Changes in Retinal Layer Thickness in the Contralateral Eye of Patients with Unilateral Neovascular Age-Related Macular Degeneration

Changes in Retinal Layer Thickness in the Contralateral Eye of Patients with Unilateral Neovascular Age-Related Macular Degeneration Muneeswar Gupta Nittala, MPhilOpt,1 Ruth E. Hogg, PhD,2 Yan Luo, MD,1 Swetha Bindu Velaga, BOpt,1 Rufino Silva, MD,3 Dalila Alves, MSc,3 Giovanni Staurenghi, MD,4 Usha Chakravarthy, MD,2 SriniVas R. Sadda, MD1,5 Purpose: To evaluate the thickness of the outer retinal layers and its relationship with visual function in fellow eyes of participants with unilateral neovascular age-related macular degeneration (AMD). Design: Longitudinal study. Participants: We enrolled 105 subjects with unilateral neovascular AMD from 3 clinical centers in Europe. Methods: The fellow eye, without advanced AMD, was selected for the study. Subjects were followed up with visits occurring every 6 months for 2 years. Spectral domain optical coherence tomography volume scans were collected at 3 clinical sites, in Belfast, Northern Ireland; Coimbra, Portugal; and Milan, Italy. Detailed manual segmentation of outer retinal layers was performed using the custom-designed and validated grading software 3D OCTOR. Thickness measurements for neurosensory retina, photoreceptor layer (PRL) outer segments, retinal pigment epithelium plus drusen (RPEþdrusen) complex, and choroidal layers from each sector of the standard macular grid were obtained. Measures of vison were distance visual acuity, near visual acuity, Smith-Kettlewell Institute low-luminance acuity score, and reading speed. Subjects were grouped based on the presence or absence of subretinal drusenoid deposits (SDDs) for further analysis. Main Outcome Measures: Change in thickness of retinal layers and change in measures of vision. Results: In all, 85 eyes were included in the analysis. The average duration of follow-up was 20.5  5.8 months. By the final visit, the RPEþdrusen complex was significantly thinner when compared with baseline (29.7 mm vs. 34.09 mm; P ¼ 0.03). Low-luminance deficit was significantly worse at the final visit (P < 0.001) and correlated with PRL outer segment thickness (r ¼ 0.33; P ¼0.02). The RPEþdrusen complex was significantly thicker in eyes with SDDs compared with that in those without SDDs (30.67 mm vs. 28.64 mm; P ¼ 0.02). PRL outer segments became significantly thinner over time in eyes with SDDs compared with those in eyes without SDDs. Conclusions: The RPEþdrusen complex layer becomes thinner over time in fellow eyes of subjects with unilateral neovascular AMD. The rate of PRL outer segment thinning was higher in eyes with SDDs than in eyes without SDDs. These findings are preliminary steps in the identification of early biomarkers for detecting and monitoring the progression of AMD. Ophthalmology Retina 2018;-:1e10 ª 2018 Published by Elsevier Inc. on behalf of the American Academy of Ophthalmology Supplemental material available at www.ophthalmologyretina.org.

In the evolution of age-related macular degeneration (AMD), the importance of morphological alterations in the retina, retinal pigment epithelium, and the choroid has been established.1e3 AMD remains a principal cause of irreversible blindness in older adults in the developed and developing world despite improvements in the management of neovascularization, a critical component of late-stage AMD.1,2 In its early stages, AMD is characterized by drusen and pigmentary changes, and subsequently atrophic patches develop in the aforementioned tissues, leading to a manifestation known as geographic atrophy (GA). Along  2018 Published by Elsevier Inc. on behalf of the American Academy of Ophthalmology

with neovascularization in the macular tissues (neovascular AMD), GA contributes to the burden of severe visual impairment and blindness, that are the clinical consequences of late-stage AMD.4 Advances over the past decade in noninvasive retinal imaging using spectral domain optical coherence tomography (SD-OCT) has made it easier to visualize and understand longitudinal changes in the chorioretinal microanatomy that occur before the development of overt disease .1,5,6 It has been noted that the progression of GA is associated with the thinning of the outer retinal layers6e8

https://doi.org/10.1016/j.oret.2018.09.017 ISSN 2468-6530/18

1

Ophthalmology Retina Volume -, Number -, Month 2018 and remodeling of the inner retinal layers. Increases in the inner nuclear layer thickness associated with significant thinning of the retinal pigment epithelium (RPE)ephotoreceptor complex (Ebneter et al6 and Sadiq et al7) and increased autofluorescence at the junctional zone of GA lesions corresponding to thickening of the RPE (Hariri et al3), have previously been demonstrated to occur in the setting of GA. Clinicalepathological correlational studies have shown that there is thinning of the photoreceptor layer (PRL) in AMD.9,10 Also, the choroid was found to be thinner in persons with GA2 and in eyes with reticular pseudodrusen, a further manifestation of the aging eye.11 Pappuru et al1 and Gella et al12 showed a moderate correlation between the center subfield outer retinal layer thickness and visual acuity in nonexudative AMD. To date, studies on layer segmentation to identify markers for progression from early to late AMD have been limited to small sample sizes and have reported findings that have been localized to areas of interest around GA7 or to the foveal center subfield and have not examined relationships with function.1,7,8 It is known that patients with exudative AMD in 1 eye are at very high risk of developing both GA and neovascular AMD.4 To date, there has been no systematic study at defined intervals of follow-up using high-resolution SD-OCT in such highrisk patients. We report the functional and morphological changes in the contralateral eye of patients with an established diagnosis of choroidal neovascularization due to AMD in the first eye.

the appropriate reading addition over the protocol refractive correction. Charts were translated into the appropriate language for each study site. Reading speed was measured using the BaileyLovie reading-speed charts presented as transparences with black text on an illuminated light box. The charts used a mix of 12 to 30 unrelated words consisting of 4, 7, or 10 letters on each line. For any given reading-speed chart, a single print size was used and ranged from 0.7 to 1.6 logarithm of the minimum angle of resolution. The size of print chosen was 0.1 logarithm of the minimum angle of resolution larger than the threshold NVA and the reading speed was expressed in units of words read per minute. Lowluminance visual acuity was tested using the Smith-Kettlewell Institute low-luminance (SKILL) acuity chart.14 The SKILL acuity card consists of 2 letter charts mounted back to back. One side is a low-contrast chart with black letters on a dark gray background (10% of the reflectance of white paper); the other is a high-contrast (>90%) black-on-white letter chart with a different letter order. The card is held approximately 40 cm from the eye with the appropriate reading addition in place and the number of letters read represents the score. The difference between the scores obtained on high- and low-contrast conditions is the SKILL deficit.

SD-OCT Image Acquisition The SD-OCT images were obtained using a Spectralis HRAþOCT (Heidelberg Engineering, Heidelberg, Germany) with a macular cube protocol of 37 B-scans using an automatic real-time tracking level of 5 images in the 30  20 scan area centered on the foveal center. SD-OCT scans were obtained every 6 months during the 2-year follow-up period using follow-up mode. Color fundus photographs were obtained using a digital color fundus camera (Topcon 50X; Topcon, Tokyo, Japan) with pupil dilation (1 drop tropicamide 1% and 1 drop phenylephrine 1%, given 10 minutes before the first image acquisition). The minimum pupil dilation accepted was 6 mm.

Methods We included the data from 85 of 105 participants with unilateral neovascular AMD who were enrolled in 3 centers (Milan, Italy; Coimbra, Portugal; and Belfast, Northern Ireland) as part of a longitudinal, observational, nonrandomized, noninterventional study of the contralateral eyes (free of the late AMD features of choroidal neovascularization and GA and henceforth referred to as the study eye). The primary aim of the study was to identify markers of progression from early to neovascular AMD in the study eye. In the United Kingdom center, ethics approval was obtained from a national research ethics committee, and in the other 2 sites, institutional review boards approved the study design. Details of the inclusion and exclusion criteria for study participation have been previously described.13 Imaging data were transferred to the Doheny Image Reading Centre (DIRC) through a materials transfer agreement and were approved by the institutional review board of the University of CaliforniaeLos Angeles. The study adhered to the tenets of the Declaration of Helsinki on research using human volunteers.

Study Methods All participants underwent standardized measurements of visual function both in the designated study eye and the eye with neovascular AMD. The study procedures have been described previously13 but in brief, best-corrected distance visual acuity (DVA) was measured at 4 meters using an Early Treatment Diabetic Retinopathy Study (ETDRS) chart, near visual acuity (NVA) was measured with a modified Bailey-Lovie near-reading chart using

2

Grading Procedure For all OCT retinal layer analyses, previously described and validated software (3D OCTOR) developed by the DIRC was used to perform quantitative assessments on 37 B-scans imported from the Spectralis HRAþOCT instrument for each case.1,15 The 3D OCTOR software effectively allows the grader to manually draw or correct for automated segmentation errors on multiple boundaries to define structures of interest. Using the boundary definitions and segmentation of each B-scan image, the calculated pixels are converted into micrometers to yield thickness measurements at each location. The thickness at unsampled locations between the line scans are interpolated using linear approximation to yield a thickness map. Mean thickness values can be generated for any zone within the 9 ETDRS subfields. For the purpose of this study, retina (neurosensory retina), PRL outer segments, RPEþdrusen complex, and choroid were automatically segmented using 3D OCTOR. On each B-scan, the following boundaries were drawn automatically and manually corrected for segmentation errors by a single experienced and certified DIRC grader (MGN): internal limiting membrane, ellipsoid zone, outer border of the PRL, inner and outer borders of the RPE, surface of Bruch’s membrane, and the inner and outer borders of the choroid (Figure 1). The space between the internal limiting membrane and the outer border of the PRL was defined as “retina” (none of the study cases had epiretinal membrane or vitreoretinal interface disease; thus, these did not confound our study assessments); “photoreceptor layer outer segments (PRL outer segments)” was defined as the space between the ellipsoid zone and the outer border of the photoreceptors (the areas of subretinal drusenoid

Nittala et al



Retinal Layer Thickness Change Analysis in Nonneovascular AMD

Figure 1. A, B, Spectral domain optical coherence tomography B-scan of a study subject with segmentation of the following boundaries: internal limiting membrane (ILM), ellipsoid zone, photoreceptor layer (PRL) outer segment, retinal pigment epithelium (RPE)eBruch’s membrane (inner and outer), and choroid (inner and outer).

deposits [SDDs] were excluded from segmentation to avoid these deposits interrupting the outer segment thickness measurements); space between inner and outer borders of the RPEeBruch’s membrane was defined as “RPEþdrusen complex” (SDD was included in this space by adjusting the segmentation appropriately); and the space between the inner and outer choroid border was defined as the “choroid.” The mean thickness measurements in 9 ETDRS zones were then computed for these defined spaces (layers). Also, thickness measurements were generated in the foveal central subfield (1 mm from the center of the fovea), the inner ring (>1 to 3 mm from the center of the foveal), and the outer ring (>3 to 6 mm from the center of the fovea). Thickness measurements of the retinal layers and choroid were presented in micrometers. The reproducibility of these boundary segmentations and measurements has been reported previously (for photoreceptor outer segments intraclass correlation coefficient 0.924 [95% confidence interval: 0.752e0.977]; for RPE thickness intraclass correlation coefficient 0.713 [95% confidence interval: 0.061e0.913]).1

Statistical Analysis Of the 105 participants, data were available on 85 study eyes in participants who had 1 follow-up visit and images gradable by 3D OCTOR at baseline and follow-up. Thickness measurements were computed for each of the ETDRS grid segments as follows: foveal center subfield, inner ring, and outer ring. A paired t test was used for comparative change analysis between baseline and lastvisit thickness measurements. Retinal layer thickness change over time was compared between groups after categorization of subjects for presence or absence of SDD. Descriptive statistics were generated for all measures of vision at each time point, along with change from baseline to the final visit. As the SKILL acuity test measures the visual acuity at simulated low luminance (defined as low-luminance SKILL scores), we computed the SKILL deficit (difference between SKILL score measured under high luminance and the low-luminance SKILL score). Analysis of variance was performed for associations between age, body mass index, and visual function assessments. Correlations between functional vision parameters and retina layer thickness measures were analyzed using Spearman correlations. All statistical analyses were performed using SPSS statistical software (IBM SPSS version 18,

IBM, Armonk, NY). P < 0.05 was considered statistically significant.

Results Of the 85 participants, all had completed 1 year of follow-up (3 visits, including baseline) and 57 completed 2 years. The average number of follow-up visits was 4.4, and the average duration of follow-up was 20.5  5.8 months. The mean age of the 85 subjects at the first visit was 75.6  7.3 years (range: 52e93 years); 39 subjects (46%) were female. The mean body mass index was 26.89  4.09 (range: 18.51e40.98). Baseline DVA ranged from 71 to 96 letters, with an average and standard deviation of 82  6 letters. Over time, there were no statistically significant differences in best-corrected visual acuity, NVA, or reading speed (Table 1). The low-luminance deficit increased markedly over time, and this change was highly significant (P < 0.001) (Table 1). The average SKILL acuity score at low luminance showed a slight increase over time and was of marginal statistical significance (P ¼ 0.04). Table 2 shows the association between measures of vision and patient factors. At baseline. best-corrected visual acuity and SKILL acuity score at high luminance were not associated with gender, smoking status and systemic factors (cardiovascular disease, hypertension, and hyperlipidemia). Reading speed was lower in female participants and in patients with a history of cardiovascular disease, hypertension, or hyperlipidemia (Table 2). SKILL at low luminance was also numerically worse in persons with systemic disorders, and this difference was statistically significant in those with hypertension. Retinal layer thickness changes from baseline to the final visit is shown in Table 3. Over time, thickness metrics reduced, with the full thickness retina measurement reaching statistical significance (P < 0.02) (Table 3). On averaging across the entire ETDRS grid, the RPEþdrusen complex layer was significantly thinner by the final visit (P ¼ 0.03) and the volume of this layer likewise reduced significantly. The trends in layer thickness change over time in the total grid area and foveal central subfield areas are shown in Figure 2.

3

Ophthalmology Retina Volume -, Number -, Month 2018 Table 1. Change in Visual Function between Baseline and Last Visit Visual Function Parameter Distance visual acuity, letters, mean  SD (range) Near visual acuity, logMAR, mean  SD (range) SKILL score low, letters, mean  SD (range) SKILL deficit, letters, mean  SD (range) Reading speed, words/min, mean  SD (range)

Baseline Visit 82.116.06 0.250.1 44.9915.81 36.3911.96 71.4928.99

(71e96) (0.1e0.5) (0e73) (0e75) (7.5e163.64)

Last Visit 80.276.25 0.250.1 49.620.94 84.1215.81 65.9124.42

P

(65e93) (0.1e0.5) (0e93) (46e110) (6.5e128.57)

0.08 0.62 0.04* <0.001* 0.1

LogMAR ¼ logarithm of the minimum angle of resolution; SD ¼ standard deviation; SKILL ¼ Smith-Kettlewell Institute low-luminance acuity test. *Statistically significant value.

Correlation Analysis There were no significant correlations between best-corrected visual acuity, NVA, and SKILL at high luminance with retinal thickness or by segmented layer at the baseline visit. The SKILL score under low luminance was statistically significantly correlated with the RPEeBruch’s complex layer thickness in all ETDRS sectors. The SKILL deficit was negatively correlated with the thickness of the RPEeBruch’s complex but only in the central region of the ETDRS grid. The correlation coefficients and significance of association are shown in Table 4.

Presence of SDDs and Retinal Layer Thickness SDDs were present in 34 of the 85 eyes (40%) in this study group. These deposits were more prevalent in females than in males (of the 34 eyes with SDD, 20 [59%] were female). The mean thickness values between eyes with SDD and with no SDD at each ETDRS

ring and total grid area are shown in Table 5. RPEþdrusen complex layer was significantly thicker in eyes with SDD (P < 0.05) compared with that in eyes with no SDD. Analysis of longitudinal change in the retinal layer was performed to determine the rate of retinal layer thickness change in eyes with and with no SDD (Tables S1 and S2, available at www.ophthalmologyretina.org). Faster thinning of measured retinal layers was observed in SDD compared with that in eyes with no SDD. Details of mean change in thickness and volume from baseline to the last visit are shown in Table 6. The PRL outer segments in all ETDRS grid locations were significantly thinner at the last visit in eyes with SDD compared with that in eyes with no SDD (total grid area: 2.59  2.96 vs. 1.3  2.46 m; P ¼ 0.001). The rates of thickness change in the neurosensory retina, RPEþdrusen complex, and choroid were higher in eyes with SDD than in eyes with no SDD, but the difference between the groups was not statistically significant.

Table 2. Association between Baseline Measures of Vision and Gender, Smoking Status, and Presence of Cardiovascular Disease, Hypertension, and Hyperlipidemia

Gender Male Female P* Smoking status Nonsmoker Former smoker Current smoker Pz Cardiovascular disease history Yes No P* Hypertension history Yes No P* Hyperlipidemia history Yes No P*

DVA, Letters, Mean ± SD (range)

SKILL Low, Mean ± SD (range)

SKILL High, Mean ± SD (range)

RSP, Mean ± SD (range)

82.96.4 (71e96) 81.25.5 (71e92) 0.21

45.315.0 (0e73) 44.616.9 (0e70) 0.87

80.58.9 (62e100) 79.012.8 (48e104) 0.56

77.831.9 (7.5e163.6) 64.323.7 (24e120) 0.03y

81.35.4 (71e92) 83.16.6 (71e96) 83.47.3 (72e93) 0.36

45.715.2 (0e70) 41.316.5 (0e63) 48.017.6 (6e73) 0.49

80.711.2 (48e104) 76.38.5 (60e89) 82.412.5 (53e100) 0.25

67.325.5 (15.4e120) 81.034.1 (7.5e163.6) 69.929.9 (32.1e120) 0.18

82.45.5 (72e93) 81.86.2 (71e96) 0.68

44.613.6 (10e70) 46.817.5 (0e73) 0.59

77.913.1 (48e100) 81.59.5 (62e104) 0.19

62.125.3 (15.4e112.5) 76.530.8 (7.5e163.6) 0.04y

81.45.9 (71e96) 83.65.9 (72e94) 0.16

43.116.7 (0e70) 53.111.0 (36e73) 0.03y

78.911.9 (48e104) 82.68.8 (66e100) 0.243

67.028.5 (7.5e163.6) 83.430.3 (28.1e128.6) 0.03y

81.56.0 (71e96) 82.36.0 (71e94) 0.54

44.915.7 (0e70) 46.416.4 (0e73) 0.70

79.112.8 (48e104) 80.79.1 (66e100) 0.58

63.527.4 (7.5e112.5) 77.729.9 (24e163.6) 0.03y

DVA ¼ distance visual acuity; RSP ¼ reading speed; SD ¼ standard deviation; SKILL ¼ Smith-Kettlewell Institute low-luminance test. *Student t test. y Statistically significant value. z Analysis of variance.

4

Nittala et al



Retinal Layer Thickness Change Analysis in Nonneovascular AMD

Table 3. Retinal Layer Thickness Differences between Baseline and Last Visit in Total Study Sample Retinal Layer Thickness at Baseline Visit, mm, Mean ± SD (range) FCS region Retina PRL outer segments RPEþdrusen complex Choroid Inner ring Retina PRL outer segments RPEþdrusen complex Choroid Outer ring Retina PRL outer segments RPEþdrusen complex Choroid Total grid area thickness Retina PRL outer segments RPEþdrusen complex Choroid Total grid area volume Retina PRL outer segments RPEþdrusen complex Choroid

Retinal Layer Thickness at Last Visit, mm, Mean ± SD (range)

P

247.0222.53 27.315.2 42.7137.93 164.6151.62

(203.9e313.9) (7.2e39.2) (23.6e280.9) (42.1e322)

241.0422.34 24.468.95 39.6730.37 175.6259.62

(180.7e276.9) (0.1e39.2) (9.4e200.6) (103.4e314.3)

0.02* 0.24 0.26 0.96

295.1516.02 25.184.35 38.328.78 161.2845.84

(261.3e336.7) (11.2e39.7) (22.8e201.1) (59.3e294.4)

290.5424.33 23.266.97 33.6816.15 172.6553.52

(232.2e334.6) (3.5e38.2) (13.6e116.7) (105.1e298)

0.01* 0.56 0.09 0.84

255.3416.48 25.363.32 32.4318.43 153.835.66

(208.6e293) (19.1e33.7) (23.2e143.6) (69.6e256.2)

254.319.18 24.874.33 28.164.16 166.8844.61

(202.9e295.9) (17.5e32.8) (22.6e40.3) (98.5e291.5)

0.09 0.58 0.05 0.54

263.9715.64 25.393.42 34.0921.02 155.7738.08

(223.1e300.6) (17.3e35.1) (23.8e146.5) (66.5e266.5)

26219.3 24.54.7 29.76.65 168.4346.67

(217.3e303.8) (15.5e32) (22.3e59.3) (100.5e293.6)

0.03* 0.53 0.03* 0.61

(6.19e8.74) (0.44e0.91) (0.63e1.69) (2.86e8.47)

0.27 0.6 0.04* 0.58

7.430.8 0.720.11 0.920.4 4.391.19

(2.09e8.41) (0.25e0.9) (0.66e3.75) (1.26e7.55)

7.540.56 0.710.14 0.860.19 4.851.36

FCS ¼ foveal center subfield; PRL ¼ photoreceptor layer; RPE ¼ retinal pigment epithelium; SD ¼ standard deviation. *Statistically significant value.

Figure 2. Line plots of thickness of retina, photoreceptor layer (PRL) outer segments, retinal pigment epithelium plus drusen (RPEþdrusen) complex, and choroid layers over a 2-year period in the total Early Treatment of Diabetic Retinopathy Study (ETDRS) grid area and foveal center subfield (FCS).

5

Ophthalmology Retina Volume -, Number -, Month 2018 Table 4. Univariate Analysis of Baseline OCT Thickness Parameters in Early Treatment of Diabetic Retinopathy Study Fields with Visual Function Parameters Correlation Coefficient r (P) Retinal Layer FCS region Retina PRL outer segments RPEþdrusen complex Choroid Inner ring Retina PRL outer segments RPEþdrusen complex Choroid Outer ring Retina PRL outer segments RPEþdrusen complex Choroid Total grid area thickness Retina PRL outer segments RPEþdrusen complex Choroid Total grid area volume Retina PRL outer segments RPEþdrusen complex Choroid

Distance Visual Acuity

Near Visual Acuity

SKILL Low

SKILL Deficit

Reading Speed

0.15 0.17 0.01 0.13

(0.23) (0.17) (0.95) (0.32)

0.21 0.07 0.003 0.01

(0.10) (0.56) (0.98) (0.97)

0.20 0.42 0.24 0.04

(0.15) (0.002)* (0.08) (0.79)

0.06 0.34 0.02 0.19

(0.66) (0.01)* (0.88) (0.17)

0.04 0.03 0.03 0.06

(0.73) (0.79) (0.79) (0.63)

0.14 0.06 0.04 0.08

(0.27) (0.61) (0.74) (0.54)

0.11 0.01 0.16 0.05

(0.40) (0.95) (0.33) (0.71)

0.33 0.34 0.17 0.06

(0.02) (0.009)* (0.24) (0.68)

0.25 0.37 0.07 0.17

(0.07) (0.006)* (0.64) (0.22)

0.13 0.08 0.05 0.03

(0.30) (0.52) (0.64) (0.80)

0.03 0.06 0.15 0.01

(0.81) (0.63) (0.23) (0.92)

0.03 0.06 0.20 0.06

(0.80) (0.61) (0.12) (0.63)

0.13 0.29 0.06 0.07

(0.37) (0.04)* (0.69) (0.60)

0.15 0.27 0.7 0.15

(0.27) (0.05) (0.63) (28)

0.01 0.09 0.26 0.01

(0.99) (0.44) (0.04)* (0.94)

0.06 0.07 0.12 0.03

(0.62) (0.59) (0.34) (0.79)

0.06 0.05 0.18 0.06

(0.64) (0.71) (0.15) (0.66)

0.18 0.33 0.04 0.07

(0.19) (0.02)* (0.78) (0.62)

0.18 0.31 0.02 0.16

(0.20) (0.02)* (0.87) (0.26)

0.03 0.09 0.20 0.004

(0.82) (0.47) (0.11) (0.98)

0.01 0.03 0.13 0.02

(0.99) (0.82) (0.31) (0.90)

0.09 0.09 0.10 0.03

(0.50) (0.48) (0.42) (0.82)

0.14 0.33 0.01 0.04

(0.32) (0.02)* (0.93) (0.77)

0.08 0.11 0.11 0.07

(0.60) (0.42) (0.45) (0.62)

0.24 0.11 0.01 0.10

(0.06) (0.47) (0.93) (0.44)

FCS ¼ foveal center subfield; OCT ¼ optical coherence tomography; PRL ¼ photoreceptor layer; RPE ¼ retinal pigment epithelium; SKILL ¼ SmithKettlewell Institute low-luminance test. *Parameter that showed significant association with visual function parameters.

Presence of Soft Drusen and Retinal Layer Thickness Of the 85 eyes included in the analysis, 15 had soft drusen. There were no significant differences in the change in retinal thickness in eyes without soft drusen when compared with those with soft drusen for the entire retina, RPEþdrusen complex, and choroid. The change in the PRL thickness was significantly greater in eyes with soft drusen compared with those without soft drusen in all ETDRS sectors except the central zone where it just failed to reach significance (Table 7).

Discussion We report the findings from a systematic longitudinal study on retinal layer thickness change in the uninvolved fellow eyes of subjects with unilateral neovascular AMD. We also analyzed the relationship between visual function assessments and retinal layer thickness. With the advantage of high-resolution SD-OCT imaging, retinal layer thickness now can be measured in vivo with precision to quantify and track the longitudinal changes that occur in progressive retinal diseases such as AMD. We observed thinning of the RPEþdrusen complex that was statistically significant, and in addition, there was decreasing PRL outer segment thickness and increasing choroidal thickness over 2 years of

6

follow-up that did not reach significance. These findings resonate with observations previously made on the pathogenesis of neovascular AMD in eyes at high risk. The pathogenesis of AMD has been linked to chronic inflammation,16 oxidative stress,17 and dysfunctional flow in the choroid. The RPEþdrusen complex is affected by all of these processes.3 We observed that the RPEþdrusen complex was thicker at the initial visits18; but over the 2-year follow-up period, gradual thinning was found and is concordant with the hypothesis that apoptosis of the RPE occurs over time.19,20 Our finding that the RPE complex was thicker at the fovea and was thinner off-center is consistent with previously reported normative data.21 Multimodal imaging studies correlating SD-OCT with confocal scanning laser ophthalmoscopy findings have suggested that reticular pseudodrusen (RPD) are SDDs located between the RPEeBruch’s band and the innersegment ellipsoid band.22,23 In this study, 40% of the patients had RPD. The previously reported increased prevalence in female individuals was also noted in this cohort.24 This study showed a thicker RPEþdrusen complex in eyes with SDD than in those without SDD, because these deposits were located not under but above the RPE, and hyperreflective aggregates22 could cause thickening of the RPEþdrusen complex. The material seemed to be hypoautofluorescent on fundus autofluorescence imaging, implying a paucity of retinoids. Interestingly, we found

Nittala et al



Retinal Layer Thickness Change Analysis in Nonneovascular AMD

Table 5. Retinal Layer Thickness Measurements Between Subjects with and without Subretinal Drusenoid Deposits at Baseline SDD Absent (n [ 51 eyes) FCS region Retina PRL outer segments RPEþdrusen complex Choroid Inner ring Retina PRL outer segments RPEþdrusen complex Choroid Outer ring Retina PRL outer segments RPEþdrusen complex Choroid Total grid area thickness Retina PRL outer segments RPEþdrusen complex Choroid Total grid area volume Retina PRL outer segments RPEþdrusen complex Choroid

SDD Present (n [ 34 eyes)

P

248.624.3 27.744.51 33.677.38 172.4853.5

(203.9e313.9) (16.8e39.2) (23.1e54.2) (42.1e322)

245.2420.6 26.825.92 36.610.67 155.7248.74

(206.2e282.6) (7.2e38.2) (22.4e61.8) (87.8e265.3)

0.55 0.49 0.12 0.19

296.4817.1 24.763.69 30.333.43 169.2248

(273.1e336.7) (11.2e31.2) (23.2e38.8) (59.3e294.4)

293.6414.84 25.655.01 34.286.56 152.3142.25

(261.3e320.1) (16.1e39.7) (25.3e50.5) (85.7e265.4)

0.47 0.42 0.003* 0.13

255.6514.92 25.313.32 27.932.91 159.9938.04

(230.1e293) (19.1e31.9) (23.5e36.3) (69.6e256.2)

25518.32 25.423.37 29.383.26 146.831.93

(208.6e292) (20e33.7) (25.1e38.2) (100.9e226.5)

0.87 0.89 0.05 0.13

264.5414.94 25.263.25 28.642.71 162.3940.38

(239.1e300.6) (17.3e31.6) (24.4e35) (66.5e266.5)

263.3216.62 25.523.64 30.673.72 148.2934.42

(223.1e294.3) (19.7e35.1) (25.2e39.6) (97.1e236.2)

0.76 0.76 0.02* 0.03*

(2.09e8.41) (0.25e0.89) (0.24e1.09) (1.26e6.73)

0.13 0.37 0.38 0.008*

7.580.4 0.730.1 0.820.08 4.661.21

(6.85e8.25) (0.51e0.9) (0.69e1) (1.89e7.55)

7.271.08 0.70.13 0.850.16 4.091.12

FCS ¼ foveal center subfield; PRL ¼ photoreceptor layer; RPE ¼ retinal pigment epithelium; SDD ¼ subretinal drusenoid deposits. *Statistically significant value.

Table 6. Change in Retinal Layer Thickness at Last Visit in Eyes with and without Subretinal Drusenoid Deposits

FCS region Retina PRL outer segments RPEþdrusen complex Choroid Inner ring Retina PRL outer segments RPEþdrusen complex Choroid Outer ring Retina PRL outer segments RPEþdrusen complex Choroid Total grid area thickness Retina PRL outer segments RPEþdrusen complex Choroid Total grid area volume Retina PRL outer segments RPEþdrusen complex Choroid

SDD Absent (n [ 51 eyes)

SDD Present (n [ 34 eyes)

2.587.29 0.73.65 1.123.68 2.9316.64

(22.6 to 6.81) (6.4 to 6.8) (6 to 8.6) (29.1 to 34.7)

5.579.6 3.575.77 3.314.44 4.0843.96

(29.6 (12.7 (47.7 (78.4

to to to to

9.8) 5.41) 8.8) 83.3)

0.36 0.02* 0.55 0.59

1.145.37 1.692.82 0.442.82 3.0717.06

(18.3 to 5.11) (3.6 to 8.1) (4.7 to 4.7) (25.5 to 33.2)

9.8812.13 3.34.66 3.277.17 1.6638.83

(33.7 (12.6 (23.2 (72.7

to to to to

4.2) 2.4) 3.5) 90)

0.01* 0.003* 0.2 0.69

0.073.84 1.222.45 0.271.81 4.1617.85

(7.91 to 5.31) (3 to 7.3) (3.7 to 3.7) (29.4 to 33.2)

3.325.09 2.352.97 1.422.22 0.6629.56

(12.4 to 3.81) (9.9 to 1.1) (5.5 to 2.7) (50.7 to 63.1)

0.07 0.002* 0.13 0.71

0.363.98 1.32.46 0.331.71 3.9217.36

(10.6 to 5) (3 to 7.5) (3.9 to 2.6) (25.3 to 33.2)

4.86.18 2.592.96 1.93.02 0.0131.59

(17.6 to 4) (8.6 to 1.3) (8.4 to 2.9) (56 to 69.5)

0.04* 0.001* 0.11 0.69

0.020.25 0.050.08 0.010.06 0.130.52

(0.34 to 0.78) (0.1 to 0.21) (0.11 to 0.08) (0.8 to 1)

0.140.23 0.080.09 0.060.09 0.010.89

(0.54 (0.25 (0.24 (1.45

0.08 0.001* 0.08 0.65

to to to to

0.25) 0.03) 0.09) 2)

P

FCS ¼ foveal center subfield; PRL ¼ photoreceptor layer; RPE ¼ retinal pigment epithelium; SDD ¼ subretinal drusenoid deposit. *Statistically significant value.

7

Ophthalmology Retina Volume -, Number -, Month 2018 Table 7. Change in Retinal Layer Thickness at Month 24 Visit in Eyes without Soft Drusen and with Soft Drusen* Soft Drusen Absent (n [ 70 eyes) FCS region Retina PRL outer segments RPEþdrusen complex Choroid Inner ring Retina PRL outer segments RPEþdrusen complex Choroid Outer ring Retina PRL outer segments RPEþdrusen complex Choroid Total thickness Retina PRL outer segments RPEþdrusen complex Choroid Total volume Retina PRL outer segments RPEþdrusen complex Choroid

Soft Drusen Present (n [ 15 eyes)

P

3.388.70 0.384.53 2.2810.20 0.8825.14

(29.60 (11.50 (47.70 (78.40

to to to to

9.8) 6.8) 8.60) 41.40)

7.005.34 6.156.22 0.656.78 4.9563.41

(11.30 to 0.50) (12.70 to 2.30) (7.20 to 8.80) (67.60 to 83.30)

0.2 0.06 0.98 0.98

3.829.13 0.383.77 1.425.32 0.5719.49

(33.70 (10.30 (23.20 (45.10

to to to to

5.10) 8.10) 4.70) 33.20)

12.0812.47 6.034.90 3.235.24 3.9566.76

(22.60 to 2.90) (12.60 to 1.00) (8.20 to 3.10) (72.70 to 90.00)

0.27 0.004y 0.43 0.98

1.094.45 0.412.52 0.501.94 2.3518.74

(12.40 to 5.30) (4.80 to 7.30) (4.20 to 3.70) (35.30 to 33.20)

3.955.77 5.103.23 2.502.02 4.5047.87

(9.10 to 1.90) (9.90 to 3.20) (5.50 to 1.10) (50.70 to 63.10)

0.2 0.001y 0.09 1

1.745.26 0.382.67 0.752.39 1.8918.79

(17.60 to 5.00) (5.60 to 7.50) (8.40 to 2.90) (38.70 to 33.20)

5.805.96 5.352.54 2.652.52 4.3851.82

(12.00 to 2.10) (8.60 to 2.50) (5.80 to 0.20) (56.00 to 69.50)

0.2 0.001y 0.16 0.98

0.040.25 0.010.08 0.020.07 0.050.55

(0.54 (0.17 (0.24 (1.05

0.130.23 0.150.08 0.070.07 0.161.43

(0.40 (0.25 (0.16 (1.45

0.75 0.003y 0.16 0.98

to to to to

0.78) 0.21) 0.08) 1.00)

to to to to

0.08) 0.06) 0.01) 2.00)

FCS ¼ foveal center subfield; PRL ¼ photoreceptor layer; RPE ¼ retinal pigment epithelium. *Presence of soft drusen at baseline visit. y Statistically significant value.

significantly more thinning of PRL outer segments in eyes with SDD. The long term accumulation of cellular debris and/or hyperreflective material in the subretinal space, leading to thinning of outer segments over time, may cause a chronic inflammatory stimulus to late AMD progression.22 As reported in previous studies,25e27 we also found that SDD was associated strongly with the thinning of the choroid. Switzer et al28 showed a lower subfoveal choroidal thickness in eyes with RPD. We found a similar significant difference in choroidal thickness, and the choroid was thinner farther from the center of the fovea in eyes with SDD than in those without SDD.29 Longitudinal change analysis revealed that the choroidal thickness was decreasing faster (50%) in eyes with SDD than in eyes without SDD (20%). Eyes with SDD may show a dynamic structural change in the choroid along with the progression of AMD.30 We observed that the PRL became thinner over time, particularly in eyes with soft drusen. Increasing size of soft drusen has been commonly reported in the nonneovascular stages of AMD, and the collapse of these drusen has been shown to be associated with subsidence and loss of the outer nuclear layer and PRL inner and outer segments. This study showed a significant reduction in PRL thickness over time in eyes with soft drusen compared with that in those without. Schuman et al,10 in a cross-sectional small study sample, found that the PRL was thinned over drusen. Nivison-Smith et al,31 in a study comprising around 50

8

subjects, observed that the PRL became noticeably thinner over regions with drusen growth. Although we did not specifically record drusen growth in this study, our findings of thinning of the PRL are consistent with these prior observations. In evaluating the relation between outer retinal substructures and DVA, Pappuru et al1 showed that PRL inner segments seem to have predictive value. In this study, we observed a significant relationship between outer segment thickness and low-luminance deficit. However, we found no correlation between changes in the chorioretinal microanatomy and the other clinical measures of vision such as DVA and NVA. Our findings support the view that testing function under low-luminance conditions is a more sensitive measure of outer retinal dysfunction and thus provides better information on early abnormalities before loss of DVA or NVA occurs.13 A previous study from our group13 evaluating the relation between the presence of RPD and functional vision impairment, showed that global assessments of macular function such as NVA and the low-luminance deficit were lower in eyes with RPD than in eyes without RPD. The SKILL test measures spatial vision under conditions of reduced contrast and luminance.14 It has been postulated that the accumulation of material deposition over the RPE may lead to abnormal adaptation processes and may affect photopigment regeneration; thus, it may be directly related to the SKILL functional vision assessment rather than to clinical methods of measuring DVA and NVA.

Nittala et al



Retinal Layer Thickness Change Analysis in Nonneovascular AMD

There are limitations to this study that should be considered. All study participants had neovascular AMD in their nonstudy eye and thus represent a more homogenous study cohort than that of previous studies. We studied only PRL outer segments and the RPEþdrusen complex as part of outer retinal layer analysis. We did not have 2 years of follow-up on all study subjects. The strengths of this study include its prospective longitudinal design; the use of a standardized DIRC grading procedure with expert, certified graders and quantitative analysis rather than qualitative assessments of layers; and multiple metrics of visual function measured using strict research protocols. In summary, we observed significant alterations in outer retinal and choroid layers and their association with visual function. In subjects with SDD, the RPEþdrusen complex layer is significantly thicker and the choroid is significantly thinner when compared with those in participants without SDD. The PRL outer segments seem to thin over time in eyes with SDD compared with those in eyes without SDD. These findings may be useful as early markers for detecting and monitoring the progression of AMD and for creating reading center protocols for future therapeutic trials in AMD. The information on longitudinal morphological change in the outer retinal layers can potentially serve as an SD-OCT biomarker for predicting late AMD.18,32 References 1. Pappuru RR, Ouyang Y, Nittala MG, et al. Relationship between outer retinal thickness substructures and visual acuity in eyes with dry age-related macular degeneration. Invest Ophthalmol Vis Sci. 2011;52:6743e6748. 2. Lu L, Xu S, He F, et al. Assessment of choroidal microstructure and subfoveal thickness change in eyes with different stages of age-related macular degeneration. Medicine (Baltimore). 2016;95:e2967. 3. Hariri AH, Nittala MG, Sadda SR. Quantitative characteristics of spectral-domain optical coherence tomography in corresponding areas of increased autofluorescence at the margin of geographic atrophy in patients with age- related macular degeneration. Ophthalmic Surg Lasers Imaging Retina. 2016;47:523e527. 4. Chakravarthy U, Bailey CC, Johnston RL, et al. Characterizing disease burden and progression of geographic atrophy secondary to age-related macular degeneration. Ophthalmology. 2018;125:842e849. 5. Verma A, Rani PK, Raman R, et al. Is neuronal dysfunction an early sign of diabetic retinopathy? Microperimetry and spectral domain optical coherence tomography (SD-OCT) study in individuals with diabetes, but no diabetic retinopathy. Eye (Lond). 2009;23:1824e1830. 6. Ebneter A, Jaggi D, Abegg M, et al. Relationship between presumptive inner nuclear layer thickness and geographic atrophy progression in age-related macular degeneration. Invest Ophthalmol Vis Sci. 2016;57:OCT299eOCT306. 7. Sadiq MA, Hanout M, Sarwar S, et al. Structural characteristics of retinal layers adjacent to geographic atrophy. Ophthalmic Surg Lasers Imaging Retina. 2015;46:914e919. 8. Wolf-Schnurrbusch UEK, Enzmann V, Brinkmann CK, Wolf S. Morphologic changes in patients with geographic atrophy assessed with a novel spectral OCT-SLO combination. Invest Ophthalmol Vis Sci. 2008;49:3095e3099.

9. Curcio CA, Medeiros NE, Millican CL. Photoreceptor loss in age-related macular degeneration. Invest Ophthalmol Vis Sci. 1996;37:1236e1249. 10. Schuman SG, Koreishi AF, Farsiu S, et al. Photoreceptor layer thinning over drusen in eyes with age-related macular degeneration imaged in vivo with spectral-domain optical coherence tomography. Ophthalmology. 2009;116:488e496. e2. 11. Querques G, Querques L, Forte R, et al. Choroidal changes associated with reticular pseudodrusen. Invest Ophthalmol Vis Sci. 2012;53:1258e1263. 12. Gella L, Raman R, Sharma T. Imaging drusens using spectral domain optical coherence tomography. Saudi J Ophthalmol. 2015;30:88e91. 13. Hogg RE, Silva R, Staurenghi G, et al. Clinical characteristics of reticular pseudodrusen in the fellow eye of patients with unilateral neovascular age-related macular degeneration. Ophthalmology. 2014;121:1748e1755. 14. Haegerstrom-Portnoy G, Brabyn J, Schneck ME, Jampolsky A. The SKILL card: An acuity test of reduced luminance and contrast. Invest Ophthalmol Vis Sci. 1997;38: 207e218. 15. Velaga SB, Nittala MG, Konduru RK, et al. Impact of optical coherence tomography scanning density on quantitative analyses in neovascular age-related macular degeneration. Eye (Lond). 2017;31:53e61. 16. Donoso LA, Kim D, Frost A, et al. The role of inflammation in the pathogenesis of age-related macular degeneration. Surv Ophthalmol. 2006;51:137e152. 17. Shen JK, Dong A, Hackett SF, et al. Oxidative damage in agerelated macular degeneration. Histol Histopathol. 2007;22: 1301e1308. 18. Karampelas M, Sim DA, Keane PA, et al. Evaluation of retinal pigment epithelium-Bruch’s membrane complex thickness in dry age-related macular degeneration using optical coherence tomography. Br J Ophthalmol. 2013;97:1256e1261. 19. Dunaief JL, Dentchev T, Ying G-S, Milam AH. The role of apoptosis in age-related macular degeneration. Arch Ophthalmol. 2002;120:1435e1442. 20. Folgar FA, Yuan EL, Sevilla MB, et al. Drusen volume and retinal pigment epithelium abnormal thinning volume predict 2-year progression of age-related macular degeneration. Ophthalmology. 2016;123:39e50. e1. 21. Christensen UC, Kroyer K, Thomadsen J, et al. Normative data of outer photoreceptor layer thickness obtained by software image enhancing based on Stratus optical coherence tomography images. Br J Ophthalmol. 2008;92:800e805. 22. Zweifel SA, Spaide RF, Curcio CA, et al. Reticular pseudodrusen are subretinal drusenoid deposits. Ophthalmology. 2010;117:303e312. e1. 23. Greferath U, Guymer RH, Vessey KA, et al. Correlation of histologic features with in vivo imaging of reticular pseudodrusen. Ophthalmology. 2016;123:1320e1331. 24. Klein R, Meuer SM, Knudtson MD, et al. The epidemiology of retinal reticular drusen. Am J Ophthalmol. 2008;145:317e326. 25. Ueda-Arakawa N, Ooto S, Ellabban AA, et al. Macular choroidal thickness and volume of eyes with reticular pseudodrusen using swept-source optical coherence tomography. Am J Ophthalmol. 2014;157:994e1004. 26. Garg A, Oll M, Yzer S, et al. Reticular pseudodrusen in early age-related macular degeneration are associated with choroidal thinning. Invest Ophthalmol Vis Sci. 2013;54:7075e7081. 27. Thorell MR, Goldhardt R, Nunes RP, et al. Association between subfoveal choroidal thickness, reticular pseudodrusen, and geographic atrophy in age-related macular degeneration. Ophthalmic Surg Lasers Imaging Retina. 2015;46:513e521.

9

Ophthalmology Retina Volume -, Number -, Month 2018 28. Switzer DW, Mendonça LS, Saito M, et al. Segregation of ophthalmoscopic characteristics according to choroidal thickness in patients with early age-related macular degeneration. Retina. 2012;32:1265e1271. 29. Zheng F, Gregori G, Schaal KB, et al. Choroidal thickness and choroidal vessel density in nonexudative age-related macular degeneration using swept-source optical coherence tomography imaging. Invest Ophthalmol Vis Sci. 2016;57:6256e6264. 30. Querques G, Canouï-Poitrine F, Coscas F, et al. Analysis of progression of reticular pseudodrusen by spectral domain-optical

coherence tomography. Invest Ophthalmol Vis Sci. 2012;53: 1264e1270. 31. Nivison-Smith L, Wang H, Assaad N, Kalloniatis M. Retinal thickness changes throughout the natural history of drusen in age-related macular degeneration. Optom Vis Sci. 2018;95: 648e655. 32. Abdelfattah NS, Zhang H, Boyer DS, et al. Drusen volume as a predictor of disease progression in patients with late agerelated macular degeneration in the fellow eye. Invest Opthalmol Vis Sci. 2016;57:1839e1846.

Footnotes and Financial Disclosures Originally received: June 8, 2018. Final revision: August 24, 2018. Accepted: September 14, 2018. Available online: ---. Manuscript no. ORET_2018_237. 1 Doheny Eye Institute, Los Angeles, California.

HUMAN SUBJECTS: Human subjects were included in this study. The human ethics committees at Queen’s University Belfast, University of Milan and University of California Los Angeles approved the study. All research adhered to the tenets of the Declaration of Helsinki. All participants provided informed consent.

2

Center for Experimental Medicine, Queen’s University Belfast, Belfast, United Kingdom.

No animal subjects were included in this study.

3

Conception and design: Chakravarthy, Sadda

Association for Innovation and Biomedical Research on Light and Image (AIBILI), Coimbra, Portugal.

4

Department of Clinical Science Luigi Sacco’ University of Milan, Milan, Italy.

5

Department of Ophthalmology, David Geffen School of Medicine at University of CaliforniaeLos Angeles, Los Angeles, California.

Funding: This study was funded by educational grants from Roche (R8821CEM) and Pfizer (A9011051). Financial Disclosure(s): The author(s) have made the following disclosure(s): U.C.: Grants e Roche Pharmaceuticals; Personal fees e Roche Pharmaceuticals. S.R.: Grants e Roche Pharmaceuticals; Personal fees e Allergan, Alimera, Bayer, Novartis, THEA. S.S.: Grants e Carl Zeiss Meditec, Optos, Allergan; Personal fees e Carl Zeiss Meditec, Optos, Allergan, Iconic Therapeutics, Novartis, Thrombogenics, Genentech, Heidelberg, Topcon, CenterVue, NightstaRx; Research instruments – Topcon, Heidelberg, Nidek, Carl Zeiss Meditec, CenterVue. G.S: Grants e Roche Pharmaceuticals; Personal fees e Roche Pharmaceuticals.

10

Author Contributions: Analysis and interpretation: Nittala, Hogg, Luo, Velaga, Silva, Alves Data collection: Nittala, Luo, Velaga Obtained funding: Silva, Staurenghi, Chakravarthy, Sadda Overall responsibility: Sadda Abbreviations and Acronyms: AMD ¼ age-related macular degeneration; DIRC ¼ Doheny Image Reading Centre; DVA ¼ distance visual acuity; ETDRS ¼ Early Treatment Diabetic Retinopathy Study; FCS ¼ foveal center subfield; GA ¼ geographic atrophy; ILM ¼ internal limiting membrane; NVA ¼ near visual acuity; PRL ¼ photoreceptor layer; RPD ¼ reticular pseudodrusen; RPE ¼ retinal pigment epithelium; SD ¼ standard deviation; SDD ¼ subretinal drusenoid deposits; SD-OCT ¼ spectral domain optical coherence tomography; SKILL ¼ Smith-Kettlewell Institute lowluminance. Correspondence: SriniVas R. Sadda, MD, Doheny Eye Institute, 1355 Sanpablo Street, Los Angeles, CA 90086. E-mail: [email protected].