Editors Selection IGR 20-2

Clinical Examination Methods: Rim Loss Quantification by Artificial Intelligence

Kouros Nouri-Mahdavi

Comment by Kouros Nouri-Mahdavi on:

79328 Macula Vessel Density and Thickness in Early Primary Open-Angle Glaucoma, Hou H; Hou H; Hou H; Moghimi S; Zangwill LM et al., American Journal of Ophthalmology, 2019; 199: 120-132

Find related abstracts

Hou et al. compared the performance of central macular OCT thickness measures to those of OCTA (vascular density) for detecting preperimetric glaucoma or early perimetric glaucoma (MD > -6 dB). Whether macular OCTA parameters can identify eyes with early glaucoma more effectively than macular thickness measures remains a controversial topic and data are scant especially in eyes with early glaucoma. In this manuscript, preperimetric glaucoma was defined as eyes with glaucomatous appearing optic nerve, in which the achromatic visual field still remained normal.

The investigators used both absolute measured parameters and percent change from average normal values to fully explore potential differences in performance of the two structural measures. Overall the findings confirm some of prior published research that OCTA measures do not necessarily perform better for detection of early glaucoma. One interesting finding was that in very early glaucoma, i.e., in the preperimetric stage, a reduction in both OCT thickness and vascular density was observed in the central macula; however, thickness measurements continued to evolve and further thin out as glaucoma advanced whereas the reduction in vascular density slowed down as glaucoma progressed.

It is promising that the structure-structure correlations were reasonably strong in the early perimetric group (R2 = 31-32%). In the only example shown for a preperimetric glaucoma, the areas of decreased vascular density and reduced thickness were not very consistent. It would have been interesting to see how much central functional damage if anything would be observed on the 10-2 VFs. Glaucoma was defined based on disc and 24-2 VFs. As evidence of central loss can be variable in early glaucoma, this was likely a heterogeneous group of patients with variable levels of central damage and this might have led to an underperformance of both macular structural measures.

Overall, the findings add to the currently available evidence that measuring OCTA derived vascular density is a potentially useful measure for detection of early glaucoma although by itself, it may not be superior to GCC thickness measures. However, the potential additive value of OCTA measures beyond the GCC parameters is yet to be determined. Also, there is still significant variations in the imaging algorithms and post-acquisition processing of the OCTA images and therefore, the findings may not necessarily be generalizable to other OCTA devices at this point.

Issue 20-2

Select Issue