Placental gene signatures may predict which premature infants relapse after anti-VEGF eye treatment
Placental RNA sequencing flagged nine genes tied to VEGF signaling that were elevated in preterm infants whose retinopathy returned after ranibizumab.
Summary
Retinopathy of prematurity (ROP) is a leading cause of childhood blindness, and anti-VEGF injections like ranibizumab can treat it. But in some infants the disease reactivates, and doctors cannot currently predict which ones. Korean researchers sequenced RNA from stored placentas of 12 preterm infants treated with ranibizumab, comparing 4 whose ROP reactivated with 8 whose ROP did not. They found 329 differentially expressed genes, including 15 upregulated hub genes linked to VEGFA-VEGF receptor 2 signaling. Nine of these (ACTG1, BSG, ETS1, FN1, FOXO4, IGFBP3, NUMB, MDM2, SMARCA2) showed an AUC above 0.7 for distinguishing reactivation. Seven are tied to angiogenesis. The findings suggest placental gene expression could one day serve as an early biomarker, though the sample is very small and the results need validation.
Detailed Summary
Retinopathy of prematurity (ROP) is a major cause of vision loss in premature infants. Anti-VEGF drugs such as ranibizumab are widely used, but the disease sometimes reactivates after treatment, requiring prolonged follow-up and further intervention. No reliable way exists to predict which infants are at risk.
This retrospective case series asked whether the placenta, which reflects the fetal environment, carries a molecular signature of reactivation. Researchers enrolled preterm infants initially treated with 0.2 mg ranibizumab whose maternal placentas were stored in a biobank. They ran RNA sequencing on placentas from infants with reactivated ROP (n=4) and without (n=8), then applied functional enrichment, protein-protein interaction networks, and hub gene identification. Genes with ROC values of at least 0.7 were considered potentially predictive.
The analysis found 329 differentially expressed genes (207 up, 122 down). Fifteen hub genes, all upregulated, were linked to VEGFA-VEGF receptor 2 signaling and trophoblast cell surface antigen 2 regulatory signaling. Nine (ACTG1, BSG, ETS1, FN1, FOXO4, IGFBP3, NUMB, MDM2, SMARCA2) had an AUC above 0.7. Apart from FOXO4 and SMARCA2, the other seven were associated with angiogenesis, including VEGF pathway and endothelial cell activity.
The authors propose that heightened pro-angiogenic signaling in the placenta may mark infants more prone to ROP relapse, and that these genes could become biomarkers for risk stratification after anti-VEGF therapy.
Important caveats apply. Only 12 infants were analyzed, so the findings are exploratory and prone to overfitting, especially with an AUC threshold as low as 0.7. There was no independent validation cohort, and the abstract does not describe adjustment for confounders such as gestational age or maternal conditions. Placental tissue is also not available for prospective screening in a form that is clinically practical yet.
Key Findings
- RNA sequencing of placentas identified 329 differentially expressed genes (207 up, 122 down) in infants with reactivated ROP.
- Fifteen upregulated hub genes were linked to VEGFA-VEGF receptor 2 signaling and trophoblast cell surface antigen 2 signaling.
- Nine hub genes, including FN1, ETS1, MDM2 and IGFBP3, showed AUC above 0.7 for predicting ROP reactivation.
- Seven of the nine predictive genes are associated with angiogenesis and endothelial cell activity.
- Study compared only 4 reactivated cases with 8 controls, so results are preliminary.
Methodology
Retrospective case series using stored placentas from preterm infants initially treated with 0.2 mg ranibizumab for ROP. Bulk RNA sequencing compared reactivated ROP (n=4) with controls (n=8), followed by enrichment, protein-protein interaction, hub gene analysis, and ROC screening (threshold ≥0.7).
Study Limitations
The very small sample (4 vs 8) and lack of external validation make the AUC results unreliable and likely overfit. Confounding factors, multiple-testing handling, and tissue sampling details are not clear from the abstract. Placenta-based prediction is also only feasible at birth and would need confirmation by protein-level or targeted assays.
Enjoyed this summary?
Get the latest longevity research delivered to your inbox every week.
Enter your email to subscribe: