We thank Drs Chan, Papakostas, and Vavvas for their thoughtful comments. We agree on the importance and need of a qualitative and quantitative analysis of Bruch membrane changes for comparison with choroidal thickness measurements. However, there has never been a validation of whether suggested measures of Bruch membrane disease, such as number, location, length, etc of angioid streaks, would indeed reflect the extent of Bruch membrane alterations. In our experience, there is a large variability in those parameters between patients with otherwise similar disease manifestations. This may be explained by additional factors influencing the development of angioid streaks, including mechanical stress and individual susceptibility. In our study on choroidal changes associated with Bruch membrane pathology in Pseudoxanthoma elasticum (PXE), we considered the similar spatial distribution of choroidal thinning and Bruch membrane calcification to be a good indicator for a relation between these 2 parameters.
The rationale to classify PXE patients into 3 distinct groups is based on the observation that the natural endpoint of the disease is chorioretinal atrophy. Choroidal neovascularizations, which may develop during the disease course, have been reported to be associated with choroidal thinning in other diseases. Overlap between groups is a general difficulty in biological systems, which we attempted to address herein by weighting phenotypic characteristics. Evidence for a choroidal neovascularization as the predominant feature resulted in classification as Group 2, even in the presence of minor choroidal atrophy.
Chan and associates also raised various issues when imaging and quantifying distinct layers of the choroid. With the current commercially available techniques, we consider the delineation of different choroidal layers still to be challenging and experimental and, therefore, did not aim to further subdivide choroidal layers in our study. This may certainly be overcome with the introduction of novel imaging technologies in clinical routine application, such as long-wavelength high-penetration optical coherence tomography (OCT). To illustrate structural changes of the inner choroid, we chose to reconstruct en face views 60 μm below the retinal pigment epithelium. At this depth, choroidal en face views appeared relatively uniform in age-matched healthy controls. As mentioned by Chan and associates, the result is influenced by different degrees of thickness, which is exactly what we intended to show.
There is still an ongoing debate whether to use section scans or volume scans for quantification of choroidal thickness. With the imaging device available in our study, we decided to use the section method for the following reasons:
The sclerochoroidal border is more reliably visualized on single scans using enhanced depth imaging OCT with up to 100 images averaged. The lower rate of averaged images when recording volume scans results in inferior quality of single OCT scans.
We aimed at correlating choroidal thickness with the peripheral spread of Bruch membrane pathology. Demonstration of potential associations appeared superior when using a 1-dimensional (eg, eccentricity) compared to a topographic (eg, area) analysis.