IGF-1 for Premature Infant Eye Disease Shows No Clear Benefit in Small Trials

Retinopathy of prematurity is a potentially blinding disorder affecting the developing retinas of extremely premature infants. Normal retinal vascularization depends heavily on insulin-like growth factor-1 (IGF-1), which is supplied transplacentally during fetal life. When infants are born before 28 weeks' gestation, IGF-1 levels plummet, and the resulting growth factor deficiency is hypothesized to disrupt the orderly progression of retinal vessel growth — setting the stage for the pathological neovascularization that defines ROP. The therapeutic rationale for postnatal IGF-1 supplementation is therefore biologically compelling: restore circulating IGF-1 to in utero levels and potentially interrupt ROP pathogenesis before it begins.

This 2026 Cochrane systematic review (updated from a 2018 protocol) searched CENTRAL, MEDLINE, Embase, CINAHL, Epistemonikos, and major clinical trial registries through March 10, 2025. Eligibility was restricted to randomized controlled trials, cluster-RCTs, and quasi-RCTs comparing IGF-1 to standard care or placebo in preterm infants. Only two studies met inclusion criteria: a two-center RCT enrolling 19 infants and a 20-center RCT enrolling 121 infants, both conducted between 2011 and 2016 in Europe and North America. Both trials enrolled infants at 23 weeks to 27 weeks 6 days gestational age and administered intravenous IGF-1 as mecasermin rinfabate starting on day one of life, with follow-up extending through 40 weeks' postmenstrual age and, in some cases, five to six years of age.

For the primary outcome of Type 1 ROP or ROP requiring treatment, pooled analysis of 116 infants showed a risk ratio of 0.94 (95% CI 0.38–2.35; P=0.90; I²=0%), meaning IGF-1 treatment had no detectable effect compared to standard care. For ROP of stage 3 or greater, the RR was 1.27 (95% CI 0.61–2.65; P=0.52), and for ROP of any severity the RR was 1.30 (95% CI 0.94–1.80; P=0.12). None of these results reached statistical significance, and all were rated very low certainty by GRADE assessment.

Safety signals were more concerning. Pooled analysis of serious adverse events (SAEs) across both studies (140 infants) yielded an RR of 1.18 (95% CI 0.94–1.47; P=0.15; I²=65%), which was not statistically significant. However, a pre-specified post-hoc sensitivity analysis restricted to the larger 121-infant RCT found a statistically significant increase in SAEs with IGF-1: RR 1.28 (95% CI 1.01–1.62; P=0.05), risk difference 0.17 (95% CI 0.01–0.33; P=0.04), with a number needed to harm of approximately 5.9. Mortality data from the larger trial showed an RR of 1.69 (95% CI 0.71–3.99; P=0.23), and hypoglycemia rates were identical between groups (RR 1.00, 95% CI 0.62–1.63; P=0.99).

The authors assessed all outcomes as very low certainty using GRADE, citing two primary concerns: the extremely small total enrollment (140 infants across both trials) and high risk of bias in two domains for each included study. Both trials were also industry-funded, raising additional concerns about potential conflicts of interest. The review concludes that current evidence is insufficient to draw any conclusions about IGF-1's efficacy or safety for ROP prevention or treatment in preterm infants, and that larger, well-designed, independently funded trials are needed before this intervention can be recommended or definitively ruled out.