Dear Editor,

We thank Dr. Uzun et al. for their interest in our paper. Most of their concerns could perhaps be explained in terms of inherent limitation of retrospective study. We did not analyze axial length in our patients due to limited data availability. However, we found no significant difference in refractive error between good-responders and poor-responders (+0.41 ± 1.55 diopter vs. -0.25 ± 1.24 diopter, P = 0.157). Since axial length and refractive error are highly correlated [1, 2], we may safely assume that there would be no significant difference in axial length between the two groups. Therefore, we do not think that a lack of axial length analysis should affect our finding.

We evaluated the baseline choroidal thickness as the predictor of treatment response. Thus, the possibly raised intraocular pressure from intravitreal injections and their effects on choroidal thickness should not be a relevant issue in our study. There was no difference in baseline intraocular pressure between the two groups (data not shown).

The diurnal variation and other local or systemic conditions that may affect the choroidal thickness and the use of two different drugs (ranibizumab and bevacizumab) are a possible bias in our retrospective study, as discussed in the manuscript. The subgroup analysis of 47 patients who received ranibizumab only confirmed the significantly thinner choroidal thickness in good responders (225 ± 94 μm vs. 284 ± 125 μm, P = 0.037). Further prospective studies that include analyses suggested by Uzun et al. are warranted to validate our findings.