We thank Sengupta and associates for the interest shown in our recent article, and we appreciate their comments.

Our paper showed a high sensitivity and specificity in detecting polypoidal choroidal vasculopathy (PCV) based on at least 3 criteria from among the following: (1) sharp pigment epithelium detachment (PED) peak; (2) PED notch; (3) visible hyporeflective lumen within hyperreflective lesions adherent to the outer surface of the retinal pigment epithelium; and (4) multiple PEDs seen using spectral-domain optical coherence tomography (SD OCT). Following SD OCT analysis, we observed that hard exudates were also present in a high percentage of our cases (86.5%).

Using these criteria to analyze our SD OCT scans, we failed to detect 2 PCV lesions that were then identified by indocyanine green angiography (ICGA). Those 2 cases both presented with a sharp PED peak and a tomographic notch, but no hyporeflective lumen was identified under the PED while hard exudates were present in both cases.

Moreover in the first case, even though the patient was new to our department, he had undergone ranibizumab injections in the past, which probably altered the tomographic appearance.

The case we failed to detect in the choroidal neovascularization (CNV) group using only SD OCT criteria presented with multiple PEDs only; hence we did not classify it as PCV, but believed it was instead a fibrovascular PED.

The 95% confidence intervals for sensitivity and specificity were 94.6% (87.3, 100.0%) and 92.9% (79.5, 100.0%), respectively.

None of our patients presented with coexisting classic CNV and PCV, but they all presented with either occult CNV or PCV. Also, in our study population, we excluded all the cases with classic CNV.

There is an open debate on whether PCV represents a subtype of occult exudative CNV in age-related macular degeneration or a completely different entity. Variations in PCV and CNV features could also be a consequence of different ethnicities. Furthermore, although a classification of PCV has recently been proposed in the Asian population, analyzing in detail the different presentations of these entities was outside the purpose of our study.

Finally, even though we used the Spectralis OCT (Heidelberg, Germany) in our study, which showed a very high sensitivity and specificity for PCV, we believe that using the same tomographic signs as we detected in our series will also allow a high index of suspicion of PCV to users of other commercially available SD OCT machines.