Introduction
Glaucoma is a leading cause of blindness, which affects more than 2 million individuals in the United States and 60 million worldwide.[12] It was a chronic optic neuropathy in combination with characteristic visual field (VF) defects. Open-angle glaucoma (OAG) can be classified into primary OAG (POAG) and normal tension glaucoma (NTG) based on elevated intraocular pressure (IOP). Studies have demonstrated the difference of optic disc damage,[3456] VF progression,[7] as well as ocular and systemic vascular function[8] between POAG and NTG.
Thanks to the development of optical coherence tomography angiography (OCTA), which provides a noninvasive assessment and measurement of optic nerve head and retinal perfusion. It has ability to detect the movement of red blood cells and eliminate other information based on the split spectrum amplitude decorrelation algorithm.[9] In recent years, OCTA studies have reported reduced vessel density (VD) in peripapillary retina and optic disc among glaucomatous eyes[1011] and its correlation with VF defect among POAG and NTG patients.[121314]
Several studies have investigated the difference of microvasculature of peripapillary and parafoveal area between POAG and NTG.[1516171819] However, there is no established consensus on the sectoral microvascular change with different subtypes of OAG based on VF examination.[20] Our previous study suggested that early structural changes in radial peripapillary capillaries (RPCs) may help clinicians assess disease severity more accurately, which could inform – but not directly determine – target IOP in NTG.[12] Hence, the purpose of the present study was to investigate whether peripapillary and parafoveal microvascular changes are associated with disease severity in different stages of glaucoma.
Materials and Methods
Study population
The study protocol and a waiver for consent to participation were approved by the Institutional Review Board of Chang Gung Memorial Hospital (Approval No.: 202002149B0) and followed the tenets of the Declaration of Helsinki.
In this retrospective cohort study, a total of 960 days of OCTA imaging data were collected from POAG and NTG patients at Chang Gung Memorial Hospital in Keelung, Taiwan, between March 2018 and November 2020. The inclusion criteria were POAG and NTG. POAG patients required a history of pretreatment elevated IOP (>21 mm Hg) either with or without treatment, typical glaucomatous optic neuropathy (defined as focal or diffuse neuro-retinal rim thinning, focal or diffuse retinal nerve fiber layer (RNFL) loss, or an inter-eye vertical cup-to-disc ratio asymmetry of >0.2 not explained by differences in disc size), glaucomatous VF defects on standard automated perimetry (defined as a cluster of three adjacent points depressed by at least 5 dB from normal age values, with one of these points depressed by at least 10 dB from normal age values), and open angles on gonioscopy.
NTG patients had identical criteria except for a history of untreated peak IOP of 21 mm Hg or less.
The exclusion criteria were as follows: (1) presence of angle-closure or secondary glaucoma, comorbid retinal vascular disease, other cause of optic neuropathies (2) eyes with epiretinal membrane, macular degeneration, significant dry eye, pterygium, or any pathology affecting image quality (3) spherical equivalent error <−6.0 diopters (D) or >+3.0 D (4) axial length more than 26.5 mm (5) unreliable VF test results (i.e., fixation errors >20%, false positives >15%, and false negatives >33%), and (6) OCTA signal strength index (SSI) <50%.
For subjects with two eligible eyes of POAG or NTG, both eyes were included in the analysis. The database contained complete ophthalmic examination, including best-corrected visual acuity, central corneal thickness (CCT), slit-lamp biomicroscopy, gonioscopy, and iCare® PRO rebound tonometry (iCare, Tiolat Oy, Helsinki, Finland), standard automated perimetry (Swedish interactive threshold algorithm standard 30-2 test; Humphrey field analyzer II; Carl Zeiss Meditec, Dublin, CA, USA), and OCTA measurements (Optovue Inc., Fremont, CA, USA). Only patients with standard automated perimetry performed within 6 months of imaging were included.
The severity of glaucomatous damage was classified as early stage (MD≥−6 dB), moderate stage (−12 dB ≤MD <−6 dB), or severe stage (MD <−12 dB) based on Hodapp-Parrish-Anderson criteria.[21]
Baseline ophthalmic evaluations included color fundus photography and standard automated perimetry. For representative illustration, Figure 1a shows a typical case of primary open-angle glaucoma (POAG), with a color fundus photograph demonstrating superotemporal and inferotemporal rim thinning, while Figure 1b presents the corresponding 30-2 pattern deviation plot showing dense superior and inferior arcuate scotomas. Likewise, Figure 2a depicts a representative case of normal-tension glaucoma (NTG), showing a vertically enlarged cup with inferotemporal rim thinning on color fundus photography, and Figure 2b shows the associated 30-2 pattern deviation plot demonstrating superior and inferior arcuate defects.
Optical coherence tomography angiography measurement
OCTA scans were acquired using the RTVue-XR Avanti (Optovue Inc., Fremont, CA, USA). An 840 nm wavelength light source emitted from the Avanti system scans ocular microstructures and can provide the data on the RNFL thickness and ganglion cell complex (GCC) thickness at a diameter of 3.45 mm around the center of the optic disc [Figures 1c, g and 2c, g]. 3.0 mm × 3.0 mm OCTA scans centered on the fovea and 4.5 mm × 4.5 mm OCTA scans centered on the optic nerve head were obtained.
The global RPCs was measured in the entire 4.5 mm × 4.5 mm image. RNFL thickness and RPCs at each Garway-Heath sector were measured in the region defined as a 750-μm–wide elliptical annulus extending from the optic disc boundary [Figure 1d and 2d].
Images of the fovea microvasculature were produced using the Optovue OCTA device. The globally superficial and deep layers of parafoveal VD were measured in the entire 3.0 mm × 3.0 mm image. Sectorally, superficial and deep layers of parafoveal VD and retinal thickness were also observed.
The boundary of the superficial retinal layer extends from the internal limiting membrane (ILM) to the inner plexiform layer; the boundary of the deep retinal layer extends from inner to outer plexiform layers [Figures 1e and f and 2e and f].
Statistical analysis
Statistical analyses were performed using the commercial software (SPSS, ver. 20.0; IBM Corp., Armonk, NY, USA). Data were expressed as mean ± standard deviation for the continuous variables and percentage for the categorical variables.
Categorical variables were compared using the Chi-square test, whereas continuous variables were compared using an independent t-test. P < 0.05 was considered statistically significant.
Results
The charts of 515 eyes of glaucoma imaged with OCTA were reviewed. After excluding eyes based on the criteria mentioned above, a total of 312 eyes from 198 patients (154 POAG from 100 patients and 158 NTG from 98 patients) were included. A total of 117 eyes of POAG and 86 eyes of NTG were excluded.
Among 117 eyes excluded, 42 were excluded due to poor image quality (SSI <50), with similar exclusion rates in POAG and NTG groups.
Among 312 eyes of glaucoma, there were 154 and 158 eyes in the POAG and NTG group [Table 1]. The mean age was 62.2 ± 10.3 years in the NTG group and 51.6 ± 16.5 years in the POAG group (P < 0.001). Axial length was similar between two groups (P = 0.417).
CCT was 557.8 ± 34.0 μm in the POAG group and 538.3 ± 34.0 μm in the NTG group (P < 0.001).
Without stage classification, there were no significant differences between the POAG and NTG group among global RPCs, RNFL thickness, parafoveal VD, parafoveal deep retinal thickness, and GCC thickness [Table 2]. According to the Garway-Heath map, RPCs in nasal-inferior sector, RNFL thickness in temporal-superior and nasal-inferior sectors were found lower in the POAG group compared to NTG group significantly (42.3% ± 7.5% vs. 43.1% ± 6.0%, P = 0.006; 71.7% ± 16.5% vs.
76.2% ± 37.9%, P = 0.035; 76.0% ± 18.4% vs. 78.9% ± 17.7%, P = 0.007, respectively).
For patient with early VF defect, global RNFL thickness was significantly lower in the NTG group than POAG group (87.5 ± 11.3 μm vs. 92.4 ± 10.5 μm, P = 0.034) [Table 3]. RNFL thickness in superior-nasal sector was also lower in the NTG group compared to POAG group significantly (111.7 ± 28.4 μm vs.
122.7 ± 26.8 μm, P = 0.030).
For patient with moderate VF defect, there was no significant difference among all global and sectoral microvasculature parameters between the POAG and NTG group [Table 4].
For patient with severe VF defect, there were no significant difference between the POAG and NTG group among global RPCs, RNFL thickness, parafoveal VD, parafoveal deep retinal thickness, as well as GCC thickness. RPCs in nasal-superior and nasal-inferior sectors, as well as RNFL thickness in temporal-superior sector were significantly lower in the POAG group compared to NTG group (32.1% ± 9.8% vs. 39.2% ± 8.8%, P = 0.044; 32.9% ± 10.6% vs.
40.0% ± 6.9%, P = 0.012; 53.0 ± 15.0 μm vs. 69.9 ± 21.7 μm, P = 0.019, respectively) [Table 5].
Discussion
OCTA has become popular in recent years owing to its benefits in the detection of microvascular changes in glaucomatous eyes. Previous studies have investigated the difference of microvasculature of peripapillary and parafoveal area between POAG and NTG.[1516171819] The current study analyzed that global and superior-nasal sector of RNFL thickness was lower in NTG eyes than POAG eyes in the early stage of glaucoma. RPCs in nasal-superior and nasal-inferior sectors, as well as RNFL thickness in temporal-superior sector were found significantly lower in the POAG group compared to NTG group in severe stage of glaucoma.
Clinically, this sectoral differentiation may be useful in distinguishing early NTG from POAG, allowing more tailored surveillance and treatment strategies according to disease subtype and severity. To the best of our knowledge, this is the first study using OCTA to explore the association of sectoral peripapillary and parafoveal microvasculature between POAG and NTG based on different severity of glaucoma.
Many studies have shown that NTG patients are more prone to vascular dysregulation, which may be associated with increased prevalence of migraine and optic disc hemorrhage.[22232425] In fact, it is still controversial regarding to the change of RPCs and optic disc perfusion between normal patients and both POAG and NTG.[161726] Usually, RPCs declines as glaucoma progressed. Our result showed that lower RPCs in nasal-superior and nasal-inferior sectors in POAG than NTG patients was only found in severe stage, which is not totally consistent to previous studies.[16] Scripsema’s et al. reported lower RPCs in POAG than NTG patients in all different stage of glaucoma based on the Hodapp-Anderson-Parrish VF Severity Score.[16] Large percentage of POAG patients with medication such as timolol and difference in the underlying pathophysiology between NTG and POAG may provide an explanation.[2728] However, Onishi et al. found that only POAG eyes instead of NTG eyes showed significantly decreased RPCs compared to healthy eyes.[17] Furthermore, Bojikian et al. observed that there was no difference of optic disc perfusion between POAG and NTG groups with similar levels of VF MD damage using optical microangiography-based OCTA technique.[26] From a mechanistic standpoint, NTG is thought to be driven more by systemic vascular dysregulation, while POAG damage is more pressure-driven. This could explain why NTG eyes may show more diffuse RNFL thinning, whereas POAG eyes present with localized sectoral microvascular loss.
Clinically, these differences suggest that vascular protection may be especially important in NTG, while aggressive IOP lowering remains central for POAG.
As for microvascular change among OAG cases, current studies have not been universally consistent.[1517] In our analysis, there was no significant difference of globally and sectorally parafoveal VD between POAG and NTG in all glaucoma stage. However, Onishi et al. observed quantitatively lower parafoveal and peripapillary VD in POAG eyes compared to NTG.[17] In contrast, Xu et al. found increased parafoveal full retinal thickness VD in POAG eyes within a Chinese population.[15] Such discrepancies may reflect differences in ethnic backgrounds, disease stage distribution, or systemic vascular comorbidities. This highlights the need for personalized interpretation of OCTA metrics in clinical practice.
In general, these controversy may be related to the multi-factorial etiology of NTG or differences of severity between study populations, which needed to be clarified in future.
Our study has some limitations, including retrospective design, cross-sectional analysis, and single ethnicity of included patients, which might influence the results. In addition, in order to obtain the images with adequate signal strength, patients must have good fixation and relatively good central acuity. However, the acquisition time required for the OCTA imaging may influenced by advanced glaucoma or poor acuity due to other conditions, such as cataracts or macular lesions.
Furthermore, the effect of glaucoma eye drops on microvascular change may offers a potential confounding variable that we were unable to control. The relatively small sample size in moderate and severe subgroups also limits generalizability. We acknowledge several methodological limitations of this study.
First, the inclusion of both eyes from some individuals may introduce inter-eye correlation, which is not fully addressed by t-tests. Second, the significant age difference between diagnostic groups may act as a confounder, particularly for vascular and structural OCTA parameters. Third, we did not stratify patients by medication use, which may influence ocular perfusion measurements.
These factors should be addressed in future studies using age-matched cohorts and statistical models that account for repeated measures and pharmacological effects. Clinically, awareness of these limitations is crucial when interpreting OCTA findings; physicians should be cautious about overgeneralizing results and should integrate OCTA data with structural OCT and VF progression in decision-making. In general, further long-term and large-scale prospective studies of OCTA in different stages of glaucoma are needed in the future.
Andrade De Jesus et al. reported a combination OCTA modeling of multilayer and multisector microvascular, which seems to provide a higher discriminative power than the RNFL thickness to discriminate between severity levels of glaucoma.[20] The observed differences in sectoral microvasculature may reflect underlying differences in pathophysiology. For instance, POAG may involve more pressure-dependent axonal damage, while NTG may reflect vascular dysregulation with different regional susceptibility. These mechanistic insights emphasize that OCTA not only holds diagnostic value but also has potential therapeutic implications: Strategies targeting vascular dysregulation could be prioritized for NTG, while pressure-lowering therapies remain the cornerstone for POAG.
In summary, current study provided evidence that sectoral peripapillary and parafoveal microvasculature significantly differs between POAG and NTG with different stages. In addition to RNFL thickness, RPC may be a useful indicator of glaucoma progression.
Data availability statement
The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Conflicts of interest
Dr. Chi-Chun Lai, an editorial board member at Taiwan Journal of Ophthalmology, had no role in the peer review process of or decision to publish this article. All authors declared no conflicts of interest in writing this paper.