Age-related macular degeneration (AMD) is a degenerative macular disorder and a major cause of visual impairment in individuals over 50 years of age in industrialized countries.1 2 AMD can be subdivided into early and late stages. Early AMD is characterized by soft intermediate drusen (larger than 63 μm) and areas of hyperpigmentation or hypopigmentation of the retinal pigment epithelium (RPE).3 Drusen are attributed to distinct deposits between the inner collagenous layer of Bruch’s membrane and the basement membrane of the RPE exceeding those of the normal aging process.4 Ultrastructural findings of material deposited in the inner Bruch’s membrane are in accordance with the assumption that the material predominantly derives from the RPE cells. All human RPE cells show an age-related accumulation of lipofuscin granula within the lysosomal compartment, as a byproduct of the permanent phagocytosis of lipid-rich distal photoreceptor outer segments.5 Lipofuscin is considered a biomarker for cellular aging and a cumulative index for oxidative damage. Lipofuscin contains the dominant fluorophores responsible for the in-vivo fundus autofluorescence (FAF) phenomenon as determined by fundus spectrophotometric measurements.6 7

With the advent of confocal scanning laser ophthalmoscopy it is possible to visualize FAF and its spatial distribution in vivo easily, and therefore to evaluate RPE during the aging process and in ocular disease.6 8^–12 It is believed that lipofuscin accumulation precedes photoreceptor degeneration, and is responsible for measurable visual deficits, abnormal retinal sensitivity measurements in the central visual field, and slower rates of dark adaptation associated with age-related maculopathy.13^–17 A novel fundus-related perimetry technique called microperimetry has recently been introduced into clinical practice in order to evaluate macular function by exactly correlating fundus lesions to retinal sensitivity.18^–23

The aim of this study was to compare macular microperimetry with fundus lesions and FAF (increased or decreased) in patients with early AMD.

MATERIALS AND METHODS

In this cross-sectional study, 13 consecutive and eligible bilateral early AMD patients were enrolled. All patients were recruited from the Medical Retina Clinic at the Department of Ophthalmology, University of Padova. The inclusion criteria were the presence of small drusen (⩽63 μm) and at least five or more intermediate soft drusen (>63 μm) or large (or confluent) soft drusen (>125 μm) localized within 3000 μm from the centre of the fovea and associated with pigment abnormalities. Visual acuity was 20/20 in all patients. Exclusion criteria were significant media opacities, clinical evidence of geographical atrophy (defined as hypopigmentation ⩾125 μm), choroidal neovascularization, any evidence of diabetic retinopathy, myopia greater than 6 D, glaucoma, and previous macular laser treatment. One eye of each patient, randomly selected, was studied.

This study was conducted in accordance with the tenets of the Declaration of Helsinki, and with the approval of our institutional ethical committee.

After a detailed explanation of the purpose of the study, a written consent form was signed by all enrolled patients. The ophthalmological examination consisted of: refraction and best corrected visual acuity determination; anterior segment examination; 90 dioptres lens biomicroscopy; fundus photography; microperimetry; and FAF.

Best corrected distance visual acuity for each eye was measured at 4 m distance using standard early treatment diabetic retinopathy study protocols with a modified early treatment diabetic retinopathy study distance chart transilluminated with a chart illuminator (Precision Vision, Bloomington, Illinois, USA).24 Visual acuity was scored as the total number of letters read correctly and expressed as the Snellen equivalent.

Microperimetry

All patients underwent at least 5 mm pupil dilation with tropicamide 0.5% and 15 minutes mesopic conditions adaptation before examination.

Microperimetry was performed on all subjects using a new automatic fundus-related perimeter: the MP1 Microperimeter (Nidek Technologies, Padova, Italy). This instrument has previously been described in detail.18 Briefly, for the purpose of this study, the following parameters were used: a fixation target consisting of a red ring 2° in diameter; white, monochromatic background at 4 asb, stimulus size Goldman III with 200 ms projection time; customized radial grid of 61 stimuli covering the central 10° (centred onto the fovea); stimuli were projected randomly onto the 0°, 30°, 60°, 90°, 120°, 150° axes, 1° apart (fig 1). A 4–2–1 double staircase strategy was used with an automatic eye tracker that compensates for eye movements.18 25 In order to allow for better clinical correlation between microperimetric data and retinal details, functional results are always displayed onto a colour digital retinograph, acquired by the charge-coupled device colour camera of the MP1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 A microperimetric customized radial grid pattern of 61 stimuli covering the central 10° (centred onto the fovea). The stimuli are projected randomly onto the 0°, 30°, 60°, 90°, 120°, 150° axes, 1° apart.

Autofluorescence

FAF was recorded using a confocal scanning laser ophthalmoscope (Heidelberg Retinal Angiograph, HRA 2; Heidelberg Engineering, Heidelberg, Germany). The optical and technical principles of the HRA have previously been described in detail.26 27 In order to amplify the autofluorescence signal of the final image, 10 acquired images were aligned and a mean image was calculated from these, after detection and correction of eye movements were performed by image analysis software. Digital images were saved on hard disc for further analysis and processing. FAF images were graded for different patterns (normal, increased and decreased). Increased or decreased FAF in the central retinal 10° were assumed to be present on the basis of the intensity of the background FAF pattern encountered just immediately peripheral to the superior temporal arcade.

Colour photos

Digital colour images were taken by a Topcon TRC 50IA 30 degree fundus camera (Topcon, Tokyo, Japan), on the same day as the MP1 and autofluorescence images were taken and immediately saved in JPEG format.

Image analysis

For every single retinal point tested with the MP1 on each eye, the presence and type of drusen (small, intermediate, large), pigment (present, absent) and FAF pattern (increased FAF, decreased FAF, normal FAF) were graded. A microperimetry sensitivity map was also overlaid onto the FAF images using dedicated MP1 software. This software allows for the exact superimposition of sensitivity data to different images separately (FAF, fluorescein angiography, indocyanine green). The superimposition is obtained by the semi-automatic detection of three identical anatomical landmarks on both images (fig 2). This software has previously been tested for validity and repeatibility.28 Thereafter, in every single pair of images correlation between the drusen type, RPE pigment and RPE atrophy (smaller than 125 μm) and distinct FAF pattern was evaluated.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Fundus autofluorescence and colour images of the same patient with large soft drusen and pigment abnormalities with the overlapped sensitivity map. The numerical map (values of sensitivity expressed in decibels) shows the sensitivity decrease over areas with increased fundus autofluorescence (on the left) that correspond to drusen and pigment (on the right).

Independent, combined grading of colour fundus images and FAF images was carried out by two trained retinal specialists. In the case of disagreement, arbitration grades were awarded by a different retinal specialist and these results were taken as final.

Statistical analysis

All sensitivity data were analysed using an analysis of variance test for repeated measures analysis of variance and a Bonferroni post-hoc test. FAF data were analysed using the chi-square test. To assess interobserver variability, kappa statistics were calculated.29

We assumed that data were significant if p was less than 0.05. The SAS statistical package on a personal computer (version 8.4; SAS Institute Inc., Cary, North Carolina, USA) was used for all analyses.

RESULTS

Of the 13 patients examined, three were men and 10 were women, with seven left and six right eyes included. The mean age was 76.2 ± 6.55 years and mean visual acuity was 20/20. Anterior segment examination was unremarkable, with no significant media opacities.

Of 793 total retinal points examined (61 for each patient), 310 points presented with drusen (39.1%), 483 points had no drusen (60.9%), whereas 161 points had pigment alterations (20.3%) and 632 were without pigment alterations (79.7%); 697 points had normal FAF (87.9%), 30 decreased FAF (3.8%) and 66 increased FAF (8.3%) (see table 1).

Retinal sensitivity was significantly influenced by both drusen and RPE pigment abnormalities (analysis of variance p<0.0001). In particular, retinal sensitivity was not significantly decreased over small or intermediate drusen (p>0.05), whereas it was significantly decreased over large drusen when compared with an absence of drusen (15.0 ± 4.5 versus 11.2 ± 5.6 dB, Bonferroni test p<0.0001). Retinal sensitivity was also decreased when pigment abnormalities were found compared with their absence (15.0 ± 4.5 versus 13.1 ± 3.6 dB, p<0.0001). If RPE pigment abnormalities were associated with large soft drusen sensitivity reduction was even higher if compared with its association with small or intermediate drusen (9.6 ± 4.3 versus 12.3 ± 0.6 dB, p<0.0001) (table 2). In all eyes fixation was central and stable.

The association between the FAF pattern (normal FAF, decreased FAF and increased FAF) and the presence of drusen was significant (p<0.0001). Increased and decreased FAF was found in only 27 of the examined points without drusen (6.4%), one of the examined points with small and intermediate drusen (1.7%) and in 32 of the examined points associated with large soft drusen (20.4%). Of 67 hyperpigmented points without drusen, seven (10.4%) presented with decreased FAF, four (6.0%) with increased FAF and 56 (83.6%) with a normal FAF pattern. When large soft drusen were associated with hyperpigmentation, 24 (26.4%) presented with increased FAF, one (1.1%) with decreased FAF and 66 (72.5%) with a normal FAF pattern. There were no points with small drusen and pigment abnormalities, and there were only three points with intermediate drusen associated with pigment abnormalities characterized by normal FAF (table 3).

The mean sensitivity was significantly decreased over areas with altered FAF (increased and decreased) when compared with areas with normal FAF (9.7 ± 5.5 and 10.2 ± 6.4 dB versus 14 ± 4.7 dB, p<0.0001).

The overall interobserver variability resulted in a kappa of 0.74 (95% confidence interval 0.64–0.84).

DISCUSSION

In the past, macular function in early AMD was studied by means of visual acuity, macular recovery function, central visual field sensitivity and spatiotemporal contrast sensitivity determinantion.30 More recent studies used short wavelength automated perimetry for the study of cone sensitivity, and a photopic and scotopic fine matrix mapping technique for the study of cone and rod sensitivity.31 32 Retinal sensitivity loss was documented regardless of good visual acuity in patients with early AMD. The fine matrix mapping technique detected more rod than cone sensitivity decreases.32 Both studies agreed that functional alterations correlate with funduscopic abnormalities (large soft drusen and pigmentary RPE changes).31 32 Conventional (automatic) perimetry has been used in the past to study early AMD, but the quantification of retinal thresholds over small and descrete retinal lesions were beyond the possibilities of the technique.33 Microperimetry, also known as fundus-related perimetry, allows for a detailed examination of macular function, giving a highly reproducible evaluation of retinal sensitivity at selected points, by a continuous eye tracking system.18 The semiautomatic MP1 software developed for this study allowed us an exact, point-to-point correlation between fundus lesions and FAF and sensitivity values. We found a significant sensitivity reduction over the areas characterized by large soft drusen using microperimetry, although this is not in accordance with previously published data obtained with traditional perimetry.34 Sensitivity deterioration was greater when pigment abnormalities were associated with drusen. Large soft drusen and RPE pigment alterations were mutually independent, and large soft drusen had a greater impact on retinal sensitivity than RPE pigment abnormalities alone. This fact can be explained by RPE and photoreceptor degeneration as a result of the presence of drusen. Johnson et al.35 reported that both inner and outer rod and cone segments overlying drusen were physically deflected. Outer photoreceptor segments overlying drusen were shorter than nearby outer segments not overlying drusen. This was correlated with drusen size.35 Johnson et al.36 also observed a consistent reduction in the density of photoreceptors over drusen, with an approximate reduction of 30% in photoreceptor numbers. These retinal abnormalities are correlated with drusen size and can be functionally documented with microperimetry. We found a more relevant decrease of macular sensitivity over large soft drusen compared with small or intermediate drusen, despite normal visual acuity.

We found that the FAF pattern was altered (increased or decreased) in only 20.4% of the points examined that were associated with large soft drusen. Previous studies have shown that not all types of drusen have the same FAF pattern, which might be explained by different fluorophores contained by the lipofuscin and consequent different stages of RPE degeneration.37 The material deposited might also be at different levels in the retina and therefore not always detectable by means of FAF.38

Our data show a significant and inverse correlation between retinal sensitivity and an increased FAF pattern, confirming that lipofuscin accumulation in the RPE cells has a functional correlate in patients with early AMD.39 Both clinical fundus FAF and experimental findings show that lipofuscin contains toxic biomolecules that may interact with normal cell function. The major retinoid fluorophore found in lipofuscin is A2E, which is responsible for the inhibition of lysosomal enzymes, detergent and phototoxic properties within the RPE cells.4 All these molecular pathophysiological effects increase apoptotic mechanisms with the consequent death of RPE cells. Clinical entities characterized by the accumulation of lipofuscin thus point to RPE cell dysfunction with secondary functional correlates. Therefore, comparing fundus morphological features with functional data may allow for a better understanding of the mechanisms underlying this pathological process, and may suggest a more appropriate follow-up for patients.

Further prospective studies on larger samples are needed in order to obtain a better understanding of the evolution of large soft drusen and pigment abnormalities characterized by increased FAF and decreased sensitivity, and whether patients with these characteristics are at greater risk of developing advanced stages of AMD when compared with areas with clinical features of early AMD and normal FAF.