Evaluation of ganglion cell-inner plexiform layer thickness in the diagnosis of pre-perimetric glaucoma and comparison to retinal nerve fiber layers

Abstract

Purpose:

Evaluation of ganglion cell-inner plexiform layer thickness in the diagnosis of pre-perimetric glaucoma (PPG) and comparison to retinal nerve fiber layers.

Methods:

This study was a prospective hospital-based study. A total of 30 PPG and control patients were studied for retinal nerve fiber layer thickness (RNFL) and ganglion cell-inner plexiform layer complex (GC-IPL) by spectral-domain optical coherence tomography. PPG was defined as eyes with a normal visual field and one or more localized RNFL defects that were associated with a typical glaucomatous disc appearance. Diagnostic abilities of GC-IPL, optic nerve head (ONH), and RNFL parameters were computed using area under receiver-operating curve (AUROC), sensitivity, and specificity.

Results:

GC-IPL parameters showed significant changes in PPG cases as compared to normal subjects in each region (P value < 0.001). RNFL parameters also differed significantly from normal subjects in all quadrants (P value 0.003 to < 0.001). Within GC-IPL parameters, the superotemporal region had the maximum area under the curve (AUC), followed by inferior, superior, and inferotemporal regions. Within RNFL parameters, the inferior quadrant had the maximum AUC, followed by superior and nasal quadrants. the GC-IPL parameters in PPG showed that the AUC of the GC-IPL parameters was much higher than those of the ONH and RNFL values.

Conclusion:

Although both the parameters RNFL and GC-IPL showed significant changes in PPG patients compared to healthy subjects, a higher AUC of GC-IPL points toward the higher sensitivity of GC-IPL than RNFL for detecting glaucoma in early stages.

The glaucoma is a collection of progressive optic neuropathies that can lead to irreversible damage to retinal ganglion cells (RGCs) and their axons and eventual loss of vision, if inadequately treated. In the early stages, the disease is largely asymptomatic and it is estimated that only half of glaucoma patients are aware that they have the disease. Clinically, glaucoma patients show characteristic optic nerve head (ONH) and retinal nerve fiber layer (RNFL) changes. The early detection of nerve fiber layer (NFL) changes is crucial for all patients with glaucoma. Once a visual field deficit is detectable, the disease has already caused irreversible visual loss.[1,2] Compared with the average number of RGCs in the healthy group, glaucomatous eyes had an average RGC loss of 28.4%.[3]

It has been shown that damage to the RNFL precedes visual field loss.[4] Thus, RNFL assessment has emerged as an important parameter for pre-perimetric diagnosis of glaucoma and may aid ophthalmologists in making an accurate and early diagnosis. Various imaging technologies have been introduced for an objective assessment of RNFL.[5,6] Spectral-domain optical coherence tomography (SD-OCT) with high resolution and reduced spectral noise can better objectively measure the optic disc, RNFL, and macular ganglion cell-inner plexiform layer complex (GC-IPL). Imaging systems can offer the prospect of detecting progression early so that changes to a patient’s treatment plan can be made to prevent visual impairment. Imaging tests also provide quantitative measurements of glaucoma relevant structures and can be compared from visit to visit. They are intended to provide information that can be used in conjunction with clinical examinations.[7,8]

Methods

This study was carried out from December 2021 to December 2022. Patients were enrolled at a tertiary care center, Prayagraj. The study was conducted in accordance with the ethical standards as approved by the institutional ethics review board after having obtained informed consent from all participants.

It was a prospective hospital-based, comparative study. The sample size was calculated using the formula for the comparison of means, discussed in ‘A. Indrayan, basic method in medical research’ by putting standard deviation. Considering the 95% confidence interval and 90% power of study, a total of 30 pre-perimetric glaucoma (PPG) patients and 30 normal subjects were enrolled in the study for the evaluation of RNFL and ganglion cell-inner plexiform layer patterns in PPG patients, which was designed to investigate the roles of RNFL and GC-IPL imaging for diagnosis and monitoring of PPG. PPG patients were defined as eyes with a normal visual field and one or more localized RNFL defects that were associated with a typical glaucomatous disc appearance. Inclusion and exclusion criteria are described in Table 1.

Table 1.

Inclusion and exclusion criterias

Inclusion Criteria	Exclusion Criteria
- Eyes with glaucomatous change in optic disc - Age >18 years, - best corrected visual acuity 6/12 or better.	- Patients unable to perform reliable perimetry - Glaucoma patients with visual field changes, - Patients with significant cognitive impairment, - History of intraocular surgery (except for uncomplicated cataract surgery or glaucoma surgery), - Systemic or ocular diseases other than glaucoma that are known to affect visual field, - Optical media opacities, -Patients with angle closure glaucoma,

All subjects underwent a full ophthalmic examination comprising visual acuity, Goldman applanation tonometry, gonioscopy, fundus examination with slit lamp biomicroscopy, oculus twinfield perimetry, and SD-OCT (Topcon 3D OCT-1 Maestro2).

OCT procedure

The GC-IPL analysis available on the Topcon 3D OCT-1 Maestro2 software measured the combined thickness of RNFL and GC-IPL in a 4.8 mm × 4.0 mm oval with a longer horizontal axis. It provided measurements in six wedge-shaped sectors after excluding the central foveolar region (1 mm in diameter) along with a pseudo-color scheme for the GC-IPL thickness. A deviation map also flagged abnormally thin areas as yellow (P < 5%) or red (P < 1%) superpixels. The parameters identified were average GC-IPL, minimum GC-IPL, and sector measurements (superonasal, superior, superotemporal, inferonasal, inferior, and inferotemporal). The optic disc cube 200 × 200 consisted of 40,000 axial scans (in 6 mm × 6 mm × 2 mm cube) centered on the optic disc. Average RNFL thickness and RNFL thickness in quadrants on a measurement circle 3.46 mm in diameter were calculated.

Statistical analysis

Statistical analysis was performed using IBM SPSS V 23.0 for window software. The results were analyzed using descriptive statistics and making comparisons among the various groups. Categorical data were summarized as in proportions and percentages (%), while discrete in mean and standard deviation (SD). The two groups were compared using unpaired t test and Chi square test. Receiver-operating characteristic (ROC) analysis curves were used to assess the overall diagnostic performance and to compare the performances of the two methods. The P value was taken significant when less than 0.05 (P < 0.05).

Observations and Results

The mean age of patients in PPG patients was 36.97 ± 13.30 yr, and that of healthy subjects was 36.37 ± 9.26 yr. The male to female ratio in PPG was 3:2.3, and that in control was 3:2.6. The cup-disc ratio and intra-ocular pressure were significantly higher in the PPG group [Table 2].

Table 2.

Demographic and clinical characteristics of study patients

Characteristics	Subgroups	Control	PPG		P
Age (Mean±SD)		36.37±9.26	36.97±13.30		0.840
Sex (No/%)	Male	16 (53.3%)	17	56.7%	0.795
	Female	14 (46.7%)	13	43.3%	0.796
IOP (Mean±SD)		17.48±1.96		19.93±3.91	<0.001
C:D Ratio (Mean±SD)	Vertical	0.38±0.07		0.59±0.12	<0.001
	Horizontal	0.34±0.06		0.58±0.12	<0.001

IOP (intra-ocular pressure), C:D ratio (cup-disc ratio), PPG (preperimetric glaucoma)

The mean RNFL thickness of superior, inferior, nasal, and temporal regions in PPG patients was 121.75 ± 19.83 (µm), 126.88 ± 18.44 (µm), 80.20 ± 13.56 (µm), and 66.52 ± 10.95 (µm), and in control subjects, it was 137.70 ± 6.65 (µm), 134.73 ± 7.56 (µm), 91.37 ± 8.68 (µm), and 85.38 ± 8.53 (µm), respectively. A significant difference was found in all quadrants in PPG patients in comparison to control subjects (P value 0.003 to < 0.001) [Table 3, Fig. 1].

Table 3.

Intergroup comparison of RNFL thickness

RNFL thickness (um)	Control		PPG		Unpaired t-test
	Mean	SD	Mean	SD	t	P
Superior	137.70	6.65	121.75	19.83	-5.91	<0.001
Inferior	134.73	7.56	126.88	18.44	-10.53	<0.001
Nasal	91.37	8.68	80.20	13.56	-5.37	<0.001
Temporal	85.38	8.53	66.52	10.95	-3.05	0.003

Figure 1.

SD-OCT peripapillary RNFL image of a PPG patient with slight changes in superotemporal RNFL thickness

The mean GC-IPL thickness of superotemporal, superior, superonasal, inferotemporal, inferior, and inferonasal regions in PPG patients was 67.68 ± 4.75 (µm), 67.80 ± 4.86 (µm), 71.77 ± 5.94 (µm), 69.28 ± 4.74 (µm), 64.57 ± 4.43 (µm), and 69.18 ± 6.05 (µm), and in control subjects, it was 89.23 ± 8.24 (µm), 87.02 ± 8.94 (µm), 86.30 ± 9.47 (µm), 87.60 ± 9.17 (µm), 85.85 ± 9.48 (µm), and 85.03 ± 10.09 (µm), respectively. A significant difference was found in all quadrants in PPG patients in comparison to control subjects (P value < 0.001) [Table 4, Fig. 2].

Table 4.

Intergroup comparison of GC-IPL thickness

GC-IPL thickness (um)	Control		PPG		Unpaired t-test
	Mean	SD	Mean	SD	t	P
Superotemporal	89.23	8.24	67.68	4.75	-17.55	<0.001
Superior	87.02	8.94	67.80	4.86	-14.62	<0.001
Superonasal	86.30	9.47	71.77	5.94	-10.06	<0.001
Inferotemporal	87.60	9.17	69.28	4.74	-13.74	<0.001
Inferior	85.85	9.48	64.57	4.43	-15.75	<0.001
Inferonasal	85.03	10.09	69.18	6.05	-10.44	<0.001

Figure 2.

SD-OCT macular GC-IPL image of the same patient with GC-IPL thinning in superotemporal and inferotemporal quadrants

Within RNFL parameters, the inferior quadrant had the maximum area under curve (AUC), followed by superior and nasal quadrants [Graph 1, Table 5]. Within GC-IPL parameters, the superotemporal region had the maximum AUC, followed by inferior and superior regions [Graph 2, Table 5]. The same results were observed at 2 months, 4 months, and 6 months of follow-up as there was no change in RFNL and GC-IPL thickness during follow-up.

Graph 1.

ROC curve for sensitivity and specificity of RNFL thickness parameters

Table 5.

AUCs, sensitivities, specificities (with 95% CI in parentheses), values of C: D ratio, RNFL, and GC-IPL parameters

Parameter	AUROC	Sensitivity	Specificity
C:D Vertical	0.935 (0.892-0.979)	90 (82.4-97.6)	86.7 (78.1-95.3)
C: D Horizontal	0.958 (0.925-0.992)	86.7 (78.1-95.3)	96.7 (92.2-101.2)
S - RNFLT	0.790 (0.702-0.878)	63.3 (51.1-75.5)	98.3 (95-101.6)
I - RNFLT	0.902 (0.847-0.956)	73.3 (62.1-84.5)	98.3 (95-101.6)
N - RNFLT	0.741 (0.651-0.830)	41.7 (29.2-54.2)	98.3 (95-101.6)
T - RNFLT	0.652 (0.547-0.758)	48.3 (35.7-60.9)	96.7 (92.2-101.2)
ST - GC IPLT	0.998 (0.994-1.000)	98.3 (95-101.6)	96.7 (92.2-101.2)
S - GC IPLT	0.994 (0.985-1.000)	95 (89.5-100.5)	98.3 (95-101.6)
SN - GC IPLT	0.942 (0.902-0.981)	93.3 (87-99.6)	83.3 (73.9-92.7)
IT - GC IPLT	0.991 (0.981-1.000)	100 (100-100)	86.7 (78.1-95.3)
I - GC IPLT	0.995 (0.988-1.000)	95 (89.5-100.5)	98 (94.5-101.5)
IN - GC IPLT	0.938 (0.900-0.977)	88.3 (80.2-96.4)	83.3 (73.9-92.7)

Graph 2.

ROC curve for sensitivity and specificity of GC-IPL thickness parameters

Discussion

The clinical RNFL evaluation has been recorded using a variety of different methodologies over the course of its history. Ophthalmoscopic exams that make utilization of red-free light and NFL photography have been utilized for a significant amount of time, but conducting and interpreting these assessments can be challenging. The slit-lamp biomicroscopic examination of the RNFL and the photographic evaluation of the RNFL are both subject to individual interpretation and are dependent on the examiner’s degree of expertise (and also the photographer, as in the case of RNFL photography).

The thickness of the NFL can also be measured with other methods, such as scanning laser polarimetry, which has been developed in recent years. This apparatus functions by using shifts in the polarization of light and retardation, both of which are influenced by the thickness of the NFL. It is unfortunate that the NFL is not the only structure in the eye that possesses a birefringent property. However, it is unknown how the effects of corneal and lenticular birefringence, as well as ageing, affect polarimetric observations. This is because the NFL is located in the central portion of the retina. Polarimetric readings are susceptible to error when peripapillary atrophy or chorioretinal scars are present. Peripapillary atrophy and chorioretinal scars can appear in eyes that are otherwise healthy or in eyes that have diseases other than glaucoma.[9]

The direct assessment of NFL thickness is made possible using OCT, which also provides cross-sectional viewing. The microstructure of the retina may be seen with a resolution of 10 µm when low-coherence light is delivered with a device that does not need touch or invasiveness. Measurements of the NFL obtained using OCT have been found to have a strong connection with histopathologic examinations.[10] With the newer SD-OCT, we get better resolution and quality of images and greater repeatability of measurements. This has led to renewed interest in macular measurements of GC-IPL for detection of glaucoma and existing PPG measurements, that is, RNFL. There are studies[11] that have shown that RNFL and GCC reduce with increased axial length and negative spherical equivalent; hence, we excluded eyes with refractive errors outside -6DS and +4DS and both the groups we had taken were age- and sex-matched.

RNFL Thickness

In our study, a statistically significant difference (P < 0.001) is seen in mean RNFL thinning in PPG patients in comparison to the control group. Maziar Lalezary et al.[12] compared normal subjects to glaucoma suspects, and they found a significant difference in RNFL thinning in their glaucoma suspects in comparison to normal subjects and 20% of their glaucoma suspects developed glaucomatous field changes during the course of 4.2 years.

In our study, the inferior quadrant showed the highest AUC and greater sensitivity for RNFL thinning within the RNFL parameters, followed by the superior, nasal, and temporal quadrants. Because laminar pores tend to be larger in superior and inferior areas of the lamina cribrosa and in the pore, fibers are closely packed and more in density, so these axons are more susceptible to compressibility.[13] Similar to our study, Christopher K. S. Leung et al.[14] reported that inferior peripapillary RNFL thickness had the best AUCs within RNFL parameters in discriminating glaucoma and glaucoma suspects. Renato Lisboa et al.[15] compared normal subjects to PPG subjects and found similar results as in our study. In this research, thinner retinal quadrants in the superior and inferior regions were predictive of glaucomatous changes. However, the nasal and temporal quadrants were not indicative of glaucomatous changes. According to Wang YX et al.,[16] RNFL parameter thinning was the greatest in the inferior quadrant, followed by the superior, temporal, and nasal regions.

GC-IPL thickness

More than 50% of all RGCs are concentrated and multi-layered in the macular area. The exact location is 4–5 mm from the center of the fovea,[17] and the density reaches its peak at 750–1100 µm from the foveal center. Curcio CA et al.[17] hypothesized that quantitative detection of glaucomatous damage at the macula using retinal thickness mapping may provide a unique method for the early detection and monitoring of early glaucomatous tissue loss.

In our study, a statistically significant difference (P < 0.001) in mean GC-IPL thinning was seen in PPG patients in comparison to the control group. A considerable difference in the macular GC-IPL thickness was observed by Na JH et al.[18] between healthy control eyes and eyes with PPG. They hypothesized that macular GCL-IPL thickness could serve as an early predictor of glaucomatous structural damage. This study has shown that the inferior, inferotemporal GC-IPLs are the regions that demonstrate early thinning and these 27% of PPG patients developed glaucomatous abnormalities during the course of 2.13 years.

In our study, within the parameters of the GC-IPL, the superotemporal region showed the highest AUC, followed by the inferior, superior, inferotemporal, superonasal, and inferonasal areas. The inferotemporal area was shown to have a higher level of sensitivity than the superotemporal, inferior, or superior region. However, the superonasal and inferonasal areas are not nearly as prominent as the other regions. Mwanza et al.[11] compared normal subjects to PPG patients, and they found similar results as in our study. They found the inferotemporal area of GC-IPL thickness was the most important factor. They also reported that the ability of GC-IPL parameters to diagnose early glaucoma was high and comparable to that of RNFL and ONH parameters. Kotera Y. Hangai et al.[19] reported that the inferortemporal region of GC-IPL thickness showed the greatest AUCs and hypothesized that the macular GC-IPL is useful for detecting early glaucomatous changes. In the research carried out by Huang et al.,[20] all GC-IPL parameters demonstrated a satisfactory level of glaucoma diagnostic abilities and the AUC for the average, superior, and inferior GC-IPL thickness decreased as the glaucomatous damage became more severe.

GC-IPL Versus RNFL

In our study, it has been demonstrated that the AUCs of GC-IPL parameters have more diagnostic value than the values of RNFL parameters. When it comes to the diagnosis of PPG, the GC-IPL thickness has a higher sensitivity than the RNFL. Deshpande GA et al.[21] found similar results as in our study; they compared the control group to the PPG group and reported that AUCs of GC-IPL parameters were significant compared to RNFL parameters (P < 0.001). However, studies by Seong et al., Rao et al., and Kim et al.[22,23,24] found GC-IPL comparable to RNFL. In contrast to our study, K. NouriMahdavi et al.[25] reported that RNFL thickness measurements were superior to GC-IPL thickness values for detection of early glaucoma according to the AUC values. Another study by Viquar U. Begum et al.[26] found that the diagnostic ability of GC-IPL parameters was similar to that of the ONH and peripapillary RNFL parameters in perimetric glaucoma. However, in PPG, the diagnostic ability of GC-IPL parameters was significantly lower than that of the ONH and RNFL parameters (P < 0.001).

As the OCT facilities are increasing in India, we recommend to perform the OCT-based diagnostic test in cases of glaucoma suspects and also emphasize on the study with a larger sample size with longer duration to decrease the burden of irreversible blindness due to glaucoma. The limitation of our study was the small sample size, and so, large population-based studies are required for generalization of the results.

Conclusion

Confirmed diagnosis of glaucoma is made on characteristic disc changes and corresponding visual field defects at a time when 28–30% retinal nerve fibers have already been damaged. Newer SD-OCTs have the potential to provide objective and quantitative measurements of the thickness of the GC-IPL and the RNFL, and both are necessary for the early diagnosis and monitoring of PPG. GC-IPL with a higher AUC curve has greater sensitivity than RNFL in detecting PPG.

Financial support and sponsorship:

Nil.

Conflicts of interest:

There are no conflicts of interest.