In this retrospective cohort study, Nishida and colleagues evaluated the association between the rate of macular vessel density (VD) loss during an initial follow-up and the rate of visual field (VF) loss during an extended follow-up period in 38 eyes suspected of having glaucoma and 86 eyes with primary open-angle glaucoma. The rate of VD loss was derived from macular whole-image (3 × 3 mm2) vessel density values from 3 optical coherence tomography angiography (OCTA) scans early during the study. The rate of VF loss was calculated from the visual field mean deviation (MD) during the entire follow-up period (mean, 4.0 years) after the first OCTA visit. The authors found that faster VD loss during the initial follow-up period was associated with faster concurrent and subsequent rates of VF loss during the extended period.One may speculate that macular VD loss is a result of macular ganglion cell complex (GCC) thinning. However, the data on the association of VF MD loss rate with VD loss rate and GCC thinning is interesting, because the association was stronger for the OCTA model than for the OCT model.There are some limitations to this study: (1) the 3x3 mm2 scan area seems to be too small to represent the global index (VF MD) loss; (2) the results (from a relatively small sample size and short follow-up period) need to be confirmed by future longitudinal follow-up studies.
The current paper raises the promising possibility that fast progressors identified by OCTA are at higher risk of functional loss and may need more intensive observation and treatment
However, the current paper raises the promising possibility that fast progressors identified by OCTA are at higher risk of functional loss and may need more intensive observation and treatment. A future prospective follow-up study could further clarify the utility of OCTA for assessment of VF progression in patients with glaucoma. In addition, even though it was not in the scope of this study, there is a possibility that training of OCTA images by deep learning will further enhance the effectiveness of VD loss as a predictor of future VF loss.