Abstract
Purpose.:
To characterize changes in macular sensitivity during geographic atrophy (GA) progression using microperimetry.
Methods.:
Retinal sensitivity in the macular area was evaluated by microperimetry in 10 patients with bilateral GA, with adequate data obtained in 9 of 10 patients (n = 18 eyes). Patients had been enrolled in an interventional trial in which one eye had been randomized to treatment and the other eye observed. No treatment effect with regard to GA growth and microperimetric measurements was detected, and all eyes were analyzed. Microperimetric assessments of the central 20° of the macula were performed every 6 months over 24 months. Parameters analyzed included number of scotomatous points, mean retinal sensitivity of responding points, and fixation stability. Autofluorescence imaging and fundus photography were also obtained.
Results.:
Microperimetric parameters demonstrated statistically significant changes as a function of time. Mean number of scotomatous points increased significantly with time (P = 0.004) at a rate of 4.4 points/year. Mean retinal sensitivities of all points, all responding points, and all perilesional points all decreased significantly with time (P < 0.003), as did fixation quality within the 2° and 4° circles (P < 0.002). The growth of GA lesion area was associated with the changes in the number of scotomatous points (P = 0.01) but not with changes in the other microperimetric parameters.
Conclusions.:
Macular sensitivity and fixation quality undergo progressive change during the GA progression, reflecting alterations in macular function extending beyond the GA lesion proper. Microperimetric measurements may provide useful functional outcome measures for the clinical study of GA.
Central geographic atrophy (GA) is the advanced atrophic form of age-related macular degeneration (AMD) that is a cause of progressive moderate and severe vision loss,
1,2 estimated to affect 3.8 million people in the United States by 2050.
3 GA is characterized by central areas of atrophy of the retina, retinal pigment epithelium (RPE), and choroid that enlarge and coalesce with time.
4,5 Because there is currently no effective treatment for GA to prevent either its onset or progression,
6,7 clinical trials employing useful and relevant outcome measures in the discovery of a treatment are of important public health significance.
8
The natural history of the anatomic changes associated with GA progression has been closely examined using a variety of fundus imaging modalities including fundus photography,
9 autofluorescence imaging,
10 and optical coherence tomography.
11 Because the gradual time-dependent increase in the total area of atrophy of the retina-RPE-choriocapillaris complex is a key feature of GA progression, change in area of the atrophic lesion has constituted a key anatomic outcome measure in interventional studies of GA.
7 At the same time, it is important to demonstrate concurrent development of visual function loss associated with the increase in area of GA. Best corrected visual acuity, as determined using current standard clinical protocols,
12 often does not correlate well with GA progression,
13 is influenced by the visual function of the contralateral eye,
14 and may not comprehensively reflect visual disability of patients with GA.
15 Previous studies have revealed that patients with GA demonstrate low luminance visual dysfunction and reduced reading speed,
16,17 but how these measures change with time, or correlate with anatomic outcome measures such as change in the area of GA lesion, have yet to be fully investigated.
Microperimetry is a testing modality that can be used to measure and map central retinal sensitivity in macular diseases such as GA.
18,19 The MP-1 microperimeter (Nidek, Padua, Italy) is a commercially available device that performs fundus tracking and automated image alignment for eye movements, permitting a precise and repeatable mapping of retinal sensitivities of discrete points in the macula.
20 This device has been previously used to map retinal sensitivity in various macular disorders, including central serous choroidopathy, macular telangiectasia, diabetic macular edema, and multiple evanescent white dot syndrome.
21 –27 In the present study, we have used the MP-1 perimeter to follow retinal sensitivities in the central macula in eyes with GA over a follow-up period of 24 months. Our results demonstrate that microperimetric parameters can reveal significant aspects of functional decline in GA, both within and beyond the GA lesion, and may provide potential functional outcome measures for future clinical trials for GA.
In the present study, we have characterized functional changes in GA progression using the MP-1 microperimeter and have analyzed parameters that reflect retinal sensitivity in the macula, both within and around GA lesions and fixation quality. By testing eyes with GA at regular time intervals using a standard microperimetric protocol over 24 months, we have detected patterns of progressive change that reveal the nature, rate, and location of functional decline inherent during GA progression.
In our study, we discovered that a number of microperimetric measures undergo significant change as a function of time. One of these is the total number of scotomatous (or nonresponding) points, which increased progressively with follow-up time. This finding indicated the ability of the MP-1 microperimeter to detect and measure the expansion of the functional scotomata expected from the anatomic enlargement of the GA lesion. We had calculated a mean annual rate of increase of scotomatous points of +4.4 points/year, which we estimated to correspond to an increase in the area of nonresponsive retina by 1.85 mm
2/year (estimating the tested 20° field to be 6000 μm in diameter and 28.3 mm
2 in area, and the approximate area covered by each tested point to be 28.3/68 = 0.42 mm
2), a value that is comparable to previous estimates for the annual growth of GA area.
6,13,30 –32 We also found that the change in the number of scotomatous points could be statistically associated with the increase of GA lesion area as measured by CFP and HRA FAF. This relationship confirmed the ability of MP-1 microperimetry to commensurately quantify changes in scotoma size occurring as a result of GA lesion expansion.
We also examined microperimetric parameters relating to the sensitivity of macula areas outside the GA lesion. In study eyes, the retinal sensitivity measured in areas outside the GA lesion were in general lower than those found in equivalent areas in healthy aged subjects,
20 and that the mean sensitivity, considering all responding points, declined significantly as a function of time. This result indicated that functional decline in GA involved not only an increase in the size of an absolute scotoma but also a general decrease in macular sensitivity in the tested area.
Segmentation of these responding points into perilesional points and extralesional points demonstrated that this time-related decrease in sensitivity occurred to a greater extent in areas closer to, rather than farther away from, the GA lesion borders. We found that perilesional points were consistently less sensitive than extralesional points in multiple microperimetric assessments and declined in sensitivity more rapidly with follow-up time. Previous cross-sectional studies have documented sensitivity loss by perimetry in the so-called junctional zone of GA lesions,
18,19 relating this to the excessive accumulation of lipofuscin in RPE cells in areas bordering the GA lesion. Autofluorescence imaging data in GA had indicated that perilesional autofluorescence features may influence the GA progression rate, although this may still be controversial.
32,33 Although many of the eyes analyzed in the present study contained areas of increased fundus autofluorescence (FAF) signal in the perilesional zones, the relatively low density of testing loci used here precluded an analysis of the correlation between macular sensitivity and abnormal FAF signal. Finally, we did not detect a correlation between the magnitude of change in retinal sensitivity and the size of increase in GA lesion area. The absence of a detectable relationship may have resulted from imprecision in the quantification of retinal sensitivity using a relatively sparse array of tested points; alternatively, it may be that the more general loss of macular sensitivity in areas around GA lesion occurs as a separate process from the local expansion of the GA lesion itself. Future studies may investigate further the loss of retinal sensitivity in GA as a separate functional outcome variable and its potential correlation with other causative influences.
34,35
We found that fixation stability, as quantified by the MP-1 microperimeter, decreased significantly as a function of time. Previous studies have reported on the changes in the nature and location of fixation patterns in GA patients over time,
17,36 but quantification of fixation quality over time in GA has not been previously reported. Previous investigators have shown that although MP-1 parameters for fixation may grossly correlate with reading speed, measurements of fixation with the bivariate contour ellipse formula may correlate better.
37,38 Despite this limitation, the fixation quantification method used by the MP-1 microperimeter provides a rapid recording of fixation ability that agrees with previous measurement methods
39 and an accessible method for detecting progressive change in fixation ability. Our results indicate that fixation quality may indeed decline progressively during GA progression, although refinements in methodology and further validation may be necessary to establish this as an outcome measure in GA.
The present study contains a number of limitations, including a relatively small number of participants, an exposure to a novel topical agent in an interventional study,
28 and a reliance on the accuracy of both the device and the study participant. The study relies on the accuracy of the eye-tracking software used by the MP1 device to return consistently to the tracked retinal loci between examinations. We have also not quantified the test–retest variability in our patient population; however, previous studies examining this issue in healthy volunteers
40 and patients with macular disease
41 have reported that mean measures, such as those used in this study, have the lowest variability and may be recommended for monitoring macular function. These mean or “global” measures have the shortcoming of not providing spatial information regarding the location of interval changes. On the other hand, the present study had the advantage of employing MP-1 assessments that were collected using a standard testing pattern at regular 6-month intervals over a period of 2 years, increasing the likelihood that actual changes in retinal sensitivity are being detected and measured. Also, corresponding stereoscopic fundus photographs and autofluorescence images were obtained on the same visit as the MP-1 assessments, with the GA lesion areas being quantified by masked graders at a reading center, promoting the accurate correlation of MP-1 parameters to GA lesion area change. We had considered a priori that there may be a confounding “learning” effect in the subjective performance of microperimetry, in which participants gain more familiarity and thus “improve” with repeated testing.
42 However, all microperimetric parameters (except for macular sensitivity of extralesional points) demonstrated statistically significant longitudinal trends that were in the opposite direction of any potential learning effect and cannot be explained as arising from repeated testing effects.
Taken together, our results indicate that the progression of functional decline in GA involves not only an expansion of an absolute scotoma, as can be expected from the growth of a central atrophic lesion, but also a more widespread decrease in retinal sensitivity, particularly in perilesional areas, as well as a loss of fixation stability. These changes are apparent in individual cases over a relatively short period of 24 months and are statistically associated with follow-up time when considered as a group. These trends may be relevant to the development of outcome measures in clinical trials of GA. As visual acuity in GA is not an outcome measure that correlates well with follow-up time or GA lesion expansion, there is a need for relevant and feasible functional outcome measures that relate to GA progression. As the microperimetric findings described here are relevant to the visual function of patients and correlate highly with the course of progression of GA, we propose that microperimetry may deserve further consideration for development as a useful functional outcome measure for future clinical trials of GA.
Supported by funding from the National Eye Institute Intramural Research Program.
Disclosure:
A.
D.
Meleth, None;
P.
Mettu, None;
E.
Agrón, None;
E.
Y.
Chew, None;
S.
R.
Sadda, Carl Zeiss Meditech (F), Heidelberg Engineering (C), Topcon Medical Systems (P);
F.
L.
Ferris, None;
W.
T.
Wong, None