Intravitreal aflibercept for proliferative diabetic retinopathy – Authors' reply

• Published in: The Lancet (British edition) Vol. 390; no. 10108; pp. 2141 - 2142
• Main Authors: Sivaprasad, Sobha, Prevost, A Toby, Vasconcelos, Joana, Riddell, Amy, Hykin, Philip
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 11-11-2017; Elsevier Limited
• Subjects: Angiogenesis Inhibitors; Bevacizumab; Diabetes; Diabetes mellitus; Diabetic Retinopathy; Edema; Humans; Receptors, Vascular Endothelial Growth Factor; Recombinant Fusion Proteins; Retinopathy; Statistical models; Vascular endothelial growth factor; Vascular Endothelial Growth Factor A; Visual impairment; United Kingdom > UK; England; Wales
• ISSN: 0140-6736; 1474-547X
• DOI: 10.1016/S0140-6736(17)32422-4

The median duration of diabetes in the aflibercept group was 20 years (IQR 14·0-26·0) and 20 years (13·3-27·0) in the PRP group. [...]the discussion of our study conclusions remains unaltered. With regard to the points raised by David M Brown and Charles C Wykoff, the statistical model for the primary outcome was predefined before the availability of 52-week data and included the refracted best-corrected visual acuity (BCVA) visit timepoints at baseline, 12, and 52 weeks. Since this was a non-inferiority study with the lower inferiority margin set at −5 letters, we prespecified that the model should exclude temporary substantial decreases in BCVA due to vitreous haemorrhage within 3 months of the week 12 visit and week 52 visit that would cause transient, wide CIs, defined as more than 3 SD below the mean, to avoid these values interfering with the ability to answer the research question. No patients in the aflibercept group developed vitreous haemorrhage that required exclusion from the model at 12 weeks. [...]we do not believe that we can conclude that pharmacological vascular endothelial growth factor (VEGF) blockade did not prevent severe visual loss on the basis of these two episodes of vitreous haemorrhage.