Prediction of severe retinopathy of prematurity using the winrop algorithm in a cohort from malopolska. A retrospective, single-center study

Retinopathy of prematurity (ROP) is a developmental vascular proliferative disease. It is a serious problem among the preterm infant population. The frequency of ROP is estimated between 10% to 21% among children under 1,250 g, however, it can reach even up to 68% [1, 2, 3]. Retinopathy of prematurity should be effectively and quickly detected, because if left untreated, it can lead to visual impairment or even vision loss [4]. The early detection of problems provides an opportunity to apply proper prevention. The effective screening programme should identify infants with ROP that require treatment. Currently the gold standard test for the diagnosis of ROP is ophthalmological examination, but it is a stressful and painful procedure [5]. In contrast to that, the WINROP algorithm is a non-invasive, simple and inexpensive method of predicting ROP risk.

The online WINROP (weight, insulin-like growth factor [IGF], neonatal, retinopathy of prematurity [ROP]) algorithm (https://www.winrop.com) can be used to identify infants with increased risk for developing severe ROP [4]. It was developed in Gothenburg, Sweden based on their studies of IGF-1 showing a correlation between a prolonged period of low-level serum IGF-1 and ROP [6, 7]. WINROP calculated on the basis of birth weight, postnatal weight gain, and gestational age determines the approximate IGF-1 levels and in this way identifies the risk of developing sight-threatening ROP [8].

The WINROP surveillance system has been validated in several studies. In highly developed countries (Sweden and United States) the accuracy of the test was higher, as opposed to less and moderately developed countries (Taiwan, Turkey, Korea) [8, 9, 10]. ROP newborns in those countries were more mature and the WINROP algorithm omitted more infants with severe ROP.

The vascular endothelial growth factor (VEGF) and insulin-like growth factor 1 are human proteins, which are necessary for the normal development of retinal blood vessels. Prolonged low levels of serum IGF-1, which usually affect preterm infants, may finally lead to the development of ROP [6]. Gestational age and birth weight cannot be considered in isolation from other risk factors of ROP [8]. Retinopathy of prematurity is also associated with other risk factors, such as assisted ventilation for longer than one week, surfactant therapy, cumulative illness severity, low caloric intake, hyperglycemia and insulin therapy, sepsis, fluctuations in blood gas measurements, intraventricular hemorrhage (IVH), bronchopulmonary dysplasia (BPD) and early administration of erythropoietin for the treatment of anemia of prematurity [11, 12, 13, 14, 15, 16, 17, 18].

The aim of this study was to validate the diagnostic accuracy of the WINROP algorithm for detection of severe ROP in a single-centre cohort of preterm infants from Malopolska, Poland.

In our analysis the WINROP algorithm correctly identified 22 out of 27 (81.5%) preterm infants, who developed type 1 ROP and required treatment according to ETROP criteria. Five newborns who developed sightthreatening ROP were omitted by that online surveillance system (false negative results). In several studies from Sweden and the United States the sensitivity of the WINROP algorithm in detecting proliferating ROP reached 100% [22, 23, 24]. The sensitivity of WINROP in different studies is heterogeneous, with a range of 64.7% to 100% [8, 9, 10, 24]. In our cohort group the sensitivity of 81.5% is comparable to the value estimated in Turkey or Taiwan [8, 9]. There are some reasons that could have had an impact on those discrepancies between the studies.

The disparity possibly came from the fact that the WINROP algorithm was developed and validated in children from Sweden and the United States. The sensitivity validated on a population from less developed countries was significantly lower [9].

Previous study outcomes suggested that larger, more mature infants develop severe ROP in lower developed countries compared with highly developed countries [25]. In the results of those studies it was said that the mean birth weight of premature patients was lower in highly developed countries, ranging from 737 to 763g than in less developed countries, where it ranged from 903 to 1527 g. The gestational age of infants with severe ROP was also distinct: 25.3 to 25.6 weeks in highly developed countries compared with 26.3 to 33.5 weeks in less developed countries. Various mixes of cases, neonatal care, survival rates and screening practices could have impact on those conditioning. In our study the values of those parameters were heterogeneous. The gestational age of infants not identified by WINROP was 23-28 weeks and the birth weight 720-1270 grams.

The gestational age and birth weight are major risk factors significantly associated with ROP. There are also multiple contributing risk factors, including: bronchopulmonary dysplasia, intraventricular hemorrhage, patent ductus arteriosus, mechanical ventilation and sepsis. Surfactant therapy in multivariate analysis included 38 possible risk factors which were also associated with higher incidence of ROP. Surfactant therapy decreased the mortality of extremely low-birth-weight infants, i.e. patients most at risk for retinopathy of prematurity. Selecting this factor has contributed to enhancing WINROP accuracy [11, 17, 26].

Several studies from Korea and Turkey have shown that patients who were not identified by the WINROP algorithm had additional risk factors, including BPD, IVH, necrotizing enterocolitis (NEC) or sepsis, which added to the analysis and significantly improved the sensitivity of the WINROP algorithm results [8, 10]. In our study 20 out of 27 patients with type 1 ROP and 4 out of 5 patients omitted by WINROP required surfactant therapy [27]. The modified algorithm included surfactant therapy as an additional risk factor and increased sensitivity to 96.3% [26, 27].

Patients admitted to the Neonatal Intensive Care Unit of the Department of Pediatrics, Jagiellonian University, were a high-risk population with significant disorders including congenital defects, surgical disorders and severe prematurity complications. Infants with serious postnatal problems were transported from referring hospitals. These aspects may be especially relevant to confirming that the infant population is highly specific. Screening examination of premature infants for retinopathy of prematurity is regularly performed due to the significant number of hospitalized patients with these diseases. The low frequency of antenatal steroid therapy among hospitalized patients (47/79) is a remarkable risk factor of ROP occurring. Antenatal dexamethasone or betamethasone are associated with a reduced risk for ROP [28, 29].

The study was limited by a comparatively small number of patients and confirmation of the disease only by a single ophthalmologist, without inter-observer variability assessment. Diagnosis and documentation of ophthalmoscopic findings may be heavily subjective. It was documented that the agreement on plus disease diagnosis among pediatric ophthalmologists is low [30].

WINROP is a useful noninvasive screening surveillance system which can predict proliferative retinopathy of prematurity for very low birth weight infants in highly developed countries. As a result of repetitive falsely negative outcomes, it demonstrates insuficient sensitivity to becoming a screening test in a Polish cohort group. Modification of the WINROP algorithm by taking different factors into account, like surfactant therapy after birth, increases the test’s sensitivity and only individual cases remain falsely negative. The accuracy of the WINROP algorithm should be validated in large multi-center prospective studies in the Polish population of preterm infants.