Ranibizumab versus laser therapy for the treatment of very low birthweight infants with retinopathy of prematurity (RAINBOW): five-year outcomes of a randomised trial

Abstract

Background: Concerns remain over the long-term safety of vascular endothelial growth factor (VEGF) inhibitors to treat retinopathy of prematurity (ROP). RAINBOW is an open label randomised trial comparing intravitreal ranibizumab (in 0.2 mg and 0.1 mg doses) with laser therapy in very low birthweight infants (<1500 g) with ROP.

Methods: Of 201 infants completing RAINBOW, 180 were enrolled in the RAINBOW Extension Study. At 5 years, children underwent ophthalmic, development and health assessments. The primary outcome was visual acuity in the better-seeing eye. The study is registered with ClinicalTrial.gov, NCT02640664.

Findings: Between 16-6-2016 and 21-4-2022, 156 children (87%) were evaluated at 5 years. Of 32 children with no acuity test result, 25 had a preferential looking test, for 4 children investigators reported low vision for each eye, and in 3 further children no vision measurement was obtained. 124 children completed the acuity assessment, the least square mean (95% CI) letter score in the better seeing eye was similar in the three trial arms-66.8 (62.9-70.7) following ranibizumab 0.2 mg, 64.6 (60.6-68.5) following ranibizumab 0.1 mg and 62.1 (57.8-66.4) following laser therapy; differences in means: ranibizumab 0.2 mg v laser: 4.7 (95% CI: -1.1, 10.5); 0.1 mg v laser: 2.5 (-3.4, 8.3); 0.2 mg v 0.1 mg: 2.2 (-3.3, 7.8). High myopia (worse than -5 dioptres) in at least one eye occurred in 4/52 (8%) children following ranibizumab 0.2 mg, 8/55 (15%) following ranibizumab 0.1 mg and 11/45 (24%) following laser therapy (0.2 mg versus laser: odds ratio: 3.99 (1.16-13.72)). Ocular and systemic secondary outcomes and adverse events were distributed similarly in each trial arm.

Interpretation: 5-year outcomes confirm the findings of the original RAINBOW trial and a planned interim analysis at 2 years, including a reduced frequency of high myopia following ranibizumab treatment. No effects of treatment on non-ocular outcomes were detected.

Funding: Novartis Pharma AG.

Keywords: Anti-VEGF treatment; Infant; Neurodevelopment; Preterm; Randomised controlled trial; Retinopathy of prematurity.

Fig. 1

Consort diagram for children who completed RAINBOW and entered the RAINBOW extension study.

Fig. 2

Visual acuity for children evaluated at 5 years in the RAINBOW Extension Study. Individual results are shown as dots in study groups as ranibizumab 0.2 mg (RBZ0.2 mg: light grey), ranibizumab 0.1 mg (RBZ0.1 mg: mid grey), and laser (dark grey). Results are shown for the better and worse seeing eyes, and separately for children who received only a Cardiff Card binocular assessment; LV indicates a small number of children were recorded as having low or no vision without Early Treatment Diabetic Retinopathy Study (ETDRS) chart Letter Score results or Cardiff Acuity Card results. In three further children (not shown) no vision could be measured. Y axes show ETDRS Letter score (left y axis), equivalent LogMAR value or Snellen chart equivalent (right y axis and dotted lines).

Fig. 3

T-scores (top graphs) and Raw scores (bottom graphs) from the Mullen Scales of Early Learning at 5 years among children evaluated at 5 years in the RAINBOW extension study. The dotted line represents standardization population mean; error bars show the median with 25th and 75th percentiles.