Predicting Visual Field Progression by Optical Coherence Tomography Angiography and Pattern Electroretinography in Glaucoma : Journal of Glaucoma

Secondary Logo

Journal Logo

New Understandings of Glaucoma: Original Studies

Predicting Visual Field Progression by Optical Coherence Tomography Angiography and Pattern Electroretinography in Glaucoma

Lee, Mee Yon MD, PhD*,†; Park, Hae-Young Lopilly MD, PhD†,‡; Kim, Seong Ah MD†,‡; Jung, Younhea MD, PhD†,§; Park, Chan Kee MD, PhD†,‡

Author Information
doi: 10.1097/IJG.0000000000002088
  • Free

Abstract

BACKGROUND

Glaucoma is an optic neuropathy characterized by the increased rate of retinal ganglion cells (RGCs) degeneration, which results in progressive loss of visual function. Before the irreversible death of RGCs that can be detected by structural measurements, such as the retinal nerve fiber layer (RNFL) thickness from optical coherence tomography (OCT) measurements, RGCs may exhibit a dysfunctional state that can be detected by functional measurements. However, structural changes in the optic disc or RNFL are detected before the appearance of visual field (VF) damage, as measured by standard automated perimetry (SAP). One contributing factor is the limited ability of SAP to detect early functional changes, specifically, SAP cannot detect VF abnormalities until 25%−35% of the RGCs die.1 In addition, SAP depends on the patient’s subjective response and may not fully represent objective RGC function. Most patients are unfamiliar with SAP at the time of diagnosis, which likely influences the reliability and reproducibility of VF tests. Therefore, a method is needed to objectively measure visual function in glaucoma patients, and to estimate the proportion of dysfunctional RGCs that may impact future glaucoma progression.

RGC function can be evaluated by other means of functional testing, including pattern electroretinography (PERG). PERG is a unique type of electroretinography that uses a stimulus of contrast-reversal gratings, rather than uniform flashes of light, to evaluate RGC function.2 RGC death and/or RGC dysfunction can alter the PERG waveform.3 An important characteristic of PERG is that it requires physiological integrity among viable RGCs, because a reduced PERG amplitude may reflect the effects of both RGC death and viable RGC dysfunction.4,5 Therefore, baseline PERG parameters may better represent the proportion of dysfunctional RGCs, compared with SAP or OCT.6 In addition, dysfunctional RGCs may have a lower oxygen demand, resulting in reduced blood flow.7 This regulation, known as neurovascular coupling, indicates alterations in local perfusion in response to neuronal activity.8,9 Therefore, vessel imaging of the retina and the optic disc using OCT angiography (OCT-A) may also show the status of RGCs.

In the present study, we aimed to identify the baseline demographics and ocular characteristics that predict future progression in glaucoma patients, including the associated PERG and OCT-A parameters, to establish the clinical significance of detecting dysfunctional RGCs in glaucoma patients in terms of PERG and OCT-A.

METHODS

Participants

This study was a component of the Catholic Medical Center Glaucoma Progression Study (CMC-GPS), which is a prospective study commenced in 2009 at Seoul St. Mary’s Hospital, Seoul, South Korea. The work was approved by the Institutional Review Board of Seoul St. Mary’s Hospital. All relevant tenets of the Declaration of Helsinki were followed. We enrolled all consecutive eligible patients who were willing to participate, and all provided written informed consent. Among the enrolled patients, patients who underwent both OCT-A and PERG and had at least 5 subsequent VF testing were included in the present study.

All open angle glaucoma (OAG) patients enrolled in the CMC-GPS underwent a complete ophthalmic examination, including a review of their medical history, measurement of best-corrected visual acuity, refraction assessment, slit-lamp biomicroscopy, gonioscopy, Goldmann applanation tonometry, measurement of central corneal thickness through ultrasound pachymetry (Tomey Corp., Nagoya, Japan), measurement of axial length with ocular biometry (IOL Master; Carl Zeiss Meditec, Dublin, CA), dilated stereoscopic examination of the optic disc, red-free fundus photography (Canon, Tokyo, Japan), Cirrus OCT examination (Carl Zeiss Meditec), and Humphrey VF examination using the Swedish interactive threshold Standard 24-2 algorithm (Carl Zeiss Meditec). Beginning in 2017, patients underwent additional PERG (Neuro-ERG, Neurosoft, Ivanovo, Russia) and OCT-A (DRI OCT Triton; Topcon, Tokyo, Japan) examinations. All patients were followed up at intervals of 1–3 months with color disc and fundus photography. VF and OCT examinations were performed at 6-month intervals. All disc hemorrhages (DHs) detected on color disc and fundus photography during follow-up were recorded. A DH was defined as an isolated flame-shaped or splinter-like hemorrhage on the optic disc or in the parapapillary area, extending to the disc border. Alternative causes of hemorrhage (eg, ischemic optic neuropathy, papillitis, retinal vein occlusion, diabetic retinopathy, and posterior vitreous detachment) were diagnostically excluded. Intraocular pressure (IOP) was recorded at each visit. The mean IOP throughout the follow-up period was calculated by averaging all such measurements. Baseline IOP was measurement before initiation of glaucoma medication. Eyes with a glaucomatous VF defect in either or both hemifields within 24 points of a central 10 degrees of fixation, and with no VF abnormality in the nasal periphery outside 10 degrees of fixation, were considered to have isolated central scotoma. The criteria for central scotoma were the presence of three or more points with P<5%, one of which was P<1%, among 12 points on the pattern deviation plot.

OAG was defined by the presence of a glaucomatous optic disc (exhibiting diffuse or localized rim thinning, a notch in the rim, or a vertical cup-to-disc ratio ≥0.2 compared with that of the other eye); a VF finding consistent with glaucoma (a cluster of ≥3 nonedge points on the pattern deviation plot, with a probability of <5% of the normal population, and with one of these points carrying a probability of <1%), a pattern standard deviation (PSD) with a P-value <5%, or a Glaucoma Hemifield Test result consistently outside the normal limits on two VF examinations, as confirmed by two glaucoma specialists (H.Y.P. and C.K.P.); and an open angle evident on gonioscopy. Additional inclusion criteria were the presence of at least 5 reliable VF tests (false negatives <15%, false positives <15%, and fixation losses <20%) with a minimum follow-up period of 3 years; a best-corrected visual acuity ≥20/40; a spherical refraction within ±6.0 diopters (D); a cylinder correction within ±3.0 D; and a mean deviation (MD) better than −30.00 decibels (dB). Exclusion criteria were a history of any retinal disease, including diabetic or hypertensive retinopathy; a history of eye trauma or surgery, with the exception of uncomplicated cataract surgery or any patients who underwent cataract surgery during the study period; any optic nerve disease other than glaucoma; and a history of systemic or neurological disease that might affect the VF. If incisional or laser glaucoma treatment was performed during follow-up, only data obtained before treatment were analyzed. If both eyes of an enrolled patient met all inclusion and exclusion criteria, 1 eye was randomly chosen for inclusion in the study.

Electroretinography

One trained examiner performed the PERG examinations. Two Ag/AgCl ground electrodes were placed on both earlobes, and reference electrodes were placed on the lower eyelid of the ipsilateral side. A detailed description of the examination was explained in our previous paper.10,11 Briefly, patients with proper optical correction and undilated pupils were seated in front of a display. A checkerboard pattern with a mean luminance of 105 cd/m2 was reversed at a rate of 4 reversals per second at a distance of 60 cm. The stimulus display covered 48×33 degrees of the VF with each check size of 1.8-degree visual angle. Patients focused on the red fixation point at the center of the display, and both eyes were examined simultaneously. The reproducibility of the ERG results was identified in previous studies by the intraclass correlation coefficients of randomly selected measurements.

OCT-A

The optic nerve head and the parapapillary region were imaged using a commercial, swept-source OCT-A device (DRI OCT Triton; Topcon). The central wavelength was 1050 nm, the acquisition speed was 100,000 A-scans/s, and the axial and transversal resolutions were 7 and 20 μm, respectively. Cubes 4.5 ×4.5 mm in size were scanned; each cube consisted of 320 clusters of four repeated B-scans centered on the macula and optic disc.

The automatically segmented retinal vascular plexus consisted of a superficial layer and a deep layer. The macular vessel density (VD) in this study was regarded as the VD of the superficial layer, which was calculated automatically after applying the built-in software projection removal algorithm for deep retinal and choriocapillary layers. The superficial vascular plexus extended from 2.6 μm below the internal limiting membrane to 15.6 μm below the junction of the inner plexiform layer and inner nuclear layer, according to the default settings of the device. The deep-layer parapapillary microvasculature in the optic disc region was evaluated on en-face images generated through automated layer segmentation of signals from the retinal pigment epithelium that extended to the outer border of the sclera. Microvasculature dropout (MvD) was defined as focal, sectoral capillary dropout within the visible microvascular network. MvD was identified when the dropout width was more than 2-fold greater than the width of the visible juxtapapillary microvessels. Two independent observers (H.Y.P. and S.A.K.), who were masked to clinical data, independently identified all MvDs. Disagreements were resolved by a third author (C.K.P.). Only clear images with quality scores >70 that did not exhibit blurring attributable to motion were analyzed.

Determination of Progression

The parameters of VF progression were determined using linear regression analysis of the MD, PSD, and visual field index (VFI) values from the five SAP tests of the Humphrey VF examination. The MD, PSD, and VFI slopes were compared between groups. VF progression was determined by event-based analysis using the Glaucoma Progression Analysis (GPA) software from the Humphrey Field Analyzer. For each individual VF test location, the GPA compared the sensitivity on follow-up testing with the sensitivity for the same location obtained from averaging 2 baseline tests. VF progression was defined as a significant decrease from the baseline (2 VFs) pattern deviation at 3 or more of the same test points on 2 or 3 consecutive VF tests. The software classified VF progression as “possible progression” or “likely progression.” Only the “likely progression” condition was considered to have glaucoma progression.

Statistical Analysis

We used Student t test to compare continuous variables, and the χ2 test to compare categorical variables. The к coefficient was calculated to evaluate interobserver agreement in determining the presence of MvD. Univariate and multivariate logistic regression analyses were used to identify factors associated with glaucoma progression. The dependent variable was “likely progression” of GPA results or MD slope from SAP tests. The independent variables were age, sex, medication history, best-corrected visual acuity, axial length, central corneal thickness, baseline and mean follow-up IOPs, baseline peripapillary RNFL and macular ganglion cell/ inner plexiform layer thicknesses, macular VD and the presence of MvD from OCT-A, baseline VF and PERG parameters, the presence of a DH, disc area, vertical cup-to-disc ratio, and follow-up duration. Independent variables yielding P values <0.10 in the univariate model were included in the multivariate model. A P-value <0.05 was considered to indicate statistical significance. All statistical analyses were performed with SPSS software (version 16.0; SPSS Inc., Chicago, IL).

RESULTS

In total, 161 eyes of 161 OAG patients were included in the study who had both baseline OCT-A and PERG and at least 5 subsequent VF tests. Of the 161 eyes, 12 (7.5%) were excluded because the OCT-A images were of poor quality or had motion artifacts. An additional 9 eyes (5.6%) were excluded due to unreliable SAP or PERG results. Ultimately, 140 eyes of 140 OAG patients were included in further analyses. Interobserver agreement in terms of MvD detection was excellent (к=0.892; 95% confidence interval, 0.816–0.975; P<0.001).

The baseline patient characteristics are listed in Table 1. Of the 140 eyes, 107 (76.4%) were eyes with normal tension glaucoma (NTG). The mean patient age was 55.92±10.06 years, and the mean spherical equivalent was −2.86±3.13 D. The mean baseline IOP was 16.93±3.29 mmHg. All patients received glaucoma medication and showed a mean IOP during follow-up of 14.38±2.92 mmHg. Sixty-seven eyes (47.9%) had MvD on OCT-A and 17 eyes (12.1%) had DH during the follow-up period. The mean MD slope was 0.02±1.08 dB/year during the total follow-up period of 3.15±0.96 years.

TABLE 1 - Baseline Demographics and Ocular Characteristics of 140 Eyes of 140 Glaucoma Patients
Variables Description
Demographics
 Age at diagnosis, y 55.92±10.06
 Female, n (%) 88 (62.9)
 Normal tension glaucoma, n (%) 107 (76.4)
Systemic demographics
 Medication of DM, n (%) 9 (6.4)
 Medication of HTN, n (%) 15 (10.7)
Ocular demographics
 Best-corrected visual acuity 0.92±0.16
 Axial length, mm 25.38±3.02
 Central corneal thickness, μm 541.51±36.41
 Spherical equivalent, diopter −2.86±3.13
 Pseudophakic, n (%) 14 (10.0)
Glaucoma medications
 ß-blockers, n (%) 30 (21.4)
 Carbonic anhydrase inhibitors, n (%) 25 (17.9)
 α2-agonist, n (%) 15 (10.7)
 Prostaglandin, n (%) 50 (35.7)
IOP parameters
 Baseline IOP, mmHg 16.93±3.29
 Mean follow-up IOP, mmHg 14.38±2.92
OCT parameters
 Baseline average pRNFL thickness, μm 73.04±12.83
 Baseline average mGC/IPL thickness, μm 70.59±8.93
OCT angiography parameters
 Average superficial macular VD, % 45.72±5.56
 Presence of MvD, n (%) 67 (47.9)
VF parameters
 Baseline MD of SAP, dB −5.83±6.25
 Baseline PSD of SAP, dB 5.54±4.15
 Baseline VFI, % 85.97±18.71
 Presence of isolated central scotoma at baseline, n (%) 51 (36.4)
 MD slope, dB/y 0.02±1.08
 PSD slope, dB/y 0.18±0.81
 VFI slope, %/y −0.68±2.44
Baseline PERG parameters
 N35 latency, ms 22.63±5.30
 P50 latency, ms 49.50±4.07
 N95 latency, ms 100.81±8.72
 N35-P50 amplitude, μV 2.81±1.03
 P50-N95 amplitude, μV 4.81±1.79
Disc parameters
 Presence of DH, n (%) 17 (12.1)
 Disc area, mm2 1.88±0.42
 Vertical cup-to-disc ratio 0.73±0.11
 Follow-up duration, y 3.15±0.96
Data are mean±SD unless otherwise indicated.
dB indicates decibel; DH, disc hemorrhage; DM, diabetes mellitus; HTN, systemic hypertension; IOP, intraocular pressure; MD, mean deviation; mGC/IPL, macular ganglion cell-inner plexiform layer; MvD, microvascular dropout; OCT, optical coherence tomography; PERG, pattern electroretinography; pRNFL, peripapillary retinal nerve fiber layer; PSD, pattern standard deviation; SAP, standard automated perimetry; VD, vessel density; VF, visual field; VFI, visual field index.

Of the 140 eyes, 57 (40.7%) exhibited glaucoma progression as defined by the GPA of the Humphrey VF (Table 2). The progressors exhibited more frequent presence of MvD on OCT-A at baseline (82.0%) and DH during follow-up (19.3%), compared with nonprogressors (40.6%, P<0.001; 7.2%, P=0.031, respectively). The MD slopes were −0.43±1.11 dB/year in the progressors and 0.59±1.27 dB/year in the nonprogressors; the difference was statistically significant (P<0.001). Both the PSD slope (P<0.001) and VFI slope (P<0.001) significantly differed between progressors and nonprogressors. The progressors included significantly more patients with isolated central scotoma at baseline (25.7%), compared with the nonprogressors (18.1%, P=0.011). Among PERG parameters, the reduction in the P50-N95 amplitude significantly differed between progressors (4.46±1.53 μV) and nonprogressors (5.01±1.81 μV, P=0.045).

TABLE 2 - Comparison Between Progressor and Nonprogressors
Variables Progressor (n=57) Nonprogressor (n=83) P
Demographics
 Age at diagnosis, y 56.13±11.86 53.49±9.89 0.173*
 Female, n (%) 35 (61.4) 53 (63.9) 0.452
Systemic demographics
 Medication of DM, n (%) 3 (5.3) 6 (7.2) 0.462
 Medication of HTN, n (%) 6 (10.5) 9 (10.8) 0.591
Ocular demographics
 Best-corrected visual acuity 0.91±0.21 0.94±0.16 0.393*
 Axial length, mm 25.72±3.96 25.07±1.65 0.376*
 Central corneal thickness, μm 541.86±38.73 539.73±34.19 0.746*
 Spherical equivalent, diopter −3.17±3.03 −3.27±3.34 0.854*
IOP parameters
 Baseline IOP, mmHg 17.00±3.69 16.73±3.36 0.666*
 Mean follow-up IOP, mmHg 14.11±2.96 14.55±3.10 0.390*
OCT parameters
 Baseline average pRNFL thickness, μm 72.25±11.93 91.72±9.54 0.482*
 Baseline average mGC/IPL thickness, μm 70.11±8.48 70.58±8.83 0.910*
OCT angiography parameters
 Average superficial macular VD, % 45.21±4.24 46.52±3.45 0.686*
 Presence of MvD, n (%) 41 (82.0) 26 (40.6) <0.001
VF parameters
 Baseline MD of SAP, dB −5.23±5.56 −6.24±6.31 0.331*
 Baseline PSD of SAP, dB 5.62±3.48 5.76±4.59 0.848*
 Baseline VFI, % 86.47±16.63 85.63±18.21 0.783*
 Presence of isolated central scotoma at baseline, n (%) 36 (25.7) 15 (18.1) 0.011
 MD slope, dB/y −0.43±1.11 0.59±1.27 <0.001 *
 PSD slope, dB/y 0.62±0.99 0.07±0.97 <0.001 *
 VFI slope, %/y −1.64±2.98 0.63±3.76 <0.001 *
Baseline PERG parameters
 N35 latency, ms 21.77±5.64 20.64±5.92 0.252*
 P50 latency, ms 48.29±4.00 48.91±5.41 0.441*
 N95 latency, ms 102.06±8,84 99.69±9.81 0.141*
 N35-P50 amplitude, μV 2.73±1.00 2.93±1.09 0.292*
 P50-N95 amplitude, μV 4.46±1.53 5.01±1.81 0.045 *
Disc parameters
 Presence of DH, n (%) 11 (19.3) 6 (7.2) 0.031
 Disc area, mm2 1.91±0.43 1.86±0.40 0.559*
 Vertical cup-to-disc ratio 0.74±0.14 0.72±0.10 0.424*
 Follow-up duration, year 7.54±1.32 7.35±1.21 0.781*
Data are mean±SD unless otherwise indicated.
Data are mean±SD unless otherwise indicated.
Factors with statistical significance are shown in bold.
*Student t test.
χ2 test.
dB indicates decibel; DH, disc hemorrhage; DM, diabetes mellitus; HTN, systemic hypertension; IOP, intraocular pressure; MD, mean deviation; mGC/IPL, macular ganglion cell-inner plexiform layer; MvD, microvascular dropout; OCT, optical coherence tomography; PERG, pattern electroretinography; pRNFL, peripapillary retinal nerve fiber layer; PSD, pattern standard deviation; SAP, standard automated perimetry; VD, vessel density; VF, visual field; VFI, visual field index.

We used regression analysis to identify the factors associated with glaucoma progression. Age at diagnosis (P=0.008), baseline N35-P50 amplitude (P=0.034), and baseline P50-N95 amplitude (P=0.004) were significantly associated with MD slope in the univariate analysis (Table 3). Among these factors, age at diagnosis (P=0.038) and baseline P50-N95 amplitude (P=0.019) showed significant associations with the MD slope in multivariate analyses. The presence of MvD on OCT-A (P<0.001), the presence of an isolated central scotoma at baseline (P=0.014), baseline P50-N95 amplitude (P=0.045), and the presence of a DH (P=0.038) were significantly associated with VF progression, as defined by GPA in univariate analyses (Table 4). In the multivariate analysis, the presence of MvD on OCT-A (P<0.001) and baseline P50-N95 amplitude (P=0.037) were significantly associated with VF progression.

TABLE 3 - Factors Associated With Mean Deviation Slope in Glaucoma Patients
Univariate Multivariate
Variables β (95% CI) P β (95% CI) P
Age at diagnosis −0.027 (−0.047 to −0.007) 0.008 −0.027 (−0.052 to −0.001) 0.038
Female 0.272 (−0.178 to 0.722) 0.234
Medication of DM 0.573 (−0.313 to 1.459) 0.203
Medication of HTN −0.277 (−0.982 to 0.428) 0.439
Best-corrected visual acuity −0.451 (−1.682 to 0.779) 0.470
Axial length −0.043 (−0.156 to 0.071) 0.459
Central corneal thickness −0.004 (−0.011 to 0.002) 0.184
Baseline IOP −0.048 (−0.110 to 0.015) 0.133
Mean follow-up IOP 0.040 (−0.032 to 0.113) 0.277
Baseline average pRNFL thickness 0.015 (−0.003 to 0.032) 0.102
Baseline average mGC/IPL thickness 0.007 (−0.021 to 0.034) 0.625
Average superficial macular VD 0.008 (−0.033 to 0.049) 0.711
Presence of MvD −0.479 (−1.005 to 0.046) 0.074 −0.259 (−0.761 to 0.244) 0.310
Baseline MD of SAP −0.007 (−0.044 to 0.029) 0.689
Baseline PSD of SAP 0.005 (−0.007 to 0.018) 0.404
Baseline VFI −0.012 (−0.064 to 0.041) 0.659
Presence of isolated central scotoma 0.186 (−0.313 to 0.685) 0.462
Baseline N35 latency −0.034 (−0.071 to 0.003) 0.074 −0.045 (−0.093 to 0.004) 0.071
Baseline P50 latency −0.014 (−0.059 to 0.031) 0.535
Baseline N95 latency 0.016 (−0.0007 to 0.039) 0.178
Baseline N35-P50 amplitude 0.222 (0.017–0.426) 0.034 −0.058 (−0.396 to 0.280) 0.735
Baseline P50-N95 amplitude 0.182 (0.057–0.306) 0.004 0.255 (0.043–0.468) 0.019
Presence of DH −0.508 (−1.171 to 0.156) 0.133
Disc area 0.242 (−0.285 to 0.770) 0.366
Vertical cup-to-disc ratio −1.895 (−3.861 to 0.071) 0.059 −0.419 (−2.654 to 1.817) 0.711
Follow-up duration −0.025 (−0.084 to 0.047) 0.565
Factors with P<0.1 in univariate analysis were included in multivariate analysis.
Data are mean±SD unless otherwise indicated.
Factors with statistical significance are shown in bold.
CI indicates confidence interval; DH, disc hemorrhage; DM, diabetes mellitus; HTN, systemic hypertension; IOP, intraocular pressure; MD, mean deviation; mGC/IPL, macular ganglion cell-inner plexiform layer; MvD, microvascular dropout; OR, odds ratio; pRNFL, peripapillary retinal nerve fiber layer; PSD, pattern standard deviation; SAP, standard automated perimetry; VD, vessel density; VFI, visual field index.

TABLE 4 - Factors Associated With Visual Field Progression Determined by Glaucoma Progression Analysis Software From the Humphrey Field Analyzer in Glaucoma Patients
Univariate Multivariate
Variables β (95% CI) P β (95% CI) P
Age at diagnosis 1.023 (0.991–1.056) 0.160
Female 0.901 (0.449–1.807) 0.901
Medication of DM 0.713 (0.171–2.976) 0.713
Medication of HTN 0.967 (0.324–2.885) 0.967
Best-corrected visual acuity 0.419 (0.063–2.800) 0.370
Axial length 1.097 (0.920–1.307) 0.301
Central corneal thickness 1.002 (0.992–1.011) 0.737
Baseline IOP 1.022 (0.928–1.126) 0.658
Mean follow-up IOP 0.952 (0.851–1.065) 0.391
Baseline average pRNFL thickness 0.990 (0.963–1.018) 0.485
Baseline average mGC/IPL thickness 0.994 (0.951–1.038) 0.779
Average superficial macular VD 1.016 (0.960–1.075) 0.586
Presence of MvD 6.658 (2.770–16.006) <0.001 5.001 (2.005–12.471) <0.001
Baseline MD of SAP 1.029 (0.971–1.091) 0.330
Baseline PSD of SAP 0.992 (0.914–1.076) 0.847
Baseline VFI 1.003 (0.983–1.022) 0.781
Presence of isolated central scotoma 2.644 (1.217–5.746) 0.014 2.150 (0.815–5.673) 0.122
Baseline N35 latency 1.035 (0.975–1.098) 0.255
Baseline P50 latency 0.974 (0.908–1.045) 0.464
Baseline N95 latency 1.027 (0.990–1.064) 0.150
Baseline N35-P50 amplitude 0.841 (0.607–1.165) 0.298
Baseline P50-N95 amplitude 0.822 (0.768–0.926) 0.045 0.741 (0.559–0.983) 0.037
Presence of DH 3.069 (1.064–8.855) 0.038 2.942 (0.718–11.881) 0.130
Disc area 1.282 (0.567–2.899) 0.550
Vertical cup-to-disc ratio 1.271 (0.156–2.623) 0.426
Follow-up duration 1.005 (0.977–1.012) 0.683
Factors with P<0.1 in univariate analysis were included in multivariate analysis.
Data are mean±SD unless otherwise indicated.
Factors with statistical significance are shown in bold.
CI indicates confidence interval; DH, disc hemorrhage; DM, diabetes mellitus; HTN, systemic hypertension; IOP, intraocular pressure; MD, mean deviation; mGC/IPL, macular ganglion cell-inner plexiform layer; MvD, microvascular dropout; OR, odds ratio; pRNFL, peripapillary retinal nerve fiber layer; PSD, pattern standard deviation; SAP, standard automated perimetry; VD, vessel density; VFI, visual field index.

Scatterplots were created to evaluate the relationship between the baseline MD of the VF and the PERG parameters (Fig. 1). As the MD value worsened, the P50 and N95 latency increased; notably, the effect on P50 latency was more pronounced. In addition, the N35-P50 and P50-N95 amplitudes decreased as the MD value worsened; this effect was more evident in the P50-N95 amplitude. The relationship between MD slope and the PERG parameters showed that the baseline P50-N95 amplitude was significantly related to the MD slope (Fig. 2).

F1
FIGURE 1:
Scatterplots were created to evaluate the relationship between the baseline mean deviation (MD) of the visual field and the parameters of pattern electroretinography. P50 and N95 latencies show negative relationship with MD. N35-P50 and P50-N95 amplitudes show positive relationship with MD. dB indicates decibel.
F2
FIGURE 2:
Scatterplots were created to evaluate the relationship between the slope of mean deviation (MD) of the visual field and the parameters of pattern electroretinography. N35 and P50 latencies show negative relationship with MD slope. N35-P50 and P50-N95 amplitudes show positive relationship with MD. dB indicates decibel

A representative patient is shown in Figure 3. A 47-year-old woman with NTG had a localized inferotemporal RNFL defect and corresponding VF defect in the central 10-degree region. At baseline, the deep choroidal map from OCT-A shows the presence of MvD in the inferotemporal region and a decreased P50-N95 amplitude, compared with the contralateral eye. This patient had recurrent DHs and showed progression of the RNFL defect and the central scotoma.

F3
FIGURE 3:
A 47-year-old woman with normal tension glaucoma had a localized inferotemporal retinal nerve fiber layer (RNFL) thinning and corresponding visual field defect in the central 10-degree region. At baseline, the deep choroidal map from optical coherence tomography angiography shows the presence of microvasculature dropout in the inferotemporal region (yellow dotted line) and a decreased P50-N95 amplitude in the right eye, compared with the contralateral eye. This patient had recurrent disc hemorrhages and showed progression of the RNFL defect and the central scotoma during follow-up (Top, baseline; Bottom, last visit).

DISCUSSION

The findings of this study support the use of PERG and OCT-A to predict glaucoma progression. Reduced P50-N95 amplitude and the presence of MvD at baseline were significant factors associated with MD slope and VF progression, as detected by the GPA of the Humphrey Field Analyzer. Glaucoma patients who showed VF progression had MvD features on OCT-A, an isolated central scotoma, a DH, and reduced P50-N95 amplitude at baseline. Thus, progression tended to occur in glaucoma patients with vascular features, who presented with a central scotoma, or possessed RGC dysfunction detected by PERG.

The baseline status of RGC function may be important for predicting glaucoma progression. Regardless of a similar amount of RGC loss at the time of glaucoma development due to the glaucomatous process, the degree to which RGC dysfunction occurs during the developmental stage could affect the disease course thereafter. Baseline IOP as one of the most important risk factor for glaucoma progression and blindness. Recently, the United Kingdom Glaucoma Treatment Study Group reported that a higher baseline IOP is a risk factor for glaucoma deterioration.12 This finding was evident in both placebo and treatment groups with hypotensive medication, thus indicating that the contribution of baseline IOP is independent of the IOP-lowering effect during the follow-up period. The baseline IOP was also an independent risk factor in the Early Manifest Glaucoma Trial.13 A higher baseline IOP implies greater RGC dysfunction, regardless of similar RGC loss due to glaucoma and remaining RGCs. This dysfunctional RGC subset dies slowly and may contribute to glaucoma progression, despite glaucoma medication and IOP control. In addition, dysfunctional RGCs may be vulnerable to other stresses (eg, aging, IOP fluctuation, and reduced blood flow), which comprise known risk factors for glaucoma progression. Therefore, considering the degree of RGC dysfunction and RGC loss may be important in determining the treatment and follow-up evaluation of glaucoma patients.

The detection of functional changes using SAP occurs in a later stage, compared with the structural analysis; thus, RGC dysfunction cannot be fully evaluated using SAP. To overcome the limitations of SAP, several efforts to assess objective visual function in early glaucoma using electrophysiological tests have been performed. Ventura and colleagues identified PERG changes after IOP reduction in glaucoma patients, except for patients with severely impaired visual function. Thus, dysfunctional but viable RGCs exist in early glaucoma.14 A study by Jafarzadehpour and associates described the various PERG patterns of dysfunctional RGCs before cell loss.15 Another study suggested that the reduction in RGC electrical activity was more pronounced, compared with the number of lost RGCs in early glaucoma; dysfunctional RGCs were identified using PERG.5 Our group previously reported that the N95 amplitude of PERG was reduced in preperimetric glaucoma, compared with normal controls; this implied early functional changes in RGCs, which were not detected by SAP.10 PERG amplitudes have been correlated with disc morphology and several parameters of glaucoma in patients with suspected glaucoma; this suggest that RGC dysfunction can be detected before structural damage, using PERG to resolve early changes.11 Furthermore, the reduced P50-N95 amplitude at baseline was used to predict VF progression in the present study, which was not significantly associated with SAP or OCT parameters. Thus, the amount of RGC dysfunction in a patient detected by PERG is associated with future deterioration of VF when glaucomatous damage is similar in terms of the SAP or RNFL. Studies on PERG are summarized on Table 5. There are several studies related to PERG and glaucoma progression comparing with the VF or RNFL changes.11,16–18 Correlation between VD and OCT-A and comparing diagnostic performances has been published.7,19,20 However, there was not any study looking into progression with both PERG and OCT-A parameters yet.

TABLE 5 - Summary of Studies Related to PERG and OCT-A in Glaucoma
First Author (y) Design Location No. Eyes Mean Age (y) Methodology Findings
Bode (2011)11 Prospective longitudinal study Germany 120 Not provided Baseline PERG and VF Baseline PERG ratio detected glaucoma conversion 4 y before VF in glaucoma suspects
Banitt (2013)16 Longitudinal cohort study USA 201 56.1 Baseline PERG and OCT Baseline PERG amplitude predicted RNFL thickness loss in glaucoma suspects
Park (2017)17 Retrospective study South Korea 74 49.5 PERG and OCT PERG had stronger structure-function relationship with OCT parameters in early glaucoma compared with the VF
Jeon (2019)18 Cross-sectoinal study South Korea 145 56.32 PERG and OCT P50 and N95 amplitude of PERG was associated with RNFL thickness in glaucoma
Al-Nosairy (2020)7 Observational study Germany 12 Not provided PERG, OCT-A, and OCT PERG parameter shows correlation with OCT-A parameters and combining ERG and OCT-A measures improve diagnostic performance in glaucoma
Lee (2020)19 Retrospective study South Korea 95 56.9 PERG, OCT-A, and OCT N95 amplitude was associated with vessel density of OCT-A in glaucoma
Jeon (2022)20 Cross-sectoinal study South Korea 76 57.47 PERG, OCT-A, and OCT N95 amplitude had correlation with macular vessel density of OCT-A in early glaucoma
ERG indicates electroretinogram; OCT, optical coherence tomography; OCT-A, optical coherence tomography angiography; PERG, pattern electroretinography; RNFL, retinal nerve fiber layer; VF, visual field.

Among PERG parameters, the P50-N95 amplitude was significantly associated with progression in the present study. Previous reports have shown latency delay and/or a reduced amplitude of the P50 and N95 peak in glaucoma.21,22 The latencies of the peaks did not differ between progressors and nonprogressors in the present study. The N95 amplitude is presumably related to the RGC action potential, with changes clearly evident at the N95 peak. However, changes within the inner retina and nonspiking activity within the RGCs likely contribute to changes in P50, as demonstrated in other glaucoma studies.23,24 Along with the P50-N95 amplitude, the N35-P50 amplitude was correlated with the baseline MD or MD slope and had a significant association with the MD slope in univariate regression analysis in the present study. Although the N35-P50 amplitude was not a significant risk factor related to VF progression according to the GPA of the Humphrey Field Analyzer, it may indicated RGC dysfunction, because the P50 peak has been associated with the bipolar cells—RGC synapse and the current within RGC somata/dendrites.

Choroidal MvD is a new ocular finding recently revealed by OCT-A.25–27 MvD can be observed within the β-zone of parapapillary atrophy on choroidal VD maps around the optic disc generated by OCT-A. MvD has frequently been found in glaucomatous eyes and is related to the site of glaucomatous damage.26,28–30 in addition, it has been associated with glaucoma progression in terms of RNFL thinning and VF loss.31–33 Jo et al34 showed that glaucoma patients with MvD tended to have greater glaucoma severity compared with glaucoma patients lacking MvD. Lee et al35 showed that MvD was an independent predictor of faster RNFL thinning in glaucoma patients and was frequently found in glaucoma patients with DH; it also exhibited a topographical correlation with juxtapapillary choroidal thickness.36 Recent findings have shown that glaucoma patients with MvD have a greater and longer reduction in their night-time diastolic blood pressure.37 These findings suggest a vascular origin for the presence of MvD.33,38 Our group previously reported that MvD at the DH site poses a greater risk of localized RNFL progression.33 Accordingly, we hypothesize that MvD in the parapapillary region compromises the blood supply to the RNFL and the optic disc rim, triggering vascular changes that facilitate DH development and further RNFL thinning. The exact mechanism by which MvD develops remains unclear; however, it may be related to vascular insufficiency in the deep optic nerve head tissues. Hypoperfusion of the prelaminar and/or laminar region may reduce RGC function, contributing to further RGC damage caused by glaucoma. Patients who initially exhibited choroidal MvD at baseline tended to show faster VF progression in the present study. Therefore, MvD along with a DH could be an indication of vascular features that contribute to glaucoma progression.

Glaucoma patients with vascular features show VF defects within the central 10-degree region in both early and late stages of the disease.39–41 Central VF progression has been linked to vascular features, including DH, autonomic dysfunction, migraines, orthostatic hypotension, and Raynaud’s phenomenon.42,43 Recently, central VF involvement may indicate loss of VD on OCT-A.29,44–46 Kwon and colleagues showed that MvD and enlargement of the foveal avascular zone on OCT-A were significantly associated with central scotoma.47–49 Lee et al46 showed that the presence of MvD was associated with initial parafoveal scotoma. Our group demonstrated that VD in the parapapillary region on OCT-A was associated with central scotoma in myopic glaucoma patients.29 In the present study, glaucoma patients with an initial central scotoma showed faster progression, independent of the presence of MvD or DH. Central VF damage at baseline has been identified as a progression risk factor in several studies.44,50 The central macular area may be more vulnerable to ocular perfusion due to higher oxygen demand. Therefore, glaucoma patients presenting with initial central scotoma may possess vascular factors that contribute to damage in the central VF region. Early damage in the central VF region may affect the PERG outcome, compared with VF defects outside of the central 10-degree region, because PERG focuses on macular function. Therefore, our findings that baseline PERG parameters were related to glaucoma progression may also be related to patients with central scotoma.

This study had several limitations. First, the included patients had predominantly NTG, and a large proportion of patients with isolated central scotoma were included in the analyses. These patients tended to show clinical characteristics related to vascular risk factors, which may have affected our findings. Therefore, it is difficult to generalized our results or apply the results of our study to all glaucoma types. Second, the projection artifacts of OCT-A could have influenced vascular status evaluation. The VD measurements of our study were restricted to the superficial retinal layer. It is unclear whether projection artifacts affected the most anterior part of the retinal vasculature in the current study. In addition, only eyes with clear dropout were presumed to have MvD. MvD is located within the β-zone parapapillary atrophy, which has few overlying superficial retinal vessels; this may have reduced the risk of artifacts.

In conclusion, a reduced P50-N95 amplitude and the presence of MvD at baseline were significant factors associated with VF progression in predominantly NTG patients. Glaucoma patients who showed VF progression had features of MvD on OCT-A, an isolated central scotoma, a DH, and a reduced P50-N95 amplitude at baseline. In this study on predominantly NTG patients, patients with vascular features, who presented with central scotoma, or possess RGC dysfunction detected by PERG, tended to show progression and should be monitored closely.

REFERENCES

1. Kerrigan-Baumrind LA, Quigley HA, Pease ME, et al. Number of ganglion cells in glaucoma eyes compared with threshold visual field tests in the same persons. Invest Ophthalmol Vis Sci. 2000;41:741–748.
2. Berninger TA, Arden GB. The pattern electroretinogram. Eye (London, England). 1988;2:S257–S283.
3. Bach M, Poloschek CM. Electrophysiology and glaucoma: current status and future challenges. Cell Tissue Res. 2013;353:287–296.
4. Mavilio A, Scrimieri F, Errico D. Can variability of pattern ERG signal help to detect retinal ganglion cells dysfunction in glaucomatous eyes? BioMed Res Int. 2015;2015:571314.
5. Ventura LM, Sorokac N, De Los Santos R, et al. The relationship between retinal ganglion cell function and retinal nerve fiber thickness in early glaucoma. Invest Ophthalmol Vis Sci. 2006;47:3904–3911.
6. Ventura LM, Porciatti V, Ishida K, et al. Pattern electroretinogram abnormality and glaucoma. Ophthalmology. 2005;112:10–19.
7. Lee T, Seo DR, Kim JY, et al. Relationship between N95 amplitude of pattern electroretinogram and optical coherence tomography angiography in open-angle glaucoma. J Clin Med. 2020;9:3854.
8. Garhöfer G, Chua J, Tan B, et al. Retinal neurovascular coupling in diabetes. J Clin Med. 2020;9:2829.
9. Wareham LK, Calkins DJ. The neurovascular unit in glaucomatous neurodegeneration. Front Cell Dev Biol. 2020;8:452.
10. Jung KI, Jeon S, Shin DY, et al. Pattern electroretinograms in preperimetric and perimetric glaucoma. Am J Ophthalmol. 2020;215:118–126.
11. Jeon SJ, Park HL, Jung KI, et al. Relationship between pattern electroretinogram and optic disc morphology in glaucoma. PLoS One. 2019;14:e0220992.
12. Founti P, Bunce C, Khawaja AP, et al. Risk factors for visual field deterioration in the United Kingdom glaucoma treatment study. Ophthalmology. 2020;127:1642–1651.
13. Leske MC, Heijl A, Hyman L, et al. Predictors of long-term progression in the early manifest glaucoma trial. Ophthalmology. 2007;114:1965–1972.
14. Ventura LM, Porciatti V. Restoration of retinal ganglion cell function in early glaucoma after intraocular pressure reduction: a pilot study. Ophthalmology. 2005;112:20–27.
15. Jafarzadehpour E, Radinmehr F, Pakravan M, et al. Pattern electroretinography in glaucoma suspects and early primary open angle glaucoma. J Ophthalmic Vis Res. 2013;8:199–206.
16. Bode SF, Jehle T, Bach M. Pattern electroretinogram in glaucoma suspects: new findings from a longitudinal study. Invest Ophthalmol Vis Sci. 2011;52:4300–4306.
17. Banitt MR, Ventura LM, Feuer WJ, et al. Progressive loss of retinal ganglion cell function precedes structural loss by several years in glaucoma suspects. Invest Ophthalmol Vis Sci. 2013;54:2346–2352.
18. Park K, Kim J, Lee J. Measurement of macular structure-function relationships using spectral domain-optical coherence tomography (SD-OCT) and pattern electroretinograms (PERG). PLoS One. 2017;12:e0178004.
19. Al-Nosairy KO, Prabhakaran GT, Pappelis K, et al. Combined multi-modal assessment of glaucomatous damage with electroretinography and optical coherence tomography/angiography. Transl Vis Sci Technol. 2020;9:7.
20. Jeon SJ, Jung KI, Park CK, et al. Macular blood flow and pattern electroretinogram in normal tension glaucoma. J Clin Med. 2022;11:1790.
21. Turkey E, Elsanabary ZSE, Elshazly LHM, et al. Role of pattern electroretinogram in ocular hypertension and early glaucoma. J Glaucoma. 2019;28:871–877.
22. Wilsey LJ, Fortune B. Electroretinography in glaucoma diagnosis. Curr Opin Ophthalmol. 2016;27:118–124.
23. Luo X, Frishman LJ. Retinal pathway origins of the pattern electroretinogram (PERG. Invest Ophthalmol Vis Sci. 2011;52:8571–8584.
24. Kong AW, Della Santina L, Ou Y. Probing ON and OFF retinal pathways in glaucoma using electroretinography. Transl Vis Sci Technol. 2020;9:14.
25. Lee EJ, Lee SH, Kim JA, et al. Parapapillary deep-layer microvasculature dropout in glaucoma: topographic association with glaucomatous damage. Invest Ophthalmol Vis Sci. 2017;58:3004–3010.
26. Suh MH, Zangwill LM, Manalastas PI, et al. Deep retinal layer microvasculature dropout detected by the optical coherence tomography angiography in glaucoma. Ophthalmology. 2016;123:2509–2518.
27. Ichiyama Y, Minamikawa T, Niwa Y, et al. Capillary dropout at the retinal nerve fiber layer defect in glaucoma: an optical coherence tomography angiography study. J Glaucoma. 2017;26:e142–e145.
28. Shin JW, Kwon J, Lee J, et al. Choroidal microvasculature dropout is not associated with myopia, but is associated with glaucoma. J Glaucoma. 2018;27:189–196.
29. Park HL, Jeon SJ, Park CK. Features of the choroidal microvasculature in peripapillary atrophy are associated with visual field damage in myopic patients. Am J Ophthalmol. 2018;192:206–216.
30. Lin S, Cheng H, Zhang S, et al. Parapapillary choroidal microvasculature dropout is associated with the decrease in retinal nerve fiber layer thickness: a prospective study. Invest Ophthalmol Vis Sci. 2019;60:838–842.
31. Lee JY, Shin JW, Song MK, et al. An increased choroidal microvasculature dropout size is associated with progressive visual field loss in open-angle glaucoma. Am J Ophthalmol. 2021;223:205–219.
32. Kim CY, Lee EJ, Kim JA, et al. Progressive retinal nerve fibre layer thinning and choroidal microvasculature dropout at the location of disc haemorrhage in glaucoma. Br J Ophthalmol. 2020;105:674–680.
33. Park HL, Kim JW, Park CK. Choroidal microvasculature dropout is associated with progressive retinal nerve fiber layer thinning in glaucoma with disc hemorrhage. Ophthalmology. 2018;125:1003–1013.
34. Jo YH, Kwon J, Shon K, et al. Greater severity of glaucomatous damage in eyes with than without choroidal microvasculature dropout in open-angle glaucoma. Invest Ophthalmol Vis Sci. 2019;60:901–912.
35. Lee EJ, Kim JA, Kim TW. Influence of choroidal microvasculature dropout on the rate of glaucomatous progression: a prospective study. Ophthalmol Glaucoma. 2020;3:25–31.
36. Lee SH, Lee EJ, Kim TW. Topographic correlation between juxtapapillary choroidal thickness and parapapillary deep-layer microvasculature dropout in primary open-angle glaucoma. Br J Ophthalmol. 2018;102:1134–1140.
37. Shin JW, Jo YH, Song MK, et al. Nocturnal blood pressure dip and parapapillary choroidal microvasculature dropout in normal-tension glaucoma. Sci Rep. 2021;11:206.
38. Rao HL, Sreenivasaiah S, Dixit S, et al. Choroidal microvascular dropout in primary open-angle glaucoma eyes with disc hemorrhage. J Glaucoma. 2019;28:181–187.
39. Park HY, Jung KI, Na KS, et al. Visual field characteristics in normal-tension glaucoma patients with autonomic dysfunction and abnormal peripheral microcirculation. Am J Ophthalmol. 2012;154:466–475.
40. Kosior-Jarecka E, Wróbel-Dudzińska D, Łukasik U, et al. Ocular and systemic risk factors of different morphologies of scotoma in patients with normal-tension glaucoma. J ophthalmol. 2017;2017:1480746.
41. Park SC, De Moraes CG, Teng CC, et al. Initial parafoveal versus peripheral scotomas in glaucoma: risk factors and visual field characteristics. Ophthalmology. 2011;118:1782–1789.
42. Park HY, Park SH, Park CK. Central visual field progression in normal-tension glaucoma patients with autonomic dysfunction. Invest Ophthalmol Vis Sci. 2014;55:2557–2563.
43. Choi J, Lee JR, Lee Y, et al. Relationship between 24-hour mean ocular perfusion pressure fluctuation and rate of paracentral visual field progression in normal-tension glaucoma. Invest Ophthalmol Vis Sci. 2013;54:6150–6157.
44. Jo YH, Kwon J, Jeong D, et al. Rapid central visual field progression rate in eyes with open-angle glaucoma and choroidal microvasculature dropout. Scientific reports. 2019;9:8525.
45. Jeon SJ, Park HL, Park CK. Effect of macular vascular density on central visual function and macular structure in glaucoma patients. Sci Rep. 2018;8:16009.
46. Lee EJ, Kim TW, Kim JA, et al. Central visual field damage and parapapillary choroidal microvasculature dropout in primary open-angle glaucoma. Ophthalmology. 2018;125:588–596.
47. Kwon J, Choi J, Shin JW, et al. Glaucoma diagnostic capabilities of foveal avascular zone parameters using optical coherence tomography angiography according to visual field defect location. J Glaucoma. 2017;26:1120–1129.
48. Kwon J, Choi J, Shin JW, et al. Alterations of the foveal avascular zone measured by optical coherence tomography angiography in glaucoma patients with central visual field defects. Invest Ophthalmol Vis Sci. 2017;58:1637–1645.
49. Kwon J, Shin JW, Lee J, et al. Choroidal microvasculature dropout is associated with parafoveal visual field defects in glaucoma. Am J Ophthalmol. 2018;188:141–154.
50. Raman P, Suliman NB, Zahari M, et al. Baseline central visual field defect as a risk factor for NTG progression: a 5-year prospective study. J Glaucoma. 2019;28:952–957.
Keywords:

glaucoma; pattern electroretinogram; microvasculature dropout; optical coherence tomography angiography

Copyright © 2022 Wolters Kluwer Health, Inc. All rights reserved.