Visualization of Focal Thinning of the Ganglion Cell–Inner Plexiform Layer in Patients with Mild Cognitive Impairment and Alzheimer’s Disease

Abstract

Background:

A detailed analysis of the tomographic thickness of intraretinal layers may provide more information on neurodegeneration in patients with mild cognitive impairment (MCI) and Alzheimer’s disease (AD).

Objective:

The goal was to analyze tomographic thickness patterns of intraretinal layers in patients with AD andMCI.

Method:

Forty-nine patients (25 AD and 24 MCI) and 21 cognitively normal (CN) controls were imaged using ultra-high-resolution optical coherence tomography to obtain volumetric data centered on the fovea. The segmented intraretinal layers were retinal nerve fiber layer (RNFL), ganglion cell– inner plexiform layer (GCIPL), inner nuclear layer (INL), outer nuclear layer (ONL), outer plexiform layer (OPL), and retinal photoreceptor (PR), in addition to the total retinal thickness(TRT).

Results:

The thickness differences were negative (thinning) mainly in TRT, RNFL, and GCIPL in both AD and MCI groups in comparison to CN, while the thickness differences were positive (thickening) mainly in ONL and PR in AD. GCIPL of AD and MCI was thinner in superior, nasal superior, and temporal superior quadrants, compared to CN (p < 0.05). GCIPL of the inner superior, inner nasal superior, inner temporal superior, and outer nasal superior sectors was significantly thinner in AD than CN (p < 0.05). GCIPL of the outer superior, inner temporal superior, outer nasal, and temporal superior sectors was significantly thinner in MCI than CN (p < 0.05).

Conclusion:

Focal thinning of the GCIPL was visualized and quantified by detailed partitions in AD and MCI, which provides specific information about neurodegeneration in MCI and AD.

Keywords

Alzheimer’s disease ETDRS partition ganglion-inner plexiform layer hemispheric partition mild cognitive impairment retinal thickness mapping ultrahigh-resolution OCT Zeiss elliptical partition

INTRODUCTION

Alzheimer’s disease (AD), the most common form of dementia, has a prolonged preclinical period [1]. Neurodegeneration occurs decades before cognitive function declines. Mild cognitive impairment (MCI) is the transitional phase between a cognitively normal (CN) status and AD during which the individuals have cognitive function decline but are still able to maintain their independent living, and about 15% of patients with MCI convert to AD each year [2]. Currently, there are no effective treatments to cure AD, or even slow the disease progression to prevent the transformation from MCI to AD [1]. Preventing AD onset and progression at the preclinical or MCI phase is crucial to reduce the burden of AD.

Monitoring neurodegeneration in the central nervous system (CNS) provides direct information about disease progression [3]. Diffuse cerebral atrophy is associated with an advanced disease stage, whereas hippocampal and entorhinal cortex damage is often present at the early AD stage [3]. Evaluating brain atrophy using brain magnetic resonance imaging (MRI) or positron emission tomography (PET) scans is expensive and time consuming and sometimes invasive [4]. Noninvasive and cost-effective biomarkers that could characterize the neurodegeneration at the preclinical and MCI stages from normal aging are urgently needed [5].

The retina is an extension of the brain that shares similar embryological, anatomical, and physiological characteristics [6]. The transparent ocular media allows for assessment of neurodegeneration in vivo. Optical coherence tomography (OCT), a noninvasive in vivo optical “biopsy” and cross-sectional imaging modality, allows for visualization and quantitative analysis of intraretinal layers, including retinal nerve fiber layer and ganglion cells [7]. Previous OCT studies documented retinal ganglion cell loss manifested as average thinning of the ganglion cell-inner plexiform layer (GCIPL) in patients with MCI and AD [6 , 9].

Detailed analysis and visualization of the thinning patterns of intraretinal layers may provide more information for better understanding the characteristics in retinal neurodegeneration. Since the ganglion cells are not evenly distributed [10, 11], the average thickness of GCIPL may not be sensitive enough to show early neurodegeneration. In addition, patterns of retinal ganglion cell (RGC) damage in various neurodegenerative disorders are different [12]. In patients with AD, there is more evidence of magnocellular damage, which is reflected by a preferential RNFL thinning in the superior and inferior quadrants [12, 13]. We hypothesize that the tomographic thicknesses of intraretinal layers can be visualized and characteristic alteration patterns can be identified in patients with AD and MCI. The goal of this project was to visualize the tomographic thickness patterns of intraretinal layers in patients with MCI and AD and to compare them to normal controls.

METHODS

The study was approved by the research review board of the University of Miami and conducted in accordance with the Declaration of Helsinki. All subjects signed informed consent forms. Twenty-five AD and 24 MCI patients (Table 1) were recruited from the neuro-ophthalmology clinic at Bascom Palmer Eye Institute. They were referred from the Division of Cognitive Disorders at the Department of Neurology, University of Miami. A group consensus conference that included neurologists, psychiatrists, and neuropsychologists discussed and confirmed the diagnoses of AD [14] and MCI [15] based on the National Institute on Aging– Alzheimer’s Association (NIA-AA) criteria. All the recruited patients had brain MRI images which ruled out cerebrovascular disorders including lacunar strokes. The disease duration was calculated from the date of symptom onset. In addition, 21 cognitively normal (CN) control subjects with a similar range of ages as the patients in the AD and MCI groups were recruited from the Bascom Palmer Eye Institute. Those controls were either hospital staff members or patients who came for an annual eye examination.

Table 1

Demographic characteristics of the recruited subjects

	AD	MCI	CN	p
Subject	25	24	21
Eye	25	24	21
Age (y)	74±8	69±8	68±7	^*p < 0.05
Gender (M:F)	13:12	14:10	10:11	NS (χ² test)
Duration (y)	4±2	3±3	NA	NS (t-test)
MMSE	22±4	28±2	29±1	p < 0.05^*
SBP (mmHg)	132±18	129±17	140±20	NS (ANOVA)
DBP (mmHg)	79±11	81±10	82±11	NS (ANOVA)
HR (beats per minute)	64±13	69±11	62±11	NS (ANOVA)
BCVA	20/20	20/20	20/20
Color Vision	8/8	8/8	8/8
Confrontation Visual field	Full	Full	Full
Diabetes	3	3	2	NS (χ² test)
Hypertension	16	12	10	NS (χ² test)
Dyslipidemia	12	11	9	NS (χ² test)
Smoking	0	0	0

Results are presented as the mean±standard deviation. AD, Alzheimer’s disease; BCVA, best corrected visual acuity; DBP, diastolic blood pressure; F, female; HR, heart rate; M, male; MCI, mild cognitive impairment; MMSE, Mini-Mental State Examination; NA, no application; NS, not significant; p: p-value from statistical test; SBP, systolic blood pressure; SD, standard deviation; y, years. ^*p < 0.05 AD versus control and AD versus MCI. χ² test, Chi-squared test.

All subjects met the following criteria: 1) refraction diopters spherical – 6.00 to +6.00 diopters; and 2) normal eye without any ocular diseases, such as glaucoma, optic neuritis, or retinal degeneration. The exclusion criteria were as follows: 1) any ocular disorders, such as glaucoma, ocular infection, inflammation, ischemic diseases, retinal detachment, severe myopia, retinal degeneration, or diabetic retinopathy; 2) uncontrolled diabetes and hypertension or a history of myocardial infarction or stroke; or 3) any other systemic inflammatory or malignant disorders.

All subjects underwent eye exams, including best-corrected visual acuity (BCVA), color vision, intraocular pressure (IOP), and slit lamp biomicroscopy. The CN control subjects underwent Mini-Mental State Examination (MMSE) by the neurologists or trained research associates. Subjects’ medical and family histories and clinical characteristics, including age, gender, MMSE, and disease duration were recorded.

The custom built UHR-OCT device has been extensively described elsewhere [10]. Briefly, the system is a spectral domain OCT with an axial resolution of ∼3μm (in tissue) and scan speed of 24,000 A-scans per second. The three-dimensional volume in an area of 6×6 mm² centered on the fovea was acquired using a 512×128 macular cube protocol with 128 consecutive B scans and 512 A-scans per B scan [10]. The enface views of all scans were visually inspected and scans with blinking and excessive eye movement in the lateral direction were removed. One eye of each subject was imaged. The right eye was the first choice and the left eye was chosen if the right eye was not eligible.

The dataset was processed using automatic image processing software (Orion, Voxeleron LLC, Pleasanton, CA, USA) to segment 6 intraretinal layers (Fig. 1). In addition, interpolation using data points in diagonal meridians and inner, middle and outer circles of the dataset was automatically performed for axial motion compensation (http://www.voxeleron.com). The segmented intraretinal layers were retinal nerve fiber layer (RNFL), ganglion cell– inner plexiform layer (GCIPL), inner nuclear layer (INL), outer nuclear layer (ONL), outer plexiform layer (OPL), and retinal photoreceptor (PR), in addition to the total retinal thickness (TRT). The thickness map was further analyzed using ETDRS and hemispheric partition definitions (Fig. 2) [10]. In addition, all data on the thickness and coordinates of the center of the fovea were exported. To create the average thickness map of each group, the center of the fovea was aligned and the thickness in each pixel of the 512×128 pixels was averaged in the group, resulting in the average thickness maps. The thickness differentiation maps were calculated by subtracting the thickness of the control group from the AD or MCI groups in each pixel. By convention, the right eyes were used in all figures. Data from the left eyes were left– right flipped to obtain mirror-image maps. These maps were averaged and analyzed together with data from righteyes.

Fig.1

Cross-sectional retina and segmented tomographic thickness maps of intraretinal layers. The retina (A) of a healthy subject was scanned using ultrahigh-resolution optical coherence tomography (UHR-OCT) and 6 intraretinal layers (B) were segmented using the Orion software. There are seven segmented boundaries, defining six intraretinal layers, which correspond to six tomographic thickness maps. RNFL, retinal nerve fiber layer; GCIPL, ganglion cell– inner plexiform layer; INL, inner nuclear layer; ONL, outer nuclear layer; OPL, outer plexiform layer; and PR, retinal photoreceptor. Bar = 250μm. Scale unit =μm.

Fig.2

Partition methods for analyzing the tomographic thickness of the retina. The partition methods were used to define sectors of the intraretinal layers for calculating the tomographic thickness. UHR-OCT measured the thickness within a 27.48 mm² annulus (from 1 to 6 mm in diameter). In the ETDRS definition, the quadrantal division was performed using 45° and 135° medians (A and B). The central 1 mm zone of the fovea was removed. In addition, three concentric rings with diameters of 1, 3, and 6 mm were used to divide the map into nine zones (B). The inner and outer annuli were divided into four quadrants for each annulus. In the hemispheric definition, the division was performed using vertical and horizontal medians (C and D). I, inferior; II, inner inferior; IN, inner nasal; INI, inner nasal inferior; INS, inner nasal superior; IS, inner superior; IT, inner temporal; ITI, inner temporal inferior; ITS, inner temporal superior; N, nasal; NI, nasal inferior; NS, nasal superior; ON, outer nasal; OI, outer inferior; ONI, outer nasal inferior; ONS, outer nasal superior; OS, outer superior; OT, outer temporal; OTI, outer temporal inferior; OTS, outer temporal superior; S, superior; T, temporal; TI, temporal inferior; TS, temporal superior; and UHR-OCT: ultrahigh-resolution optical coherence tomography.

Continuous variables are presented as the mean±standard deviation (SD). Thickness in annulus, quadrant and sectors were averaged in each group. Multiple linear regressions were used to test for age-adjusted mean difference and trend among the groups. Spearman rank-order correlation was used to evaluate the relationship among the parameters (MMSE and disease duration) which were not normally distributed, and the Spearman’s correlation coefficient (ρ) was reported. The chi-Square test was used to assess associations for categorical variables including gender and potential confounding factors. p < 0.05 was considered statistically significant. Thickness maps were rendered in Matlab (Ver. 2014b, MathWorks, Natick, MA,USA).

RESULTS

Twenty-five AD and 24 MCI patients together with 21 CN controls were recruited. The detailed subjects’ characteristics are summarized in Table 1.

By aligning with the center of the fovea detected by the Orion software, the thickness map of each intraretinal layer was averaged in the AD, MCI and CN subjects, showing the differences in the thickness maps in the AD and MCI groups, compared to the CN group (Fig. 3). By subtracting the thickness of the CN group from that of the AD group (AD – CN) and that of the MCI group (MCI – CN), the differences in thickness maps of all intraretinal layers were rendered for visualization (Fig. 4). The thickness differences were negative (indicating thinning) mainly in TRT, RNFL, and GCIPL in both the AD and MCI groups. In contrast, the thickness differences were positive (indicating thickening) mainly in ONL and PR in AD group.

Fig.3

Averaged thickness maps of intraretinal layers. By aligning with the center of fovea detected by the Orion software, the thickness map of each intraretinal layer was averaged in AD (n = 25), MCI (n = 24), and CN control (n = 21) subjects, showing the changes in the thickness maps in AD, compared to the control group. AD: Alzheimer’s disease; CN: cognitively normal control; GCIPL: ganglion cell– inner plexiform layer; INL: inner nuclear layer; MCI: Mild cognitive impairment; ONL: outer nuclear layer; OPL: outer plexiform layer; PR: photoreceptor complex; RNFL: retinal nerve fiber layer; and TRT: total retinal thickness. Scale unit =μm.

Fig.4

Alterations of intraretinal thickness maps. By subtracting the thickness of the CN group from that of the AD group (AD – CN) and that of the MCI group (MCI – CN), the differences in thickness maps of all intraretinal layers are visualized. The thickness differences were negative (indicating thinning) mainly in TRT, RNFL, GCIPL, and ONL in both AD and MCI groups. In contrast, the differences were positive (indicating thickening) mainly in OPL and PR in AD group. Note the different bar scales for maps with negative and positive differences. AD, Alzheimer’s disease; CN, cognitively normal control; GCIPL, ganglion cell– inner plexiform layer; INL, inner nuclear layer; MCI, mild cognitive impairment; ONL, outer nuclear layer; OPL, outer plexiform layer; PR, photoreceptor complex; RNFL, retinal nerve fiber layer; TRT, total retinal thickness. Scale unit =μm.

Average thickness maps of the GCIPL showed the collapsing of the highest elevation rim in the superior region in patients with MCI and AD (Fig. 5). The thickness map showed smaller areas of the contours from 50 to 80μm in both the AD and MCI groups compared to the CN group, with apparently more profound changes in the AD group. The thinning showed a focal pattern in both the AD and MCI groups and the focal thinning was mainly located in the superior region. In addition, the AD group showed focal thinning in the inner superior sector while the MCI group showed focal thinning in the outer superior sector using ETDRSpartition.

Fig.5

Focal thinning of GCIPL in patients with AD and MCI. Average thickness maps (A) of the GCIPL showed the collapsing of the highest elevation rim in the superior region in patients with AD and MCI. The thickness contours (B) showed smaller areas of the contours from 50 to 80μm in AD and MCI groups compared to CN group with more profound changes in AD group. The thinning (C) showed a focal pattern in both groups and the location with the profound focal thinning was located in the superior region distributed in some quadrants and sectors partitioned using both ETDRS partition (C) and hemispheric partition (D). AD, Alzheimer’s disease; CN, cognitively normal control; GCIPL, ganglion cell– inner plexiform layer. Scale unit =μm.

The thickness maps were partitioned into inner and outer annuli. The thickness of the RNFL in the inner annulus of the AD group was thinner than the MCI group (p < 0.05). After age adjustment, the thickness of the RNFL in the annulus (1 to 6 mm in diameter) and outer annulus (3 to 6 mm) was thinner in the AD group than that in the CN group (p < 0.05). Compared with the CN group, the thickness of the GCIPL in the annulus (1 to 6 mm in diameter) and outer annulus (3 to 6 mm) was thinner in the MCI group after age adjustment (p < 0.05,Fig. 6).

Fig.6

Annular thicknesses of GCIPL. Compared with the CN group, the thickness of the GCIPL in the annulus (1 to 6 mm in diameter) and outer annulus (3 to 6 mm) was thinner in the MCI group after age adjustment (p < 0.05). Age was a factor responsible for the thickness differences of the GCIPL in the inner annulus (1 to 3 mm) among groups (p < 0.05). AD, Alzheimer’s disease; MCI, mild cognitive impairment; CN, cognitively normal control. Mean measurement values are listed in the center of the columns. Bars = standard errors.

Quadrantal analyses of the annulus from 1 to 6 mm in diameter using both the ETDRS and hemispheric partitions were done. After age adjustment, the thickness of the RNFL in the nasal and nasal superior quadrants was thinner in the AD group compared to the CN group (p < 0.05). The GCIPL thickness of the AD and MCI groups was thinner in the superior, nasal superior and temporal superior quadrants, compared to the CN group (p < 0.05,Fig. 7).

Fig.7

Quadrantal thicknesses of GCIPL in the annulus from 1 to 6 mm in diameter using ETDRS and hemispheric methods. GCIPL thickness of AD and MCI was thinner in superior (S), nasal superior (NS), and temporal superior (TS) quadrants, compared to the CN group (p < 0.05). AD, Alzheimer’s disease; MCI, mild cognitive impairment; CN, cognitively normal control; ETDRS, Early Treatment Diabetic Retinopathy Study. Mean measurement values are listed in the center of the columns. Bars = standard errors.

Analysis of eight sectors using the ETDRS and hemispheric partitions showed significant differences in the thicknesses of RNFL, GCIPL in both MCI and AD patients compared to CN (p < 0.05, Fig. 8). After age adjustment, RNFL thickness in the outer nasal (ON) and outer nasal superior (ONS) sectors was thinner in the AD group compared to the CN group (p < 0.05). In addition, RNFL thickness in the inner nasal (IN) and inner nasal inferior (INI) sectors was thinner in the AD group compared to the MCI group (p < 0.05). GCIPL thickness of the inner superior (IS), inner nasal superior (INS), inner temporal superior (ITS), and outer nasal superior (ONS) sectors was significantly thinner in the AD than the CN groups (p < 0.05, Fig. 8). GCIPL thickness of the outer superior (OS), inner temporal superior (ITS), outer nasal superior (ONS), and outer temporal superior (OTS) sectors was significantly thinner in the MCI group than the CN group (p < 0.05, Fig. 8). In addition, PR thickness in the outer nasal (ON) sector was thicker in MCI than CN (p < 0.05).

Fig.8

Sectorial thicknesses of GCIPL. The sectorial thicknesses of the GCIPL in the inner and outer annuli were divided using the ETDRS (A) and hemispheric (B) partitions. Analysis of eight sectors using the ETDRS partition showed significant differences in the GCIPL thickness in both MCI and AD patients compared to CN (p < 0.05). After age adjustment, GCIPL thickness of the inner superior (IS), inner nasal superior (INS), inner temporal superior (ITS) and outer nasal superior (ONS) sectors was significantly thinner in AD than CN (p < 0.05). GCIPL thickness of the outer superior (OS), inner temporal superior (ITS), outer nasal superior (ONS) and outer temporal superior (OTS) sectors was significantly thinner in MCI than CN (p < 0.05). AD, Alzheimer’s disease; MCI, mild cognitive impairment; CN, cognitively normal control. Mean measurement values are listed in the center of the columns. Bars = standard errors.

The trend analyses after age adjustment indicated statistically significant trends for reduced thickness (from control to MCI to AD) for: 1) RNFL in the annulus (1 to 6 mm, p = 0.04) and outer annulus (3 to 6 mm, p = 0.03); 2) RNFL in the nasal (p = 0.01) and nasal superior (p = 0.02) quadrants; 3) RNFL in the outer nasal (p = 0.01) and outer nasal superior (p = 0.02) sectors; 4) the GCIPL superior (p = 0.02), nasal superior quadrant (p = 0.02), and temporal superior (p = 0.04) quadrants; 5) the GCIPL inner superior (p = 0.006), inner nasal superior (p = 0.02), and outer nasal superior (p = 0.03) sectors.

In patients with AD and MCI, MMSE was significantly related to RNFL thickness in the nasal quadrant (ρ= 0.31, p < 0.05), inner nasal (ρ= 0.32, p < 0.05), and inner nasal superior (ρ= 0.31, p < 0.05) sectors. In addition, MMSE was significantly related to the GCIPL thickness in the inner nasal (ρ= 0.30, p < 0.05) and inner nasal superior (ρ= 0.29, p < 0.05) sectors. Disease duration was not correlated with the thickness measurements (p > 0.05).

DISCUSSION

Visualization and analyses of intraretinal layers revealed the alterations of these unevenly distributed thickness maps and their contribution to the changes in total retinal thickness in patients with AD and MCI in comparison to controls. The changes in the macular region were mainly found in the GCIPL. While the GCIPL changes appeared to have a diffuse pattern, the more profound changes were located in the superior region in both AD and MCI patients. Thinning of the average GCIPL in MCI and AD patients has been reported in previous OCT studies [6, 8]. The decrease of the GCIPL layer thickness of the posterior pole (i.e., macular area up to 6×6 mm²) is approximately 5-6μm in patients with AD [6 , 16]. Mapping the GCIPL thickness for the whole scanned area may help find the focal thinning in the GCIPL and facilitate the determination of thinning patterns. In contrast to previous studies [6 , 16], the present study located the region with the most profound loss of the GCIPL and determined the thinning patterns in AD and MCI patients. When the average thinning of the GCIPL reached ∼5μm, the ellipsoidal partition detected an additional 2-3μm thinning in the inferior sector in previous studies (Fig. 9) [6, 8]. In the present study, the average GCIPL thinning was only 4-5μm, whereas the ETDRS or hemispherical octave methods detected up to ∼11μm focal thinning in AD patients and ∼7μm in MCI patients, which was best described at the superior region using the GCIPL tomographic mapping. The thinning of the superior region found in the present study is consistent with previous studies [6, 17]. Cheung et al. studied a cohort of 100 AD patients and found that the most profound thinning was located in the superior nasal (GCIPL thinning – 4.99μm, Zeiss ellipsoidal partition) and superior (GCIPL thinning – 4.80μm) sectors after full adjustments for confounding factors [6]. Bayhan et al. studied 31 AD patients and found that the thinning (– 7.48μm) of the macular ganglion cell complex (GCC) was mainly in the superior region [17]. In contrast, Choi et al. found the most profound thinning of GCIPL (– 8.07μm, Zeiss ellipsoidal partition) in the inferior region [8]. In the present study, the thickness maps showed a trend of thinning in the inferior region, although the thinning in the inferior region did not reach a significant level.

Fig.9

Comparison of studies based on the GCIPL thickness in MCI and AD patients in the present study. When averaged thinning of GCIPL reached ∼5μm [6 , 16], the ellipsoidal partition detected an additional 1– 3μm in these previous studies [6, 8]. The present study detected up to 5μm more in the inner superior (IS) sector in AD and up to 3μm more in the inner temporal superior (ITS) in MCI. AD, Alzheimer’s disease; GCIPL, ganglion cell– inner plexiform layer; I, inferior; II, inner inferior; IN, inner nasal; IN*, inferior nasal; INI, inner nasal inferior; INS, inner nasal superior; IS, inner superior; IT, inner temporal; IT*, inferior temporal; ITI, inner temporal inferior; ITS, inner temporal superior; MCI, mild cognitive impairment; ON, outer nasal; OI, outer inferior; ONI, outer nasal inferior; ONS, outer nasal superior; OS, outer superior; OT, outer temporal; OTI, outer temporal inferior; OTS, outer temporal superior; S, superior; SN, superior nasal; ST, superior temporal. Note: *IT (inferior temporal) and *IN (inferior nasal) in the ellipsoidal partition are not the same as IT (inner temporal) and IN (inner nasal) in the ETDRS partition. Unit =μm.

While the GCIPL thinning pattern is supported in general by the previous studies [6, 8], our study and some others [6, 8] are not in agreement with a most recent study [9], in which den Haan et al. did not find differences between early onset AD and CN controls in GCIPL thickness using OCT volumetric scan and ETDRS partition. It is not clear whether a focal thinning occurred within the annuli because the averaged GCIPL thicknesses in the annuli were analyzed and no thickness maps were created [9]. Another explanation for this discrepancy in documenting GCIPL alterations in AD by den Hann et al. may simply be due to the different AD cohorts. den Hann et al. studied early onset AD, whereas the present study and previous studies included late onset AD patients[6, 8].

The GCIPL thinning pattern detected in the present study and other in vivo studies [6 , 18] also agrees with postmortem pathologic studies [19, 20]. The number of ganglion cells in AD patients decreased by 25% compared with control subjects in the posterior pole of the retina (center fovea/fovea/near the fovea), and the most significant changes were near the fovea (as high as 52%). At a distance of 0– 1.5 mm from the fovea, the neuronal loss rates ranged from 43% to 26% in AD patients [19]. The loss of retinal ganglion cells could be involved in the primary disease process because deposition of beta-amyloid protein was observed in the retina and was related to the cortical amyloid burden [21]. These histological findings of the loss in ganglion cells are supported by recent OCT studies with the thinning of GCIPL [6 , 22]. The thinning of RNFL in AD is more profound in the superior and inferior sectors of the optic nerve [22 –24], which may explain our finding that both the AD and MCI groups had GCIPL thinning in the superior region, and the AD group had additional thinning in the inferior region. Alternatively, GCIPL loss may be a consequence of retrograde neurodegeneration occurring in the cortical regions, as reported in other causes of cortical lesions [25 –27].

Furthermore, the correlation between the focal thinning of GCIPL thickness in INS and IN sectors and MMSE may indicate that the GCIPL in these sectors could be more sensitive in relating to neurodegeneration in AD and MCI. The relation between annular GCIPL thickness (r = 0.33, measured using 3D OCT-2000, Topcon Corp., Tokyo, Japan) and MMSE was reported in a previous AD (n = 24) study by Cunha et al. [16]. In addition, Bayhan et al. also established the relation (r = 0.549, measured using RTVue-100, Optovue Inc, Fremont CA, USA) between macular ganglion cell complex (GCC) and MMSE in AD patients (n = 31) [17]. However, with a mixed AD and MCI group (n_AD = 42 and n_MCI = 26) measured using Cirrus HD-OCT (Carl Zeiss Meditec, Inc., Dublin, CA, USA), Choi et al. did not uncover a relationship between the average GCIPL thickness and MMSE. These discrepancies among various studies in establishing a relationship between retinal GCIPL measurement and cognitive function may be due to differences in study cohorts, sample sizes and measurement systems.

There are limitations to the present study. First, the sample size was relatively small, and the average GCIPL thinning in MCI and AD patients was only ∼4μm. However, with the small sample size, the GCIPL focal thinning was evident in patients with AD and MCI. These observations may lead to further development of more sensitive UHR-OCT biomarkers in detecting early neurodegeneration. Further large-scale studies are needed to determine the relationships between the focal thinning and other clinical manifestations of AD and MCI in longitudinal studies. Second, our UHR-OCT system is custom made with sophisticated partition strategies provided by Orion software [10], and the data may not be directly comparable to previously published data collected with commercial OCT systems and the segmentation algorithms with specific partitions. Of note, the Orion software used in the present study is commercially available and can be used to process a dataset obtained using commercially available OCT systems [28, 29]. Third, in our previous study [30], we found the loss of capillary network density in patients with AD and MCI, which may indirectly affect the macular thickness measurement. However, the thickness measurement in the present study may not identify the influence of retinal vessel changes on the thickness measurement since the vessels are within the tissue. Fourth, eyes with more than 6.00 diopters may affect the results due to the difference of the field of view of the imaged area [31, 32], therefore, correction of the image field may be needed. In the present study, subjects with eyes with refraction from – 6D to +6D were recruited and the correction was not performed. Fifth, the AD and MCI diagnoses of the subjects recruited in our study, as in previous studies [6, 8], were based on clinical criteria without information on amyloid PET scans and likely resulted in some misclassification. Further studies using amyloid PET scans may further define the patient population with accurate diagnoses [33].

In conclusion, focal thinning of the GCIPL was visualized and evident in sectors partitioned by both the ETDRS and hemispheric methods in AD and MCI patients, which may provide more sensitive and specific information about the neurodegeneration in MCI and AD.

Footnotes

ACKNOWLEDGMENTS

Supported by the North American Neuro-ophthalmology Society, McKnight Brain Institute, NIH Center Grant P30 EY014801, and a grant from Research to Prevent Blindness (RPB).

Authors’ disclosures available online ().

References

Sperling

, Aisen

, Beckett

, Bennett

, Craft

, Fagan

, Iwatsubo

, Jack

Jr , Kaye

, Montine

, Park

, Reiman

, Rowe

, Siemers

, Stern

, Yaffe

, Carrillo

, Thies

, Morrison-Bogorad

, Wagster

, Phelps

(2011) Toward defining the preclinical stages of Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 280–292.

Kirova

, Bays

, Lagalwar

(2015) Working memory and executive function decline across normal aging, mild cognitive impairment, and Alzheimer’s disease. Biomed Res Int 2015 748212.

Zhang

, Song

, Zhang

, Chen

, D’Arcy

, Darvesh

, Fisk

, Rockwood

(2011) An MRI brain atrophy and lesion index to assess the progression of structural changes in Alzheimer’s disease, mild cognitive impairment, and normal aging: A follow-up study. J Alzheimers Dis 26(Suppl 3), 359–367.

Jack

Jr , Holtzman

(2013) Biomarker modeling of Alzheimer’s disease. Neuron 80, 1347–1358.

Reardon

(2015) Antibody drugs for Alzheimer’s show glimmers of promise. Nature 523, 509–510.

Cheung

, Ong

, Hilal

, Ikram

, Low

, Ong

, Venketasubramanian

, Yap

, Seow

, Chen

, Wong

(2015) Retinal ganglion cell analysis using high-definition optical coherence tomography in patients with mild cognitive impairment and Alzheimer’s disease. J Alzheimers Dis 45, 45–56.

Wang

, Jiang

, Shen

, Lam

, Debuc

, Ye

, Li

, Tao

, Shao

, Wang

(2012) Quantitative analysis of the intraretinal layers and optic nerve head using ultra-high resolution optical coherence tomography. J Biomed Opt 17, 066013.

Choi

, Park

, Kim

(2016) Macular ganglion cell -inner plexiform layer thickness is associated with clinical progression in mild cognitive impairment and Alzheimers disease. PLoS One 11, e0162202.

den

, Janssen

, van de Kreeke

, Scheltens

, Verbraak

, Bouwman

. (2018) Retinal thickness correlates with parietal cortical atrophy in early-ons heimer’s disease and controls. Alzheimers Dement (Amst) 10, 49–55.

10.

Tan

, Yang

, Jiang

, Liu

, Deng

, Lam

, Hu

, Oakley

, Wang

(2016) The measurement repeatability using different partition methods of intraretinal tomographic thickness maps in healthy human subjects. Clin Ophthalmol 10, 2403–2415.

11.

Mwanza

, Durbin

, Budenz

, Girkin

, Leung

, Liebmann

, Peace

, Werner

, Wollstein

(2011) Profile and predictors of normal ganglion cell-inner plexiform layer thickness measured with frequency-domain optical coherence tomography. Invest Ophthalmol Vis Sci 52, 7872–7879.

12.

, Di

, Carelli

, Carbonelli

(2017) Patterns of retinal ganglion cell damage in neurodegenerative disorders: Parvocellular vs magnocellular degeneration in optical coherence tomography studies. Front Neurol 8, 710.

13.

, Ross-Cisneros

, Sadun

, Carelli

(2017) Retinal ganglion cells and circadian rhythms in Alzheimer’s disease, parkinson’s disease, and beyond. Front Neurol 8, 162.

14.

McKhann

, Knopman

, Chertkow

, Hyman

, Jack

Jr , Kawas

, Klunk

, Koroshetz

, Manly

, Mayeux

, Mohs

, Morris

, Rossor

, Scheltens

, Carrillo

, Thies

, Weintraub

, Phelps

(2011) The diagnosis of dementia due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 263–269.

15.

Albert

, DeKosky

, Dickson

, Dubois

, Feldman

, Fox

, Gamst

, Holtzman

, Jagust

, Petersen

, Snyder

, Carrillo

, Thies

, Phelps

(2011) The diagnosis of mild cognitive impairment due to Alzheimer’s disease: Recommendations from the National Institute on Aging-Alzheimer’s Association workgroups on diagnostic guidelines for Alzheimer’s disease. Alzheimers Dement 7, 270–279.

16.

Cunha

, Lopes

, Costa-Cunha

, Costa

, Pires

, Almeida

, Monteiro

(2016) Macular thickness measurements with frequency domain-OCT for quantification of retinal neural loss and its correlation with cognitive impairment in Alzheimer’s disease., e, PLoS One 11, 0153830.

17.

Bayhan

, Aslan

, Celikbilek

, Tanik

, Gurdal

(2015) Evaluation of the chorioretinal thickness changes in Alzheimer’s disease using spectral-domain optical coherence tomography. Clin Exp Ophthalmol 43, 145–151.

18.

Marziani

, Pomati

, Ramolfo

, Cigada

, Giani

, Mariani

, Staurenghi

(2013) Evaluation of retinal nerve fiber layer and ganglion cell layer thickness in Alzheimer’s disease using spectral-domain optical coherence tomography. Invest Ophthalmol Vis Sci 54, 5953–5958.

19.

Blanks

, Torigoe

, Hinton

, Blanks

(1996) Retinal pathology in Alzheimer’s disease. I. Ganglion cell loss in foveal/parafoveal retina. Neurobiol Aging 17, 377–384.

20.

Hinton

, Sadun

, Blanks

, Miller

(1986) Optic-nerve degeneration in Alzheimer’s disease. N Engl J Med 315, 485–487.

21.

Snyder

, Johnson

, Lim

, Santos

, Alber

, Maruff

, Fernandez

(2016) Nonvascular retinal imaging markers of preclinical Alzheimer’s disease. Alzheimers Dement (Amst) 4, 169–178.

22.

Coppola

, Di

, Ziccardi

, Martelli

, Fadda

, Manni

, Barboni

, Pierelli

, Sadun

, Parisi

(2015) Optical coherence tomography in Alzheimer’s disease: A meta-analysis. PLoS One 10, e0134750.

23.

Kesler

, Vakhapova

, Korczyn

, Naftaliev

, Neudorfer

(2011) Retinal thickness in patients with mild cognitive impairment and Alzheimer’s disease. Clin Neurol Neurosurg 113, 523–526.

24.

Kirbas

, Turkyilmaz

, Anlar

, Tufekci

, Durmus

(2013) Retinal nerve fiber layer thickness in patients with Alzheimer disease. J Neuroophthalmol 33, 58–61.

25.

Albrecht

, Frohlich

, Hartung

, Kieseier

, Methner

(2007) Optical coherence tomography measures axonal loss in multiple sclerosis independently of optic neuritis. J Neurol 254, 1595–1596.

26.

Meier

, Maeder

, Kardon

, Borruat

(2015) Homonymous ganglion cell layer thinning after isolated occipital lesion: Macular OCT demonstrates transsynaptic retrograde retinal degeneration. J Neuroophthalmol 35, 112–116.

27.

Mitchell

, Oliveira

, Tsiouris

, Dinkin

(2015) Corresponding ganglion cell atrophy in patients with postgeniculate homonymous visual field loss, J Neuroophthalmol 35 353–359.

28.

Behbehani

, Abu Al-Hassan

, Al-Salahat

, Sriraman

, Oakley

, Alroughani

(2017) Optical coherence tomography segmentation analysis in relapsing remitting versus progressive multiple sclerosis. PLoS One 12, e0172120.

29.

Keller

, Oakley

, Russakoff

, Andorra-Ingles

, Villoslada

, Sanchez-Dalmau

(2016) Changes in macular layers in the early course of non-arteritic ischaemic optic neuropathy. Graefes Arch Clin Exp Ophthalmol 254, 561–567.

30.

Jiang

, Wei

, Shi

, Wright

, Sun

, Gregori

, Zheng

, Vaner

, Lam

, Rundek

, Wang

(2017) Altered macular microvasculature in mild cognitive impairment and Alzheimer disease., J Neuroophthalmol. doi: 10.1097/WNO.0000000000000580.

31.

Bennett

, Rudnicka

, Edgar

(1994) Improvements on Littmann’s method of determining the size of retinal features by fundus photography. Graefes Arch Clin Exp Ophthalmol 232, 361–367.

32.

Nowroozizadeh

, Cirineo

, Amini

, Knipping

, Chang

, Chou

, Caprioli

, Nouri-Mahdavi

(2014) Influence of correction of ocular magnification on spectral-domain OCT retinal nerve fiber layer measurement variability and performance. Invest Ophthalmol Vis Sci 55, 3439–3446.

33.

Ossenkoppele

, Jansen

, Rabinovici

, Knol

, van der Flier

, van Berckel

, Scheltens

, Visser

, Verfaillie

, Zwan

, Adriaanse

, Lammertsma

, Barkhof

, Jagust

, Miller

, Rosen

, Landau

, Villemagne

, Rowe

, Lee

, Na

, Seo

, Sarazin

, Roe

, Sabri

, Barthel

, Koglin

, Hodges

, Leyton

, Vandenberghe

, van

, Drzezga

, Forster

, Grimmer

, Sanchez-Juan

, Carril

, Mok

, Camus

, Klunk

, Cohen

, Meyer

, Hellwig

, Newberg

, Frederiksen

, Fleisher

, Mintun

, Wolk

, Nordberg

, Rinne

, Chetelat

, Lleo

, Blesa

, Fortea

, Madsen

, Rodrigue

, Brooks

(2015) Prevalence of amyloid PET positivity in dementia syndromes: A meta-analysis. JAMA 313, 1939–1949.