- Top of page
- MATERIALS AND METHODS
Background: The aim was to evaluate the macular structure and function in children with human immunodeficiency virus (HIV) disease without cytomegalovirus retinitis or visual symptoms.
Methods: Thirty-eight eyes of 19 HIV-positive children (Group A) were examined. Group B included 20 (40 eyes) age- and sex-matched control subjects. Each individual underwent a complete ophthalmic examination, optical coherence tomography (OCT) scan and multifocal electroretinogram (mfERG) recording.
Results: In all patients, visual acuity and colour vision testing were normal. The mean foveal thickness in groups A and B was 190.28 ± 26.58 (SD) µm and 169.47 ± 10.17 µm, respectively (p = 0.0002). In Group A, the mean retinal response density of the fovea (area 1) was 19.87 ± 10.16 nV/deg2 and the latency was 38.56 ± 1.18 ms. In the parafoveal area (area 2), the mean retinal response density was 10.82 ± 2.34 nV/deg2 and the mean latency was 36.52 ± 1.73 ms. In the perifoveal area (area 3), the mean retinal response density was 10.83 ± 0.90 nV/deg2 and the mean latency was 36.36 ± 1.90 ms. In Group B, the mean retinal response density of area 1 was 22.02 ± 0.9 nV/deg2 and the mean latency was 32.56 ± 1.25 ms. In area 2, the mean retinal response density was 12.23 ± 0.55 nV/deg2 and the mean latency was 30.84 ±1.22 ms. Finally, in the perifoveal area (area 3), the mean retinal response density was 12.74 ± 0.44 nV/deg2 and the mean latency was 29.7 ± 11.09 ms. The differences in amplitude and latency were statistically significant.
Conclusion: Increased foveal thickening and significant decrease of the electrical activity of areas 1, 2 and 3 were found in HIV-positive children. These findings suggested some subclinical dysfunction of the photoreceptors and the inner retinal layers of the fovea in HIV-positive children with normal vision and without ocular disease.
Ocular involvement is common in patients infected with human immunodeficiency virus (HIV) and includes retinal microangiopathy, opportunistic ocular infections, primarily cytomegalovirus (CMV) retinitis, conjunctival, lid and orbital involvement by Kaposi's sarcoma and lymphoma and neuro-ophthalmic lesions.1
Highly active anti-retroviral therapy (HAART) decreases the incidence of CMV retinitis among patients with acquired immunodeficiency syndrome (AIDS) by approximately 75 per cent.2,3 Previous studies4–7 have shown that adults infected with HIV without infectious retinitis demonstrate reduced sensitivity in the visual fields, abnormal colour and contrast sensitivity tests and subnormal electrophysiological responses. Further studies in HIV-positive adults without CMV retinitis have demonstrated that there is up to 50 per cent loss in nerve fibre population, notwithstanding that the optic nerve appears clinically normal.8,9 It was hypothesised that the loss of optic nerve fibres is secondary, due to the inner retinal damage caused by HIV.
To the best of our knowledge, no study has been performed in HIV-positive children to investigate the macular structure and function and alterations of the outer retinal layers. The purpose of this prospective study is to assess macular thickness and electrophysiological responses of the fovea, parafovea and perifovea by means of optical coherence tomography (OCT) and multifocal electroretinogram (mfERG) in HIV children without ocular disease.
MATERIALS AND METHODS
- Top of page
- MATERIALS AND METHODS
HIV-positive children were examined in the Department of Ophthalmology of Athens University. The study was conducted in accordance with the tenets of the Declaration of Helsinki and informed consent for imaging and data collection was obtained from the children's parents after explanation of the nature of the study.
All patients suffered from congenital HIV, none was premature and the births were vaginal without the need of caesarean section. The disease was diagnosed at the age of one to five years (mean age 2.775 ± 1.105). Patients had no history of ocular or neurological disease and ophthalmic examination was normal. The study included 78 eyes of 39 patients (17 boys, 22 girls), who were divided into two groups. Both groups were matched for age and sex. Group A included 19 HIV-positive children (38 eyes). Seven were boys and 12 were girls. The mean age of the patients was 9.64 ± 4.56 years, range three to16 years. Two patients had mild myopia (–2.00 D). In all patients visual acuity (VA) measured by standard Snellen chart was equal to 1.0 decimal (0.00 logMAR) and colour vision tested with Ishihara plates was normal. None of them presented concurrent or healed CMV retinitis and the data from their medical records showed that CD4 cell counts were never below 100 (1.0 × 109). All patients were treated with high active antiretroviral therapy prior to and at the end of examination. More precisely, the patients received slavudin and nelfinavir in combination with lamivudine. The treatment was initiated as the diagnosis of HIV disease and its duration ranged from two to 11 years. Group B included 20 HIV-negative children (40 eyes) without any systemic or ocular disease, who served as normal control subjects. Ten were boys and 10 were girls. The mean age of the patients was 8.2 ± 3.3 years, range three to 14 years.
Each individual included in the present study underwent a complete ophthalmic examination, comprising VA assessment (standard Snellen chart), colour vision testing, fundus examination and intraocular pressure measurement. For children aged between three and five years, ‘Random Es visual acuity test’ was used to measure VA. Fluorescein angiography was performed in only five patients older than 14 years and did not reveal any retinal pigment epithelial lesion or leakage from the macula vessels. An OCT scan and mfERG recording was performed twice in all patients by two different evaluators and the results were similar. CD4 and CD8 cell counts, presence of systemic infection, haemoglobin, haematocrit and serum β-microglobulin levels were also analysed in HIV patients and controls.
The OCT examination was performed with the Stratus model 3000 (Carl Zeiss, Inc., Dublin, CA, USA). Retinal mapping software was used, calculating the average retinal thickness of the central ring. All eyes were scanned in a radial spoke pattern centred on the foveola with scan length of 6.0 mm. The subjects were asked to gaze at the fixation light within the instrument and the foveolar fixation was confirmed by observing the retinal image through the infrared monitoring camera. Retinal thickness was calculated as the distance between the two boundaries along each A-scan using the attached automatic boundary detection software. The software automatically detects the vitreoretinal junction as the inner retinal boundary and the chorioretinal junction as the outer boundary.
For the recording of the mfERG, the Visual Evoked Response Imaging System (VERIS) III (Tomey, Nagoya, Japan) was used and the recording protocol was chosen according to the Clinical Electrophysiology of Vision (ISCEV) guidelines for basic mfERG.10 The size of the stimulus array was four pixels with a line width of two pixels and the stimulus matrix consisted of 61 hexagonal elements displayed on a cathode ray tube colour monitor. Each hexagon was independently alternated between white and black using a pseudo-random m-sequence with exponent 14. There were 214 -1 steps during the three minute recording period. The screen resolution was 800 × 600 pixels, the luminance of the stimulus for white was 200 cd/m2 and the contrast was 99.3 per cent. These hexagons elicit approximately equal signal amplitude at all locations on a normal retina. The stimulation technique allowed a retinal response from each stimulus. The bandwidth of the amplifier was 10 to 300 Hz (-6 bdB/oct) and the amplification was ×10.000.
The pupils of the subjects were dilated with tropicamide 0.5% and phenylephrine 5% and the eyes were optically corrected for near vision to clearly see the small fixation spot in the centre of the stimulus matrix. A bipolar contact lens was used for signal acquisition, in which the active and reference electrodes were incorporated in the contact lens. The ground electrode was attached to the ear lobe. The fellow eye was closed and the duration of the data acquisition was eight minutes divided into eight sessions of 60 seconds. The recording procedure was repeated if there were spurious potentials from eye blinks or if ocular movement was recorded. In the case of the three-year-old child, an observer continuously monitored the child's fixation during recording and responses were recorded only when the observer reported that the child was alert and looking at the centre of the monitor. If the child looked away from the centre of the display the segment was discarded and re-recorded.
The response density (amplitude per unit retinal area, nV/deg2) of each local response was estimated as the dot product between the normalised response template and each local response. The normal ranges for these amplitudes were defined by calculation of the median and the 95% confidence intervals in both eyes of 20 normal volunteers (Group B).
The mfERG stimuli locations and anatomic areas corresponded roughly as follows: ring 1 to the fovea (0 to 2°), ring 2 to the parafovea (2 to 7°), ring 3 to the perifovea (7 to 13°), ring 4 to the near periphery (13 to 22°) and ring 5 to the central part of the middle periphery (22 to 30.5°). The amplitude of each group was scaled to reflect the angular size of the stimulus hexagon, which produces the response. These averages give a more accurate view of the relative response densities of each group. The retinal response density decreases with eccentricity, although there is no further decrease from ring 4 to ring 5.
Continuous data are presented with means and standard deviations, while the Kolmogorov–Smirnov test evaluated the assumption of normality. Nested analysis of variance was used to compare mean amplitude, latency and thickness in HIV-positive and HIV-negative participants. The data analysis is preferable to other approaches, because it takes into account the fact that each subject contributes to observations, one for the right eye and one for the left eye. A p value less than 0.05 was considered to indicate significance.
- Top of page
- MATERIALS AND METHODS
Table 1 shows that the foveal thickness in HIV-positive children ranges from 140 to 255 µm with mean value 190.28 ± 26.58 µm. No intraretinal or subretinal fluid and no macula traction were detected in any HIV child (Figure 1). Also, there were no cystic changes in the retina on morphological analysis of the OCT cross-sectional scans in any of the eyes examined. In the HIV-negative control group, the mean foveal thickness was 169.47 ± 10.17 µm (Figure 2). The method of Kolmogorov and Smirnov assumed that the data in both groups follow Gaussian distributions. Nested analysis of variance identified that the foveal thickness was significantly higher in HIV-positive children than in the control group (p = 0.0002) (Table 1).
Table 1. Mean foveal thickness and mean electrical activity of areas 1, 2 and 3 of multifocal electroretinogram (mfERG) (nV/deg2) in eyes (n) of human immunodeficiency virus (HIV)-positive (Group A) and HIV-negative (Group B) children.
|Variables||HIV positive||HIV negative||p†|
|(n = 38)||(n = 40)|
|OCT (µm)||190.28 ± 26.58||169.47 ± 10.17||p = 0.0002|
|mfERG/area 1 (nv/deg2)||19.87 ± 4.59||22.02 ± 0.90||p = 0.027|
|mfERG/area 1 (ms)||38.56 ± 1.18||32.56 ± 1.25||p < 0.0001|
|mfERG/area 2 (nv/deg2)||10.98 ± 2.42||12.23 ± 0.55||p = 0.022|
|mfERG/area 2 (ms)||36.52 ± 1.76||30.84 ± 1.22||p < 0.0001|
|mfERG/area 3 (nv/deg2)||10.83 ± 0.90||12.74 ± 0.44||p < 0.0001|
|mfERG/area 3 (ms)||36.36 ± 1.90||29.71 ± 1.09||p < 0.0001|
Figure 1. Three-dimensional topographic plot of a multifocal electroretinogram in a human immunodeficiency virus-positive child (top left). Multifocal electroretinogram traces of the same eye (top right). Retinal response densities are decreased in areas 1, 2 and 3. The optical coherence tomography recording shows thickening of the fovea (212 µm) (bottom).
Download figure to PowerPoint
Figure 2. Three-dimensional topographic plot of a multifocal electroretinogram in a human immunodeficiency virus-negative child (top left). Multifocal electroretinogram traces of the same eye (top right). Retinal response densities are within normal limits in area 1 (22.35 nV/deg2), area 2 (12.41 nV/deg2) and area 3 (10.32 nV/deg2) (top right). Optical coherence tomography shows thickness of the fovea equal to 165 µm (bottom).
Download figure to PowerPoint
The mean retinal response density of the mfERG of the fovea (area 1) in the HIV-positive children was 19.87 ± 4.59 nV/deg2 and the mean latency was 38.56 ± 18 ms (Table 1). In the HIV-negative children, the mean retinal response density of area 1 was 22.02 ± 0.90 nV/deg2 and the mean latency was 32.56 ± 1.25 (Figure 1). Statistical analysis was performed using nested analysis of variance, which shows that the retinal response density in the HIV-positive children was significantly lower (p = 0.027) and the latency was significantly higher (p < 0.0001) compared with the control group (Table 1).
In the parafoveal area (area 2) of the HIV-positive children the mean value of retinal response density was 10.98 ± 2.42 nV/deg2 and the mean latency was 36.52 ± 1.76 ms (Table 1). In the HIV-negative children the mean retinal response density of area 2 was 12.23 ± 0.55 nV/deg2 and the mean latency was 30.84 ± 1.22 ms. Nested analysis of variance found that the retinal response density in HIV children was significantly lower (p = 0.022), whereas the latency was higher (p < 0.0001) compared with the control group (Table 1).
In area 3 (parafoveal area) of the HIV-positive children the mean retinal response density was 10.83 ± 0.90 nV/deg2 and the mean latency was 36.36 ± 1.90 ms (Table 1). In the HIV-negative children, the mean retinal response density of area 3 was equal to 12.74 ± 0.44 nV/deg2 and the mean latency was 29.71 ± 1.09 ms. Nested analysis of variance found that the retinal response density in HIV children was significantly lower (p < 0.0001) whereas the latency was higher (p < 0.0001) compared with the control group (Table 1).
- Top of page
- MATERIALS AND METHODS
Retinal microangiopathy associated with HIV infection is usually asymptomatic and escapes detection unless examination of the fundus of the eye is performed and evanescent cotton wool spots are present.
Advances in ocular imaging technology have made it possible to evaluate macular thickness in an objective quantifiable and reproducible fashion.11–14 Also, the mfERG allows functional mapping of the retina and contributes to the detailed evaluation of retinal function, especially in regional disorders of the outer retinal layers of the macula.15 It is suggested that the mfERG reflects not only the electrophysiological responses of the photoreceptors but also those of the inner retinal layers including the bipolar and Müller cells.16 Also, the retinal response densities in area 1 (fovea) reflect the integrity of the photoreceptors of the fovea, which are responsible for the normality of VA, as well as the electrical activity of the photoreceptors and the inner layers of the perifoveal area. Therefore, lesions of this area, not visible ophthalmoscopically, might be mapped by mfERG even in eyes with normal VA.
Using third generation OCT in the present study, a clear thickening of the fovea was found in 10 eyes, although the patients were clinically asymptomatic. The increase of foveal thickness might be attributed to a decompensation of the retinal pigment epithelium due to latent inflammation or breakdown of the inner and outer blood–retinal barrier, possibly created by the presence of sub-clinical HIV vasculopathy.17,18
Changes in retinal function resulting from regional disorders can be depicted in detail with the recording of the mfERG, which is important mainly for the investigation of macular lesions. Studies have shown that in all types of maculopathies there is a decrease or loss of central electrical activity, which is directly related to the degree and extent of the central lesion. The mfERG recording is particularly useful in cases in which the macular lesions are not visible ophthalmoscopically, the patient is asymptomatic and only OCT reveals a retinal thickening of the fovea. The occult macular dystrophy with central cone dysfunction, which was demonstrated by mfERG, is a characteristic example that supports the usefulness of this technique.19
In a previous study, no significant abnormalities of mfERG were found in HIV-positive patients without infectious retinitis.9 Also, no correlation between visual defects and mfERG results were assessed.9 In contrast, Falkenstein and colleagues20 found a significant delay in latencies of mfERG and a trend of reduced P1 amplitude in HIV-positive adults without infectious retinitis. This might suggest an early diffuse dysfunction of the inner retina, which might precede the infectious HIV retinitis. These changes are not directly related to involvement of the central nervous system. Nevertheless, dysfunction of the inner retinal layers might be an early sign of affection of the central nervous system.
Our results show that the mfERG is clearly abnormal in 10 eyes. In these cases, the mean retinal response densities and the mean latencies in the foveal, parafoveal and perifoveal areas of the mfERG are pathological. This might be attributed to the increase of the macular thickness as demonstrated by OCT. Unfortunately no electrophysiological test isolates the inner retinal elements from the photoreceptors. This makes it difficult to realise the site of early functional subclinical changes leading to structural damage. The use of P-ERG might help to isolate the inner retinal elements from the outer ones.
In conclusion, the present study demonstrates that in some HIV-positive children without infectious retinitis there is an increase in thickness of the fovea and decrease in the electrical activity of the macula. These findings might be important in the diagnosis of early subclinical HIV-associated visual impairment.
Also, we suggest that highly active anti-retroviral therapy does not protect from retinal dysfunction and visual loss. For this reason, pharmacological neuroprotection along with the systemic control of HIV disease itself might be important.20 To clarify this aspect, further investigation with a longer follow-up and larger cohort of patients is required.