Subfoveal choroidal blood flow and central retinal function in early glaucoma

Authors


Dario Marangoni, MD
Institute of Ophthalmology
Catholic University of Sacred Heart
Largo F. Vito 1
00168 Rome
Italy
Tel: 0039 06 30154929
Fax: 0039 06 3051274
Email: dariomarangoni80@yahoo.it

Abstract.

Purpose:  To assess subfoveal choroidal blood flow (ChBF) in patients with early manifest glaucoma (EMG) and to compare blood flow with functional measures of central retinal integrity, standard automated perimetry (SAP) and pattern electroretinogram (PERG).

Methods:  Subfoveal ChBF was determined by confocal, real-time laser Doppler flowmetry in 24 patients with EMG [>−6 dB mean deviation (MD), age range: 29–77 years, visual acuity: 20/25–20/20] and 23 age-matched control subjects. All patients had a therapeutically (topical beta-blockers with or without prostaglandin analogues) controlled intraocular pressure (IOP < 20 mmHg). Subfoveal choroidal blood volume (ChBVol), velocity (ChBVel) and ChBF were determined as the average of three 60 second recordings. In all patients and controls, the PERG and SAP (Humphrey 30-2), following standardized protocols, were also recorded.

Results:  In patients with EMG, reductions in average ChBVel and ChBF were roughly equal, respectively, by 30% and 33.4% (p < 0.01), when compared to control subjects, so that there was no significant difference in ChBVol between the two groups. Pattern electroretinogram amplitudes were reduced by 46% (p < 0.01) in patients compared to controls. No correlation was found between any of the ChBF parameters and PERG amplitude, or Humphrey 30-2 MD and pattern standard deviation.

Conclusion:  The results suggest a significant alteration of both ChBVel and ChBF in EMG, which does not appear to be associated with the severity of central retinal dysfunction. These findings may contribute to a better understanding of the pathophysiology of early glaucomatous damage in EMG and have implications for the treatment of this pathologic condition.

Introduction

Glaucoma is an optic neuropathy characterized by progressive dysfunction and loss of retinal ganglion cells (RGCs), resulting in typical visual field defects and excavation of the optic nerve head. Although an elevated intraocular pressure (IOP) is considered the major risk factor for the development of the glaucomatous disease, an altered ocular blood flow seems to be also involved in its pathogenesis (Hayreh et al. 1970, 1994; Grieshaber & Flammer 2005; Costa et al. 2009). Histological studies on eyes with primary open-angle glaucoma (OAG) have demonstrated a reduced choroidal thickness (Yin et al. 1997) and a decreased density of both large choroidal vessels and capillaries of the choriocapillaris layer (Spraul et al. 2002). Thanks to the development of many different methods to visualize and measure in vivo the ocular blood flow, several clinical studies have investigated choroidal and retinal blood flow in patients with glaucoma (Stankiewicz et al. 2011). Abnormalities have been found in both vascular systems (Flammer et al. 2002). In this regard, laser Doppler flowmetry (LDF) has proved to be a valid tool to measure optic nerve and choroidal blood flow (ChBF) in patients with glaucoma (Grunwald et al. 1998; Gugleta et al. 2003; Emre et al. 2004; Pournaras et al. 2004; Riva et al. 2004; Kochkorov et al. 2010; Resch et al. 2011).

A large body of evidence indicates that vascular dysregulation, leading to both low perfusion pressure and insufficient blood flow autoregulation, could be responsible for an unstable ocular perfusion, ischaemia and thereby RGC loss (Flammer et al. 1999; Grieshaber et al. 2007; Moore et al. 2008; Mozaffarieh et al. 2008). Although RGCs seem to be the main site of glaucomatous damage, experimental (Nork et al. 2000; Raz et al. 2003) and clinical studies (Holopigian et al. 1990; Kanis et al. 2010; Choi et al. 2011) have reported functional and structural changes in the outer retina, suggesting that both inner and outer retinal layers can be dysfunctional in glaucoma.

An objective estimate of retinal function can be obtained from the pattern electroretinogram (PERG), an electrophysiological signal reflecting the integrity of the outer and inner layers of the central retina (Holder 2001). The PERG has been shown to be altered in glaucoma as well as in patients with high-risk ocular hypertension (OHT) (Pfeiffer et al.1993; Ventura et al. 2005; Bach et al. 2006; Parisi et al. 2006; Falsini et al. 2008).

The aim of this study was to investigate whether subfoveal ChBF, as assessed by LDF, is altered in patients with early manifest glaucoma (EMG) and whether these alterations are associated with central retinal dysfunction, as evaluated by PERG.

Material and Methods

Subjects

Twenty-four eyes of 24 patients with EMG (mean age ± SD: 58 ± 15.2, range 29–77 years) and 23 eyes of 23 control subjects (mean age ± SD: 52 ± 12.4, range 28–72 years) were included in this study. The baseline characteristics of these patients are shown in Table 1. At the time of diagnosis, patients with EMG had elevated IOP (>21 mmHg) on two separate occasions (mean IOP ± SD: 25.4 ± 2.1 mmHg) measured by Goldmann tonometry as well as abnormal standard automated perimetry (SAP) test results (central 30-2 SITA-standard test, Humphrey Field Analyzer II, 750i model; Carl Zeiss Meditec, Dublin, CA, USA). Perimetric testing was performed according to standard procedures (Gordon & Kass 1999; Leske et al. 1999). Normal visual fields were defined as having a mean deviation (MD) and a pattern standard deviation (PSD) index within 95% confidence limits, and a Glaucoma Hemifield Test (GHT) within normal limits; abnormal visual fields, evident in one or both eyes of patients with EMG, had to show typical defects (arcuate and/or paracentral scotoma or nasal step; three or more adjacent points, not contiguous with the field borders nor with the blind spot, with a ≥ 5 dB loss, or two or more adjacent points, not contiguous with the field borders nor with the blind spot, with ≥10 dB loss; GHT outside normal limits), reproducible in at least three consecutive tests within 1 month. The average perimetric MD ± SD was −1.54 ± 1.76 dB (range −5.89 to +1.62 dB) for the selected eyes of the patients with EMG.

Table 1.   Demographic and clinical findings of control subjects and patients.
 Control subjects (n = 23)Patients with EMG (n = 24)
  1. Data are expressed as mean (range). Ocular characteristics are derived from right eyes.

  2. * Spherical equivalent.

  3. † Under medication.

  4. EMG = early manifest glaucoma; IOP = intraocular pressure.

Age (years)52 (28–72)58 (29–77)
Gender (M/F)12/1114/10
Visual acuity20/2020/25–20/20
Refractive error (diopter)*−0.20 (−3.50 to +2.75)−0.51 (−3.75 to +2.25)
IOP16 (14–17)17 (15–18)†
Vertical cup-to-disc ratio0.4 (0.3–0.5)0.7 (0.6–0.8)
Humphrey mean deviation (dB)0.19 (−0.95 to +1.65)−1.54 (−5.89 to +1.62)
IOP-lowering medicationsNoneBeta-blockers with/without prostaglandin analogues
Mean arterial blood pressure (mmHg)95 (93–99)98 (80–118)
Ocular perfusion pressure (mmHg)48 (30–61)49 (36–66)

Abnormal clinical appearance of the optic disc evaluated by slit-lamp biomicroscopy and 78-D lens was present in one or both eyes of all patients with EMG and included at least one of the following abnormalities: vertical cup-to-disc diameter ratio > 0.6, interocular cup-to-disc-ratio asymmetry ≥ 0.2, excavation, thinning of the rim, notching, haemorrhages, nerve fibre layer defects or parapapillary atrophy. The patients enrolled in the study were recruited from a larger cohort of patients evaluated at the Glaucoma Service of the Policlinico A. Gemelli, Catholic University, Rome. Additional inclusion criteria were the following: best-corrected visual acuity ≥ 20/25, normal slit-lamp biomicroscopic examination, including an open anterior chamber angle on gonioscopy, and central corneal thickness between 530 and 570 μm, as measured by a digital ultrasonic pachymeter (Altair V4, Optikon 2000, Roma, Italy). Patients with refractive error < −6.00 or > +3.00 diopters (D) spherical equivalent, astigmatism > ±1.00 D, presence of neuro-ophthalmologic diseases affecting the optic disc and the visual function or low perimetric reliability (Bickler-Bluth et al. 1989) were excluded. At the beginning of the study, all patients with EMG had a controlled IOP (mean ± SD: 16.67 ± 1.27 mmHg) by treatment with topical hypotensive drugs such as beta-blockers (Maleate Timolol 0.5%) with or without prostaglandin analogues (Latanoprost or Travoprost). In this study, IOP-lowering medications were not discontinued. We could not control simultaneously for the potential confounding effects of either IOP elevation or IOP-lowering drugs on the ChBF and PERG measurements. Therefore, we chose to control only for the former parameter (high IOP), by keeping unaltered the latter (i.e. medication), because previous studies of our group (Colotto et al. 1996) have shown that even a moderate IOP increase may affect significantly the PERG measurements in patients with early glaucoma. No similar results have been reported in the literature for the ChBF.

The influence of factors such as age, gender, menopausal status and systemic disease or drugs on the ocular blood flow has been widely investigated in many reports (Kavroulaki et al. 2010; Sekeroglu et al. 2011; Shoshani et al. 2011). To minimize the effect of these variables on ChBF parameters, patients and control subjects were matched on the basis of age, gender ratio and mean arterial blood pressure. Patients affected by systemic pathologies or taking systemic medications that could potentially influence the choroidal circulation were not included in the study.

Informed consent was obtained from every subject or patient after the procedures employed in the study were fully explained. The research followed the tenets of the Declaration of Helsinki and was approved by the Ethics Committee of the Catholic University.

Apparatus and procedure

Subfoveal ChBF

Because the LDF measurement of subfoveal ChBF has been previously described in detail (Riva et al. 2010), only a brief summary will be given here. In this study, a confocal LDF system was used with a probing beam in the near-infrared (laser diode, 785 nm) (Riva et al. 1994; Geiser et al.1999). Diameter and power of the beam at the cornea were 1.3 mm and 90 μW, respectively. This power conforms to the ANSI Z 136.1 standards for laser safety. The light scattered by the red blood cells (RBCs) in the volume sampled by the laser light was collected by a bundle of six optical fibres (core diameter of 110 μm each) and guided to an avalanche photodiode (APD). These fibres were arranged on a circle with a diameter of 180 μm, which was imaged at the retina so that the centre of the circle coincided with the focus of the incident beam.

The output current of the photodiode was sampled at a frequency of 44 kHz (bandwidth of 22 kHz) and processed with a software implemented on a PC to determine the ChBF LDF parameters in real time at a rate of 21.5 Hz, using an algorithm based on a photon diffusion theory (Bonner & Nossal 1981). These parameters are the velocity, choroidal blood velocity (ChBVel) (kHz), the number, choroidal blood volume [ChBVol, arbitrary units (A.U.)], and the relative flux, ChBF (A.U.), of RBCs within the sampled tissue region. This flux is proportional to blood flow if the haematocrit remains constant during the recordings. These parameters are related to each other through the relationship ChBF = k× ChBVel × ChBVol, where k represents an instrumental constant. The software automatically excludes the Doppler signal during blinks and artefacts caused by rapid eye movements. Only those recordings were kept where the direct signal (DC) at the output of the photodetector, which is proportional to the intensity of the light scattered by both the tissue and the RBCs in the illuminated volume and reaching the detector, did not fluctuate by more than 10% from the mean value during the recording duration (approximately 1 min). Subsequent data analysis and computation included average and standard deviation of the flow parameters over the recording time.

To improve the reproducibility, reliability, coefficient of variation, sensitivity as well as statistical power of the LDF technique in relation with ChBF measurements, Gugleta et al. (2002) using the same compact, confocal LDF instrument (Institut de recherche en Ophtalmologie, Sion, Lausanne, Switzerland) as the one in this study have introduced a new parameter called yield (defined as the ratio of DC/gain). This parameter was used to attenuate the influence of light scattering on the LDF parameters. Different light scattering properties of the tissue at the site of measurements are known to affect the flow parameters. In the processing unit of the instrument used by Gugleta et al. (2002), the gain of the detection system was adjustable. This capability was removed in our instrument so that a constant gain was used in all the recordings, eliminating therefore the need for using the yield parameter.

The LDF flowmeter (Institute for Research in Ophthalmology, Sion, Lausanne, Switzerland) was mounted on a slit-lamp support equipped with a chin rest and head holder. Subjects, seated, looked into the instrument, directly at the laser beam. The instrument was aligned by the operator in such a way that the DC signal, continuously displayed on the computer screen, was maximum. Recordings were obtained after pupil dilation with Tropicamide 1% eye drops, 30 min after each PERG recording session. In both normal subjects and patients, pupil size was at least 8 mm and no differences in mean pupil size were found between groups. For each patient and control subject, ChBVel, ChBVol and ChBF were determined as the average of three consecutive recordings; IOP, brachial blood systolic and diastolic pressures were determined at the end of each recording session. Mean arterial blood pressure and mean ocular perfusion pressure were calculated according to published formulas (Lovasik et al. 2003).

Pattern electroretinogram

Pattern electroretinograms were recorded according to a recently published paradigm (Falsini et al. 2008). Stimuli were horizontal gratings of 1.6 cycles/degree spatial frequency, modulated in counterphase at 8.14 Hz (corresponding to 16.28 contrast reversals/second), and electronically generated on a high-resolution monitor [contrast: 70%; mean luminance: 80 cd/m2; field size: 28° (width) × 16° (height)]. Subjects fixated at the centre of the stimulating field with natural pupils, whose size was measured, at a viewing distance of 57 cm. The subjects or patients wore full refractive correction for the test distance. Signals were recorded by a standard, flat-cup, 9-mm Ag/AgCl electrode taped on the skin of the lower eyelid. A similar electrode, placed over the eyelid of the contralateral, unstimulated eye, was used as reference. Responses were amplified (100 000), band-pass filtered (1–30 Hz), sampled with a resolution of 12 bits and averaged (250 events) with automatic artefact rejection. Peak-to-peak amplitude (in μV) of the Fourier analysed response 2nd harmonic was measured. Two replications were obtained for each record to verify reproducibility. In all subjects and patients as well as in all recording sessions, the responses satisfied the following criteria: intratest variability (differences between two replications) < 20% in amplitude and 2.45 ms in phase. An estimate of background noise at the 2nd harmonic was obtained in the study population in all sessions by recording PERG responses to a blank, unmodulated field kept at the same mean luminance as the stimulus. In all circumstances, the noise amplitude was ≤ 0.08 μV.

Data analysis

Laser Doppler flowmetry and PERG parameters, IOP, mean arterial blood pressure and mean ocular perfusion pressure values were compared between patients and control eyes by unpaired t-tests. The relationship between the LDF parameters of individual patients with OAG and the corresponding values of PERG amplitude, MD, PSD, IOP and systemic hemodynamic parameters was evaluated by Pearson’s correlation analysis. A p value < 0.05 was considered statistically significant. Although the data from both eyes of each patient and control subject were analysed, the main statistical comparisons and correlations were performed based on the results from the right eyes.

Results

There were no statistically significant differences in age, gender ratio, IOP and systemic hemodynamic parameters between patients with EMG and control subjects (Table 1). For all LDF parameters, within-subject intra-test variability, expressed as SD of the three consecutive measurements obtained from each eye of the patients with EMG and controls, was < 10%.

Figure 1 shows box plots of DC, ChBVol, ChBVel and ChBF (in each box, the symbol is the mean, the box indicates the median and interquartile range and the bars the 95 percentiles) obtained from control subjects and patients with EMG. In comparison with control subjects, mean ChBVel and ChBF were significantly reduced, by 30% (p = 0.007) and 33.4% (p = 0.009), respectively, in patients with EMG. As the changes in ChBVel and ChBF were roughly equal, ChBVol was not affected and did not change significantly (p = 0.87) between the two groups. The DC value was greater by about 30% (p = 0.0001) in patients with EMG compared with controls. No significant correlation was found between ChBVel and ChBF and the corresponding DC values in patients with EMG, as shown in Fig. 2, indicating that the LDF parameters were not depending upon the DC value. In Fig. 3, ChBF values are plotted as a function of the corresponding values of ChBVel in patients with EMG. The two Doppler flowmetry parameters were significantly correlated (p < 0.001), indicating that the decrease in ChBF in patients with EMG was mainly because of a reduction in ChBVel.

Figure 1.

 Box plots of subfoveal choroidal blood flow parameters in control subjects and patients with early manifest glaucoma (EMG). Compared to control values, in patients with EMG choroidal blood velocity (ChBVel) and choroidal blood flow (ChBF) were significantly reduced (p < 0.01), while direct current (DC) was increased (p < 0.01). Choroidal blood volume (ChBVol) did not change significantly between the two groups. A.U. = arbitrary units.

Figure 2.

 Scatter plots of choroidal blood velocity (left) and choroidal blood flow (right) as a function of the corresponding values of direct current (DC) in patients with early manifest glaucoma. No correlation was found between DC and laser Doppler flowmetry parameters. A.U. = arbitrary units.

Figure 3.

 Scatter plots of choroidal blood flow (ChBF) as a function of the corresponding values of choroidal blood velocity (ChBVel) in patients with early manifest glaucoma (EMG). The two Doppler flowmetry parameters were significantly correlated (p < 0.001). A.U. = arbitrary units.

In Fig. 4, PERG amplitude results are shown as box plots for control subjects and patients with EMG. Mean PERG amplitude was significantly reduced by 46% (p < 0.001) in patients with EMG, when compared to control values. In Fig. 5, PERG amplitude values are plotted as a function of the corresponding ChBVol, ChBVel and ChBF values found in patients with EMG and control subjects. No statistically significant correlation (respectively R = −0.04, p = 0.81; R = 0.24, p = 0.11 and R = 0.23, p = 0.12) was found between PERG amplitude and each of ChBVol, CHBVel and ChBF recorded in both groups.

Figure 4.

 Box plots of pattern electroretinogram (PERG) amplitude in control subjects and patients with early manifest glaucoma (EMG). Compared to control values, in patients with EMG, PERG amplitude was significantly reduced (p < 0.001).

Figure 5.

 Scatter plots of pattern electroretinogram (PERG) amplitude as a function of the corresponding values of choroidal blood volume, choroidal blood velocity and choroidal blood flow (from left to right) in patients with early manifest glaucoma (black dots) and control subjects (white dots). No correlation was found between PERG amplitude and laser Doppler flowmetry parameters. A.U. = arbitrary units.

The correlations between Humphrey MD and PSD values and the corresponding values of ChBVol, ChBVel and ChBF are shown in Fig. 6. MD tended to be less negative in eyes with higher ChBVel, although this positive correlation trend did not reach statistical significance.

Figure 6.

 Scatter plots of Humphrey mean deviation (MD) (top) and pattern standard deviation (PSD) (bottom) as a function of the corresponding values of choroidal blood volume, choroidal blood velocity and choroidal blood flow (from left to right) in patients with early manifest glaucoma (black dots) and control subjects (white dots). No correlation was found between perimetric MD or PSD and laser Doppler flowmetry parameters. A trend towards a positive significant correlation can be observed between MD and ChBVel. A.U. = arbitrary units.

No significant correlations were found between each of the LDF parameters and IOP, brachial blood systolic and diastolic pressures, mean arterial blood pressure and mean ocular perfusion pressure, measured in patients with EMG and in control subjects.

Discussion

This study was designed to evaluate whether subfoveal choroid blood flow is altered in patients with EMG and whether this alteration is associated with central retinal dysfunction as detected by PERG. Our results indicate that ChBF was significantly reduced in patients with EMG, primarily because of a decrease in ChBVel, whereas ChBVol did not appear to be affected. The DC values in patients with EMG were significantly increased on average when compared to control values. This could be due to early micro-anatomic changes involving the choriocapillaris layer of glaucomatous eyes, resulting in a general increase in the reflectance backscatter of the subfoveal tissue. DC changes likely did not affect the results of ChBF, because no correlations between DC values and each LDF parameter were found in both control subjects and patients.

Even if an impaired vascular auto-regulation is generally accepted as the main cause of ocular blood flow reduction in glaucomatous eyes (Flammer et al.1999; Grieshaber et al. 2007; Moore et al. 2008; Mozaffarieh et al. 2008), alterations in the choriocapillaris micro-vascular architecture could play an additional role in increasing the resistance and, consequently, reducing velocity and flow of the subfoveolar choriocapillaris blood stream. In his study, Spraul et al. (2002) observed a lower density of the capillaries in the macular choriocapillaris of glaucomatous eyes. Although these histological results were found in end-stage primary OAG, it can be hypothesized that minimal vasculature changes might appear even at the earliest phases of the disease, resulting in blood flow abnormalities.

Our findings are in general agreement with those of previous studies, supporting a reduced ChBF in patients affected by OHT and glaucoma, although in those studies different assessment techniques were used. A reduced pulsatile ocular blood flow (Trew & Smith 1991; Fontana et al. 1998; Fuchsjäger-Mayrl et al. 2004) and a slower choroidal hemodynamic using video fluorescein angiography (Duijm et al. 1997) have been observed in patients with OAG and normal tension glaucoma. The current results, however, differ from those of Grunwald et al. (1998) in showing a reduced ChBF in glaucoma. Indeed, by using a technique similar to that employed in this study, Grunwald et al. (1998) found no significant ChBF differences between normal subjects and patients with glaucoma. The difference between Grunwald’s and our results may be ascribed to the different clinical characteristics of the patients enrolled in the two studies or perhaps in the IOP-lowering medications. However, a systematic study on the effects of IOP-lowering medications administered to our patients was beyond the scope of our study.

A prominent part of our study was the evaluation of the relationship between the amount of blood flow reduction and the severity of central retinal dysfunction as estimated by PERG. Several studies (Pfeiffer et al.1993; Ventura et al. 2005; Bach et al. 2006; Parisi et al. 2006; Falsini et al. 2008) have reported abnormalities of the PERG in a substantial proportion of patients with EMG and OHT. As expected, PERG amplitude was significantly reduced in our patients with EMG when compared to healthy subjects, consistent mainly with a loss in the number and/or function of RGCs. The PERG amplitude did not show a significant correlation with the ChBF parameters, suggesting a lack of causal relationship between the reduction in flow and the level of central retinal dysfunction. Similarly, the quantitative correlations of the visual field with ChBF parameters were not statistically significant, except for a trend towards a positive correlation between MD and ChBVel.

The current results may have some relevance for a better understanding of the pathophysiology and treatment for early glaucoma. Indeed, based on our data, it can be presumed that the alteration in the subfoveal choroidal circulation may develop in parallel with the early stages of glaucoma, although this alteration does not appear to be a major aetiologic factor for the dysfunction of the central retina. A reduced ChBF may complicate, independently of the anterior optic nerve rim damage, the disease course by acting at the level of the outer retina, where it might induce photoreceptor/bipolar cell dysfunction. It is unknown whether a diminished subfoveal ChBF could be predictive of a faster disease progression. Longitudinal studies should be undertaken to address this question. In the meantime, a pharmacologic approach targeting choroidal circulation in patients with diminished ChBF could be reasonable and of potential value, to confer a better protection from disease progression.

Acknowledgements

This study was supported by a grant from Fondazione Cassa di Risparmio in Bologna, Italy (to the Department of Ophthalmology, EC Campos and CE Riva), and a grant from Ministero della Ricerca, Fondi di Ateneo ex 60% (to Benedetto Falsini).

Ancillary