We thank the Journal editor for forwarding the interesting points raised by Takkar and associates regarding our paper “Optical Coherence Tomography Angiography of Shallow Irregular Pigment Epithelial Detachments in Pachychoroid Spectrum Disease.”
In this study we sought to determine the diagnostic value of shallow irregular pigment epithelial detachments (PED) seen on cross-sectional structural optical coherence tomography (OCT) of eyes with type 1 neovascularization. We used OCT angiography (OCTA) and dye angiography to confirm that the proportion of these PEDs harboring neovascularization is greater than previously estimated with dye angiography alone, but the time disparity between these imaging modalities prevented either direct comparison between them or a sensitivity analysis of OCTA. Our principal conclusion was that a large proportion of shallow irregular PEDs contain neovascularization. We did not state that OCTA is more sensitive than dye angiography. To draw such a conclusion would have required a more formal analysis with masked graders and contemporaneous imaging, as we discussed.
Takkar and associates have suggested that exudation from implicitly “active” type 1 neovascularization may increase extravascular choroidal volume and thickness, pushing nonpachychoroid patients into our inclusion criteria. However, we believe it unlikely that nonpachychoroid patients were included or that such a mechanism would alter our conclusions because (1) 19 out of 22 study eyes had received antiangiogenic treatment (mean 31.3 injections), (2) 6 eyes had received photodynamic therapy (mean 2.6 treatments), and (3) we excluded eyes with alternative risk factors for neovascularization.
We have also revisited the raw data and determined that only 6 of 22 eyes were included on the basis of choroidal thickness >270 μm without corroborating pachyvessels (mean 418 μm, range 323–498 μm). The thinnest of these choroids (323 μm) would have to have been thickened by exudate to the order of 20% to have been included inappropriately, as suggested. We could not find data in the literature to support the idea that “active,” exudative type 1 neovascularization confounds choroidal pachymetry to this extent.
Takkar and associates have suggested that “the presence of new vessels themselves … would have … increased choroidal thickness.” However, we adhered to the anatomic classification of type 1 neovascularization. Choroidal thickness was measured conventionally, outward from the Bruch membrane, and PED elevation was not included within the calipers. To consider vessels behind the Bruch membrane as “new” would constitute a significant challenge to current classification, beyond the scope of this discussion.
Recently in this journal, Spaide and Ryan described the occurrence of posterior loculated choroidal fluid in patients with central serous chorioretinopathy, and Takkar and associates have questioned whether posterior loculated fluid might have confounded the choroidal thickness measurements in our study or affected the sensitivity of OCT angiography. Notably, Spaide and Ryan excluded eyes with choroidal neovascularization, although the method for exclusion was not reported. Interestingly, in 1 of the cases they showed, a shallow irregular pigment epithelial detachment was present.
Spaide and Ryan calculated that loculated fluid can account for a difference in choroidal thickness of approximately 68 μm, which, if applied as a correction to our data, would reduce our thinnest non -pachyvessel choroid from 323 μm to 255 μm. Pachychoroid would remain the most likely diagnosis.
Finally, posterior loculated fluid is unlikely to influence the sensitivity of OCT angiography in the setting of our study, as all lesions in this context resided anterior to the Bruch membrane. Conversely, neovascularization, if present, could arguably influence thickness measurements relating to loculated fluid.