Type I retinopathy of prematurity in infants with birth weight less than 1251 g: incidence and risk factors for its development in a nursery in Kuwait.

PURPOSE: To report the rate of acute retinopathy of prematurity (ROP) and Type I ROP among infants with birth weight (BW) <1251 g and identify the risk factors for the development of Type I ROP.
MATERIALS AND METHODS: A retrospective review of ROP records of infants with BW <1251 g was performed to identify infants with acute ROP and Type I ROP. Infants with Type I ROP were compared with those without Type I ROP to assess the risk factors for the development of Type I ROP. P < 0.05 was statistically significant. Multivariate analysis was performed and odds ratio (OR) and 95% confidence intervals (CI) were calculated.
RESULTS: Among the 207 infants with BW <1251 g, acute ROP occurred in 154 infants (74.4%) and Type I ROP in 95 eyes of 50 infants (24.4%). The numbers of infants with BW <750 g and BW <1000 g were 19.3% and 58.4%, respectively, and the incidences of Type I ROP were 50% and 36.4%, respectively, among them. Forty-four (46.3%) eyes were treated at stage 2+ ROP in zone I or II. All the eyes treated for Type I ROP showed complete regression. Gestational age at birth (OR 0.657, 95% CI: 0.521-0.827; P < 0.0001) and number of ventilated days (OR 1.017, 95% CI: 1.005-1.029; P = 0.006) were identified as independent risk factors for the development of Type I ROP.
CONCLUSIONS: The rate of Type I ROP in this study is higher than that in previous studies due to the higher number of infants with BW <1000 g in our cohort and the treatment of more eyes with stage 2+ ROP. However, all the treated eyes had a favorable outcome. Gestational age at birth and number of ventilated days were independent risk factors for the development of Type I ROP.

INTRODUCTION

The early treatment of retinopathy of prematurity (ETROP) study showed that the results of treating retinopathy of prematurity (ROP) at certain stages, grouped together as Type I disease in the study,1 were better than treating ROP at threshold disease as recommended by the Cryotherapy for Retinopathy of Prematurity (CRYO-ROP) study.2

If we follow the guidelines of the ETROP study,1 it is estimated that, for every two children that require treatment, one child is treated additionally.3 However, the higher rate of treatment would be rewarded by a lower risk of unfavorable anatomic and visual outcomes.3 Alme et al.4 showed that the rate of treatment almost doubled after the implementation of the ETROP guidelines of treatment.

The presence of plus disease is an important sign for the diagnosis of Type I ROP. However, there can be differences in the diagnosis between experts, leading to the possibility of under- or over-diagnosis of Type I ROP.5 Hence, it is necessary to examine the rate of Type I ROP after the implementation of the guidelines of the ETROP study.1

Though some studies have reported the rate of threshold disease from the Middle East6–9 there are only few studies that have investigated the incidence of Type I ROP.10–12 The rate of severe ROP can vary between various countries.13 Severe ROP requiring treatment has been reported in larger infants in Asian and South American countries than in the Western countries.13–15 In an earlier study before the adaptation of the ETROP study guidelines, we reported severe ROP requiring laser treatment in 47 of 599 infants screened, out of whom 45 infants had birth weight (BW) of <1251 g.9

It would be interesting to examine whether more infants with BW <1251 g develop severe ROP requiring treatment in our population. Apart from the ETROP study, no study is available regarding the incidence of Type I ROP exclusively among infants with BW <1251 g.1 There are also no studies that have identified the risk factors for development of Type I ROP in infants with BW <1251 g.

The aims of this study are to report the rate of acute ROP and Type I ROP among highly vulnerable infants with BW <1251 g and identify the risk factors for the development of Type I ROP.

MATERIALS AND METHODS

A review was performed of the medical and ROP charts of premature infants admitted in the neonatal unit of Al-Adan hospital, Kuwait from August 2005 to April 2008. Approval was obtained from the Institutional Review Board of Al-Bahar Ophthalmology Center, Kuwait for the study. Efforts were made to remain true to the guidelines of the Declaration of Helsinki Principles.

Our screening guidelines for ROP are infants born either with a BW ≤1500 g or at a gestational age (GA) ≤34 weeks. The first examination was performed at 4 weeks chronologic age. Dilatation of the pupils was achieved with cyclopentolate 0.5% eye drops. Sterile lid speculums and scleral depressors were used for each examination. The findings of indirect ophthalmoscopy were recorded in an ROP datasheet at each visit. Acute ROP was classified according to the international classification of retinopathy of prematurity.16 Follow-up examinations were performed every 1-2 weeks or earlier, depending on the presence of ROP. The infants were followed up until the retinal vessels reached within 2 disc diameters of the ora in zone 3 or the acute ROP regressed. Those who reached Type I ROP were treated by indirect diode laser treatment within 48 h. The entire avascular retina up to the ora was treated by placing moderately dense, near-confluent laser burns. Data were collected for each baby regarding the date of birth, sex, singleton or multiple pregnancies, GA at birth, and BW. We also recorded the presence of intrauterine growth retardation (IUGR), intraventricular hemorrhage (IVH), sepsis, hyaline membrane disease (HMD) and its severity, number of ventilated days till the day of laser treatment or the highest stage of ROP was reached when present or the number of ventilated days until post-menstrual age (PMA) of 40 weeks in case no ROP was present, number of units of blood transfusion, and of necrotizing enterocolitis (NEC). The stage and zone of ROP if present, the PMA at which first ROP occurred, and the number of infants who developed Type I ROP were noted.

Statistical analysis

Data were entered into Statistical Package for Social Sciences (SPSS) version 17 (SPSS Inc., IBM Corp., Armonk, NY, USA) and double-checked for any errors prior to analyses. We then used univariate logistic regression to identify the association between Type I ROP and various risk factors. Subsequently multivariate logistic regression was conducted with the stepwise (Wald) method to identify the independent predictors. P < 0.05 was statistically significant.

RESULTS

A total of 411 infants completed screening for ROP. There were 207 infants with BW <1251 g who were eligible for inclusion in this study. The average BW, GA, sex, and whether they are the result of multiple or single pregnancy for the 207 infants who form the subjects of this study are mentioned in Table 1. A number of infants had morbidities such as: IUGR in 54 (26.1%); sepsis in 111(53.4%); IVH (in 91 (44%); NEC in 35 (16.9%); and HMD of grade 3 or worse in 95 (45.9%) [Table 2]. Table 2 also presents the number of days on ventilation and the number of packed cells transfused in infants with and without Type I ROP.

Table 1

Characteristics of all infants with birth weight <1251 g and infants with and without Type I retinopathy of prematurity

Table 2

Distribution of variables among all infants with birth weight <1251 g and infants with Type I retinopathy of prematurity and infants without Type I retinopathy of prematurity

Of the 411 infants who were screened, 192 (46.7%) showed acute ROP in at least one eye. Type I ROP was observed in 99 eyes of 52 (12.7%) infants. All the 99 eyes of 52 infants were treated and all the treated eyes showed regression of the disease. Out of the 207 infants with BW <1251 g who are the subjects of this study, 154 (74.4%) developed acute ROP in at least one eye and 50 infants (24.4%) developed Type I ROP in at least one eye. Forty-five (21.7%) infants had bilateral Type I ROP. Zone I disease was noticed in 10 (3.5%) of 287 eyes with some ROP. All the 10 eyes with zone I needed to be treated. Eight eyes (10.1%) needed retreatment of skipped areas. Of the 95 eyes treated, 44 (46.3%) eyes were treated when ROP was stage 2+ in zone I or II and the rest were treated when the ROP was stage 3+ in zone I or II.

Out of the 95 eyes treated, three eyes progressed to stage IVA, which resolved spontaneously in all cases. All the 95 eyes showed complete regression of Type I ROP, with no eye developing unfavorable structural outcome as defined in the CRYO-ROP study.2 The average PMA at treatment of these 50 infants was 35 weeks (range: 31-40, SD ±2.3 weeks). The average chronological age at treatment was 8.2 weeks (range: 6-12, SD ±1.8 weeks).

The incidences of ROP among infants with BW <750 g and BW <1000 g were 97.5% and 93.4%, respectively. Similarly, the incidences of Type I ROP among infants with BW <750 g and BW <1000 g were 50% and 36.4%, respectively.

Tables 3 and 4 show the distribution of acute ROP according to GA and BW, respectively. Table 5 shows the highest stage of ROP in the right and left eyes.

Table 3

Gestational age and incidence of Type I retinopathy of prematurity

Table 4

Birth weight and incidence of Type I retinopathy of prematurity and acute retinopathy of prematurity

Table 5

Incidence of highest stage of retinopathy of prematurity seen in right and left eyes

Statistical analysis of infants with BW <1251 g with and without Type I ROP identified several risk factors to be significant for development of Type I ROP. These are listed in Tables 1 and 2. However, the multiple logistic regression analysis showed only GA at birth (odds ratio (OR) 0.657, 95% CI 0.521-0.827; P < 0.0001) and number of ventilated days (OR 1.017, 95% CI 1.005-1.029, P = 0.006) as independent risk factors for the development of Type I ROP.

DISCUSSION

Acute ROP among infants with BW <1251 g

The rate of acute ROP in infants with BW <1251 g in the literature varies from 46% to 68.8%.69–1117–19 However in our study, the rate of acute ROP was slightly higher at 74.4%. The probable reason for the higher rate of acute ROP in our study compared to other studies from the region could be due to the presence of more number of infants with lower BWs in our cohort. The numbers of infants with BW <1000 g and with BW <750 g were 58.4% and 19.3%, respectively, of our cohort of 207 infants [Table 4]. Acute ROP was seen in 97.5% of infants with BW <750 g and in 92.6% of infants with BW <1000 g in our study. In study by Mutlu et al.11 there were 77 infants with BW<1251g out of whom 37(48%) had BW<1000g. Acute ROP was seen in 81% of the infants with BW <1000 g and in 68.8% of all infants with BW <1251 g.11 Binkathen et al.8 reported the incidence of acute ROP among 166 infants, out of whom 26 infants had BW <1000 g. Twenty-two out of these 26 infants (84%) developed acute ROP.8 However, Binkathen et al.’s8 study does not provide data regarding the incidence of acute ROP among infants with BW <1251 g. Kumar et al.20 reported acute ROP among only 32.8% of infants with BW <1000 g. In their study, infants with BW <1000 g constituted 48.8% of infants with BW <1251 g20 This study does not report the incidence of acute ROP among infants with BW <1251 g or among infants with BW <750 g.

However, studies from other regions report varying rates of acute ROP among infants with BW <750 g and <1000 g. The multicenter studies of Lorenz et al., Zin et al., and the ETROP incidence study reported acute ROP in 70.4%, 79.2%, and 92.7%, respectively, among infants with BW <750 g.17–19 In these studies, the number of infants with BW <750 g formed 10.8%, 15.7%, and 24.9% of infants with BW <1251 g, respectively.17–19 Hence, the rate of acute ROP increases as more infants with lower BW are present in the cohort of infants with BW <1251 g.

Lorenz et al. reported acute ROP among 62.1% of infants with BW <1000 g, who formed 56.1% of infants with BW <1251 g.17 Zin et al., in their multicenter study, reported acute ROP among 61.9% of infants with BW <1000 g.18 In their study the infants with BW <1000 g formed 45.1% of infants with BW <1251 g.18 In the ETROP incidence study, acute ROP was seen in 82.5% of infants with BW <1000 g. Infants with BW <1000 g constituted 62.6% of the 6998 infants with BW <1251 g.19

Type I ROP in infants with BW <1251 g

The rate of Type I ROP among the 207 infants in the present study was 24.4%. Type I ROP was seen in 50% of the 40 infants with BW <750 g and in 36.4% of the 121 infants with BW <1000 g. Mutlu et al.11 reported severe ROP in 20.8% of the subgroup of infants with BW <1251 g, which is closer to the rate in our study. However, in this study the treated infants included both Type I ROP and threshold ROP depending on the period of study.11 In this study, the rate of treatment was 32.4% in the subgroup of infants with BW <1000 g, which is again comparable to our study.11 Karkhaneh et al. do not report the rate of Type I ROP among infants with BW <1251 g but they reported Type I ROP among 12.6% of 117 infants with BW <1000 g.10

In a study by Gharaibeh et al.,12 among the 91 infants studied for the incidence of acute ROP, there were only 11 infants weighing <1000 g (12.1%) compared to 58.4% of 207 infants in the present study. The authors do not mention the number of infants with BW <1251 g.12 The incidence of Type I ROP among 11 infants with BW <1000 g was 36.4% in the study by Gharaibeh et al.,12 which is similar to our study.12 Kumar et al. reported severe ROP that included Type I ROP or threshold ROP depending on the period of screening, among 8.6% of infants with BW <1251 g, which is much lower than our rate of 24%.20 The rate of Type I ROP in the ETROP incidence study, which included infants with BW <1251 g, can only be estimated to be 16% from the data in the study.19 Other studies do not provide data regarding the rate of Type I ROP among subgroup of infants with BW <1251 g.48101718 We believe there is a need to report the rate of Type I ROP among the subgroup of infants with BW <1251 g as it provides a yardstick to compare various studies regarding the incidence of Type I ROP because screening guidelines can vary between centers.

The higher rate of Type I ROP in our study could be due to several reasons. The possibility of racial predisposition and differences in neonatal care contributing to the higher rate of severe ROP have been highlighted by Gilbert et al.13 and Aralikatti et al.15 Other studies from the Middle East have also observed that larger infants develop severe ROP in their population.101214 The presence of the higher proportion of infants with BW <750 g and <1000 g in our cohort as highlighted in an earlier section compared to other studies from Asia may be another reason why the current study reports a higher rate of Type I ROP compared to previous studies1120 as low BW is an important risk factor for severe ROP.920–22

However, a Japanese study of infants with BW <1000 g reported severe ROP needing treatment in nearly 41% of the infants studied, which is comparable to 36.4% among all infants with BW <1000 g in our study.23

The average BW and GA at birth of all the screened infants in the ETROP incidence study were 907 g and 27.7 weeks, respectively,19 which are comparable to 937.8 g and 28.1 weeks, respectively, reported in the present study. The infants who were randomized for early or late treatment in the ETROP study1 had average BW and GA of 703 g and 25.3 weeks, respectively. However, in our study the average BW and GA of the 50 cases of Type I ROP were higher at 798 g and 26.6 weeks, respectively. Hence, our study shows that smaller infants in our population develop severe ROP more often than infants in the west.

Another important reason for the higher rate of treatment in our study is the higher proportion of Type I ROP with stage 2+ disease [Table 5]. In our study, 47 of the 95 eyes (49.5%) treated had stage 2+ disease in zone I or II in contrast to 43 out of 266 eyes (16.6%) in the ETROP study (from Table 9 in the study).1 None of the eyes treated at stage 2+ in our study had unfavorable structural outcome. It is difficult to state that stage 2+ disease represents a milder variety of Type I ROP and hence our study has obtained better results, as nearly 20% of the 43 eyes treated at stage 2+ in the ETROP study had unfavorable outcome. Our study shows that prompt recognition and treatment of stage 2+ disease may offer better results in the treatment of this potentially vision-threatening condition. However, zone I disease was observed in 9.1% of eyes with acute ROP in the ETROP incidence study compared to only 10 out of 287 eyes (2.9%) of infants with ROP in our study.19 All these 10 eyes in our study required treatment, constituting 10.5% of the 95 treated eyes. This is in contrast to the presence of zone I disease in nearly 40% of the eyes treated in the ETROP study.1 This may be one of the reasons why the structural outcomes were better in our study than that of ETROP studies.119

If the higher rate of treatment and better results in our study are due to the over-diagnosis of pre-plus disease as plus disease in our population is difficult to prove though all efforts were made to verify the presence of plus disease by another specialist in case of doubt.

Risk factors for Type I ROP

We found that lower BW (P < 0.0001) and lower GA at birth (P < 0.0001) were significant for the development of Type I ROP. Other risk factors found to be significant were the presence of sepsis, NEC, number of ventilated days, number of blood units transfused, need for medical or surgical treatment for PDA, and HMD of grade 3 or more [Tables 1 and 2]. Other studies too have found low BW, low GA, respiratory distress syndrome, number of blood units transfused, and number of ventilated days to be significant risk factors for the development of severe ROP requiring treatment among the screened population.7920–2224–26

We did not find sex, single or multiple pregnancies, administration of surfactant, IVH, and IUGR to be significant for the development of Type I ROP. However, Darlow et al. found male sex and IUGR to be significant factors for developing severe ROP compared with females.24

Presence of sepsis, NEC, and the need for treatment of PDA have been recognized as risk factors for the development of Type I ROP as was the case in our study.20 Number of packed cell transfusions has also been found to be a significant risk factor for the development of threshold disease in other studies.2627

However, multivariate analysis showed GA at birth and number of ventilated days to be independent risk factors for the development of Type I ROP among our cohort of infants with BW <1251 g. Akkoyun et al.26 also showed the number of ventilated days to be an independent factor for severe ROP requiring treatment.

CONCLUSIONS

The rate of Type I ROP was 24.2% in infants with BW <1251 g, which is higher than that in other studies. The number of treated eyes with stage 2 in zone I or II with plus disease (49.5%) is higher in our study. However, the higher rate of treatment and earlier treatment yielded zero incidence of unfavorable outcome in all the 99 eyes treated. GA at birth and number of ventilated days were found to be the independent risk factors for Type I ROP among infants with BW <1251 g.