Our study found that the sector macular SVD in PACG was topographically related to corresponding GCIPL thickness only in the IN sector. In addition, pRNFL and macular SVD also showed significant correlations. The FAZ circularity index showed the highest diagnostic power.
In the current study, the macular SVD of PACG eyes was significantly lower than that of healthy eyes, which were consistent with the findings reported in earlier studies 11, 12. Li et al.11and Liu et al. 12described decrease of the macular SVD in 6 ×6 mm circular area. But they didn't investigate whether there were differences between glaucomatous eyes and healthy eyes in the sectoral division. On the contrary, we found that each sector of macular SVD in PACG eyes decreased, with the exception of SN sector, than that in the control group. The greatest reduction of vessel density in glaucomatous eyes was the macular II sector. This may be due to structural reasons. The studies of glaucoma using OCT proved that the inferior macula was more easily damaged in glaucoma 13, 27.
The macular GCIPL thickness had an important correlation with the macular SVD only in the IN sector in present study which was similar to that found in POAG18. Richter et al.18's study of POAG showed that the macular GCIPL thickness and the macular vessel density were significantly correlated only II sector. Kim et al19's research on glaucoma- suspect and early NTG reported that macular GCIPL thickness and macular SVD had important correlation in the ST, IT and II regions. This difference may be explained by the different pathogenesis of different types of glaucoma28. The macular GCIPL injuries during different IOP elevation may be different. This glaucomatous injury of PACG and POAG is more closely related to IOP which is different from NTG. Depending on our current results, IN and II sectors of macular SVD showed strong topographic correlations with inferior clock-hour regions of pRNFL thickness in PACG. This may be related to the fact that most of the inferior region of the macula projects to the inferior quadrant of the disc, a region that is particularly susceptible to glaucomatous damage 13.
The FAZ is highly sensitive to ischemia. Atrophic changes of macular capillaries might first affect the shape and size of FAZ, which had been proved to have significant clinical application value in retinal vein occlusion and diabetic retinopathy15, 16, 29. However, few studies investigated the clinical value of FAZ in glaucoma, especially PACG. Liu et al.12reported that the FAZ circularity index showed high diagnostic power for detecting acute primary angle closure eyes in PACG patients after acute primary angle closure episodes. This was similar to our result. However, previous studies did not compare the diagnostic value of macular vessel density and FAZ parameters with traditional OCT measurement of pRNFL and macular GCIPL thicknesses metrics. Our study demonstrated that the diagnostic abilities of macular vessel densities were not better than the traditional pRNFL and macular GCIPL thickness measurements in PACG. However, FAZ circularity index showed the highest diagnostic power.
The advantage of the present study was that we used imageJ software to calculate the vessel density values that matched the macular GCIPL regions of each partition. In this way, each vessel density region could be accurately matched with the macular GCIPL region for correlation analysis. When to calculate the vessel density, we removed the large vessels to reduce the influence on the measurement of vessel density, because the large vessels did not participate in microcirculation perfusion. Moreover, large vessels would produce low vessel density areas around the large vessels, which might lead to false low vessel density values. Our algorithm had no such error, so it could accurately reflect the perfusion state.
The limitation was that we couldn't remove the projection artifacts of superficial vessels in deep vessels, so we couldn't investigate the correlation between deep macular vessel density and macular GCIPL thickness. Another possible limitation of current research was that most of our patients were patients with severe glaucoma. We couldn't divide into groups. Future research should evaluate the diagnostic ability of FAZ parameters for mild to moderate glaucoma.
Our research showed that OCTA might provide useful information for glaucoma. The important correlation between pRNFL, macular GCIPL thicknesses and macular SVD in topographic features could provide a basis for clinicians to comprehensively explain glaucoma damage. A better understanding of the correlation between macular SVD parameters and macular GCIPL, pRNFL thicknesses might broaden the application range of macular SVD. Perhaps these vascular parameters will gradually become important in the etiology, diagnosis, progression and treatment choice of glaucoma patients in the future. As a cross-sectional study, we couldn't assess whether vessel changed occur before structural changed in glaucomatous development. Future research should longitudinally evaluate the dynamic relationship between vessel changes and structural changes.
In conclusion, sector of macular SVD not only had a spatial positional correlation with corresponding macular GCIPL thickness, but also with clock-hour regional pRNFL thicknesses in PACG eyes. The SVD of macular area in glaucoma patients was sparser than that of control eyes. The FAZ circulation index might be a useful diagnostic parameter.