less variable than the choroidal thickness and less influenced by systemic and ocular circumstances(12). This marker is rarely studied in ROP patients. Consistent with our results, Lavric et al. reported CVI preterm children aged 5–15 years had the same CVI compared to preterm children with a history of ROP(20).

For the first time, we used the decision tree method to analyze the use of choroidal parameters to treat ROP patients. In contrast to more traditional statistics, such as linear regression, the decision tree method analyzed the nonlinear and interactive patterns between factors. This study successfully used decision tree analysis to detect the most important choroidal factor for identifying the treatment group and determining cut points for each parameter. The results presented in Fig. 2 showed that in ROP patients, the most important choroidal factor for predicting the need for treatment was CSI. All subjects, associated with CSI higher than 29.1, required treatment. This novel index represents the avascularity of the choroid. Therefore patients with the higher value might benefit from treatment to stop the vicious circle of oxidative stress in ROP patients. Further studies are required to evaluate the addition of choroidal parameters to the current practice of ROP screening regarding effectiveness and outcomes.

The current study focused on the choroidal parameters; this does not exclude the importance of retinal parameters in ROP. This study has other limitations. The handheld OCT imaging was time-consuming and prone to imaging artifacts; future handheld OCT with faster imaging acquisition would improve their imaging quality and feasibility. We measured choroidal measurements manually, and therefore, although reliability analysis showed acceptable observer agreement, the measurements might be susceptible to imprecision. We have a small number of ROPs did not require treatment. Also, the role of gestational age and preexisting systemic and neurological disorders in choroidal parameters were not evaluated, and future studies should be planned to evaluate these factors.

ROP: Retinopathy of prematurity, SD-OCT: spectral domain optical coherence tomography, IVB: Intravitreal Bevacizumab, CT: Choroidal thicknesses, CVI: Choroidal vascular index, CSI: choroidal stromal index

Financial Support: None

Conflict of Interest: No conflicting relationship exists for any author

Data availability:

All data are available from the corresponding author upon request.

Author contributions:

F.B , A.D.F and M.M.B conceived the research design. F.B and M.M.B assisted in data acquisition, performed image analysis and statistical analysis. F.S wrote the initial draf of the manuscript. C.A.H and F.GH and C.A.H provided valuable comments and advice, and revised the various drafs of the manuscript. All authors reviewed and approved the fnal version of the manuscript.

Competing interests :

The authors declare no competing interests

Blencowe H, Lawn JE, Vazquez T, Fielder A, Gilbert CJPr. Preterm-associated visual impairment and estimates of retinopathy of prematurity at regional and global levels for 2010. 2013;74(1):35-49.
Chen J, Smith LEJA. Retinopathy of prematurity. 2007;10(2):133-40.
Sapieha P, Hamel D, Shao Z, Rivera JC, Zaniolo K, Joyal JS, et al. Proliferative retinopathies: angiogenesis that blinds. 2010;42(1):5-12.
Shao Z, Dorfman AL, Seshadri S, Djavari M, Kermorvant-Duchemin E, Sennlaub F, et al. Choroidal involution is a key component of oxygen-induced retinopathy. 2011;52(9):6238-48.
Fulton AB, Hansen RM, Moskowitz A, Barnaby AMJDo. Multifocal ERG in subjects with a history of retinopathy of prematurity. 2005;111(1):7-13.
Fulton AB, Hansen RM, Moskowitz AJO, Optometry vsopotAAo. Development of rod function in term born and former preterm subjects. 2009;86(6):E653.
Moreno TA, O'Connell RV, Chiu SJ, Farsiu S, Cabrera MT, Maldonado RS, et al. Choroid development and feasibility of choroidal imaging in the preterm and term infants utilizing SD-OCT. Investigative ophthalmology & visual science. 2013;54(6):4140-7.
Erol MK, Coban DT, Ozdemir O, Dogan B, Tunay ZO, Bulut MJR. Choroidal thickness in infants with retinopathy of prematurity. 2016;36(6):1191-8.
Ulusoy MO, Kivanç SA, Kal AJJoG. Evaluation of Iridocorneal Angle, Choroidal Thickness, and Retinal Nerve Fiber Layer Thickness in Children With a History of Retinopathy of Prematurity. 2020;29(2):112-6.
Wu W-C, Shih C-P, Wang N-K, Lien R, Chen Y-P, Chao A-N, et al. Choroidal thickness in patients with a history of retinopathy of prematurity. 2013;131(11):1451-8.
Bowl W, Bowl M, Schweinfurth S, Holve K, Andrassi-Darida M, Stieger K, et al. Choroidal thickness with swept-source optical coherence tomography versus foveal morphology in young children with a history of prematurity. 2018;60(4):205-13.
Agrawal R, Gupta P, Tan K-A, Cheung CMG, Wong T-Y, Cheng C-YJSr. Choroidal vascularity index as a measure of vascular status of the choroid: measurements in healthy eyes from a population-based study. 2016;6(1):1-9.
Rivera JC, Holm M, Austeng D, Morken TS, Zhou TE, Beaudry-Richard A, et al. Retinopathy of prematurity: inflammation, choroidal degeneration, and novel promising therapeutic strategies. 2017;14(1):1-14.
Anderson M, Ramasamy B, Lythgoe D, Clark DJE. Choroidal thickness in regressed retinopathy of prematurity. 2014;28(12):1461-8.
Park KA, Oh SY. Analysis of spectral-domain optical coherence tomography in preterm children: retinal layer thickness and choroidal thickness profiles. Investigative ophthalmology & visual science. 2012;53(11):7201-7.
Mangalesh S, McGeehan B, Tai V, Chen X, Tran-Viet D, Vajzovic L, et al. Macular OCT characteristics at 36 weeks’ postmenstrual age in infants examined for retinopathy of prematurity. 2021;5(6):580-92.
Hardy P, Nuyt AM, Abran D, St-Louis J, Varma DR, Chemtob SJPr. Nitric oxide in retinal and choroidal blood flow autoregulation in newborn pigs: interactions with prostaglandins. 1996;39(3):487-93.
He Y, Pettenkofer M, Nittala MG, Sadda SR, Tsui I, Chu AJIO, et al. Early Postnatal Oxygen Exposure Predicts Choroidal Thinning in Neonates. 2021;62(9):23-.
Islam Y, Andersen L, Agarwal-Sinha SJOR. Angiographic Review of Choroidal Involution in Eyes with Retinopathy of Prematurity Treated with Bevacizumab. 2022;65(2):229-36.
Lavric A, Tekavcic Pompe M, Markelj S, Ding J, Mahajan S, Khandelwal N, et al. Choroidal structural changes in preterm children with and without retinopathy of prematurity. 2020;98(5):e611-e6.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Choroidal Measurements in Decision Making for Retinopathy of Prematurity: A Decision Tree Analysis

Status:

Version 1

Abstract

Figures

Introduction

Method

Subjects

Image processing:

Statistical analysis:

Results

Patient demographics

Decision tree analysis

Discussion

Conclusion

Abbreviations

Declarations

References

Additional Declarations

Status:

Version 1