Reply

To the Editor:

We thank Gioia et al¹ for their comments on our article, titled “Changes in subfoveal choroidal thickness after intravitreal dexamethasone implant therapy for diabetic macular edema,” published in Retina.² We are pleased with their interest in our study.

They have suggested that the subfoveal choroidal thickness (SFCT) measurement in our study might have been inaccurate because the measurement line was perpendicular to the choroidal–scleral interface and not to the retinal pigment epithelium. In addition, they claimed that the slice of the imaged choroidal tissue for SFCT measurement during follow-up could be different in patients with subretinal fluid and that it could have confounding effects on the results.

Regarding the SFCT measurement technique, we are afraid that the description of our method for measuring SFCT in the article was confusing. Indeed, we originally intended to state that SFCT measurement was performed perpendicular to the retinal pigment epithelium—going vertically from the outer border of the hyperreflective line of the retinal pigment epithelium to the choroidal–scleral junction in the center of the macula—as adopted in previous studies and suggested by Gioia et al as well.^1,3–5

The second concern raised by Gioia et al could be another limitation of this study. Since the choroid is a highly vascularized structure, and its thickness and detailed morphology keep changing—often influenced by many factors—serial slices of enhanced depth imaging-optical coherence tomography scans during follow-up might have been slightly different. However, to avoid this confounding factor, optical coherence tomography scans with poor choroidal image quality were excluded from the study, and two experienced examiners (M.K.Y. and C.S.Y.), who were blinded to the patients' clinical data, performed independent measurements and carefully selected the horizontal sections passing through the fovea for final analysis. The intraclass correlation coefficient to assess the reliability of the 2 examiners' measurements was 0.99 (95% confidence interval: 0.98–0.99), indicating excellent reliability. The eye tracking feature of the Heidelberg Spectralis system also ensured that sequential scans were obtained of the same location, enabling accurate assessment of changes in choroidal thickness. Therefore, even if a few serial optical coherence tomography scans were not completely consistent, they would have limited influence on choroidal thickness as they were examined for very close positions. Furthermore, the large sample size of this study could compensate for the possibility of such small errors. Moreover, although linear regression analysis showed that each unit change (−0.1 logarithm of the minimum angle of resolution) in best-corrected visual acuity improvement was associated with a 3.91-µm decrease in SFCT, the actual changes in the mean choroidal thickness ranged from 19 µm to 48 µm depending on the subgroup.

Once again, we thank Gioia et al for their important insight into the need for a more accurate assessment of the choroid, which is of fundamental importance in choroidal research. We look forward to future studies that suggest more precise and standardized methods for measuring the choroid.