Macular vessel density as a novel biomarker for glaucoma diagnosis in highly myopic eyes: a multicenter study
Qin Chen1, Jing Liu2, Qingwei Meng3
1Department of Ophthalmology, The First Affiliated Hospital with Nanjing Medical University, Nanjing, 210000, China. chenqin@jsph.org.cn.
Background:
To evaluate macular microvascular (superficial vessel density, SVD) and structural (ganglion cell complex thickness, GCCT) parameters in highly myopic glaucoma (HM-G) and non-highly myopic glaucoma (NHM-G), and to observe the correlations between these parameters and visual field mean sensitivity (VFMS), and to evaluate the diagnostic capabilities of macular vascular and structural parameters in myopic patients with glaucoma.
Methods:
In this multicenter study, 127 HM-G and 115 NHM-G eyes were matched to high/non-high myopia controls by 1:1 propensity score matching (PSM) based on age and axial length. Macular SVD and GCCT were quantified using optical coherence tomography angiography (OCTA). Spearman correlation and linear regression analyses were used to assess structure-function relationship. Receiver operating characteristic (ROC) curve analyses were carried out to evaluate the diagnostic performance, with Bonferroni correction for multiple comparisons.
Results:
Both HM-G and NHM-G groups exhibited significantly lower SVD, GCCT, and VFMS compared to their respective controls (p < 0.001). In stage-specific analysis, SVD showed a significant correlation with VFMS in the early stage of HM-G (average: r = 0.359, p = 0.008), whereas GCCT did not exhibit a correlation with VFMS (p = 0.056) until the moderate to late stage. SVD had stronger regression coefficients than GCCT for VFMS (HM-G: 47.7 vs. 24.9; NHM-G: 47.5 vs. 20.7), with combined models enhancing explanatory power. ROC curve analyses indicated that SVD, GCCT were sensitive indicators for diagnosing myopia with glaucoma (p < 0.001). The regions with the best diagnostic ability for SVD and GCCT were concentrated in the inferior perifoveal (peri-I) region. SVD-peri-I showed superior diagnostic performance in HM-G (AUC = 0.924 vs. 0.837/0.873 for peripapillary metrics; p < 0.01) comparing with peripapillary vessel density in superior temporal or inferior temporal region (pVD-ST/IT).
Conclusion:
Macular SVD in the inferior perifoveal sector demonstrated superior diagnostic performance to peripapillary vessel density in highly myopic glaucoma, while GCCT showed comparable global discrimination. Combined OCTA and OCT imaging significantly enhance diagnostic accuracy and functional correlation, offering critical clinical utility for glaucoma diagnosis and monitoring in highly myopic patients.
