Correlation of peripapillary retinal nerve fiber layer and macular ganglion cell–inner plexiform layer in early to moderate glaucoma using the Cirrus^® widefield analysis (PanoMap^®)

Abstract

Purpose:

To evaluate and correlate the structural changes between peripapillary retinal nerve fiber layer (RNFL) and macular ganglion cell layer (GCL) + inner plexiform layer (IPL) in different stages of glaucoma using PanoMap^® optical coherence tomography (OCT).

Design:

Retrospective observational study.

Methods:

Glaucoma diagnostic test data were collected from early to moderate open-angle glaucoma patients. The average and minimum GCL + IPL thickness, sectoral GCL + IPL thickness, and the average and sectoral RNFL thickness were correlated with the different glaucoma stages.

Results:

This study included 157 eyes from 157 glaucoma patients. Patients were grouped into pre-perimetric, early, and moderate glaucoma. The mean average RNFL thickness, RNFL thickness per sector, average GCL + IPL thickness, and minimum GCL + IPL thickness were different between the three groups (P < 0.001), except for the nasal sector (P = 0.643). The mean GCL + IPL thickness in all six sectors showed differences between the groups (P < 0.001), except the superonasal sector (P < 0.002). The inferior GCL + IPL sector is the thinnest, followed by the inferotemporal sector. There was a strong correlation between the mean average RNFL and the average GCL + IPL thickness in the pre-perimetric group (r = 0.4963, P < 0.001) and the moderate group (r = 0.6534, P < 0.001). The early glaucoma group did not show significant correlation (r = 0.2963, P = 0.0536).

Conclusion:

Peripapillary RNFL and macular GCL + IPL thinning was evident in different stages of glaucoma, with more thinning observed with increasing glaucoma severity. The peripapillary RNFL and macular GCL + IPL average thickness values were highly correlated in the pre-perimetric and moderate stages of glaucoma.

Spectral domain-optical coherence tomography (SD-OCT) has been widely accepted as a diagnostic tool for detecting and monitoring glaucoma, particularly during early to moderate stages.[1,2,3,4] SD-OCT studies have shown the importance of the inner macular layer thickness in detecting early glaucoma structural damage. The macular ganglion cell layer (GCL) contains more than 50% of the retinal ganglion cells. The inner macular layers are composed of the retinal nerve fiber layer (RNFL) and GCL, which are structurally similar to the peripapillary region. The ganglion cell complex is defined as the three innermost macular retinal layers: RNFL, GCL, and the inner plexiform layer (IPL).[2] The ganglion cell analysis algorithm of the Cirrus HD OCT (Carl Zeiss Meditec Inc., Dublin, CA, USA) automatically performs segmentation of the macular GCL + IPL layers without including RNFL, giving a more precise measurement of the inner macular region.[5]

The diagnostic ability of macular GCL + IPL thickness parameters in early glaucoma has been proven in different studies.[5,6,7] Early glaucoma affects the macular region, which may lead to undetected central visual field (VF) sensitivity losses.[6] The average and minimum GCL + IPL thickness had higher sensitivity and specificity in diagnosing early versus late glaucoma.[5] Mwanza et al.[7] demonstrated that the minimum GCL + IPL thickness was best for diagnosing early glaucoma in both pre-perimetric and perimetric groups. This study also showed that the inferotemporal and inferior macular GCL + IPL thickness sectors were the most sensitive in diagnosing early glaucoma.

Some studies have established significant correlation between GCL + IPL and RNFL parameters in detecting structural damage in glaucoma. These studies have shown that both structural parameters are comparable of detecting early glaucoma and have recommended combining both to diagnose glaucoma.[1,3,4,8,9,10,11,12] Currently, there is an emerging interest of observing both areas in diagnosing and monitoring glaucoma.

A new feature of the Cirrus^® SD-OCT is integrating both the optic disc cube and the macular cube scans into one. This combined widefield RNFL and macular thickness and deviation map (PanoMap^®) can help in easier correlation between RNFL and macular GCL + IPL changes and, hopefully, simpler identification of patterns of glaucomatous changes.[1] Recent PanoMap^® studies focused on baseline damage and progression in early glaucoma (pre-perimetric and perimetric).[1,2,8,9] Baseline glaucomatous structural damage was observed more frequently in the macular areas versus the peripapillary areas in early glaucoma.

Our study aims to provide additional data on the structural changes between peripapillary RNFL and GCL + IPL observed in different stages of glaucoma patients using Cirrus^® SD-OCT widefield analysis (PanoMap^®). Only a few studies have shown the significance of using this recently new widefield analysis feature in determining structural changes across different stages of glaucoma. The importance of a juxtaposed analysis of both structural parameters can give a more complete picture of damage that occurs at different stages of glaucoma.

Our study specifically aimed to determine and correlate the presence of peripapillary RNFL and macular thinning in various stages of glaucoma by measuring average RNFL thickness, RNFL thickness per quadrant, average and minimum GCL + IPL thickness, and GCL + IPL thickness per sector.

Methods

Study design and population

This was a retrospective observational study with data collected from glaucoma diagnostic tests done at a tertiary hospital from January 2020 to September 2022 in patients diagnosed with early to moderate open-angle glaucoma. Both eyes of patients with open-angle glaucoma were analyzed.

Glaucoma diagnostic tests were composed of optic disc stereo-disc photography, red free RNFL photography (Visucam; Carl Zeiss Meditec Inc., Dublin, CA, USA), Swedish interactive thresholding algorithm 30-2 or 24-2 perimetry (Humphrey Field Analyzer II; Carl Zeiss Meditec, Jena, Germany), and Cirrus HD-OCT (Carl Zeiss Meditec Inc., Dublin, CA, USA).

Patients with open-angle glaucoma were clinically identified by the presence of a glaucomatous optic disc (i.e., localized or diffuse neuroretinal rim thinning and increased cupping or cup-to-disc ratio difference >0.2 between the eyes) on stereo-disc photography, the presence of RNFL defect on red free fundus imaging regardless of the presence or absence of glaucomatous VF defects, an open-angle confirmed on gonioscopy, and a normal or increased intraocular pressure (IOP) measured by applanation tonometry.

Patients with open-angle glaucoma were staged according to the Hodapp Anderson Parrish criteria as follows: mean deviation (MD) ≥−6 dB for the early glaucoma group, MD of >−6 dB, but less than <−12 dB for the moderate glaucoma group (Hodapp Anderson Parrish criteria), and pre-perimetric glaucoma without glaucomatous VF defects or MD <−2 dB.[13]

Glaucomatous VF defects were defined as follows: (1) glaucoma hemifield test values outside the normal limits; (2) three or more abnormal points with a <5% probability of being normal, among which at least one point had a pattern deviation probability of <1%; and (3) a pattern standard deviation (SD) probability of <1%.[14] VF defects were confirmed using reliable tests (fixation loss rate ≤20% and false-positive and false-negative error rates ≤20%).

The patient’s data sheets from the institution’s eye institute databank of glaucoma diagnostic tests contained information on visual acuity, ocular diseases such as cataract or any media opacities, history of trauma, history of diabetes or hypertension, and past ocular surgeries and laser treatment. The clinical parameters such as IOP and gonioscopy findings were obtained from the patient records of the requesting physicians who hold clinics at the tertiary hospital. Permission was obtained from the requesting physicians to access their patients’ records. Patients with missing data were excluded from the study. This study was approved by the Institutional Scientific Review Committee and the Institutional Ethics Review Committee of the institution.

Inclusion and exclusion criteria

Patients fulfilling the following criteria were included in the study: age more than 18 years, best corrected visual acuity (BCVA) of 20/40 and better, diagnosis and staging of open-angle glaucoma based on clinical examination with clinical exam findings of normal and increased IOP and open angles based on applanation tonometry and gonioscopy and the following glaucoma diagnostic tests: optic nerve stereophotography, red free photograph, OCT of optic nerve head, peripapillary RNFL, GCL–IPL (PanoMap^®) [Fig. 1], and VF examination to be interpreted by a glaucoma specialist (staging to follow Hodapp–Anderson classification – see the “Methods” section), and no significant cataract (above NO2NC2 by Lens Opacity Classification System). The following patients were excluded from the study: patients with other ocular or systemic diseases that may affect the optic nerve function and VFs such as non-glaucomatous optic neuropathies due to neuro-ophthalmologic or retinal diseases, open-angle glaucoma caused by other conditions such as diabetes, inflammation, or postsurgery, patients with any media opacity that would significantly interfere with acquisition of OCT images and inability to obtain a high-quality OCT image (i.e. signal strength [SS] <6), patients with severe glaucoma showing an MD worse than –20 dB and with unreliable VF results (fixation loss rate ≥20% and false-positive and false-negative error rates ≥20%). Patients with history of trauma, complicated intraocular surgery, and previous glaucoma filtering surgery and glaucoma laser procedures were excluded.

Figure 1.

Zeiss Cirrus HD OCT of ONH and RNFL (left) and PanoMap^® analysis (right). OCT = optical coherence tomography, RNFL = retinal nerve fiber layer, ONH = optic nerve head

Statistical analysis

Sample size was calculated based on the F-test analysis of variance (ANOVA) one-way analysis. The total sample size calculated is 125 eyes for all groups or 41–42 eyes per group (pre-perimetric glaucoma, early glaucoma, and moderate glaucoma groups).

Descriptive statistics such as mean and standard deviation were used to present continuous variables, while frequency and percentage were used for categorical data. In comparing the three groups, one-way ANOVA was used for continuous data, while Chi-square test was used for categorical data. Pearson r coefficient was utilized in determining correlation. All statistical tests were performed using MedCalc Statistical software version 20.113, with the level of significance at 5% (α = 0.05).

Results

This study included a total of 157 eyes from 157 glaucoma patients. Sixty-one eyes had pre-perimetric glaucoma, 52 eyes had early glaucoma, and 44 eyes had moderate glaucoma. Older patients belonged to the early and moderate glaucoma group and younger patients belonged to the pre-perimetric glaucoma group. The average VF MD was worse in more advanced glaucoma [Table 1].

Table 1.

Demographic and patient characteristics

	Pre-perimetric glaucoma (n=61)	Early glaucoma (n=52)	Moderate glaucoma (n=44)	P
Age (Mean+SD)	58.4±12.8	66.2±7.8	66.2±9.8	0.001*
Males (%)	27 (44.3)	20 (38.5)	18 (40.9)	0.821
Average MD (Mean±SD)*	-0.52±1.11	-4.14±1.41	-8.51±1.77	0.001*

MD=mean deviation, SD=standard deviation. *Significant

The mean SS for both GCL + IPL and RNFL was similar among the three groups; however, SS in the pre-perimetric group was statistically significantly higher than in the early and moderate groups. The mean average RNFL thickness per group was significantly different (P < 0.001). There were differences in the mean RNFL thickness in all sectors between the three groups (P < 0.001 for the superior and inferior sectors, P < 0.004 for the temporal sector), except for the nasal sector (P = 0.643). In the pre-perimetric group, the nasal sector (65.9 ± 10.8 µm) was the thinnest sector compared to the other sectors. The temporal sector was the thinnest sector compared to the other sectors in both the early and moderate groups (63 ± 12 and 60 ± 13.3 µm, respectively) [Table 2].

Table 2.

GCL–IPL and RNFL parameters (mean±SD) between different glaucoma stages

	Pre-perimetric glaucoma (n=61)	Early glaucoma (n=52)	Moderate glaucoma (n=44)	P
Peripapillary RNFL
Signal strength	9.0±0.8	8.6±0.9	8.7±1	0.035*
Average RNFL thickness (µm)	84.4±10.5	78.4±9.5	73±12.4	0.001*
Superior (µm)	101.3±16.1	95.2±16.3	87.2±18	0.001*
Inferior (µm)	102.5±19.8	89.9±17.8	80.5±21.6	0.001*
Temporal (µm)	68.4±13.5	63±12	60±13.3	0.004*
Nasal (µm)	65.9±10.8	65.5±12	63.8±13.2	0.643
GCL+IPL
Signal strength	9.3±0.8	8.8±0.9	9.1±1	0.024*
Average GCL + IPL thickness (µm)	75±6.6	71.5±7.3	66.8±8	0.001*
Minimum GCL + IPL thickness (µm)	69.9±9.7	62.6±12.1	56.5±12.2	0.001*
Superior (µm)	74.9±7.1	73.9±10	68.1±10.5	0.001*
Superior temporal (µm)	76.3±7.4	72.2±10.5	67.8±10	0.001*
Superior nasal (µm)	77.6±9.8	76.6±8.1	70.9±11.5	0.002*
Inferior (µm)	71.4±7.9	66.3±9.5	62.6±10.9	0.001*
Inferior temporal (µm)	74.6±7.9	67.3±11.6	64.6±13.4	0.001*
Inferior nasal (µm)	75.3±8.7	72.9±9.8	67.2±9.6	0.001*

GCL+IPL=ganglion cell layer+inner plexiform layer, RNFL=retinal nerve fiber layer, SD=standard deviation. *Significant

The mean average GCL + IPL thickness and the mean minimum GCL + IPL thickness showed statistically significant differences between the three groups (P < 0.001). The mean GCL + IPL thickness in all six sectors showed statistically significant differences between each group (P < 0.001 in all sectors, except the superonasal sector = P < 0.002). Pairwise comparison of the mean GCL + IPL thickness between groups in each sector showed that the moderate group was significantly thinner than the early and pre-perimetric groups in the superior, superior nasal, and inferior nasal sectors, while the mean thickness values of the pre-perimetric and early groups were the same in the mentioned sectors. Meanwhile, the mean thickness in the pre-perimetric group was significantly higher than the other two groups in the superior temporal, inferior, and inferior temporal sectors, while early and moderate groups had the same mean in the mentioned sectors. The inferior GCL + IPL sector was the thinnest sector in all groups (71.4 ± 7.9 µm, pre-perimetric; 66.3 ± 9.5 µm, early; and 62.6 ± 10.9 µm, moderate), followed by the inferotemporal GCL + IPL sector in all groups (74.6 ± 7.9 µm, pre-perimetric; 67.3 ± 11.6 µm, early; and 64.6 ± 13.4 µm, moderate).

There was a strong positive correlation between the mean average RNFL and GCL + IPL thickness in the pre-perimetric and moderate glaucoma groups. The early glaucoma group had no significant correlation [Table 3].

Table 3.

Correlation between mean average RNFL and GCL+IPL thickness for different glaucoma stages

	Pearson r	P
Pre-perimetric glaucoma (n=61)	0.4963	0.0001*
Early glaucoma (n=51)	0.2693	0.0536
Moderate glaucoma (n=44)	0.6534	0.0001*

GCL=ganglion cell layer, IPL=inner plexiform layer, RNFL=retinal nerve fiber layer. *Significant

Discussion

This study demonstrated statistically significant differences in the peripapillary RNFL and macular GCL + IPL thickness parameters across the three glaucoma stages [Table 2]. Measurements of both the peripapillary RNFL and macular GCL + IPL parameters significantly decreased corresponding to increasing glaucoma severity, which was also observed in many other studies.[3,5,15,16] Aside from measuring the traditional peripapillary RNFL, the macular GCL + IPL thickness parameters could also be used to indicate severity of glaucoma. In our study, the inferior GCL + IPL sector followed by the inferotemporal sector were the thinnest among glaucoma patients, regardless of severity, similar to that reported in many other studies.[1,2,3,6,7,8,9,17]

A hospital-based study by Cotaoco et al.,[15] which also used a Cirrus SD-OCT, observed that the mean average RNFL thickness in glaucoma Filipino subjects without VF defects was 82.53 µm. Similarly, the average RNFL thickness of our pre-perimetric glaucoma subjects was 84.4 ± 10.5 µm. Moreover, the average and minimum GCL + IPL thickness values of our pre-perimetric group (75 ± 6.6 and 69.9 ± 9.7 µm, respectively) were lower than the average and minimum GCL + IPL values observed in the control groups in the studies of Zivkovic et al.[5] (86.01 ± 3.68 and 82.49 ± 1.80 μm, respectively) and Mwanza et al.[17] (82.1 ± 6.2 and 80.4 ± 6.4 µm, respectively). Unfortunately, the differences or similarities in values cannot be significantly confirmed because our study did not include an age-matched and population-matched normal control group.

Mwanza et al.[17] showed a strong correlation between the average RNFL thickness and the average GCL + IPL thickness. RNFL thickness was the strongest factor determining GCL + IPL thickness with a correlation coefficient (R²) of 0.42 (P < 0.001) among other factors compared, such as age (R² 0.14, P < 0.001), axial length (R² 0.14, P < 0.001), and rim area (R² 0.11, P < 0.001). Based on the regression analysis done, for every micrometer decrease in RNFL thickness, there is a 0.35 µm decrease in GCL + IPL thickness. This supports the anatomical fact that axons emerging from the retinal ganglion cells in the macula converge into the optic nerve head. However, regarding the temporal relationship, it is still unknown as to whether retinal ganglion cell loss precedes RNFL loss or vice versa.

Our study concurs with Mwanza et al.,[17] wherein a strong relationship was also shown between the average GCL + IPL and the average peripapillary RNFL thickness in both pre-perimetric and moderate glaucoma groups (r = 0.4963 and r = 0.6534, P < 0.001 respectively). This implies that a thinner or lower value of RNFL will also give a lower GCL + IPL thickness. On the contrary, however, these parameters in the early glaucoma group did not show any significant relationship. Thus, there can still be instances where both these structural parameters fail to exhibit a relationship with one another. Inconsistencies have been attributed to various reasons, namely, differences in the scope or area scanned for each parameter and concurrent retinal or macular diseases that may affect one parameter than the other, to name a few.[7]

Our study had some limitations. We did not have a normal control group, which would have been an important factor in determining if there was a significant difference between the RNFL and GCL + IPL values of the pre-perimetric group and a normal age-matched group in a homogenous population. Only assumptions have been made regarding peripapillary RNFL and macular GCL + IPL thinning of the pre-perimetric group when compared to normal values. Since this is a retrospective study spanning 2 years of data on the PanoMap^®, temporal sequence and spatial relationship of both structural parameters in glaucoma progression were not assessed.

It is also important to note that ocular diseases, particularly retinal and macular diseases, were excluded from the study, as they may affect the values obtained for RNFL thickness and, most importantly, macular GCL + IPL thickness. It may be inappropriate to give a universal conclusion for all glaucoma patients based on the results of this study. Results may differ for glaucoma patients with confounding ocular diseases. Our study included patients with a BCVA of 20/40, but did not include a specific range of refraction in the criteria. Studies have shown that increase in axial length affects the glaucoma diagnostic ability of both structural parameters.[5,6,17] Inclusion criteria in most studies include patients only with refraction less myopic than −4 D to potentially remove confounding results for both parameters. We recommend this to be added in the inclusion criteria in future studies.

Conclusion

In this first PanoMap^® study involving Filipino eyes, peripapillary RNFL and macular GCL + IPL thinning was evident in different stages of glaucoma, with more thinning observed with increasing glaucoma severity. The peripapillary RNFL and macular GCL + IPL average thickness values were highly correlated in the pre-perimetric and moderate stages of glaucoma.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.