SEARCH

SEARCH BY CITATION

Keywords:

  • HIV;
  • retinal oxygen reactivity;
  • retinal WBC flux

Abstract.

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

Purpose:  The aetiology of the apparently vasoocclusive phenomena in human immunodeficiency virus (HIV)-related retinopathy is not well understood. Several hypotheses, including infectious damage of the retinal vasculature and altered retinal haemodynamics, have been postulated. Direct measurement of oxygen tension in the retina is not possible in vivo and indirect methods have to be employed. The objective of this study was to investigate the retinal vascular response to 100% oxygen breathing in patients with HIV.

Methods:  Twelve patients infected with HIV and 12 healthy individuals, matched for age, sex and smoking habits, were studied in an open study using the blue-field entoptic technique for the measurement of retinal white blood cell (WBC) flux. Reactivity in retinal blood flow during 100% O2 breathing over 15 min was measured and expressed as percentage change over baseline.

Results:  WBC velocity during oxygen inhalation decreased over baseline by 9.0 ± 5.8% in HIV-infected patients and by 18.6 ± 5.4% in healthy participants (p < 0.04 between groups, anova). The decrease in WBC velocity was paralleled by a decrease in WBC density. This decrease tended to be more pronounced in healthy participants (13.6 ± 7.9%) than in HIV-infected patients (8.0 ± 10.8%), but the difference was not statistically significant (p = 0.1 between groups, anova). WBC flux decrease was 16.2 ± 11.4% in HIV-infected patients and 29.5 ± 9.5% in the control group and was significant between groups (p = 0.007 between groups, anova).

Conclusion:  Our results indicate a reduced reactivity of WBC flux to systemic hyperoxia in patients with HIV. Whether abnormal retinal haemodynamics in HIV-infected persons contributes to the pathogenesis of HIV-related microvascular diseases or is a consequence of the structural changes associated with the disease is unknown.


Introduction

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

Human immunodeficiency virus (HIV)-related retinopathy is a common finding in patients infected with HIV (Freeman et al. 1984). Since the introduction of highly active antiretroviral therapy (HAART), the clinical picture/disease pattern of ocular manifestations in HIV-positive patients has changed. Furthermore, the incidence of HIV-related microangiopathies and ocular opportunistic infection, such as cytomegalovirus (CMV) retinitis, has decreased dramatically (Accorinti et al. 2006). The immune recovery uveitis, a HAART-related inflammatory syndrome, is now found in a great number of HIV-infected patients with CMV retinitis (Goldberg et al. 2005). However, HIV-related retinopathy remains an important finding in immunocompromised individuals. With immunological reconstitution the survival of HIV-infected persons is increased, making it of vital importance to elucidate the mechanisms leading to microvasculopathies in this class of patients (Kempen et al. 2003).

Factors that may contribute to the development of HIV-related retinopathy are infectious damage of retinal vasculature and retinal tissue (Pomerantz et al. 1987; Skolnik et al. 1989), and altered ocular haemodynamics (Yung et al. 1996; Dejaco-Ruhswurm et al. 2001; Dadgostar et al. 2006). Hypotheses for the development of HIV-related microvasculopathies include immunoglobulin deposition, endothelial cell damage by HIV, and hyperviscosity secondary to increased red cell aggregation and fibrinogen. Faber et al. (1992) noted focal loss of capillary perfusion in HIV-infected individuals using fluorescein angiography and retinal capillary cell loss and focal occlusions of small vessels in histological examinations. A reduction in blood flow velocities may lead to vessel damage caused by ischaemia and alteration in leukocyte–endothelial cell interactions, resulting in additional vascular damage (Weiss 1989). In addition, HIV-infected individuals seem to have increased polymorphonuclear leukocyte rigidity, which might lead to direct microvascular damage by the release of protease and toxic oxygen radicals (Dadgostar et al. 2006).

However, little is known about ocular blood flow regulation in patients with HIV. In diabetic retinopathy, altered retinal vascular reactivity to oxygen breathing is an early event. In the present study we examined patients without clinical evidence of HIV-related retinopathy. We tested the hypothesis that the blood flow regulation of the retina of HIV-infected patients is deranged even before the onset of HIV-related retinopathy.

Materials and Methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

Participants

Twelve HIV-positive patients without HIV-related retinopathy with a CD4 cell count < 500 cells/μl [age range 25–45 years, mean 36.1 ± 5.9 standard deviation (SD)] and 12 healthy participants, matched for age, sex and smoking habits (age range 23–48 years, mean 36.7 ± 7.1 SD) participated in this open pilot study. The study protocol was approved by the Ethics Committee of the Medical University of Vienna and followed the tenets of the Declaration of Helsinki. The nature of the study was explained and all participants signed a written informed consent to participate. Each participant had to undergo a complete screening examination that included systemic physical assessment and an ophthalmic examination on both eyes, including slit-lamp biomicroscopy and indirect funduscopy. Demographic and medical information was obtained from each participant, including the use of antiretroviral therapy, the interval since infection diagnosis and the stage of virus infection.

Absolute CD4 lymphocyte count was determined by two-colour flow cytometry of whole blood preparations. Actual and lowest documented CD4 count before antiretroviral treatment were recorded. Viral load (HIV-1 RNA levels) was measured by polymerase chain reaction (Amplicator HIV-1 Monitor Test; Roche Diagnostics Corporation, Indianapolis, Indiana, USA).

Study design

In this open, cross-sectional study an a priori sample size calculation was performed based on the reproducibility data of the blue-field entoptic technique. This sample size was calculated based on an α-error of 0.05 and a β-error of 0.2.

All participants had to abstain from beverages containing xanthine derivates for at least 12 hr before the start of the study. To exclude diurnal variations in blood flow, measurements were performed between 10.00 and 0.00 hr.

One study day was scheduled for each participant. After steady-state conditions were reached (which was ensured by repeated blood pressure monitoring), baseline measurements of retinal white blood cell (WBC) flux were performed. All blood flow measurements were performed on the participant’s right eye. Thereafter a 15 min period of 100% O2 (AGA; Gases for Human Use, Vienna, Austria) inhalation was scheduled. The oxygen was delivered through a partially expanded reservoir bag at atmospheric pressure under nasal occlusion. During the last 7 min of the breathing period, retinal WBC flux measurements were repeated. In addition, pulse rate and systemic blood pressure were measured simultaneously with ocular haemodynamic measurements.

Methods

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

Non-invasive measurement of systemic haemodynamics and intraocular pressure

Systolic, diastolic and mean blood pressure were measured on the upper arm using an automated oscillometric device. Pulse rate was recorded automatically using a finger pulse-oxymetric device (HP-CMS Patient Monitor; Hewlett Packard, Palo Alto, California, USA). Intraocular pressure (IOP) was measured using applanation tonometry. Ocular perfusion pressure was calculated as: 2/3 × mean arterial pressure – IOP.

Blue-field entoptic technique

This non-invasive method is described in detail by Riva & Petrig (1980). We used a commercially available system for the quantification of retinal WBC movement (Oculix Blue-Field Simulator; Oculix Sarl, Arbaz, Switzerland). For determinations of the velocity and density of flying corpuscles, a simulated particle field was shown to the study participants. By comparison with their own entoptic observation, participants could adjust WBC density and mean flow velocity (WBC velocity). Retinal WBC flux was calculated as: WBC flux = WBC density × WBC velocity. These outcome parameters characterize WBC dynamics in perimacular retinal capillaries. The reproducibility of this method in our laboratory has been published previously (Luksch et al. 2008). In the present study, five matching trials were performed by each participant at the pre-study screening. Only participants with a coefficient of variation < 15% were included. This criterion was also applied on the measurements of the study day. No participant had to be excluded at the stage of the actual experiment.

Data analysis/statistics

Statistical analysis was performed using the CS: statistica software package version 6.1 (StatSoft, Tulsa, Oklahoma, USA). For data description and statistical analysis, haemodynamic parameters were expressed as percentage change from baseline (Δ%). Effects of O2 on the outcome parameters were assessed by repeated-measures anova. Differences between groups were assessed as the interaction between grouping and different measuring times. Post-hoc analysis was performed using planned comparisons. Results are given as mean ± SD.

Results

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

All included patients in both the HIV and healthy groups were male. Four participants were smokers in each group. None of the patients showed clinical evidence of HIV-related retinopathy. The mean CD4 cell count of the HIV-infected persons was 168.1 ± 65.1 cells/μl at the time of examination. Mean minimal CD4 cell count before antiretroviral treatment was 48.9 ± 31.3 cells/μl. Mean viral load was 43.3 × 26.6 × 102 copies/ml at the time of examination. All patients received the HAART regimen. The mean interval from the time of the study and the date of HIV diagnosis was 69.7 ± 30.0 months.

Systemic haemodynamic parameters and IOP

Systolic, diastolic and mean blood pressure levels as well as the pulse rate of HIV-infected persons and healthy controls are presented in Table 1. Breathing of 100% oxygen for 15 min had no influence on systemic haemodynamic parameters (other than pulse rate) in either HIV-infected patients or the healthy control group. For pulse rate, there was a slight difference in the response to oxygen breathing between healthy participants and HIV-infected patients, but none of these effects reached the level of significance (p > 0.10 each). IOP was comparable between groups. Ocular perfusion pressure tended to be lower in healthy controls, but this effect was only borderline significant (p = 0.054).

Table 1.   Systemic haemodynamic parameters and intraocular pressure (IOP) in HIV-infected persons and age-matched controls at baseline and during 100% O2 inhalation for 15 min.
 Baseline100% O2 inhalation
  1. Values expressed as mean ± standard deviation.

HIV-infected patients (n = 12)
 Mean arterial pressure (mmHg)91.8 ± 10.0 92.7 ± 13.7
 Systolic blood pressure (mmHg)129.5 ± 13.4130.2 ± 17.2
 Diastolic blood pressure (mmHg)73.3 ± 9.474.1 ± 12.8
 Pulse rate (beats/min)76.1 ± 10.678.3 ± 9.0
 IOP (mmHg)13.3 ± 1.9
 Ocular perfusion pressure (mmHg)47.9 ± 7.1
Healthy controls (n = 12)
 Mean arterial pressure (mmHg)83.8 ± 7.884.8 ± 5.3
 Pulse rate (beats/min)76.4 ± 14.771.6 ± 11.7
 Systolic blood pressure (mmHg)116.3 ± 10.1116.3 ± 10.1
 Diastolic blood pressure (mmHg)68.2 ± 9.569.0 ± 7.4
 IOP (mmHg)13.3 ± 1.5
 Ocular perfusion pressure (mmHg)42.5 ± 5.8

Ocular haemodynamic parameters

The results of the ocular blood flow measurements as assessed using the blue-field entoptic technique are presented in Table 2. All HIV-infected persons and all healthy controls showed an adequate reproducibility, as defined in the Methods section. There was no difference in WBC velocity at baseline between HIV-infected patients and healthy controls (p = 0.9). WBC flux at baseline tended to be lower in HIV-infected patients, but this difference was not statistically significant (p = 0.09). In contrast, the baseline measurements of WBC density showed a significantly lower value in HIV-infected patients compared to healthy controls (p = 0.04). In patients with HIV, none of the parameters of the blue-field entoptic technique were correlated with mean CD4 cell count.

Table 2.   Ocular haemodynamic parameters in HIV-infected persons and age-matched controls at baseline and after 100% O2 inhalation for 15 min in absolute units.
 Baseline100% O2 inhalation
  1. WBC, white blood cells.

  2. Values expressed as mean ± standard deviation.

HIV-infected patients (n = 12)
 WBC velocity (mm/second)1.0 ± 0.4 0.9 ± 0.4
 WBC density (number)91.5 ± 26.984.8 ± 30.2
 WBC flux90.7 ± 34.075.6 ± 29.8
Healthy controls (n = 12)
 WBC velocity (mm/second)1.0 ± 0.30.8 ± 0.2
 WBC density (number)119.4 ± 35.1104.9 ± 39.3
 WBC flux120.7 ± 48.587.0 ± 39.5

WBC velocity during oxygen inhalation decreased over baseline by 9.0 ± 5.8% in HIV-infected patients and by 18.6 ± 5.4% in healthy participants. This effect was less pronounced in HIV-infected patients compared to healthy controls (p < 0.04 between groups, anova). The decrease in WBC velocity was paralleled by a decrease in WBC density over baseline during oxygen inhalation. This decrease tended to be more pronounced in healthy participants (13.6 ± 7.9%) than in HIV-infected patients (8.0 ± 10.8%), but the difference was not statistically significant (p = 0.1 between groups, anova). Because both WBC velocity and WBC density decreased during oxygen breathing, WBC flux also decreased. This decrease was 16.2 ± 11.4% in HIV-infected patients and 29.5 ± 9.5% in the control group. The difference in the effect of systemic hyperoxia on WBC flux was significant between groups (p = 0.007 between groups, anova). When all participants were pooled, the response of blue-field entoptic parameters to systemic hyperoxia was not associated with baseline mean arterial blood pressure (Fig. 1) or baseline ocular perfusion pressure (data not shown). This was also the case if the correlations were calculated for HIV-infected patients and healthy controls separately.

image

Figure 1.  Linear correlation analysis between the oxygen-induced % change in white blood cell (WBC) velocity, WBC density and WBC flux and mean arterial pressure. Data are pooled for healthy participants and HIV-infected patients (n = 24).

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References

In the present study a significantly reduced reactivity of WBC flux to systemic hyperoxia was observed in HIV-infected patients compared to healthy controls. This indicates that the vasoconstrictive response of retinal vessels to oxygen in HIV-infected patients is reduced.

The pathogenesis of microvascular diseases in HIV-infected patients is still unclear and is most probably multifactorial. The combination of microvasculopathy and haemorheological abnormalities, both of which have been described in HIV-infected individuals, is assumed to result in focal ischaemia with the development of retinal haemorrhages and cotton-wool spots, leading to the clinical appearance of HIV-related retinopathy (Glasgow & Weisberger 1994). Cotton-wool spots are caused by localized accumulations of axoplasmic debris following obstruction to axoplasmic flow (Mc Leod et al. 1977). There is evidence that retinal ischaemia is the primary cause of cotton-wool spots (Newsome et al. 1984; Pepose et al. 1985, Tufail et al. 2000).

HIV-related retinal microvascular changes are ultrastructurally similar to those seen in patients with diabetic retinopathy. This HIV-related structural damage of the vessel wall could well influence blood flow regulation in the retina, which has been described extensively for diabetic retinopathy (Kohner et al. 1995; Grunwald & Bursell 1996; Schmetterer & Wolzt 1999; Ciulla et al. 2003). Indeed, a variety of haemodynamic abnormalities has been described in the ocular vascular system of HIV- infected individuals. A high association between HIV-related retinopathy and cerebral hypoperfusion argues in favour of a vascular component in the pathogenesis of the disease (Geier et al. 1992). Abnormal blood flow in HIV-infected patients has been noted in conjunctival capillaries based on biomicroscopic investigations (Engstrom et al. 1990). More importantly for the disease, reduced microvascular blood flow in the retina (as measured with the Heidelberg Retina Flowmeter) has been reported (Dadgostar et al. 2006). Other investigators provided evidence for reduced perifoveal blood flow velocities in HIV-infected patients using scanning laser ophthalmoscope (SLO) fluorescein angiography (Yung et al. 1996). In previous studies using the blue-field entoptic technique to determine whether perimacular WBC flux is altered in HIV-infected individuals, WBC density was particularly reduced (Dejaco-Ruhswurm et al. 2001; Lim et al. 2001). This is in good agreement with the results of the present trial, although WBC flux was not significantly different between groups in the present study, most likely because of the small number of participants included.

The ability of the normal retina to maintain constant oxygen tension seems to be essential for its function. In a regulatory response, changes of arterial blood oxygen are accompanied by changes in retinal blood flow (Riva et al. 1983; Tomic et al. 2005). Previous studies (Grunwald et al. 1984; Patel et al. 1994; Evans et al. 1997) showed a reduced vascular reactivity to oxygen in patients with diabetes. They interpreted these findings as being indicative of a loss of the regulatory capacities of the retina to metabolic and environmental changes, which may result in the development of retinopathy in these patients. Accordingly, abnormal regulatory responses make the retina vulnerable to these changes. This abnormal blood flow regulation was explained by the assumption that the diabetic retina is hypoxic. In healthy individuals, there is no oxygen deficit and its reaction to oxygen represents the reduction in blood flow needed to maintain retinal normoxia. However, in the presence of diabetic retinopathy, there is hypoxia before giving oxygen and therefore the response of the retina to additional oxygen is reduced to counteract the oxygen deficit (Grunwald et al. 1984). In support of this model, oxygen reactivity improved to near normality after panretinal photocoagulation for proliferative diabetic retinopathy (Grunwald et al. 1986). This is in good agreement with the hypothesis that panretinal photocoagulation improves retinal oxygenation by reducing the amount of viable retinal tissue thereby reducing oxygen demand (Stefánsson 2006, 2008).

The same may be true for HIV-infected patients. The oxygen reactivity of retinal blood flow in the HIV group versus the control group provides experimental support for some degree of retinal hypoxia. The fact that the retinal flux in the present study decreased to a smaller degree in the HIV group compared to the control group might show an attempt to keep blood flow and hence the delivery of oxygen to retinal tissue within some critical level. This interpretation may be supported by our observation that baseline WBC density tended to be lower in HIV-infected patients than in healthy controls. Because WBC density was not correlated with mean CD4 cell count in HIV-infected patients, our results may indicate constriction of retinal vessels at baseline.

While testing the density of WBCs, it may be assumed that the WBC density reflects the number of oxygenated red blood cells. A lower number of red blood cells could indicate the presence of relative hypoxia in the HIV group who had no retinopathy at the time of testing. However, in a disease like HIV one needs to be careful when proposing that a reduced number of leukocytes in the retina is also indicative of a reduced number of erythroytes. To the best of our knowledge, no laser Doppler study in HIV-infected patients is available to provide experimental insight into this question.

A blood gas analysis with verification of systemic hyperoxia was not performed. However, it is unlikely that significant differences between the two groups could have occurred, leading secondarily to the observed differences in the retinal white blood cell flux reactivity. Another limitation of the present study is related to the fact that systemic blood pressure and ocular perfusion pressure tended to be higher in HIV-infected patients than in healthy controls. Nevertheless, it is unlikely that this is related to the lower retinal response to hyperoxia in HIV-infected patients because no correlation between oxygen-induced changes in retinal haemodynamics and systemic blood pressure was seen.

The reduced response of hyperoxia could well be related to a generally reduced vascular reactivity of retinal vessels in HIV-infected patients. One mechanism could be related to the potent vasoconstrictor endothelin-1 (ET-1). ET-1 is the primary modulator of the retinal response to hyperoxia (Kohner et al. 1995; Dallinger et al. 2000). Elevated plasma levels of ET-1 have been detected in HIV-infected persons (Geier et al. 1995) as well as in several other vascular diseases, including hypertension and diabetes (Takahashi et al. 1990; Lerman et al. 1991; Stewart et al. 1991) Thus, the retinal response to hyperoxia may be blunted because of the already elevated ET-1 levels and because of desensitization of the ET-1 receptors. This is in accordance with several other articles, which hypothesize ET-1 to be responsible for decreased blood flow (Cioffi et al. 1995; Schmetterer et al. 1997) and the development of retinopathy and HIV-related microvasculopathy (Geier et al. 1995).

Of course, other vasoactive factors or metabolites, which may be inadequately removed or produced, could also be responsible for the abnormalities in the regulatory response to oxygen that has been observed in HIV-infected patients.

In conclusion, our results indicate a significantly reduced vasoconstriction response of retinal vessels to oxygen exposure in patients with HIV in comparison to a healthy control group. The hypothesis that retinal vascular regulation is abnormal in HIV-infected patients has been given further support in this study. Whether this is a sign of vascular dysregulation or is related to reduced retinal blood flow and/or hypoxia at baseline needs to be elucidated. Whether these vascular phenomena may contribute to the pathogenesis of HIV-related microvascular diseases or are a consequence of the structural changes in the disease is unknown.

References

  1. Top of page
  2. Abstract.
  3. Introduction
  4. Materials and Methods
  5. Methods
  6. Results
  7. Discussion
  8. References