Abstract
Objective
To determine the effect of perinatal and neonatal risk factors on retinopathy of prematurity (ROP) and to examine the association of fertility treatments on the risk for ROP in very low birth weight (VLBW) preterm twins.
Methods
The population-based observational study consisted of VLBW twins born at 24–29 weeks gestational age (GA). Data from the Israel national database (1995–2020) were applied. Univariate and multivariable logistic regression using the General Estimating Equation were used for assessment of risk factors.
Results
The study population comprised 4092 infants of whom 2374 (58%) were conceived following fertility treatments. ROP was diagnosed in 851 (20.8%) infants. The odds for ROP approximately doubled with each week decrease in GA: at 24 weeks, Odds Ratio (OR) 58.00 (95% confidence interval (CI) 31.83–105.68); 25 weeks, OR 25.88 (95% CI 16.76–39.96); 26 weeks, OR 12.69 (95% CI 8.84–18.22) compared to 29 weeks GA. Each decrease in one birthweight z-score was associated with 1.82-fold increased risk for ROP (OR, 1.82, 95% CI 1.59–2.08). Infertility treatments were not associated with ROP. Neonatal morbidities significantly associated with ROP were surgical necrotizing enterocolitis (NEC) (OR, 2.04, 95% CI 1.31–3.19); surgically treated patent ductus arteriosus (PDA) (OR, 1.63, 95% CI 1.12–2.37); sepsis (OR, 1.43, 95% CI 1.20–1.71) and bronchopulmonary dysplasia (OR, 1.52, 95% CI 1.22–1.90).
Conclusion
Among preterm VLBW twins, poor intrauterine growth and surgical interventions for NEC and PDA were associated with high odds for ROP. This study does not support an association of fertility treatments with increased risk for ROP.
Subject terms: Retinal diseases, Physical examination
Introduction
Retinopathy of prematurity (ROP) is a disorder of the developing retina of very low birth weight (VLBW) preterm infants; it potentially leads to blindness in a small but significant percentage of infants [1] and remains a leading cause of childhood blindness in many parts of the world. Therefore, screening these infants is crucial to identify those who could benefit from treatment and to make appropriate recommendations regarding the timing of future screening. In cases of twin births, since both babies share the same gestational age and are exposed to the same pre-natal conditions; these babies might present a varied ROP disease course depending on various factors such as birth weight and other systemic diseases they develop [2, 3]. Azad et al. [4] demonstrated that in 11 pairs of twins (20%) the ROP showed differences in zone or need for treatment, and in 3/4 pairs of twins, the heavier birth weight twin had a more severe ROP course. Wang et al. [5] also studied the effect of birth weight, gestational age, and neonatal morbidities on ROP in preterm twin pairs discordant for birth weight and reported that gestational age was a better predictor of ROP than birth weight. They did not find any relationship to neonatal morbidities. Whilst it has been suggested that fertility treatments may be a risk factor for ROP, others have concluded that it is simply the increased rates of prematurity and low birth weight which generate the higher rates of ROP [6–8], thus, it remains unclear whether fertility treatment is a risk factor for ROP. The purpose of this study was to determine the independent effect of perinatal and neonatal risk factors on ROP and to examine the association of fertility treatments with the risk for ROP in the Israeli population of VLBW very preterm twin infants.
Materials and methods
This population-based observational study was performed on data obtained from the Israel National VLBW Infant Database.
The Israel national very low birth weight infant database
This population-based observational study analysed data from the Israel National Very Low Birth Weight Infant Database on VLBW infants (≤1500 g) born in Israel from January 1995 through December 2020. All 28 neonatal departments in Israel comprising the Israel Neonatal Network participated in the data collection. (Appendix). All neonatal units in Israel have intensive care capabilities. Infants are usually treated from birth to discharge at a single centre, regardless of unit size. The data collected included: demographic details, antenatal and perinatal history, post-delivery status, neonatal diagnoses, medical and surgical treatments, and outcome at discharge. All live born infants receive a unique identification number at birth. Patient information received by the database coordinator is cross-checked with the national birth registry, and missing data for any infant is requested from the birth hospital. Hospital and patient identification subsequently remain confidential by consensus agreement of all participating centres. Data are reported on all infants until death or discharge home.
Study population
From 1995 to 2020, the database included records of 39,263 infants, comprising more than 99% of all VLBW infants born in Israel. The study population comprised twin infant pairs born at 24 to 29 weeks of gestation (Fig. 1). Excluded were infants with a gestational age (GA) of <24 or >29 weeks (n = 19,436), infants who did not undergo ROP screening, mainly due to early deaths (n = 4260), non-twin infants (n = 10,591) and twin infants where only one infant was registered in the database (n = 884). Following exclusions, the final study population comprised 4092 infants (2046 twin pairs).
Fig. 1.
The study cohort selection process is depicted.
Definitions
Definitions have been previously reported in detail [9, 10]. The following maternal demographic and perinatal characteristics were considered: infertility treatment, antenatal care, ethnicity, hypertensive disorders, diabetes mellitus, obstetric complications, including premature labour, premature rupture of membranes (PROM), chorioamnionitis, antepartum haemorrhage (APH), antenatal steroid therapy, and mode of delivery. The GA in completed weeks was defined as the best estimate of GA based on the last menstrual period, obstetric history and examination, prenatal ultrasound, or an early postnatal physical examination. Gender-specific birth weight z-scores and percentiles were determined according to the intrauterine growth charts of Kramer et al. [11]. Small for gestational age (SGA) was considered as birthweight <10th percentile for GA. Birthweight discordance in twin pair siblings was defined as a difference in birthweight of ≥20%. Percent birthweight discordance was calculated as [(birthweight of heavier twin - birthweight of lighter twin/birthweight of heavier twin) × 100]. Delivery room resuscitation included endotracheal intubation, chest compressions or epinephrine administration.
Retinopathy of prematurity (ROP) was staged according to the International Classification of Retinopathy of Prematurity [12–16], and for the purpose of this study, the diagnosis of ROP included ROP stages 2 or greater in either or both eyes. ROP treatment comprised any therapy including cryotherapy, laser coagulation or intravitreal anti-vascular endothelial growth factor, as performed by the treating ophthalmologist. Neonatal morbidities considered included: respiratory distress syndrome (RDS); sepsis defined as positive microbial growth on one or more bloodstream cultures obtained with accompanying clinical signs of sepsis and both early onset sepsis (age <72 h) and late onset sepsis (≥72 h) were included; necrotizing enterocolitis (NEC), defined as definite NEC stages 2–3 according to Bell’s criteria [17] and NEC treated surgically; patent ductus arteriosus (PDA) and PDA treated surgically; and bronchopulmonary dysplasia (BPD) defined as requirement for oxygen therapy at 36 weeks post-conceptional age.
Statistical analysis
Univariate analyses were used to compare the characteristics and outcomes of infants diagnosed with ROP to those with no ROP. The independent effect of these characteristics on ROP was assessed using multivariable logistic regression analyses. For both the univariate and multivariable analyses, the General Estimating Equation (GEE) approach was applied, taking into account the correlation between siblings in twin births. For the analyses of discordant twin pairs, conditional logistic regression models or the GEE approach were applied where appropriate, taking into account the correlation between siblings in twin births. Multivariable analyses were undertaken in two stages. Firstly, the significant demographic and perinatal factors associated with ROP were determined, and secondly major neonatal morbidities, including RDS, PDA, NEC, sepsis, and BPD were included in the model. Significant collinearity was present between birthweight and GA and therefore birthweight was not included in the models. As a measure of intrauterine growth, birthweight z-scores rather than SGA were included in the models. Results of the multivariable analyses are presented as adjusted odds ratios (OR) with 95% confidence intervals (CI). Statistical analyses were performed using the SAS statistical software version 9.4 (SAS Institute, Inc, Cary, NC).
Results
Descriptive statistics
The initial cohort included 39,263 VLBW infants. Following exclusions, the final study population comprised 4092 twin infants (2046 pairs), of whom 851 infants (20.8%) were diagnosed with ROP stages 2–4. The rates of ROP by demographic and perinatal characteristics are shown in the univariate analysis shown in Table 1. The rates of ROP decreased from 68.8% at 24 weeks gestation to 5.7% at 29 weeks (p < 0.0001) Almost 60% of the twins were conceived following infertility treatments (2374/4092). The rate of ROP among infants conceived following infertility treatments was significantly higher than in those conceived spontaneously (22.5 vs 18.6%, p = 0.016) as shown in the univariate analysis. The rates of ROP were also significantly higher in infants of mothers with premature contractions (22% vs 17.8%, p = 0.014) and amnionitis (26.7% vs 20.3%, p = 0.032). SGA infants (29.8% vs 20.1%, p < 0.0001) and infants requiring delivery room resuscitation (30.6% vs 11.3%, p < 0.0001) had significantly higher rates of ROP. The rates of ROP decreased markedly over time, from 29.4% in 1995–2000 to 13.5% in 2013–2020 (p < 0.0001).
Table 1.
Univariate analysis of perinatal and demographic characteristics of 4092 very low birthweight, very preterm twin infants with and without Retinopathy of Prematurity Stage 2 or greater.
| ROP Stages 2.3,4 | No ROP | ||||
|---|---|---|---|---|---|
| n | % | n | % | P value | |
| All Infants | 851 | (20.8) | 3241 | (79.2) | |
| Gestational age (week) | |||||
| 24 vs. 29 weeks | 77 | (68.8) | 35 | (31.2) | <0.0001 |
| 25 vs. 29 weeks | 134 | (54.5) | 112 | (45.5) | <0.0001 |
| 26 vs. 29 weeks | 215 | (39.5) | 329 | (60.5) | <0.0001 |
| 27 vs. 29 weeks | 194 | (25.9) | 554 | (74.1) | <0.0001 |
| 28 vs. 29 weeks | 155 | (13.9) | 959 | (86.1) | <0.0001 |
| 29 | 76 | (5.7) | 1252 | (94.3) | |
| Ethnicity | 0.16 | ||||
| Jewish | 670 | (21.4) | 2458 | (78.6) | |
| Non-Jewish | 181 | (18.8) | 781 | (81.2) | |
| Infertility treatment | |||||
| None | 318 (18.5) | 1400 (81.5) | |||
| Ovulation factor vs. none | 163 (27.9) | 421 (72.1) | <0.0001 | ||
| IVF vs. none | 337 (20.4) | 1311 (79.6) | 0.25 | ||
| Other treatment vs. none | 33 (23.2) | 109 (76.8) | 0.28 | ||
| Weeks starting prenatal care | 0.30 | ||||
| 0–12 weeks | 793 | (21.0) | 2975 | (79.0) | |
| >12 weeks | 58 | (17.9) | 266 | (82.1) | |
| Maternal hypertensive disorders | 0.51 | ||||
| Yes | 61 | (19.1) | 259 | (80.9) | |
| No | 789 | (20.9) | 2979 | (79.1) | |
| Diabetes mellitus | 0.16 | ||||
| Yes | 66 | (17.4) | 314 | (82.6) | |
| No | 782 | (21.1) | 2924 | (78.9) | |
| Premature contractions | 0.014 | ||||
| Yes | 638 | (22.0) | 2256 | (78.0) | |
| No | 213 | (17.8) | 983 | (82.2) | |
| PROM/Amnionitis (amnio) | |||||
| 1. No amnio, ROM No or <24 h | 670 | (20.2) | 2640 | (79.8) | |
| 2. No amnio, ROM > 24 h vs. 1 | 104 | (21.1) | 390 | (78.9) | 0.74 |
| 3. Amnionitis vs. 1 | 77 | (26.7) | 211 | (73.3) | 0.032 |
| Antepartum haemorrhage | 0.50 | ||||
| Yes | 121 | (22.2) | 424 | (77.8) | |
| No | 729 | (20.6) | 2814 | (79.4) | |
| Antenatal steroids | |||||
| None | 168 | (22.1) | 592 | (77.9) | |
| Complete vs. none | 508 | (20.1) | 2018 | (79.9) | 0.34 |
| Partial vs. none | 175 | (21.7) | 631 | (78.3) | 0.88 |
| Delivery mode | 0.06 | ||||
| Vaginal | 185 | (23.9) | 588 | (76.1) | |
| Cesarean section | 666 | (20.1) | 2653 | (79.9) | |
| Sex | 0.31 | ||||
| Male | 485 | (22.1) | 1706 | (77.9) | |
| Female | 366 | (19.3) | 1535 | (80.7) | |
| Birth weight (g) | |||||
| <750 | 229 | (53.0) | 203 | (47.0) | |
| 750–999 vs. <750 | 412 | (30.7) | 928 | (69.3) | <0.0001 |
| 1000–1249 vs. <750 | 172 | (11.0) | 1393 | (89.0) | <0.0001 |
| 1250–1500 vs. <750 | 38 | (5.0) | 717 | (95.0) | <0.0001 |
| Small for Gestational Age | <0.0001 | ||||
| Yes | 210 | (70.2) | 89 | (29.8) | |
| No | 3031 | (79.9) | 762 | (20.1) | |
| Birth weight Z score | |||||
| >75% | 54 | (17.6) | 252 | (82.4) | |
| 25–75% vs. >75% | 543 | (18.9) | 2327 | (81.1) | <0.0001 |
| <25% vs. >75% | 254 | (27.7) | 662 | (72.3) | 0.005 |
| Birth weight Z score (mean ± SD) | −0.368 ± 0.707 | −0.188 ± 0.672 | <0.0001 | ||
| Congenital anomalies | 0.65 | ||||
| Yes | 51 | (18.3) | 228 | (81.7) | |
| No | 799 | (21.0) | 3013 | (79.0) | |
| Delivery room resuscitation | <0.0001 | ||||
| Yes | 615 | (30.6) | 1396 | (69.4) | |
| No | 236 | (11.3) | 1845 | (88.7) | |
| Epoch | |||||
| 1995–2000 | 210 | (29.4) | 504 | (70.6) | |
| 2001–2006 vs. 1995–2000 | 255 | (28.0) | 657 | (72.0) | 0.60 |
| 2007–2012 vs. 1995–2000 | 202 | (18.2) | 906 | (81.8) | <0.0001 |
| 2013–2020 vs. 1995–2000 | 184 | (13.5) | 1174 | (86.5) | <0.0001 |
ROP retinopathy of prematurity, CPAP continuous positive airway pressure, IVF in vitro fertilization, PROM premature rupture of membranes, SD standard deviation, n number.
Factors associated with ROP
The multivariable analyses of perinatal factors associated with ROP is presented in Table 2. Infertility treatments were not significantly associated with an increased risk of ROP: In-vitro fertilization: OR 0.86 (95% CI 0.67–1.10) and ovulation induction OR 1.09 (95% CI 0.80–1.50) as shown in the multivariable analyses. In the final model (Table 2, model 2) including the significant variables only, the odds for ROP approximately doubled with each week decrease in GA, with an OR of 58.00 (95% CI 31.83–105.68) at 24 weeks, OR 25.88 (95% CI 16.76–39.96) at 25 weeks, and OR 12.69 (95% CI 8.84–18.22) at 26 weeks, compared to 29 weeks GA. For each decrease in 1 birth weight z-score there was a 1.82-fold increase in the odds for ROP (95% CI 1.59–2.08). The risk of ROP was significantly lower between the years 2007–2012 and 2013–2020 in comparison to the years 1995–2000 (Table 1). In addition, the risk for ROP decreased by over 3-fold in 2013–2020 compared to 1995–2000 [OR 0.29 (95% CI 0.21–0.40)], (Table 1).
Table 2.
Multivariable analyses of demographic and perinatal factors associated with Retinopathy of Prematurity Stage 2 or greater in 4092 very low birthweight, very preterm twin infants.
| Model 1 Complete Model | Model 2 Significant Variables | |
|---|---|---|
| Confounder | Adjusted OR (95% CI) | Adjusted OR (95% CI) |
| Gestational age (week) | ||
| 24 vs. 29 weeks | 61.63 (33.13–114.64) | 58.00 (31.83–105.68) |
| 25 vs. 29 weeks | 26.15 (16.73–40.87) | 25.88 (16.76–39.96) |
| 26 vs. 29 weeks | 12.31 (8.53–17.76) | 12.69 (8.84–18.22) |
| 27 vs. 29 weeks | 6.15 (4.34–8.72) | 6.13 (4.34–8.65) |
| 28 vs. 29 weeks | 2.69 (1.90–3.81) | 2.69 (1.91–3.80) |
| Ethnicity Non-Jewish | 0.85 (0.66–1.11) | |
| Fertility treatment | ||
| In Vitro Fertilization | 0.86 (0.67–1.10) | |
| Ovulation Induction | 1.09 (0.80–1.50) | |
| Other treatment | 0.87 (0.48–1.57) | |
| First trimester prenatal care | 0.97 (0.63–1.49) | |
| Maternal hypertensive disorders | 0.83 (0.55–1.26) | |
| Diabetes mellitus | 1.00 (0.66–1.49) | |
| Premature contractions | 1.17 (0.91–1.50) | |
| Amnionitis/Premature rupture of membranes (PROM) | ||
| No amnionitis, Rupture of membranes >24 h | 0.92 (0.68–1.25) | |
| Amnionitis | 1.19 (0.80–1.79) | |
| Antepartum haemorrhage | 1.01 (0.74–1.38) | |
| Antenatal steroid therapy | ||
| Partial vs. none | 0.89 (0.64–1.24) | |
| Complete vs. none | 1.00 (0.75–1.33) | |
| Cesarean section | 1.03 (0.79–1.34) | |
| Male sex | 1.08 (0.91–1.28) | |
| Birth weight Z score (one unit decrease) | 1.82 (1.59–2.13) | 1.82 (1.59–2.08) |
| Congenital Malformations | 0.94 (0.64–1.37) | |
| Delivery room resuscitation | 1.55 (1.27–1.89) | 1.54 (1.27–1.87) |
| Epoch | ||
| 2001–2006 vs. 1995–2000 | 0.98 (0.73–1.32) | 0.94 (0.70–1.26) |
| 2007–2012 vs. 1995–2000 | 0.42 (0.31–0.58) | 0.41 (0.30–0.55) |
| 2013–2020 vs. 1995–2000 | 0.29 (0.21–0.40) | 0.28 (0.21–0.38) |
The rates of ROP by major neonatal morbidities are shown in Table 3. Very high rates of ROP were present in infants with BPD (42.2% vs 16.6% for no BPD), sepsis (34% vs 14.9% for no sepsis), surgical NEC (45% vs 19.7% for no NEC) and PDA treated surgically (48.6% vs 15.4% for no PDA) (all p < 0.0001). Multivariable analyses of neonatal morbidities associated with ROP are presented in Table 4 after adjusting for the significant perinatal variables (Table 2, Model 2). Surgically treated NEC (OR 2.04, 95% CI 1.31–3.19), surgically treated PDA (OR 1.63, 95% CI 1.12–2.37), sepsis (OR 1.43, 95% CI 1.20–1.71) and BPD (OR 1.52, 95% CI 1.22–1.90) were all associated with significant increased odds for ROP.
Table 3.
Univariate analysis of major neonatal morbidities in 4092 very low birthweight, very preterm twin infants with and without Retinopathy of Prematurity Stage 2 or greater.
| ROP Stages 2.3,4 | No ROP | ||||
|---|---|---|---|---|---|
| n | % | n | % | P value | |
| All Infants | 851 | (20.8) | 3241 | (79.2) | |
| RDS | <0.0001 | ||||
| Yes | 786 | (22.3) | 2738 | (77.7) | |
| No | 65 | (11.4) | 503 | (88.6) | |
| BPD 36 weeks | <0.0001 | ||||
| Yes | 288 | (42.2) | 395 | (57.8) | |
| No | 559 | (16.6) | 2816 | (83.4) | |
| NEC | |||||
| No NEC | 756 | (19.7) | 3075 | (80.3) | |
| NEC, grade 2 vs. no NEC | 46 | (30.9) | 103 | (69.1) | 0.038 |
| NEC, grade 3 vs. no NEC | 49 | (43.8) | 63 | (56.3) | <0.0001 |
| NEC and Surgery | |||||
| No NEC | 756 | (19.7) | 3075 | (80.3) | |
| NEC, no surgery vs. no NEC | 45 | (30.0) | 105 | (70.0) | 0.055 |
| NEC, surgery vs. no NEC | 50 | (45.0) | 61 | (55.0) | <0.0001 |
| Early-onset sepsis | 0.015 | ||||
| Yes | 29 | (37.2) | 49 | (62.8) | |
| No | 822 | (20.5) | 3192 | (79.5) | |
| Late-onset sepsis | <0.0001 | ||||
| Yes | 415 | (34.1) | 802 | (65.9) | |
| No | 436 | (15.2) | 2439 | (84.8) | |
| Any Sepsis | <0.0001 | ||||
| Yes | 431 | (34.0) | 837 | (66.0) | |
| No | 420 | (14.9) | 2404 | (85.1) | |
| PDA | <0.0001 | ||||
| Yes | 507 | (27.4) | 1345 | (72.6) | |
| No | 344 | (15.4) | 1895 | (84.6) | |
| PDA & treatment | |||||
| No PDA | 344 | (15.4) | 1895 | (84.6) | |
| PDA, no treatment vs. no PDA | 96 | (17.6) | 448 | (82.4) | 0.46 |
| PDA, medical treatment vs. no PDA | 322 | (28.6) | 803 | (71.4) | <0.0001 |
| PDA, surgery vs. no PDA | 89 | (48.6) | 94 | (51.4) | <0.0001 |
ROP retinopathy of prematurity, RDS respiratory distress syndrome, BPD bronchopulmonary dysplasia, NEC necrotizing enterocolitis, PDA patent ductus arteriosus.
Table 4.
Multivariable analyses of major neonatal morbidities associated with Retinopathy of Prematurity Stage 2 or greater in 4092 very low birthweight, very preterm twin infants.
| Confounder | *Adjusted OR (95% CI) |
|---|---|
| Respiratory Distress Syndrome | 1.02 (0.74–1.41) |
| Bronchopulmonary Dysplasia | 1.52 (1.22–1.90) |
| Necrotizing Enterocolitis, no surgery | 1.02 (0.66–1.57) |
| Necrotizing Enterocolitis, surgical treatment | 2.04 (1.31–3.19) |
| Sepsis | 1.43 (1.20–1.71) |
| Patent Ductus Arteriosus, no treatment | 1.00 (0.75–1.34) |
| Patent Ductus Arteriosus, medical treatment only | 1.19 (0.97–1.46) |
| Patent Ductus Arteriosus, surgical treatment | 1.63 (1.12–2.37) |
*Adjusted for Gestational age, birthweight z-score, delivery room resuscitation and epoch (From Table 2, Model 2).
ROP treatment
A similar secondary analysis was performed assessing the factors associated with ROP treatment. Of the 4092 infants, 197 (4.8%) underwent treatment to one or both eyes. Results of the multivariable analyses of perinatal factors and neonatal morbidities associated with ROP treatment are shown in Supplementary Table 1. Treated ROP was similarly associated with lower birthweight z-scores and delivery room resuscitation (OR’s 1.82, 95%, CI 1.39–2.38, and 3.05, 95% CI 1.86–5.01, respectively. Sepsis (OR 1.75, 95% CI 1.27–2.43) and BPD (OR 1.74, 95% CI 1.17–2.58) were associated with significant increased odds for ROP treatment. However, in contrast to the morbidities associated with ROP grades 2 or greater analysis, treated ROP was not significantly associated with surgically treated NEC or PDA.
Discordant twin pairs
Subgroup analyses were performed in two different cohorts of discordant twin pairs: (a) 307 twin pairs discordant for the presence of ROP stage 2 or greater and (b) 329 twin pairs discordant for birthweight (≥20% weight difference).
The univariate analysis of perinatal and demographic characteristics and neonatal morbidities in 307 pairs of twin infants, discordant for ROP are shown in the supplementary Table 2. The twin infants with ROP were of significantly lower mean birthweight (922 ± 200 g vs, 969 ± 197 g; p = 0.004) and birthweight z-scores (−0.417 ± 0.754 vs, −0.224 ± 0.664; p < 0.0001). Multivariable analyses of demographic and neonatal morbidities associated with ROP are presented in Supplementary Table 5. For each decrease in 1 birth weight z-score there was a 2.25-fold increase in the odds for ROP (95% CI 1.55–3.28). Surgically treated PDA was associated with greater than 4-fold odds for ROP in this group of twin pairs.
In the 329 pairs of twins with a birthweight discordance of ≥20%, the mean birthweight and z-scores of the smaller twin were 792 ± 163 g, and −1.313 ± 0.632 respectively, compared to 1130 ± 201 g. and 0.055 ± 0.647 in the larger twin. ROP stage 2 or greater was present in 80/329 (24.3%) of the smaller infants and 45/329 (11.7%) of the larger twins (OR 2.03, 95% CI 1.49–2.77; p < 0.0001).
Discussion
This large population-based observational study evaluated the association between perinatal and neonatal risk factors and ROP in VLBW twins born at 24 to 29 weeks of gestation. In the multivariable analyses adjusted for confounding effects such as demographic and perinatal characteristics, no significant effects of the different fertility treatments on the risk for ROP were found. However, surgical treatments for both NEC and PDA were associated with 2.04 and 1.63-fold increased odds for stages 2 or greater ROP, respectively.
The ROP rate in our cohort, 20.8% (851/4,092), was similar to previous reports. ROP occurred in 25.9% (459/1772) of a Polish cohort [18] and 19.2% (324/1688) of a Dutch cohort [19]. A higher percentage of ROP, 43.1% (3224/7483) was demonstrated in the USA and Canadian Postnatal Growth and Retinopathy of Prematurity (G-ROP) study cohort [20]. This may reflect the inclusion of stage 1 ROP, the higher proportion of extremely preterm infants in their cohort, as well as the earlier time period of the study that was conducted between the years 2006 and 2011. As shown in our study, the incidence of ROP declined strongly with the years, ROP diagnoses between the years 2007 and 2012 in our study were 18.2% vs. 13.5% between the years 2013 and 2020. Similar to our finding, in the multicentre study of cryotherapy for ROP each 100-g increase in BW decreased the odds of reaching threshold ROP by 27%, and each week increase in GA decreased the odds of reaching threshold disease by 19% [21]. These results have been replicated in multiple subsequent studies [22–25].
The discordant twin-pair sub group analyses in our study showed that the twin infants with ROP had a significantly lower mean birthweight. Furthermore, ROP stage 2 or greater was present in 24% of the smaller infants and in only 11% of the larger twins. These findings are similar to those in the Paris population [26], further suggesting that intrauterine growth and low BW for gestational age are significant factors contributing to the development of ROP. Similar findings were found in the Netherlands population of identical twin models [27]. In contrast, the study of Wang et al. found no significant differences in ROP between larger and smaller infants in the Chinese twin population [5]. This difference may be due to structural differences between the populations; Chinese babies are significantly smaller than European babies [28].
An association between fertility treatments and ROP has been studied with inconclusive evidence [7, 29–31]. The study of Tsumi et al. [32] found a significantly higher cumulative incidence of ROP following assisted reproductive technology. A higher risk of stage 3 ROP in singletons born following in vitro fertilization (IVF) treatments was found in the meta-analysis of Gao et al. [33]. In contrast, in the study of Wikstrand et al. [34], no association between intracytoplasmic sperm injection pregnancies and ROP was found. Israel has one of the world’s most extensive IVF system [35] and among the twin infants in our study, almost 60% were conceived following infertility treatments, two-thirds of whom were by IVF. Thus, it might be expected that a high incidence of ROP would be found in the Israeli twin population. Nevertheless, no significant independent effects of the different fertility treatments on ROP diagnosis were found in our study when adjusting for demographic and perinatal confounders. Improvements to the IVF procedure were found to lower the risk of multiple pregnancies, preterm births, and low BW and consequently reduce the ROP rate in the United Kingdom [36].
Surgically treated NEC and PDA were associated with increased odds of ROP. These findings are consistent with Fundora et al.‘s [37] secondary data analysis of 7,483 preterm infants from the G-ROP study. In addition, the study of Malviya et al. [38] reported a significant increase in ROP stages 3 and 4 in the PDA surgical group compared to the indomethacin group. Other observational studies have failed to find a similar association [39]. One possible explanation for the higher ROP incidence in infants requiring surgical treatment for these diseases could be that the duration of the waiting time and transport to another surgical department or even a different facility could adversely affect outcomes. Secondly, the perioperative care of these ill infants could be different, as infants suffering from a more severe disease and infants in need of resuscitation may be treated with more supplemental oxygen, a different oxygen concentration, a longer duration of oxygenation, and even prolonged mechanical ventilation, all of which are among the most frequently identified risk factors for severe ROP. Thirdly, infants who need surgical treatment and or suffer from sepsis are either more critically ill or have a developmental profile that leads to increased morbidities, including ROP. Furthermore, and most importantly, the effects of inflammatory mediators, including cytokines such as interleukin (IL)-6, IL-8, tumour necrosis factor (TNF)-α l and pro-angiogenic factors, in critically ill patients are likely to contribute to the prophase and development of ROP as demonstrated in previous studies [40–42]. Surgical intervention also adds to a higher incidence of surgical complications such as pneumothorax or chronic lung disease that increase the likelihood of ROP by producing inflammatory and pro-angiogenic factors [40–42]. Surgical intervention also means a higher incidence of surgical complications such as pneumothorax or chronic lung disease that increase the likelihood of ROP by producing inflammatory and pro-angiogenic factors [43, 44]. In the Söderström et al. [45] cohort of extremely preterm infants, the incidence of ROP requiring treatment was halved following the implementation of lower oxygen saturation targets, yet an increased incidence of NEC was found in these infants. Interestingly, regression analysis in this study revealed low saturation targets and high GA to be protective factors for treatment requiring ROP, while NEC surgery and severe BPD were associated with an increased risk for ROP, similar to the results in our study. Thus, it can be suggested that caution should be taken regarding the restrictive oxygen approach. Accordingly, the European Guidelines advise against setting saturation targets below 90% [46]. Some of the limitations of the study include its observational design. The database does not include information on the chorionicity or zygosity of the twin pairs, or the possible aetiology of impaired foetal growth, such as twin to twin transfusion, which may further assist in evaluating the relationship between impaired foetal growth and the excess risk for ROP. Another limitation is that the Israel VLBW infant database only collects data solely until discharge from hospital, and the possible impact of perinatal and neonatal risk factors on the long-term developmental outcome of ROP could not be assessed, furthermore, this study potentially omits infants who develop ROP or are treated for ROP after discharge. The strength of this study is that the database includes almost the entire infant population of VLBW infants in Israel. This unique setting enables the formation of a reliable and comprehensive database of both maternal and infant medical information. In conclusion, our study shows that intrauterine growth and low BW for gestational age are significant factors contributing to the development of ROP. In addition, our findings support the higher risk of ROP in infants undergoing surgical treatment for PDA and NEC. Future studies should clarify the mechanisms by which the surgical treatment affects infants’ health and results in ROP. This knowledge may aid in understanding how morbidity related to these treatments might be avoided in the future.
Summary
What was known before
Previous studies have shown that birth weight, and gestational age in preterm twin pairs that are discordant for birth weight are predictors of ROP.
In some studies, surgical treatments for NEC and PDA were associated with an increased risk of ROP. Other studies have failed to find an association with ROP.
Previous studies have suggested that fertility treatments may be a risk factor for ROP, but others have concluded that it is simply the increased rates of prematurity and low birth weight that generate the higher rates of ROP.
What this study adds
This large population-based observational study demonstrates that poor intrauterine growth and surgical interventions for NEC and PDA are associated with high odds for ROP in preterm VLBW twin infants.
This study does not support an association of fertility treatments with increased risk for ROP.
Understanding how morbidity relates to different treatments might be effective in reducing ROP rates in the future.
Supplementary information
Appendix
The Israel Neonatal Network, which compiles the Israel national VLBW infant database, consists of these participating centers: Assaf Harofeh Medical Center, Rishon Le Zion; Assuta Hospital, Ashdod; Barzilay Medical Center, Ashkelon; Bikur Holim Hospital, Jerusalem; Bnei Zion Medical Center, Haifa; Carmel Medical Center, Haifa; English (Scottish) Hospital, Nazareth; French Hospital, Nazareth; Hadassah University Hospital Ein-Karem, Jerusalem; Hadassa University Hospital Har Hazofim, Jerusalem; Haemek Medical Center, Afula; Hillel Yafe Medical Center, Hadera; Italian Hospital, Nazareth; Kaplan Hospital, Rehovot; Laniado Hospital, Netanya; Maayanei Hayeshua Medical Center, Bnei-Brak; Meir Medical Center, Kefar Saba; Misgav Ladach Hospital, Jerusalem; Poriah Hospital, Tiberias; Rambam Medical Center, Haifa; Rivka Ziv Hospital, Zefat; Schneider Children’s Medical Center of Israel and Rabin Medical Center (Beilinson Campus), Petach- Tikva; Shaare-Zedek Hospital, Jerusalem; Sheba Medical Center, Tel-Hashomer; Soroka Medical Center, Beer-Sheva; Sourasky Medical Center, Tel-Aviv; Western Galilee Medical Center, Nahariya; Wolfson Medical Center, Holon; Yoseftal Hospital, Eilat; Coordinating center, The Women and Children’s Health Research Unit, Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center, Tel Hashomer.
Supplemental material is available at Eye’s website
Author contributions
RS cleaned and analysed the data and drafted and revised the paper, TS collected the data and revised the paper, IZP wrote the statistical analysis plan and analysed the data and revised the paper, LLG analysed the data and revised the paper, BR cleaned and analysed the data, wrote the statistical analysis plan and revised the paper, TWJ initiated the project, collected and analysed the data, and revised the paper.
Funding
The Israel National VLBW infant database is partially funded by the Israel Ministry of Health and the Israel Center for Disease Control.
Competing interests
The authors declare no competing interests.
Ethics approval
This study was conducted in accordance with prevailing ethical principles and approved by the Human Research Committee of the Sheba Medical Center (SMC-9443-22).
Footnotes
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
A list of authors and their affiliations appears at the end of the paper.
Contributor Information
Tamara Wygnanski-Jaffe, Email: Tamara.Wygnanski@sheba.health.gov.il.
in collaboration with the Israel Neonatal Network:
Iris Morag, Omer Globus, Shmuel Zangen, Tatyana Smolkin, Yaron Nave, Arieh Riskin, Karen Lavie-Nevo, Zipora Strauss, Clari Felszer, Hussam Omari, Smadar Even Tov-Friedman, Smadar Even Tov-Friedman, Amit Hochberg, Nizar Saad, Calanit Hershkovich Shporen, Aryeh Simmonds, Bernard Barzilay, Sofia Bauer, Amir Kugelman, Eric Shinwell, Gil Klinger, Yousif Nijim, Yaron Nave, Eilon Shani, Dror Mandel, Vered Fleisher-Sheffer, Anat Oron, and Lev Bakhrakh
Supplementary information
The online version contains supplementary material available at 10.1038/s41433-023-02801-8.
References
- 1.American Academy of Pediatrics Section on Ophthalmology. Screening examination of premature infants for retinopathy of prematurity. Pediatrics. 2013;131:189–95. doi: 10.1542/peds.2012-2996. [DOI] [PubMed] [Google Scholar]
- 2.Friling R, Rosen SD, Monos T, Karplus M, Yassur Y. Retinopathy of prematurity in multiple-gestation, very low birth weight infants. J Pediatr Ophthalmol Strabismus. 1997;34:96–100. doi: 10.3928/0191-3913-19970301-08. [DOI] [PubMed] [Google Scholar]
- 3.Blumenfeld LC, Siatkowski RM, Johnson RA, Feuer WJ, Flynn JT. Retinopathy of prematurity in multiple-gestation pregnancies. Am J Ophthalmol. 1998;125:197–203. doi: 10.1016/S0002-9394(99)80092-0. [DOI] [PubMed] [Google Scholar]
- 4.Azad R, Chandra P, Patwardhan SD, Gupta A. Profile of asymmetrical retinopathy of prematurity in twins. Indian J Ophthalmol. 2010;58:209. doi: 10.4103/0301-4738.62645. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Wang ZH, Li YY, Liu ZM. Birth weight and gestational age on retinopathy of prematurity in discordant twins in China. Int J Ophthalmol. 2014;7:663. doi: 10.3980/j.issn.2222-3959.2014.04.14. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Watts P, Adams GG. In vitro fertilization and stage 3 retinopathy of prematurity. Eye. 2000;14:330–3. doi: 10.1038/eye.2000.82. [DOI] [PubMed] [Google Scholar]
- 7.Friling R, Axer-Siegel R, Hersocovici Z, Weinberger D, Sirota L, Snir M. Retinopathy of prematurity in assisted versus natural conception and singleton versus multiple births. Ophthalmology. 2007;114:321–4. doi: 10.1016/j.ophtha.2006.11.010. [DOI] [PubMed] [Google Scholar]
- 8.Han G, Lim DH, Kang D, Cho J, Guallar E, Chang YS, et al. Association between retinopathy of prematurity in very-low-birth-weight infants and neurodevelopmental impairment. Am J Ophthalmol. 2022;244:205–15. doi: 10.1016/j.ajo.2022.08.010. [DOI] [PubMed] [Google Scholar]
- 9.Riskin A, Riskin-Mashiah S, Lusky A, Reichman B, Israel Neonatal Network. The relationship between delivery mode and mortality in very low birthweight singleton vertex-presenting infants. BJOG. 2004;111:1365–71. doi: 10.1111/j.1471-0528.2004.00268.x. [DOI] [PubMed] [Google Scholar]
- 10.Grisaru-Granovsky S, Reichman B, Lerner-Geva L, Boyko V, Hammerman C, Samueloff A, et al. Population-based trends in mortality and neonatal morbidities among singleton, very preterm, very low birth weight infants over 16 years. Early Hum Dev. 2014;90:821–7. doi: 10.1016/j.earlhumdev.2014.08.009. [DOI] [PubMed] [Google Scholar]
- 11.Kramer MS, Platt RW, Wen S, Joseph KS, Allen A, Abrahamowicz M, et al. A new and improved population-based Canadian reference for birth weight for gestational age. Pediatrics. 2001;108:E35. doi: 10.1542/peds.108.2.e35. [DOI] [PubMed] [Google Scholar]
- 12.International Committee for the Classification of Retinopathy of Prematurity. The international classification of retinopathy of prematurity revisited. Arch Ophthalmol. 2005;123:991–9. doi: 10.1001/archopht.123.7.991. [DOI] [PubMed] [Google Scholar]
- 13.An International Classification of Retinopathy of Prematurity. The committee for the classification of retinopathy of prematurity. Arch Ophthalmol. 1984;102:1130–4. doi: 10.1001/archopht.1984.01040030908011. [DOI] [PubMed] [Google Scholar]
- 14.Multicenter Trial of Cryotherapy for Retinopathy of Prematurity. One-year outcome–structure and function. Cryotherapy for retinopathy of prematurity cooperative group. Arch Ophthalmol. 1990;108:1408–16. doi: 10.1001/archopht.1990.01070120056029. [DOI] [PubMed] [Google Scholar]
- 15.Good WV, Early Treatment for Retinopathy of Prematurity Cooperative Group. Final results of the early treatment for retinopathy of prematurity (ETROP) randomized trial. Trans Am Ophthalmol Soc. 2004;102:233–50. [PMC free article] [PubMed] [Google Scholar]
- 16.Chiang MF, Quinn GE, Fielder AR, Ostmo SR, Chan RP, Berrocal A, et al. International classification of retinopathy of prematurity. Ophthalmology 2021;128.
- 17.Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, et al. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Ann Surg. 1978;187:1–7. doi: 10.1097/00000658-197801000-00001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Chmielarz-Czarnocińska A, Pawlak M, Szpecht D, Choręziak A, Szymankiewicz-Bręborowicz M, Gotz-Więckowska A. Management of retinopathy of prematurity (ROP) in a Polish cohort of infants. Sci Rep. 2021;11:1–8. doi: 10.1038/s41598-021-83985-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Van Sorge AJ, Termote JUM, Simonsz HJ, Kerkhoff FT, Van Rijn LJ, Lemmens W, et al. Outcome and quality of screening in a nationwide survey on retinopathy of prematurity in The Netherlands. Br J Ophthalmol. 2014;98:1056–60. doi: 10.1136/bjophthalmol-2013-304493. [DOI] [PubMed] [Google Scholar]
- 20.Quinn GE, Ying GS, Bell EF, Donohue PK, Morrison D, Tomlinson LA, et al. Incidence and early course of retinopathy of prematurity: secondary analysis of the postnatal growth and retinopathy of prematurity (G-ROP) study. JAMA Ophthalmol. 2018;136:1383–9. doi: 10.1001/jamaophthalmol.2018.4290. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Watts P, Adams GGW. In vitro fertilisation and stage 3 retinopathy of prematurity. Eye. 2000;14:330–3. doi: 10.1038/eye.2000.82. [DOI] [PubMed] [Google Scholar]
- 22.Enomoto H, Miki A, Matsumiya W, Honda S. Evaluation of oxygen supplementation status as a risk factor associated with the development of severe retinopathy of prematurity. Ophthalmologica. 2015;234:135–8. doi: 10.1159/000433565. [DOI] [PubMed] [Google Scholar]
- 23.Alajbegovic-Halimic J, Zvizdic D, Alimanovic-Halilovic E, Dodik I, Duvnjak S. Risk factors for retinopathy of prematurity in premature born children. Med Arch. 2015;69:409–13. doi: 10.5455/medarh.2015.69.409-413. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Rasoulinejad SA, Montazeri M. Retinopathy of prematurity in neonates and its risk factors: a seven year study in Northern Iran. Open Ophthalmol J. 2016;10:17–21. doi: 10.2174/1874364101610010017. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ying GS, VanderVeen D, Daniel E, Quinn GE, Baumritter A. Risk score for predicting treatment-requiring retinopathy of prematurity (ROP) in the telemedicine approaches to evaluating acute-phase ROP study. Ophthalmology. 2016;123:2176–82. doi: 10.1016/j.ophtha.2016.06.037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Zloto O, Goldfinger LM, Mazkereth R, Spierer A, Yinon Y. Retinopathy of prematurity in discordant twins: is the small twin at increased risk? Graefe’s Arch Clin Exp Ophthalmol. 2020;258:893–8. doi: 10.1007/s00417-019-04597-4. [DOI] [PubMed] [Google Scholar]
- 27.Spekman JA, Emrani SEI, Schalij-Delfos NE, Slaghekke F, van Klink JMM, Lopriore E, et al. Association between fetal growth-restriction and retinopathy of prematurity using a unique identical twin model. Pediatr Res. 2023:1–6. [DOI] [PubMed]
- 28.Janssen PA, Thiessen P, Klein MC, Whitfield MF, MacNab YC, Cullis-Kuhl SC. Standards for the measurement of birth weight, length and head circumference at term in neonates of European, Chinese and South Asian ancestry. Open Med. 2007;1:e74. [PMC free article] [PubMed] [Google Scholar]
- 29.Barker L, Bunce C, Husain S, Adams GG. Is artificial reproductive technology a risk factor for retinopathy of prematurity independent of the generation of multiple births? Eur J Ophthalmol. 2017;27:174–8. doi: 10.5301/ejo.5000832. [DOI] [PubMed] [Google Scholar]
- 30.Minasian M, Fielder A. IVF babies with ROP at higher gestational age and birth weight: implications of changing screening criteria. Br J Ophthalmol. 2005;89:1066. doi: 10.1136/bjo.2004.062935. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chan RV, Yonekawa Y, Morrison MA, Sun G, Wong RK, Perlman JM, et al. Association between assisted reproductive technology and advanced retinopathy of prematurity. Clin Ophthalmol. 2010;4:1385–90. doi: 10.2147/OPTH.S15587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Tsumi E, Lavy Y, Sheiner E, Barrett C, Harlev A, Bal MH, et al. Assisted reproductive technology and long-term ophthalmic morbidity of the offspring. J Dev Origins Health Dis. 2021;12:627–31. doi: 10.1017/S2040174420000938. [DOI] [PubMed] [Google Scholar]
- 33.Gao L, Shao W, Li N, Tian C, Jia H, Peng X, et al. The risk of retinopathy of prematurity in the infants following assisted reproductive technology: a meta-analysis. Biomed Res Int. 2019:1–11. [DOI] [PMC free article] [PubMed]
- 34.Wikstrand MH, Strömland K, Flodin S, Bergh C, Wennerholm UB, Hellström A. Ophthalmological findings in children born after intracytoplasmic sperm injection. Acta Ophthalmol Scand. 2005;84:177–81. doi: 10.1111/j.1600-0420.2005.00568.x. [DOI] [PubMed] [Google Scholar]
- 35.Shenhav‐Goldberg R, Brym R, Lenton‐Brym T. Family welfare effort, total fertility, and in vitro fertilization: explaining the israeli anomaly. Can Rev Sociol/Revue Can Sociol. 2019;56:421–38. doi: 10.1111/cars.12255. [DOI] [PubMed] [Google Scholar]
- 36.Funnell CL, Dabbs TR. Assisted conception and retinopathy of prematurity: 8-year follow-up study. Eye. 2007;21:383–6. doi: 10.1038/sj.eye.6702215. [DOI] [PubMed] [Google Scholar]
- 37.Fundora JB, Binenbaum G, Tomlinson L, Yu Y, Ying GS, Maheshwari A, et al. Association of surgical necrotizing enterocolitis and its timing with retinopathy of prematurity. Am J Perinatol. 2021. [DOI] [PMC free article] [PubMed]
- 38.Malviya MN, Ohlsson A, Shah SS. Surgical versus medical treatment with cyclooxygenase inhibitors for symptomatic patent ductus arteriosus in preterm infants. Cochrane Database Syst Rev. 2013. [DOI] [PMC free article] [PubMed]
- 39.Chorne N, Leonard C, Piecuch R, Clyman RI. Patent ductus arteriosus and its treatment as risk factors for neonatal and neurodevelopmental morbidity. Pediatr. 2007;119:1165–74. doi: 10.1542/peds.2006-3124. [DOI] [PubMed] [Google Scholar]
- 40.Lee J, Dammann O. Perinatal infection, inflammation, and retinopathy of prematurity. Semin Fetal Neonatal Med. 2012;17:26–9. doi: 10.1016/j.siny.2011.08.007. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Silveira RC, Fortes Filho JB, Procianoy RS. Assessment of the contribution of cytokine plasma levels to detect retinopathy of prematurity in very low birth weight infants. Invest Ophthalmol Vis Sci. 2011;52:1297–301. doi: 10.1167/iovs.10-6279. [DOI] [PubMed] [Google Scholar]
- 42.Jang JH, Kim JG, Lee YH, Bae JG, Park JH. The association between amniotic fluid-derived inflammatory mediators and the risk of retinopathy of prematurity. Medicine. 2022;101. [DOI] [PMC free article] [PubMed]
- 43.Laughon M, O’Shea MT, Allred EN, Bose C, Kuban K, Van Marter LJ, et al. Chronic lung disease and developmental delay at 2 years of age in children born before 28 weeks’ gestation. Pediatrics. 2009;124:637–48. doi: 10.1542/peds.2008-2874. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 44.Lad EM, Hernandez-Boussard T, Morton JM, Moshfeghi DM. Incidence of retinopathy of prematurity in the United States: 1997 through 2005. Am J Ophthalmol. 2009;148:451–8. doi: 10.1016/j.ajo.2009.04.018. [DOI] [PubMed] [Google Scholar]
- 45.Söderström F, Normann E, Holmström G, Larsson E, Ahlsson F, Sindelar R, et al. Reduced rate of treated retinopathy of prematurity after implementing lower oxygen saturation targets. J Perinatol. 2019;39:409–14. doi: 10.1038/s41372-018-0300-y. [DOI] [PubMed] [Google Scholar]
- 46.Sweet DG, Carnielli V, Greisen G, Hallman M, Ozek E, Te Pas A, et al. European consensus guidelines on the management of respiratory distress syndrome–2019 update. Neonatology. 2019;115:432–50. doi: 10.1159/000499361. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.