Introduction The ellipsoid region of the photoreceptor inner segments is easily identified in the latest generation of spectral domain OCT as the first (of 3) hyper reflective bands located on the outer aspect of the neurosensory retina; this band has been termed ellipsoid zone (EZ)1. Seventy-five percent of the volume of the ellipsoid region is accounted for by mitochondria2, providing the energy essential for phototransduction processes. It is therefore not surprising that the integrity of the EZ seen on OCT is related to retinal sensitivity3. The Radial Shape Discrimination (RSD) test is a global hyperacuity visual function test4 requiring images of the target to fall on a structurally and functionally intact areas of the retinal photoreceptor mosaic. We investigated the relationship between EZ hyper-reflectivity over the foveal area and performance on a handheld version of the RSD5 test in participants with age-related macular degeneration (AMD). We hypothesised that participants with more disruption to the EZ band would show worse RSD performance.

Methods

Results Participants (n=85, 49 females) were 77±8 years old (mean±SD; range 57 to 91 years). OCT data from all but 7 (8%) was gradable (Table 1). Of 78 graded, 45 (53%) had some EZ disruption at the fovea.

References 1 Staurenghi G et al Ophthalmology 2014;121(8):1572-8.

2 Spaide RF, Curcio CA Retina. The Journal of Retinal and Vitreous Diseases 2011;31(8):1609-19.

3 Wu ZC et al Ophthalmology 2014;121(7):1445-52.

4 Wilkinson F et al Vision Research 1998;38(22):3555-68.

5 Wang Y-Z et al Investigative ophthalmology & visual science 2013;54(8):5497-505.

6 Leuschen JN et al Ophthalmology 2013;120(1):140-50.

npitre@liverpool.ac.uk

Conclusions • A statistically significant decrease in RSD scores, but

not BCVA, was observed in eyes with greater ellipsoid zone disruption.

• The RSD test is more sensitive to dysfunctional photoreceptors than BCVA in eyes with early AMD at risk of developing nAMD.

• The RSD test, as used here, is easy to operate and handle and, unlike BCVA, can be used in unsupervised self monitoring of vision.

• Given the emergence of effective treatments for nAMD and the need for early detection, we are currently investigating the ability of the RSD test to detect early nAMD (EDiMaD project).

Relationship between foveal ellipsoid zone integrity and central visual function in age related macular degeneration

Noelia Pitrelli Vazquez1, Simon P. Harding1,2, Heinrich Heimann2, Paul C. Knox1

1Department of Eye and Vision Science, University of Liverpool, UK, 2St Paul’s Eye Unit, Royal Liverpool University Hospital, Liverpool, UK

Optical coherence tomography (OCT) • Data obtained from routine clinic scans • Heidelberg Spectralis OCT operated by trained

ophthalmic technicians • 19 single B-scans (1 volume, 20x15°) centred at fovea.

Participants Clinical sample of participants attending AMD service Study eye: • BCVA better than 0.4 logMAR • Any stage of early/intermediate AMD but no

neovascular (nAMD) or central geographic atrophy Fellow eye: • nAMD, being monitored/treated.

Discussion While drusen assessment in colour fundus photographs is the current gold standard to assess change in severity of early AMD, foveal drusen alone do not correlate well with visual function6. The advent of high resolution OCT has given rise to new potential markers of relevance to AMD, one of which is the integrity of the EZ. It has been reported that EZ integrity is related to retinal sensitivity measured by microperimetry3, suggesting that a loss of EZ hyper reflectivity indicates loss of functional photoreceptors (although not necessarily their death).

Theoretically, normal thresholds for RSD (in the hyperacuity range) are only achievable with a relatively intact photoreceptor mosaic over the area of retina on which the image of the RF pattern falls. Our results are consistent with loss of functional photoreceptors in early AMD causing a decrease in the ability to globally process radial frequency patterns, leading to worse RSD thresholds.

Acknowledgements This work was supported by a project grant from The Dunhill Medical Trust (grant number: R283/0213). We are also grateful to Vital Art & Science Inc for software and support and Sarah Houston for OCT grading protocol development.

Radial Shape Discrimination test5 • Presented on an Apple iPod Touch • Three radial frequency patterns5 presented

simultaneously (spatial 3AFC). One (the target) sinusoidally modulated (frequency of 8 cycles; Fig 1)

• Hyperacuity thresholds presented as a logMAR value (more negative implies a better result).

BCVA • ETDRS 4m R+L charts with updated refraction • Chart moved to 1m if first 20 letters not seen.

Fig 4 Example of AMD changes seen on OCT with sections of abnormal EZ hyper reflectivity (arrows).

Analysis • OCT graded for quality (good, fair, ungradable), foveal

EZ disruption, foveal drusen and foveal thickness • 1 blinded grader (NPV) • SPSS v 22.

Fig 1 Example of stimuli used in the RSD test. The modulations

on the RF pattern decreased in size. The task was to touch the

shape that was modulated (arrows).

Fig 6 Relationship between the amount of EZ disruption and central visual function (RSD, left and BCVA, right) for study eyes and fellow, nAMD eyes, covering a wider sample of the spectrum of AMD. A stronger association (r=0.67 vs 0.45) and steeper linear regression slope (0.0054±0.0005 and 0.0029±0.0004) were seen for RSD compared to BCVA.

Fig 7 Examples of EZD (arrows) occurring on its

own (top) and with large drusen (bottom).

Table 1 Overall OCT quality.

Procedures RSD and BCVA were tested in clinic with appropriate up to date optical correction, always before OCT. Both eyes were always attempted.

Fig 2 Foveal area graded covered all B-scans falling within the central ring of the ETDRS grid (highlighted yellow), centred at fovea.

Central ring

Fovea

Rating Definition N(%)

Good Correct scan positioning, no evidence of significant artefacts or noise 67 (79%)

Fair Poorer than “Good”, but enough evidence to proceed with the grading 11 (13%)

Ungradable Quality too poor to proceed with grading 7 (8%)

Fig 5 Radial shape discrimination (RSD) and best corrected visual acuity (BCVA) for study eyes of participants with intact ellipsoid zone (EZD=0%), less than 20% and more than 20% of EZ disruption. A statistically significant difference was seen in mean RSD (p=0.01) but not in mean BCVA (p=0.89). Error bars indicate ±95%CI. Boxes above show groups age, mean RSD and mean BCVA.

Intact EZ (n=33, 42%) Age: 77±8 years RSD: -0.62±0.15 logMAR BCVA: 0.05±0.14 logMAR

<20% EZD (n=17, 22%) Age: 79±7 years RSD: -0.57±0.16 logMAR BCVA: 0.05±0.11 logMAR

≥20% EZD (n=28, 36%) Age: 76±7 years RSD: -0.51±0.15 logMAR BCVA: 0.05±0.13 logMAR

Ellipsoid Zone Disruption (EZD) Foveal EZD was defined as a continuous section of a B-scan (falling within the central ring of the ETDRS grid; Fig 2) where the EZ (Fig 3) was either absent or hyporeflective (yet not completely missing); and which might be accompanied by drusen (Fig 4). The lengths of all B-scans where the EZ was intact and disrupted were summed and transformed into a percentage of the total. The level of EZD disruption was categorised into 3 groups: “intact EZ”, “less than 20%” and “more than 20%” disruption.

Fig 8 Relationship between RSD and CST was not statistically significant (p=0.25).

Fig 3 Normal OCT scan cutting through the fovea with the boundaries of the EZ highlighted in yellow.

Ellipsoid zone Fovea

Drusen and foveal thickness 64% of participants with EZD also had large drusen at the fovea (defined as 70µm in maximum height, including the RPE). No statistically significant difference in RSD was observed between those with EZD only and those with EZD and drusen (mean difference in RSD: 0.05 logMAR, F(1,42)=1.03, p=0.32). Mean±SD CST was 282±23 (95%CI 277 to 287) µm. No statistically significant correlation was seen between CST and RSD (Spearman’s r=-0.12, p=0.25) after the effect of age was accounted for.

Ellipsoid Zone Disruption (EZD) The level of EZD in study eyes influenced RSD scores (p=0.01) but did not affect BCVA (p=0.89, Fig 5). While the proportion of total RSD variance explained by EZD was small (Partial ŋ2=0.12) there was a clear statistically significant linear trend in the data (Polynomial contrasts, p=0.003). The effect of EZD was much more noticeable when the fellow, nAMD eyes were included in the analysis, covering more of the AMD spectrum (Fig 6). A stronger correlation was seen for RSD (Spearman’s r=0.67) than for BCVA (Spearman’s r=0.45). The slopes of the regression lines for RSD and BCVA (0.0054±0.0005 and 0.0029±0.0004 respectively) were statistically significantly different (p=0.0005). Results remained unchanged when 0% and 100% EZD were excluded from the analysis and when study eyes and fellow eyes were randomised in order to avoid both eyes of the same participant included in the analysis.

Poster pdf

Links Please feel free

to scan

EDiMaD project