Abstract
Purpose:
Identifying the most sensitive functional measure in intermediate age-related macular degeneration (iAMD) could help select an appropriate test for monitoring disease progression and evaluating the efficacy of novel interventions for the early stages of AMD. The purpose of the study was to determine which commonly used visual function test is the most discriminatory when comparing individuals with iAMD to normal participants.
Methods:
In this prospective observational study, iAMD cases and healthy controls underwent visual function testing (best corrected visual acuity (BCVA), low luminance visual acuity (LLVA), mesopic microperimetry, dark adaptation, and scotopic perimetry following photobleach), clinical eye examination, and multimodal retinal imaging in a single study visit. The data of each functional parameter were converted into z-score so that all the parameters had a common scale to allow a direct comparison between different functional parameters.
Results:
Forty-eight subjects (23 normal control, 25 iAMD) participated. Although all five parameters showed a significant reduction in function in iAMD eyes compared to controls (P ≤ 0.003), the rod intercept time (RIT) detected the greatest reduction in function followed by the scotopic sensitivity, mesopic sensitivity, BCVA, and LLVA, with the absolute mean z-score of 4.5, 2.2, 1.0, 1.0, and 1.2, respectively.
Conclusions:
Among the five visual function parameters commonly used, RIT is the most discriminatory functional parameter in the early stages of AMD.
Translational Relevance:
The RIT could be considered for assessing visual function and evaluating efficacy of novel interventions aimed at improving retinal function in eyes with early stages of AMD.
There are now a number of treatments aimed at preserving vision once neovascular AMD is present and there are several promising interventions being trialed which aim to slow the growth of geographic atrophy lesions. However, other than supplements and lifestyle advice, there is still no specific treatment proven to slow conversion of the earlier stages of AMD to late stage complications.
1,2 Designing trials that aim to intervene early to slow disease progression is difficult as traditionally they have been required to be large and long-term to ensure adequate power to detect a difference in conversion to late stage AMD. Much work is underway to better define earlier anatomical endpoints, now that multimodal imaging has allowed for earlier signs of disease progression to be identified.
3,4 However, it would be preferable to be able to show a benefit in a functional outcome, yet many functional tests are fairly normal or only moderately reduced in this earlier disease population. The difficulty then is to choose the most appropriate functional test that is likely to show a difference in the actively treated group compared to a placebo over a time frame of only a few years at most.
Several functional parameters, measured under different lighting conditions, have been investigated for their abilities to detect deficits in the early stages of AMD and to monitor functional changes over time. These parameters includes best corrected visual acuity (BCVA), low luminance visual acuity (LLVA), mesopic microperimetry, scotopic microperimetry, and dark adaptation (DA).
5–20 Reduced LLVA,
5–7 mesopic and scotopic sensitivity,
8–11 and increased rod intercept time (RIT)
12–18 during dark adaptation have been reported in eyes with large drusen. In addition, prolonged RIT has been shown to be worse in eyes with large drusen if they also have the reticular pseudodrusen (RPD) phenotype when compared to AMD eyes without RPD.
9,10,21
Although various levels of functional changes detected by these parameters have been reported, a direct comparison of these parameters, within the same eye, at the same time, in their ability to detect functional deficits in early stages of AMD have not been investigated. Hence, the purpose of this study was to compare functional results in these commonly performed tests to determine the magnitude of functional abnormality, in a cohort with bilateral large drusen when compared to normal participants; with all participants performing all of the tests in the same visit. It was hoped that the results will provide the evidence to help the decision making around which tests to include when designing interventional trials aiming to slow progression of the earlier stages of AMD.
All participants underwent an interview for systemic and eye history followed by all the functional testing, including BCVA, LLVA, dark-adapted chromatic perimetry, and mesopic microperimetry within a single visit. A comprehensive eye examination and multimodal retinal imaging were then performed to determine the AMD status. Only one eye, which was the eye with better BCVA, was selected as the study eye to undertake the mesopic and dark-adapted chromatic perimetry. If both eyes had the same BCVA, the right eye was selected as the study eye. The fellow nonstudy eye was patched during perimetric testing. Details about each of these functional tests have been described previously and briefly described as follows.
To make a direct comparison on the performance of different perimetric and DA parameters in detecting a visual dysfunction, only data at the four test points corresponding to identical locations in both the MAIA and DACP were included in the analysis. These test points were located at the 4° ring in the inferior, nasal, superior, and temporal retina (
Fig. 1B). The data of each test point were used to compare the performance between perimetric and DA parameters using a point-wise approach (four data points per eye). The comparison was performed using multilevel mixed-effects linear regression models with the study groups were considered fixed effects, test points nested within the participants were random effects, and age was a covariate. To compare the perimetric and DA parameters with visual acuity, we calculated the average sensitivity and RIT of the four data points for each subject (one data point per eye). Linear mixed models were used for the comparison in the performance between perimetric and DA parameters and visual acuity with age as a covariate. The fitness of the models was assessed by visual inspection of the distribution of the residuals. In a simple term, the linear regression model for each functional parameter can be described as the following equation:
\begin{eqnarray*}&&Functional\;parameter\\
&&\quad= {\beta _0} + \;{\beta _1} \times Group + {\beta _2} \times Age + {\zeta _0} + \varepsilon \;\end{eqnarray*}
Where β
0 through β
2 represent the fixed effects associated with the intercept, the study groups (control, no RPD, and RPD) and age, ζ
0 represents the random test locations nested within participant effect (only applicable to the perimetric and DA parameters), and ε represents the residual.
To overcome the potential problems with different dynamic ranges and scales among the functional tests, we used the data of the control group to calculate the z-score (standard deviation, SD) for each data point of each test for each subject. By using the z-score, measurements of all functional parameters were converted to a common scale, which allowed for a direct comparison between tests. An individual z-score within ±2 (<2 SD) was considered as within the normal range. The average z-score and 95% confidence interval (CI) of each functional parameter and study group were calculated and compared. All analyses were conducted using Stata software version 16.0 (Stata Corp, College Station, TX).
There are several functional parameters that are available and can be used to detect and monitor changes in function seen in the early stages of AMD. While most of these functional assessments are not routinely performed in clinical practice, they are often used in research settings to monitor AMD progression or assess the efficacy of potential interventions for early stages of AMD. However, when considering the design of an interventional study which aims to show a different trajectory in functional loss, it remains unclear as to which functional parameters are the most appropriate to use. In this study, all functional testing was performed in the same visit by the same AMD and control participants to allow a direct comparison among commonly used functional parameters to identify which of these parameters detects the greatest functional deficit in eyes with AMD. We found that the RIT parameter yielded the greatest functional deficits followed by the scotopic sensitivity. However, scotopic sensitivity had a relatively low discriminatory power between control and AMD eyes. The performance of the mesopic sensitivity, BCVA, and LLVA were similar in that they were not particularly sensitive in separating the function of AMD eyes without RPD from AMD eyes with RPD nor AMD eyes without RPD from controls.
Our findings suggest that tests performed in scotopic conditions are more sensitive and yield a greater functional deficit than tests performed in mesopic or photopic conditions. This was consistent with the literature which reports that rod dysfunction can be identified earlier than cone dysfunction in the early stages of AMD, and that eyes with RPD have worse rod dysfunction than eyes with only large drusen.
9,10,12,21 Hence, rod-mediated function such as the RIT and scotopic sensitivity, seem to be the most appropriate parameters for assessing functional changes in early stages of AMD. While all functional tests appeared worse in AMD eyes without RPD compared to controls the difference was not sufficient to differentiate from normal variation in many of the tests.
It is important to note that scotopic perimetry may not always reflect a rod-mediated response. The cellular contribution to the scotopic sensitivity depends upon the health status of both the rods and cones.
10,30 In the early stages of AMD when rod function is mildly reduced but cone function is normal, the scotopic sensitivity will be mediated by rods. However, when rod function is severely abnormal and potentially cone function remains relatively normal, the sensitivity under scotopic conditions could be mediated by cones. In normal eyes, rods are approximately 2.5 log units more sensitive than cones when measured with the 505 nm stimulus.
31 Hence, if cones are normal, this is the maximum magnitude of rod abnormality that could be detected using a scotopic perimeter with a full dynamic range of stimulus luminance. However, many current scotopic perimeters have a limited dynamic range, thus, subtle rod dysfunction may be missed, and the magnitude of rod dysfunction may not be reliability determined.
It is also worth noting that the scotopic sensitivity in this study was obtained after a photobleach. This was different to many of the previous studies in which no photobleach was applied prior to scotopic perimetry testing. Photobleach was used to minimize the variation in the level and duration of ambient light exposure among the participants and to activate the dynamic aspects of rod function, which has been shown to be more affected than static function in early stages of AMD.
32 We also have reported that scotopic sensitivity of AMD cases without RPD was indistinguishable from controls when tested without photobleach.
10 A difference in rod function between AMD without RPD and the control group was only observed when tested with preceding photobleach. Thus, scotopic perimetry performed after bleaching provides a greater ability to detect rod dysfunction in early stages of AMD.
In our study the participants performed all the tests in the same session. This is not usually the case where different cohorts perform different tests in the previous studies. Hence, the data collected in this study are unique and allow us to comment on which parameters appear most informative when considering dysfunction in the early stages of AMD. The RIT returned with the greatest functional deficit and as such appears to be the most appropriate parameter to test when assessing the efficacy of novel interventions aiming at slowing the decline in visual function in eyes with early stages of AMD, or indeed possibly even improving function. Furthermore, it has been shown that the greatest abnormal RIT is at 4° from the fovea, and it is relatively normal beyond 12° from the fovea,
10 thus measuring the RIT at multiple locations could help assessing both the safety and efficacy of novel interventions. It is recognized that dark adaptation is one of the hardest and least feasible tests to perform in a clinical trial setting, especially if needing this test to be performed at many sites on many participants. However, it does appear that the value in these results may outweigh the clinical difficulty and no number of noninformative tests, especially of photopic function will deliver the same informative data.
The strength of this study was that the mesopic and scotopic sensitivity and dark-adaptation parameters were collected at the same retinal locations, during the same session on the same patients, allowing a robust comparison among these functional parameters. There was also a large range of functional deficit which allowed a better estimation on the relationship between functional parameters. Although the number of subjects recruited for the study was relatively small, there were four data points obtained in each eye for the perimetric data and thus the total number of data points in each group were sufficient for the analysis. A potential limitation of the study was that AMD participants were older than the control subjects. However, all participants performed the same sequence of tests and thus the difference in the magnitude of dysfunction observed between tests should remain valid.
In conclusion, a direct comparison of commonly used tests for detecting a functional deficit in eyes with the early stages of AMD was performed and the RIT appears to be the most sensitive parameter in detecting the presence and magnitude of functional abnormality in eyes with AMD. Testing RIT at multiple locations in the retina is likely to deliver the best data to help show a difference in progression between arms in an intervention study in iAMD. The practicalities of conducting such tests at many sites on many participants remain a disincentive to include these tests. However, selecting an additional noninformative test will not compensate for the missing data that can only be obtained under scotopic conditions.
Supported by Ryan Initiative for Macular Research, Australia Awards Scholarship (RST), National Health and Medical Research Council (NHMRC) Fellowship (GNT1103013, RHG). The Centre for Eye Research Australia (CERA) receives Operational Infrastructure Support from the Victorian Government.
Disclosure: R.H. Guymer, None; R.S. Tan, None; C.D. Luu, None