Retinopathy of prematurity (ROP) is a leading cause of childhood blindness in the United States and around the world.1,2 Major risk factors for developing ROP requiring treatment include younger gestational age, lower birth weight, and prolonged oxygen therapy.3 Myriad other risk factors have been implicated, such as a patent ductus arteriosus (PDA), indomethacin treatment, and blood transfusions.4–9
With the results of the Early Treatment for Retinopathy of Prematurity study, plus disease has become the primary indication for laser treatment.10 Children with congenital heart disease are also known to have dilated and torturous retinal vessels,11,12 which we hypothesize may lead to the overdiagnosis of plus disease in ROP. The purpose of this study was to investigate PDA and indomethacin as independent risk factors for plus disease or ROP requiring treatment in a large cohort of premature infants.
Patients and Methods
This project received Institutional Review Board approval and followed The Health Insurance Portability and Accountability Act of 1996 Privacy and Security Rules. A retrospective chart review was performed of infants screened for ROP at the neonatal intensive care unit (NICU) over a 5-year period (January 2006 through December 2010). Patient information was confirmed from the birth certificate, paper chart, hospital-wide electronic medical record, and a NICU-specific multi-user data system (NeoData, Lisle, IL).
Infants were screened for ROP if they were born at gestational age less than 32 weeks, if their birth weight was less than 1,500 g, or if they had an unstable clinical course at the discretion of the consulting NICU attending physician. Birth history was recorded, including multiple births, gestational age, birth weight, birth head circumference, and birth length. Maternal prenatal steroid use was reviewed.
ROP information included the worst stage, worst zone, presence or absence of plus disease, need for laser treatment, and treatment failure.
Bronchopulmonary dysplasia was defined based on standard classification criteria.13 Briefly, patients with mild bronchopulmonary dysplasia required respiratory support on or after 28 days and not at 36 weeks of post-gestational age; patients with moderate bronchopulmonary dysplasia required respiratory support after 36 weeks with less than 30% oxygen; and patients with severe bronchopulmonary dysplasia required respiratory support after 36 weeks with more than 30% oxygen. The most advanced respiratory support was also recorded (ie, room air/nasal cannula, continuous positive airway pressure, ventilator, and oscillator).
The presence of a PDA and its treatment (observation, indomethacin, and surgical ligation) were recorded. All clinically suspected PDAs were confirmed with echocardiograms. Symptomatic PDAs were treated with a maximum of two courses of indomethacin treatments and then surgical ligation if necessary. Other complications of prematurity reviewed included gastrointestinal issues (necrotizing enterocolitis, bowel perforation, and gastrointestinal surgery), neurological issues (intraventricular hemorrhage grades 1 to 4, periventricular leukomalacia, and ventriculoperitoneal shunt), and blood transfusions (none, one, and more than one) received during the hospital admission. Definite sepsis was defined as intravenous antibiotic or antifungal treatment for 5 or more continuous days; possible sepsis was recorded if the diagnosis was stated in the chart and the patient was treated for less than 5 continuous days.
Discharge information such as discharge weight and length of stay were noted. Deaths during the admission were also recorded.
The chi-square or Fisher’s exact test was used to determine the association of PDA with other risk factors (ie, gestational age, birth weight, etc.). Multivariate logistic regression was performed to study the effect of PDA on plus disease and ROP requiring treatment. Additional adjustment for all other risk factors was achieved by including those variables in the logistic regression models. A predictor was considered as a potential candidate for the final model if the log-rank test of equality across strata had a P value less than .2 in the univariate logistic regression. During final multivariate logistic regression model building, all of the potential predictors were included in the model and the Wald test was used to measure the statistical significance of hazard ratios. All predictors other than PDA with P values less than .05 were considered as a potential risk factor for plus disease or ROP requiring treatment.
In a 5-year period, 541 infants met screening criteria for ROP as inpatients. Thirty-one infants who met screening criteria died before ROP examination and were excluded. Thirteen charts could not be found, and these infants were excluded. A total of 497 infants were eligible for inclusion in the study and, after randomly excluding twins and triplets, data from 450 infants were used for statistical analysis.
Of 450 infants included in this study, 217 (48%) were male. There were 367 (82%) singleton births, 73 (16%) twins, and 10 (2%) triplets. Mean gestational age was 28.7 weeks (range: 23 to 40 weeks, standard deviation [SD]: 2.8). Mean birth weight was 1,145 g (range: 380 to 3,825 g, SD: 438). Mean birth head circumference was 26.0 cm (range: 19 to 36 cm, SD: 2.8). Mean birth length was 36.9 cm (range: 26 to 50 cm, SD: 4.3). Stratification of birth data can be found in Table 1. There was antenatal maternal steroid use in 388 of 450 (87%) infants.
Table 1: Univariate Logistic Regression for ROP Requiring Treatment and Plus Disease
ROP documentation could be found in 448 of 450 (99%) infants. The 2 infants with missing ROP records had laser therapy. By worst stage, 40 of 448 (9%) infants had fully vascularized retina, 179 of 448 (40%) infants had stage 0 or non-perfused peripheral retina, 110 of 448 (25%) infants had stage 1 ROP, 77 of 448 (17%) infants had stage 2 ROP, and 42 of 448 (9%) infants had stage 3 ROP. No infants had stage 4 or 5 ROP. Pre-plus disease was documented in 34 of 448 (8%) infants and definite plus disease in 19 of 448 (5%) infants. Fifty-eight of 450 (13%) infants reached ROP requiring treatment. Mean gestational age at the time of laser therapy was 37.3 weeks (range: 32.6 to 42.6 weeks, SD: 2.1). Seven of 58 (13%) infants did not regress with a single laser session and repeat laser treatment was performed.
The most advanced respiratory support needed by infants included continuous positive airway pressure in 191 of 497 (42%) infants, mechanical ventilator in 141 of 497 (31%) infants, and respiratory oscillator in 92 of 497 (20%) infants. The frequency of other comorbidities can be found in Table 1. PDA was significantly associated with indomethacin, surgical ligation, bronchopulmonary dysplasia, neurologic comorbidities, sepsis, and blood transfusions (P < .0001).
On univariate analysis, gestational age, birth weight, birth head circumference, birth length, bronchopulmonary dysplasia, neurologic comorbidities, PDA and its treatments, gastrointestinal comorbidities, and sepsis were significantly correlated to plus disease and ROP requiring treatment (Table 1). Multivariate step-wise logistic regression with type 3 analysis showed that only gestational age and bronchopulmonary dysplasia were independently associated with plus disease or ROP requiring treatment (Table 2).
Table 2: Multivariate Analysis (Stepwise Logistic Regression) of Risk Factors for ROP Requiring Treatment and Plus Disease
Five of 450 (1%) infants died prior to discharge from the hospital. Mean length of hospital admission was 69 days (range: 7 to 240 days, SD: 31). Mean discharge weight was 2,754 g (range: 1,050 to 5,640 g, SD: 652).
This study sought to answer the question of whether PDA and indomethacin were independently associated with plus disease or ROP requiring treatment in a mid-sized, urban NICU. Although PDA and indomethacin were found to be associated with plus disease on univariate analysis, they were not independently significant after multivariate analysis with stepwise logistic regression. A prior study that drew a similar conclusion between PDA and plus disease only examined 54 patients and they were each from multiple births.14 Our study adds to this conclusion by studying a larger number of patients with risk factors collected and confirmed by multiple data sources.
There is less known about the effects of indomethacin on ROP. Indomethacin treats PDA by blocking the production of prostaglandin E2, which in turn decreases the muscular drive that keeps the ductus arteriosus patent.15 Animal models have suggested that indomethacin may increase the severity of ROP through a vascular endothelial growth factor (VEGF)-driven pathway.16,17 Future studies should consider the temporal effect of indomethacin on VEGF and ROP activity. If there is a causal relationship, anti-VEGF intravitreal injections could be considered in appropriate infants.
Prior studies have shown a close relationship between PDA with respiratory distress syndrome and hemodynamic instability, hypothesizing these factors as the cause of worsening ROP.18,19 In our study, PDA was strongly associated with bronchopulmonary dysplasia and blood transfusions. Retinal vascular tortuosity in patients with congenital heart disease has also been attributed to hypoxia and anemia.20 Therefore, it may not be possible to distinguish retinal vascular tortuosity in patients with congenital heart disease and plus disease in ROP.
A strength of this study is the fact that all patients were cared for at a single institution, thereby reducing variation in treatment. The retrospective nature of the study limits the ability to draw conclusions about causation, but there remains an association between PDA, indomethacin, blood transfusions, and plus disease in ROP.
PDA and indomethacin treatment were risk factors for ROP on univariate analysis, but they did not have an independent effect on multivariate analysis. PDA was closely associated with bronchopulmonary dysplasia and blood transfusions, which may explain its influence on ROP. Future studies should consider the timing of indomethacin and its effect on VEGF when examining PDA and ROP.
- Gilbert C. Retinopathy of prematurity: a global perspective of the epidemics, population of babies at risk and implications for control. Early Hum Dev. 2008;84:77–82 doi:10.1016/j.earlhumdev.2007.11.009 [CrossRef] .
- Tasman W. Retinopathy of prematurity: do we still have a problem?Arch Ophthalmol. 2011;129:1083–1086 doi:10.1001/archophthalmol.2011.192 [CrossRef] .
- Raghuveer TS, Bloom BT. A paradigm shift in the prevention of retinopathy of prematurity. Neonatology. 2011;100:116–129 doi:10.1159/000322848 [CrossRef] .
- Kumar P, Sankar MJ, Deorari A, et al. Risk factors for severe retinopathy of prematurity in preterm low birth weight neonates. Indian J Pediatr. 2011;78:812–816 doi:10.1007/s12098-011-0363-7 [CrossRef] .
- Purohit DM, Ellison RC, Zierler S, Miettinen OS, Nadas AS. Risk factors for retrolental fibroplasia: experience with 3,025 premature infants. National Collaborative Study on Patent Ductus Arteriosus in Premature Infants. Pediatrics. 1985;76:339–344.
- Bourla DH, Gonzales CR, Valijan S, Yu F, Mango CW, Schwartz SD. Association of systemic risk factors with the progression of laser-treated retinopathy of prematurity to retinal detachment. Retina. 2008;28(3 suppl):S58–S64. Erratum in: Retina. 2009;29:127 doi:10.1097/IAE.0b013e31815075b0 [CrossRef] .
- Ohlsson A, Walia R, Shah S. Ibuprofen for the treatment of a patent ductus arteriosus in preterm and/or low birth weight infants. Cochrane Database Syst Rev. 2003:CD003481.
- Procianoy RS, Garcia-Prats JA, Hittner HM, Adams JM, Rudolph AJ. Use of indomethacin and its relationship to retinopathy of prematurity in very low birthweight infants. Arch Dis Child. 1980;55:362–364 doi:10.1136/adc.55.5.362 [CrossRef] .
- Jegatheesan P, Ianus V, Buchh B, et al. Increased indomethacin dosing for persistent patent ductus arteriosus in preterm infants: a multicenter, randomized, controlled trial. J Pediatr. 2008;153:183–189 doi:10.1016/j.jpeds.2008.01.031 [CrossRef] .
- Fielder AR. Preliminary results of treatment of eyes with high-risk prethreshold retinopathy of prematurity in the early treatment for retinopathy of prematurity randomized trial. Arch Ophthalmol. 2003;121:1769–1771 doi:10.1001/archopht.121.12.1769 [CrossRef] .
- Petersen RA, Rosenthal A. Retinopathy and papilledema in cyanotic congenital heart disease. Pediatrics. 1972;49:243–249.
- Granstrom KO. Retinal changes in coarctation of the aorta. Br J Ophthalmol. 1951;35:143–148 doi:10.1136/bjo.35.3.143 [CrossRef] .
- Ryan RM. A new look at bronchopulmonary dysplasia classification. J Perinatol. 2006;26:207–209 doi:10.1038/sj.jp.7211449 [CrossRef] .
- Garcia-Serrano JL, Ramirez-Garcia MC, Pinar-Molina R. Retinopathy of prematurity in multiple births: risk analysis for plus disease [article in Spanish]. Arch Soc Esp Oftalmol. 2009;84:191–198.
- Tynan M. The ductus arteriosus and its closure. N Engl J Med. 1993;329:1570–1572 doi:10.1056/NEJM199311183292111 [CrossRef] .
- Parys-Van Ginderdeuren R, Malcolm D, Varma DR, Aranda JV, Chemtob S. Dissociation between prostaglandin levels and blood flow to the retina and choroid in the newborn pig after nonsteroidal antiinflammatory drugs. Invest Ophthalmol Vis Sci. 1992;33:3378–3384.
- Beharry KD, Modanlou HD, Hasan J, et al. Comparative effects of early postnatal ibuprofen and indomethacin on VEGF, IGF-I, and GH during rat ocular development. Invest Ophthalmol Vis Sci. 2006;47:3036–3043 doi:10.1167/iovs.06-0057 [CrossRef] .
- Gaugler C, Beladdale J, Astruc D, et al. Retinopathy of prematurity: 10-year retrospective study at the University Hospital of Strasbourg [article in French]. Arch Pediatr. 2002;9:350–357 doi:10.1016/S0929-693X(01)00792-8 [CrossRef] .
- Fehlmann E, Tapia JL, Fernandez R, et al. Impact of respiratory distress syndrome in very low birth weight infants: a multicenter South-American study [article in Spanish]. Arch Argent Pediatr. 2010;108:393–400.
- Mansour AM, Bitar FF, Traboulsi EI, et al. Ocular pathology in congenital heart disease. Eye (Lond). 2005;19:29–34 doi:10.1038/sj.eye.6701408 [CrossRef] .
Univariate Logistic Regression for ROP Requiring Treatment and Plus Disease
|Risk Factor||No. (%)||ROP Requiring Treatment||P||Plus Disease||P|
|Gestational age (wk)|
| < 28||183 (41%)||58||51|
| 28–32||224 (50%)||0||< .0001||2||< .0001|
| > 32||43 (10%)||0||0|
|Birth weight (g)|
| < 750||68 (15%)||36||30|
| 750–1,500||315 (70%)||21||< .0001||22||< .0001|
| > 1,500||67 (15)||1||1|
|Birth head circumference (cm)|
| < 32||438 (97%)||58||< .0001||53|
| 32+||12 (3%)||0||0||< .0001|
|Birth length (cm)|
| < 45||436 (97%)||58||< .0001||53||< .0001|
| 45+||14 (3%)||0||0|
| No||61 (14%)||11||.20||7||.96|
| Yes||388 (87%)||47||46|
| Male||217 (48%)||25||.40||20||.11|
| Female||233 (52%)||33||33|
| No||367 (82%)||51||.41||44||.95|
| Yes||83 (18%)||7||9|
| Mild BPD||103 (23%)||2||< .0001||3||< .0001|
| Moderate BPD||44 (10%)||3||5|
| Severe BPD||130 (29%)||51||43|
| Grade 1–2 IVH||172 (38%)||33||< .0001||28||< .0001|
| Grade 3–4 IVH||29 (6%)||11||11|
| PVL||18 (4%)||7||5|
| VP shunt||12 (3%)||6||5|
| Present||205 (46%)||49||< .0001||44||< .0001|
| Indomethacin||69 (15%)||21||< .0001||18||< .0001|
| Ligation||67 (15%)||32||< .0001||30||< .0001|
| NEC||63 (14%)||18||.0002||13||.0017|
| Bowel perforation||27 (6%)||12||8|
| GI surgery||33 (7%)||13||10|
| One||75 (17%)||0||.99||0||.0052|
| Multiple||264 (59%)||58||52|
| Possible||176 (39%)||4||< .0001||3||< .0001|
| Yes||247 (55%)||52||48|
Multivariate Analysis (Stepwise Logistic Regression) of Risk Factors for ROP Requiring Treatment and Plus Disease
|Risk Factor||ROP Requiring Treatment||Plus Disease|
|P||Odds Ratio||95% CI||Type 3 Analysis||P||Odds Ratio||95% CI||Type 3 Analysis|
|GA||< .0001||0.4||0.3, 0.5||< .0001||< .0001||0.5||0.4, 0.6||< .0001|
|Mild BPDa||.31||0.3||0.04, 2.9||.60||0.6||0.09, 4.0|
|Moderate BPD||.99||1.0||0.1, 7.9||< .0005||.40||2.2||0.4, 13.9||.0094|
|Severe BPDa||.066||5.0||0.9, 27.0||.08||4.2||0.8, 22.0|
|PDA||.58||0.8||0.3, 2.0||.5771||.48||0.7||0.3, 1.8||.4759|