Ganglion Cell Complex Analysis in Glaucoma Patients: What Can It Tell Us?

Abstract

Glaucoma is a group of optic neuropathies characterized by a progressive degeneration of retina ganglion cells (RGCs) and their axons that precedes functional changes detected on the visual field. The macular ganglion cell complex (GCC), available in commercial Fourier-domain optical coherence tomography, allows the quantification of the innermost retinal layers that are potentially involved in the glaucomatous damage, including the retinal nerve fiber (RNFL), ganglion cell and inner plexiform layers. The average GCC thickness and its related parameters represent a reliable biomarker in detecting preperimetric glaucomatous damage. The most accurate GCC parameters are represented by average and inferior GCC thicknesses, and they can be associated with progressive visual field loss. Although the diagnostic accuracy increases with more severe glaucomatous damage and higher signal strength values, it is not affected by increasing axial length, resulting in a more accurate discrimination of glaucomatous damage in myopic eyes with respect to the traditional RNFL thickness. The analysis of the structure-function relationship revealed a good agreement between the loss in retinal sensitivity and GCC thickness. The use of a 10-2° visual field grid, adjusted for the anatomical RGCs displacement, describes more accurately the relationship between RGCs thickness and visual field sensitivity loss.

Introduction

Glaucoma is a heterogenous group of progressive neurodegenerative optic neuropathies characterized by the degeneration of retinal ganglion cells (RGC) and their axons in addition to significant remodeling of the lamina cribrosa of the optic nerve head.^1,2

Standard automated perimetry (SAP) has been widely used to assess visual function in glaucomatous eyes for staging and monitoring the disease progression.^3–7 However, it has been estimated that at least 25% to 35% of RGCs must be lost before producing significant abnormalities on the visual field.⁸ Several clinical studies confirmed that the decline in RGCs precedes functional changes detected on SAP in glaucomatous eyes, confirming that a combination of structural and functional tests might offer an optimal assessment of the neural damage and its progression.^9–13

Optical coherence tomography (OCT) allows for noninvasive imaging of glaucomatous structural damage involving the optic nerve, peripapillary retinal nerve fiber layer (RNFL) and the macular region. Of these, the quantification of circumpapillary RNFL represents the most commonly used OCT parameter because it has been considered a useful method in assessing the structural loss of RGCs in glaucoma.^14–17 However, this method analyzes only the axonal portion of the RGCs without considering the cell bodies and dendrites, which are also affected in glaucoma and reside in the ganglion cell layer (GCL) and inner plexiform layer (IPL), respectively.^18–20

In addition to the conventional peripapillary RNFL, macular thickness can be affected in glaucomatous eyes, suggesting that it may represent a surrogate indicator of RGCs damage, given the prominent distribution of these cells within the macular region.^19,21–23 However, the indirect estimation of RGCs with macular thickness was inaccurate due to influence from the outer retinal layers, which constitute approximately 65% to 70% of the total retinal thickness and contribute to high variability and less discriminant power.¹⁸

With the development of newer OCT, the automatized segmentation of inner retinal layers has become possible. Macular ganglion cell complex (GCC) includes all three innermost retinal layers potentially involved in the glaucomatous damage (RNFL, GCL and IPL); in addition, the segmentation of the GCC thickness is traced from the inner limiting membrane and the outer IPL boundary.^18,24 The GCC thickness was confirmed to be superior to macular thickness in detecting glaucomatous eyes.¹⁸

The purpose of the present review is to provide an excursus of the existing literature on the role of GCC in glaucomatous disease, reporting how to interpret the GCC parameters in the early diagnosis, progression and prognosis.

Macular Ganglion Cell Complex Imaging

Topographic Distribution of RGCs in Normal and Glaucomatous Retina

Ganglion cells peak density ranges between 31,600 and 37,800 cells/mm,² and the highest cells density in a horizontally oriented elliptical ring of 0.4–2 mm from the foveal center, with approximately 50% of the RGCs overall, is located within 4.5 mm of the foveal center.²⁵ A topographical analysis of RGC in cadaveric glaucomatous eyes also revealed that RGC density was 10 times higher in the foveal region with respect to more peripheral retina.⁸

The RGCs first appeared at 150–250 μm from the foveal center, and the total displacement declined to zero at the optic nerve head (approximately 3.4 mm nasal to the foveal center) but was still detectable up to 4.5 mm temporal to the foveal center. The Henle fiber layer length is the most important contributor to the receptoral and post-receptoral displacement and was found to be greater than the one previously estimated.^25–28

Different animal models of experimental glaucoma have been developed to simulate glaucomatous optic neuropathy. It has been confirmed that sustained intraocular pressure elevation is accompanied by a progressive RGCs loss associated with visual acuity and contrast sensitivity decline.^29–31 More importantly, as demonstrated in adult rhesus macaques, the RNFL birefringence declines prior to and more quickly than the RNFL thickness, suggesting that RGCs cytoskeleton damage precedes RNFL thinning.³²

In experimentally induced glaucoma, a selective loss of large RGCs was demonstrated in the foveal region and the midperipheral retina.³³ The foveal region demonstrated a loss of ganglion cells similar to other peripheral retinal areas in experimentally induced glaucoma in monkeys.³⁴ The greater susceptibility of the large RGCs to glaucomatous damage was also confirmed in a similar experimental model, analyzing three different degrees of glaucomatous damage: mild (10–40% cell loss), moderate (40–60% cell loss) and severe (60–90% cell loss). In this model, the inferior retina appeared to be more affected than superior areas.³⁵ In human retina, greater susceptibility of the large RGCs to glaucomatous damage was confirmed, but although the inferior retina presented smaller RGCs cells, no significant differences were found in the RGCs distribution between hemifields.¹³

The RGC damage in experimentally induced ocular hypertension in rodents was primarily sectorial in a “pie-shaped” or “triangular” configuration, with selective damage of the RGC cells and relative sparing of the non-RGC neurons as presumed displaced amacrine cells. These triangular sectors of RGC loss, localized preferentially on the dorsal retina, were oriented with their base toward the peripheral retina and their apex toward the optic disc.^36,37

Quantification of GCC Using Optical Coherence Tomography

With the introduction of Fourier-domain OCT technology, several manufacturers proposed the evaluation of the ganglion cell layer using different protocols. Table 1 summarizes spectral-domain (SD) commercially available OCT protocols used to analyze macular ganglion cells.^18,38–44

Table 1.

Overview of the Ganglion Cell Layer Imaging Performed by Different Spectral-Domain OCT Devices

Instrument	Macular Scan Protocol	Macular Thickness Measurements
RTVue-100 (Optovue Inc., Fremont, CA)	7×7-mm square area consisting of a horizontal 7-mm line scan (467 A-lines) and 15 vertical 7-mm line scans (400 A-lines) spaced at 0.5-mm intervals, centered 0.75−1 mm temporal to the fovea. The GCC thickness is calculated within a 6-mm diameter circular macular area.	Ganglion cell complex (GCC) GCC= RNFL + GCL +IPL layers
Cirrus HD-OCT (Carl Zeiss Meditec, Inc., Dublin, CA)	Macular cube 200 × 200 protocol (6 × 6 mm square grid) containing 200 B-scans (200 A-scan lines) or Macular Cube 512 × 128 protocol (128 B-scans, 512 A-scans per B-scan). The macular thickness data are calculated on 14.13 mm2 elliptical annulus area centered on the fovea.	Ganglion cell analysis (GCA) Macular RNFL GCL+IPL GCC
RS-3000 SD-OCT (Nidek, Gamagori, Aichi, Japan)	30° × 30° square area (~ 9 × 9 mm square area) consisting of a vertical B scan (512 A-scans) × 128 B-scans horizontally.	Macular GCC Superior/inferior (S/I) semicircle map 8-sector map or GCchart
Spectralis (Heidelberg Engineering, Inc., Heidelberg, Germany)	30° × 25° volume scan consisting of 61 horizontal B scans (1024 A scans/line) within the central 20°. Retinal map- 8 x 8 grid composed by 64 sectors with colorimetric scale	Posterior pole asymmetry analysis (PPAA)
3D OCT (Topcon, Inc., Tokyo, Japan)	Macular 3D raster scan composed by 512 × 128 A scans covering a cube of 6 × 6 × 2 mm	Macular RNFL GCL/IPL GCC

Abbreviations: RNFL, retinal nerve fiber layer; GCL, ganglion cell layer; IPL, inner plexiform layer.

The RTVue-100 software (Optovue, Inc., Fremont, CA) acquires 26,000 A-scans per second with an axial resolution of 5 μm in the tissue.^18,38 The GCC is defined as the sum of RNFL, GCL and IPL thickness measured from the inner limiting membrane to the IPL boundaries (Figure 1). The GCC scan pattern is centered 0.75 mm temporal to the fovea, covering a square grid of 7 × 7 mm on the central macula and consisting of 15 B-scans spaced 0.5 mm from one another.¹⁸

Figure 1.

Ganglion cell complex (GCC) analysis and segmentation. (A) Thickness map, automatically calculated using RTVue Fourier-domain optical coherence tomography (FD-OCT); (B) Cross-sectional FD-OCT B-scan with segmentation of the GCC. The inset (dotted square) depicts the three anatomical layers included in the GCC analysis: nerve fiber layer (NFL), ganglion cell layer (GCL) and inner plexiform layer (IPL). The traced boundaries for the GCC scan (white arrows) pass from the inner limiting membrane and outer IPL.

After the acquisition, the software automatically computes three color-coded maps: a) a thickness map, color-coded with cooler colors (blue and black), indicating more severe GCC loss; b) a deviation map, which displays the percentage of thickness loss with respect to the normative database; and c) a significance map, which displays statistically significant thickness changes from the normal.

Two pattern-based diagnostic indexes are also calculated by the analysis software. The focal loss volume (FLV) indicates the average amount of focal GCC loss divided by the map area. The global loss volume (GLV) provides the sum of the negative fractional deviation.^18,45 (Figure 2)

Figure 2.

Quantification of ganglion cell complex (GCC) using Fourier-domain optical coherence tomography (FD-OCT) in the case of severe glaucomatous ganglion cell loss. Three different colorimetric maps were automatically computed by the FD-OCT software (RTVue-100 software; Optovue, Inc., Fremont, CA). (A) Thickness map, displaying diffuse blue and black colors indicating severe and diffuse loss of GCC thickness. The central black spot indicates the fovea where the ganglion cells are missing. (B) Deviation map, illustrating the percentage of thickness loss with respect to the normative database. Black coloring indicates a GCC loss of 50% or greater. (C) Significance map, displaying the significant thickness change variation from normal, where green represents values within the normal range (p-value 5–95%), yellow indicates borderline results (< 5%) and red represents results outside the normal limits (< 1%); (D) Quantitative indexes calculated by the software, including average thickness, hemifield thicknesses (superior, inferior and intra-eye) and the two pattern-based diagnostic indexes (i.e., focal loss volume [FLV] and global loss volume [GLV]); (E) Cross-sectional B-scan displays the segmentation used for GCC analysis.

The pattern-based parameters (i.e., FLV and GLV) are considered to have better diagnostic values than the mean GCC loss, likely because a focal GCC loss can be overlooked in eyes with above-average GCC thickness.^18,45

Cirrus HD-OCT (Carl Zeiss Meditec, Inc., Dublin, CA) achieves 27,000 A-scans per second with an axial image resolution of 5 μm.⁴⁶ The ganglion cell analysis algorithm automatically delineates the outer boundary of the macular RNFL and the IPL, obtaining the ganglion cell-inner plexiform layer (GCIPL).⁴¹

The Cirrus HD-OCT software calculates the average, minimum and sectoral GCIPL thickness (superotemporal, superior, superonasal, inferonasal, inferior, inferotemporal);⁴⁷ for further details, see Table 1. The minimum GCIPL thickness represents the lowest thickness over a single meridian crossing the anulus. The software also provides a thickness map, a color-coded topographic map matching a normative database and a deviation map. Values within the normal range (P = 5–95%) are depicted in green, borderline values (1% < P < 5%) in yellow and values outside normal limits (P < 1%) in red.^41,47

The 3D OCT-2000 (Topcon, Inc., Tokyo, Japan) device uses a scan speed of 50,000 A-scans per second with a depth resolution of 6 μm. The macular inner retinal layers analyzed by the software include GCC, GCL+IPL (mGCL+) and macular RNFL; for further details, see Table 1.^48,49 The macular inner retinal layer analysis software divides the macular square into a 6 x 6 grid containing 100 cells (0.6 x 0.6 mm each). The thickness calculated in each cell is compared with a normative database. Similar to the other color-coded maps, values within the normal range are represented as green, borderline as yellow and outside normal limits as red.⁴⁸

Spectralis (Heidelberg Engineering, Inc., Heidelberg, Germany) posterior pole asymmetry protocol (Table 1) calculates the total retinal thickness in a central 8 x 8 grid, which displays the asymmetry between the eyes and between the superior and inferior hemispheres of each eye, using darker cells to represent regions of greater asymmetry.^43,44

Diagnostic Accuracy, Reproducibility and Clinical Validation of Macular Ganglion Cell Complex Analysis

The diagnostic accuracy of GCC is significantly influenced by disease severity—with increasing accuracy in more severe glaucomatous damage—whereas the optic disc size does not influence the GCC parameters.⁵⁰ Furthermore, scan quality (i.e., signal strength index) was considered an additional factor that influences the diagnostic performance of GCC parameters in glaucoma.⁵¹

The RTVue-100 device is widely used in clinical practice, demonstrating diagnostic performance similar to other SD-OCT devices in identifying glaucomatous disease by using average peripapillary RNFL and GCC thickness.³⁸ The reproducibility of GCC measurements with the RTVue-100 OCT has been demonstrated to be superior to the previous Stratus OCT devices and is satisfactory on both healthy and glaucomatous eyes with moderate to severe damage.⁵²

The GCIPL thickness parameters are also less influenced by axial length or refractive errors, representing a potential advantage in the Asian population, known to exhibit a high prevalence of myopia.⁵³ The inner retinal layers measured by 3D-OCT demonstrated significant variations with gender and age but had no significant relationship with axial length.⁵⁴

The presence of SD-OCT artifacts caused by segmentation or acquisition errors can limit the diagnostic ability of GCC analysis, and the segmentation artifacts were considered the most frequent in both healthy and glaucoma patients.³⁹

Other factors might influence the interpretation of GCC results; in particular, the epiretinal membrane can cause errors in segmentation, but also retinal disorders and optic neuropathies can alter the GCC thickness. However, these conditions typically do not exhibit specific patterns of GCC damage, but it is important to consider other potential confounding influencing factors when interpreting GCC maps in glaucomatous eyes.⁵⁵

Commercially available OCT devices have an integrated normative database limited to individuals aged 18 years or older, thus limiting their application in the pediatric population.^56–58 The evaluation of glaucomatous damage in monitoring childhood glaucoma can be challenging in the future,^59,60 even though children with glaucoma demonstrated decreased RNFL thickness and inner retina macular layers.^61,62 The GCL-IPL values were found to be thicker in the pediatric population with respect to adults, confirming the necessity of using a pediatric reference database to avoid false estimation with the adult normative database.^63,64

The Utility of Ganglion Cell Complex in Early Diagnosis and Glaucomatous Disease Progression

Macular Ganglion Cell Complex in Preperimetric Glaucoma

In cadaveric eyes of glaucoma patients, it has been estimated that 20% of RGCs loss corresponded to a 5 dB decrease in sensitivity in the central 30 degrees, and likewise an RGCs loss of 40% was associated with a 10 dB decrease in sensitivity.¹³ Macular GCC parameters are theoretically superior over peripapillary RNFL parameters in early diagnosis because it has been experimentally demonstrated that the RGCs abnormalities precede the axonal loss.^32,45

The GCC thickness was significantly affected in eyes with RNFL defects and normal SAP. In particular, the superior GCC thickness presented predictive value similar to that of peripapillary RNFL and optic nerve head (horizontal cup: disc ratio).^11,12

The most informative OCT parameters in differentiating preperimetric glaucoma included GCC average thickness, inferior quadrant RNFL thickness, inferior GCC thickness and superior GCC thickness.^11,65,66 Kim et al⁴⁵ suggested that GCC thickness appeared to be a better diagnostic biomarker for early glaucoma than did RNFL thickness. Moreover, the OCT-derived macular GCC parameters exhibited high sensitivity in differentiating glaucomatous from normal eyes. In particular, the GLV predicted preperimetric glaucoma (odds ratio [OR]:1.74) and early glaucoma (OR = 1.22), and the FLV demonstrated greater predictive value in detecting advanced glaucoma (OR = 2.32).⁶⁷ In addition, Naghizadeh et al⁶⁸ hypothesized that pattern-derived GCC parameters (i.e., GLV and FLV) were better able to detect early structural glaucomatous changes with respect to other GCC parameters in addition to RNFL thickness and optic nerve head analysis.

Comparing the ability of GCC in detecting early glaucoma with respect to other OCT parameters, the area under the curve (AUC) was significantly higher in the GCC-related thicknesses than in the corresponding GCL/IPL or RNFL thicknesses.^69,70 Other studies reported similar diagnostic ability of the GCC and GCL/IPL parameters in discriminating early glaucomatous changes, comparable with the peripapillary RNFL thickness.^38,40

Sectorial GCC parameters with the highest AUCs for discriminating between healthy and glaucomatous eyes using Cirrus SD-OCT included infero-temporal, inferior and supero-temporal sectors in addition to the average GCC.⁴⁰ Furthermore, both minimal and inferior GCIPL thicknesses demonstrated satisfactory AUC (> 0.80), supporting their potential role as biomarkers for glaucoma screening.⁷¹ Likewise, macular parameters (GCL+IPL and GCC) and peripapillary RNFL obtained by 3D-OCT exhibited similar diagnostic accuracy in detecting preperimetric glaucoma.⁷²

Regarding the posterior pole asymmetrical analysis provided by the Spectralis software, a retinal thickness asymmetry between hemifields occupying the nasal macular area has been demonstrated, suggesting a possible role for diagnosing early glaucoma.⁴⁴

By comparing different SD-OCT devices (RTVue, Cirrus and 3D-OCT), the diagnostic performance of the average GCC thickness in diagnosing early and all stages of glaucoma was similar among the SD-OCT devices. However, RTVue exhibited better AUC for the superior hemifield GCC thickness with respect to either Cirrus and 3D-OCT, likely due to differences in the scanning protocols.³⁸

Recently, associations between localized functional defects and GCC thickness have been reported in eyes with ocular hypertension and glaucoma.⁷³ Cluster-based trend analysis of computerized visual field, using a dedicated software (EyeSuite, Octopus perimetry, Haag-Streit, Switzerland), allows for testing specific sectors grouped along nerve fiber bundles.⁷⁴ Eyes with a normal visual field (VF) but abnormal GCC thickness exhibited the best correspondence between cluster analysis VF and GCC defects (100%). Although the purpose of this study was to demonstrate the usefulness of VF cluster analysis in discriminating early focal glaucomatous defects, it also confirmed the ability of GCC to predict early-stage glaucoma before the VF global indexes (mean defect, MD and loss of variance, LV) are altered.⁷³

Functional Correlates of GCC Loss in Perimetric Glaucoma

The diagnostic ability of GCC in advanced glaucoma may be inferior to RNFL parameters because only 50% of the RGCs occupy the macular region.⁴⁵ However, both RNFL and GCC thickness demonstrated a similar ability to diagnose moderate and severe glaucoma, defined according to the modified Hodapp-Anderson-Parrish grading scale.^45,75 The best GCC parameters for detecting perimetric glaucoma were average GCC thickness and inferior GCC thickness.^18,76,77

The regional assessment of the macular inner retinal layer using 3D-OCT-2000 has been considered a better indicator of paracentral scotoma than peripapillary RNFL parameters.⁴⁸

In glaucomatous eyes, either GCC or RNFL parameters significantly correlated with retinal mean sensitivity, offering a structure-functional relationship. The GCC thickness offered the best correlation with the retinal sensitivities within the central 10° of the macula, whereas the RNFL correlated with only the periphery of the central 10° after adjusting for RGCs displacement.⁷⁸

The most common SAP test protocol in glaucomatous eyes included the 24-2 or 30-2 protocols of the Humphrey Field Analyzer (Zeiss, Inc), in which the test points are spaced 6° apart.^79,80 However, given the RGCs bodies displacement from their receptive fields in the macular region, such VF test patterns failed to accurately demonstrate a direct relationship between VF defects and RGCs damage to the macula.^79,81–83 In this regard, the use of a 10–2 test with a 2° grid and considering the RGCs displacement demonstrated a better agreement between RGCs thinning and VF damage.^79,81–84

The RGCs damage most severely involved the inferior retina (superior visual field), which projected to the macular vulnerability zone located in the inferior quadrant of the optic disc.^79,81,84 Furthermore, a small centrocecal region of the inferior macula and all the superior macula (inferior VF) projected to the temporal quadrant, which was found to be less vulnerable to glaucomatous damage.^79,81 Beyond the deep and localized damage involving the macular vulnerability zone, a diffuse and shallow defect can be identified with a widespread loss of 10–2 sensitivity associated with thinning of macular RGCs and RNFL.⁸⁵

The quantitative structure-function relationship between GCL/IPL thickness and visual sensitivity markedly decreased by approximately outside 6° of the VF (corresponding to ~7.2° on the retina), likely due to the physiological thinning of the GCL/IPL layer in the more peripheral regions of the macula.⁸²

Predictive GCC Parameters of Visual Field Progression

Advanced imaging for glaucoma study (AIGS) demonstrated that the focal loss of GCC was the strongest single predictor for VF progression (hazard ratio: 3.1), followed by the focal loss of peripapillary RNFL. The combination of GCC focal loss with age and central corneal thickness, constituting a composite index called the “Glaucoma composite progression index,” increased the predictive value for disease progression. More interestingly, intraocular pressure and VF parameters were not found to be significant predictors in the multivariate model.¹⁰

In the most recent AIGS, the focal loss of RNFL and GCC were both considered the strongest OCT predictors for glaucoma progression, with superiority of the RNFL FLV index. However, the detection of RNFL-FLV is not available in current commercial devices, making the GCC-FLV the most important predictor for glaucomatous disease. The VF progression was estimated using the MD and VF index, considering VF rapid progression an MD slope > −0.5 dB/year or VF index >−1%/year.⁸⁶ In addition, the inferior GCC thickness was found to be significantly thinner in the fast progression group with respect to slow progressors.⁸⁷

Progressive structural changes in RNFL and GCC thicknesses should be interpreted also considering the age-related loss of RGCs, as previously reported.^68,88–91 It has been estimated that the mean age-related rate of GCL+IPL loss was −0.318 μm/year, whereas the inner retina (i.e., from the inner RNFL boundary to the outer IPL boundary) decreased by −0.245 μm/year. After accounting for age-related changes, the rate of progression decreased from 50% to 14.7% for the GCL+IPL thickness and from 50% to 20% for the inner retinal thickness.⁹⁰ Medeiros et al⁹² estimated rates of RGCs loss in progressive glaucoma by interpolating structural and functional parameters. The percent rate of RCG loss was −4.4%/year in eyes exhibiting faster RGC loss than the expected age-related decline. The percentage of eyes with faster RGC loss was higher (22.1%) than those progressing based on OCT average thickness (14.6%) or VF (8.5%).

Specific Patterns of Macular Ganglion Cell Complex Alterations

Macular GCC in Primary Open-Angle Glaucoma and Normal Tension Glaucoma

Most studies conducted on macular GCC alterations in glaucomatous eyes considered primary open-angle glaucoma (POAG).^{10,68,93–95} The GCC thickness gradually decreased, according to the severity of POAG, with widening preexisting defects.^95,96

In addition, some reports aimed to identify GCC structural parameters potentially able to discriminate between normal tension glaucoma (NTG) and POAG.^24,97,98 The hypothesis of a different pattern of macular GCC loss was initially supported by the evidence of deeper and closer-to-fixation scotomas in the NTG with respect to POAG.⁹⁹

The peripapillary RNFL demonstrated strong correlation with mGCC thickness in both early-stage NTG and with a VF defect.⁹⁸ Comparison of the NTG and POAG groups revealed that the POAG group presented lower parameters in both RFNL and macular GCC thickness with respect to the NTG subgroup.^97,100 More interestingly, the GLV parameter was significantly higher in the POAG subgroup. Considering these evidences together identified that the GCC loss was more diffuse in the POAG, with a more prominent difference than peripapillary RNFL, and the mean GCC measurements and GLV represented the most important parameters in discriminating between POAG and NTG.⁹⁷

Recently, Edlinger et al¹⁰¹ reported data on structural parameters in NTG and high-tension glaucoma with perimetric and preperimetric glaucoma. Eyes with high-tension glaucoma presented a more pronounced loss in thickness for all the parameters considered (peripapillary RNFL, and macular RNFL, GCL, IPL and GCC). Global RNFL, macular inner temporal GCL, outer temporal GCC, outer inferior RNFL and inner temporal IPL presented the highest sensitivities for discriminating perimetric glaucoma. Moreover, peripapillary RNFL and macular inner temporal GCL, outer inferior GCC, outer inferior RNFL and inner temporal IPL achieved the highest sensitivities for preperimetric glaucoma.

Diagnostic Ability of GCC to Detect Glaucoma in Highly Myopic Eyes

The diagnosis of glaucoma in highly myopic eyes can be challenging considering the optic disc and retina are deformed due to elongation of the eye.¹⁰² In addition, myopia is considered a risk factor for the development of glaucoma, but it remains controversial whether myopic eyes progress more quickly than emmetropic eyes.^103–108

Macular GCC thickness was reported to be comparable or superior to the RNFL evaluation in the assessment of glaucoma in highly myopic patients.^53,109–112

The diagnostic parameters with the largest AUC for the detection of glaucomatous eyes among myopic patients was GLV (AUC: 0.957), which performed significantly better than the cup/disc vertical (AUC = 0.844) and the RNFL vertical (AUC = 0.826), as reported by Shoji et al⁵³. Furthermore, considering that the average GCC demonstrated 0.968 specificity at 80% sensitivity, it has been hypothesized that it could be considered one of the best parameters for glaucoma detection in highly myopic eyes.

The diagnostic performances of GCC and peripapillary RNFL in detecting glaucomatous damage in highly myopic eyes were similar among the various OCT devices (Cirrus, RTVue and 3D OCT).¹¹¹

Other OCT parameters found to be effective in discriminating between normal and glaucomatous eyes were inferior GCC thickness among highly myopic patients and FLV in the non-myopic subgroup.¹¹⁰

Conclusion

Glaucoma is characterized by the progressive loss of RGCs and their axons constituting the RNFL. The macular region contains more than 50% of RGCs easily estimable via commercially available OCT devices.

Experimental animal models demonstrated a loss of RGCs occupying the foveal region, similar to other peripheral retinal areas, further suggesting the importance of studying this region in glaucomatous eyes. Such models demonstrated also that the RGCs loss is nonuniformly distributed across the retina, and thus, an understanding of the location and cellular types susceptible to glaucomatous damage may help in developing more accurate biomarkers in human eyes.

A significant reduction of RGCs can be appreciated before the VF defect becomes detectable. The GCC parameters (i.e., mean thickness, GLV, FLV, superior and inferior thickness) are superior in discriminating early glaucomatous changes compared with RNFL thickness, since it has been experimentally demonstrated that the RGC dysfunction occurs earlier than does RNFL thinning.

Macular GCC parameters can also be considered important prognostic indicators in cases of non-glaucomatous RNFL thinning characterizing, for instance, highly myopic eyes.

The diagnostic accuracy of GCC analysis increases with disease severity, but it is also directly related to an increase in signal strength values.

An additional point to consider when using GCC analysis is the relationship between GCC thinning and VF damage. Considering the anatomical RGCs displacement, the conventional SAP protocol has resulted in inaccurately mapping the macular glaucomatous damage. In this regard, the use of the 10–2 test appears to demonstrate greater structural-functional agreement. Despite this, the derived parameter of focal GCC loss (i.e., FLV) represents the strongest predictor of VF progression in glaucomatous eyes. In conclusion, the estimation of glaucoma-induced RGCs loss by using GCC parameters represents an ideal predictor in assessing early glaucoma and offers a more accurate tracking of glaucoma progression.

Disclosure

The authors report no conflicts of interest in this work.