| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 From the Department of Ophthalmology, Indiana University School of Medicine, Indianapolis; 2 Division of Vision Sciences, Aston University, Birmingham, United Kingdom; and 3 Medical Sciences Program, Indiana University, Bloomington.
Abstract
PURPOSE. Although glaucomatous visual field defects are more common in the superior field than in the inferior field, microaneurysms are more frequent in the superior than in the inferior retina in diabetic retinopathy. The authors hypothesized that differences in vascular hemodynamics in the two areas might contribute to these phenomena.
METHODS. The blood flow response to hyperoxia and hypercapnia was evaluated in peripapillary retinal tissue superior and inferior to the optic nerve head using confocal scanning laser Doppler flowmetry. In 14 young, healthy persons, blood flow was measured while breathing room air and during isocapnic hyperoxia (100% O2 breathing) and isoxic hypercapnia (PCO2 increased 15% above baseline). Histograms were generated from pixel-by-pixel analysis of retinal portions of superior and inferior temporal quadrants of the entire image.
RESULTS. Baseline blood flow in the inferior temporal quadrant was significantly greater than in the superior temporal quadrant (P < 0.05). However, the inferior region failed to increase in perfusion during hypercapnia and experienced significant mean blood flow reduction; flow reduction in the pixels at the 25th, 50th, 75th, and 90th percentile of flow; and an increased percentage of pixels without measurable flow, during hyperoxia (each P < 0.05). In contrast, in the superior temporal region, hyperoxia failed to reduce blood volume, velocity, or flow, whereas hypercapnia significantly increased mean flow; increased flow in the pixels at the 25th, 50th, 75th, and 90th percentile of flow; and reduced the percentage of pixels without measurable flow (each P < 0.05).
CONCLUSIONS. The inferior temporal quadrant of the peripapillary retina is, in comparison with the superior temporal region, less responsive to vasodilation and more responsive to vasoconstriction. These differences could contribute to different susceptibility to visual field defect or vascular dysfunction in the superior and inferior retina.
Defects in the superior visual field are more common than in the inferior visual field in glaucoma.1 Further, narrowing of the retinal arteries and veins, a development that occurs in proportion to disease severity, is most pronounced inferiorly.2 In contrast, in diabetic retinopathy, microaneurysms, and acellular capillaries are more than twice as common in the superior than in the inferior retina.3 Additionally, enlargement of the retinal veins, a change that correlates with the severity of disease and the magnitude of hyperglycemia, is also most pronounced superiorly.4 The mechanisms that give rise to these distinct, disease-specific regional differences in vascular and visual field defect have not been defined.
It is possible that differences in vulnerability to ischemic insult or to hyperglycemic damage arise from inherent retinal regional differences in vasoreactivity. To test this hypothesis, we compared the blood flow responses of the inferior and superior retinal regions to vasoconstrictor and vasodilator stimuli. Hyperoxia, which provokes cerebral and whole-retinal vasoconstriction,5 and hypercapnia, which dilates arteries and arterioles within the brain and the eye,6 were used as vasoprovocative stimuli. Differing responses of the superior and inferior retina to these stimuli would support the hypothesis that these retinal regions differ in their susceptibility to ischemic or hyperglycemic insult.
Materials and Methods
Subjects
Fourteen healthy volunteers (7 men, 7 women; mean age, 27 ±
6 years; age range, 18–40 years) participated in the study. Subjects
had normal eye examinations, with corrected visual acuity 20/30 or
better, intraocular pressure (IOP) below 22 mm Hg, refractive error
between -6.00 and +2.00 diopters, and astigmatism less than 1.50
diopters cylinder. Subjects were free from heart or lung disease, had
no family history of glaucoma or diabetes, and were not pregnant or
anticipating pregnancy at the time of study. All procedures conformed
to the tenets of the Declaration of Helsinki and were reviewed and
approved by an institutional review board, with subjects signing
informed consent.
Experimental Design
Two experimental sessions, one involving hyperoxia and the other
hypercapnia, took place on separate days. The mean interval between
experiments was 10 ± 5 days (range, 2–19 days). Ocular blood
flow was measured using confocal scanning laser Doppler flowmetry
(cSLDF, Heidelberg Retinal Flowmeter; Heidelberg Engineering,
Heidelberg, Germany). Heart rate and arterial oxygen saturation were
monitored using pulse oximetry. After 5 minutes acclimation, baseline
recordings were made with the subject breathing room air. End-tidal
CO2 and O2 were monitored
continuously from a mouthpiece (Pulse Oximeter and End-Tidal Gas
Analyzer: POET II model 602-3, Criticare Systems, Milwaukee,
WI).
Hyperoxia and Hypercapnia
Hyperoxia was induced by adding 100% O2 to
the inspired gas mixture; CO2 was added to
maintain isocapnia during this procedure. Five minutes after the
end-tidal O2 fraction exceeded 70%, ocular blood
flow measurements were made. In experiments involving hypercapnia,
end-tidal PCO2 was elevated 15%
above baseline levels for 5 minutes before measurement of ocular blood
flow. Elevation of PCO2 was
accomplished by addition of 4 to 6% CO2 to the
inspiratory gas mixture.
Measurement of Ocular Blood Flow: cSLDF
One eye from each subject was randomly selected for study. With
undilated pupils, subjects fixed on a static target 3 feet away. Using
a 10° field, two mapped images (superior and inferior) were obtained
across the optic nerve head for each experimental condition. Images
were focused on the superficial retina, with focus setting and anatomic
location constant across all images from a given subject.
One observer reviewed all images (Fig. 1) . Unlike conventional analysis that uses a single 10 pixel x 10 pixel sample, each qualifying pixel from the superficial peripapillary retinal field within the 256 x 64 pixel image was included in the analysis (Fig. 1) . This methodology significantly increases the reproducibility of measurements of blood velocity, volume, and flow.7 Excluded were pixels from the cup and rim (which were poorly focused), pixels from major blood vessels, image areas interrupted by movement saccades, and areas without acceptable levels of brightness (any brightness [DC] value <70 or >200). To ensure that the same retinal locations were used for each image, a transparent overlay was used to map the retinal vasculature of each optic nerve head. This template was then overlaid on the image, and the temporal peripapillary area was separated into inferior and superior divisions. To produce a histogram, the total number of pixels for all images was determined, the average was calculated, and a normalized pixel count was calculated that gave equal weight to each subject. Pixels with less than one arbitrary unit of flow were counted as "zero flow" pixels. The number of zero flow pixels and the flow, volume, and velocity in pixels at the 25th, 50th, 75th, and 90th percentile in each category were used for analysis.7
|
= 0.05, while providing the capability for detection of a 12%
difference in blood flow. Results
Hyperoxia and Hypercapnia
Hyperoxia increased end-tidal
PO2 from 103 ± 5 (mean ±
SD) to 562 ± 80 mm Hg (P < 0.05); hypercapnia
increased end-tidal PCO2 from 32 ± 2
to 39 ± 4 mm Hg (P < 0.05). Hyperoxia reduced
heart rate from 84 ± 9 to 78 ± 7 beats/min
(P < 0.05) at constant arterial blood pressure, while
hypercapnia increased systolic blood pressure from 112 ± 8 to
115 ± 8 mm Hg (P < 0.05) at constant heart rate
and diastolic pressure.
Pixel Counts from HRF images
Group mean pixel counts did not differ between images taken from
the superior and inferior peripapillary retina (Tables 1 2
3)
. Pixel counts also did not differ between baseline images and those
taken during hyperoxia or hypercapnia (Tables 1
2
3) . The overall
mean pixel count averaged 1156 pixels/image, a quantity 11.5 times
larger than the default 10 pixel x 10 pixel box.
|
|
|
The two baseline readings, when compared over the same anatomic retinal regions, showed similar volume, velocity, and flow recordings except in regard to superior temporal blood flow at the 90th percentile pixel. This flow value was higher in the prehyperoxia recording (751 ± 34; range, 499–906 arbitrary units) than in the prehypercapnia reading (693 ± 34; range, 499–930 arbitrary units; P < 0.05).
Response to Hyperoxia and Hypercapnia: Superior Temporal
Peripapillary Retina
In the superior temporal peripapillary retina, hyperoxia failed to
change mean blood flow (Table 2)
. Hyperoxia also failed to change the
percentage of zero flow pixels, or the volume, velocity, or flow in
either the 25th, 50th, 75th, or 90th percentile pixel (Table 2)
. In
contrast, hypercapnia significantly increased mean blood volume,
velocity, and flow; decreased the percentage of zero flow pixels; and
increased blood volume, velocity, and flow within pixels at the 25th,
50th, 75, and 90th percentiles of each of these variables (Table 2)
.
Response to Hyperoxia and Hypercapnia: Inferior Temporal
Peripapillary Retina
In contrast to results found in the superior retina, in the
inferior temporal peripapillary retina, hyperoxia significantly reduced
mean blood flow (Table 3)
. Although the percentage of zero-flow pixels
was unchanged by hyperoxia, mean blood volume and blood volume within
pixels at the 25th, 50th, 75th, and 90th percentiles of this variable
was reduced (Table 3)
. In addition, mean velocity was reduced within
the 90th percentile pixel (Table 3)
. Blood flow was reduced during
hyperoxia in the 75th and 90th percentile pixels (Table 3)
.
Hypercapnia, however, failed to change any measured aspect of
hemodynamics in the inferior temporal quadrant. During
CO2 elevation, mean volume, velocity, and flow
were unaffected, the percentage of zero-flow pixels was unchanged, and
at each percentile of the cSLDF histogram, volume, velocity, and flow
were unaltered (Table 3)
.
Discussion
Higher baseline perfusion in the inferior temporal compared with the superior temporal retina may arise from several possibilities. There is evidence that the normal inferior temporal artery and vein are larger than the analogous superior vessels.2 Past authors have speculated that these vessels are increased in diameter in the inferior retina in part because the fovea is located slightly inferior to a horizontal midline and in part because the neuroretinal rim is broader inferiorly than superiorly.2 Our findings suggest in addition that the inferior temporal retina receives a slightly greater capillary perfusion. It is unknown if greater basal perfusion is associated with increased density of ganglion cells, of cells of the intermediate cell layer, or is instead linked to other factors. In other regions of the brain, although differences in metabolic activity are directly tied to differences in oxygen consumption, these differences are only roughly related to differences in blood flow.8
A number of previous studies have used hyperoxia and hypercapnia as stimuli for vasoconstriction and vasodilation. Hyperoxia reduces total cerebral blood flow, whereas hypercapnia increases bulk cerebral perfusion, with these effects mediated locally via changes in arteriolar diameter and unrelated (as again seen in this study) to any changes in systemic arterial blood pressure. However, changes in both PO2 and PCO2 give rise to substantial cerebral inter-regional variation in the constrictor or dilator response.9 This regional variation exists also within the retina, with our results showing the most anatomically localized inter-regional difference yet described in cerebral circulatory responsiveness to vasoactive stimuli.
The mechanisms that underlie regional vasoresponsiveness within the retina remain ill-defined. In the brain as a whole, it is clear that the visual cortex, for example, maintains a relatively high blood flow and high oxygen extraction in comparison to the sensorimotor cortex,8 but the factors regulating this difference are not known. Cerebral capillary perfusion appears to be controlled at least in part by neuronally derived nitric oxide: inhibition of neuronal nitric oxide synthase blocks the cerebral capillary perfusion increase provoked by hypoxia.10 However, this generalization does not explain why various tissues (or regions within a relatively homogeneous tissue) may respond differently to vasoactive stimuli. There is evidence, however, that chronically increased or decreased perfusion of regions within a single organ leads to subsequent alterations in vasoreactivity, due to changes in both neuronal and endothelially mediated vascular regulatory processes.11
The significance of intraretinal differences in vasodilatory and vasoconstrictor responsiveness also is as yet undefined. There are several possibilities, however, by which these differences might be linked to susceptibility to disease. Reduced vasodilator reserve, as seen in the temporal inferior retina, is associated in other organs with diminished capacity to withstand reductions in perfusion pressure or arterial oxygen content or increases in tissue metabolic demands.12 For the retina, reduced ocular perfusion pressure, as induced by either elevated IOP or reduced arterial pressure, could create risk for ischemic damage.13 Although most authors presume that the short posterior ciliary arterial supply of the laminar and prelaminar optic nerve head is of primary importance for the development of ischemia-induced damage in glaucoma, it is indeed in the inferior retina that at least indirect evidence for ischemia is most pronounced in glaucomatous ocular disease.2
Although the loss or absence of vasodilator reserve may exacerbate the risk of ischemic damage to tissue, an inability to generate autoregulatory vasoconstrictor responses also may expose cells to potential damage. In diabetes mellitus, a number of primary nonvascular mechanisms, ranging from the induction of angiogenic growth factors to the actions of advanced glycation end products, may underlie development of retinopathy.14 However, the microvascular complications of chronic hyperglycemia do include impaired vasoconstriction in response to endogenous endothelin-1, neuropeptide Y, and to nonspecific stimuli such as cold exposure.15 Loss of normal vascular smooth muscle contractile responses could expose tissue to hyperperfusion and subsequent microvascular damage.15 These cellular level changes may account for the loss of overall retinal vasoconstrictor responsiveness to hyperoxia that proceeds in proportion to progression of diabetes.16 The relative lack of normal vasoconstrictor responsiveness in the superior temporal retina could explain why that region is more susceptible to development of microaneurysms and acellular capillaries in diabetes3 and why dilation of the retinal veins, which also occurs in proportion to diabetes severity, is most prominent superiorly.4
In summary, variation in vasoreactivity within the healthy retina may help explain differences in regional susceptibility to a number of retinal and optic nerve head diseases. Localized retinal under- or over-perfusion, as induced by a wide range of physiological or pathophysiological perturbations, may occur predictably on the basis of normal tissue autoregulatory capacity.
Footnotes
Supported in part by Grant EY10801 (AH) from the National Institutes of Health, Bethesda, Maryland; and by an unrestricted grant from Research to Prevent Blindness, New York, New York. AH is a recipient of the William and Mary Greve Award from Research to Prevent Blindness.
Submitted for publication November 13, 1998; revised April 2, 1999; accepted April 21, 1999.
Proprietary interest category: N.
Corresponding author: Alon Harris, Department of Ophthalmology, Indiana University School of Medicine, 702 Rotary Circle, Indianapolis, IN 46202. E-mail: alharris@indiana.edu
References
This article has been cited by other articles:
|
|
 |
|
  F. Tayyari, S. T. Venkataraman, E. D. Gilmore, T. Wong, J. Fisher, and C. Hudson The Relationship between Retinal Vascular Reactivity and Arteriolar Diameter in Response to Metabolic Provocation Invest. Ophthalmol. Vis. Sci., October 1, 2009; 50(10): 4814 - 4821. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. L. Trick, P. Edwards, U. Desai, and B. A. Berkowitz Early supernormal retinal oxygenation response in patients with diabetes. Invest. Ophthalmol. Vis. Sci., April 1, 2006; 47(4): 1612 - 1619. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Feigl, B. Brown, J. Lovie-Kitchin, and P. Swann Adaptation Responses in Early Age-Related Maculopathy Invest. Ophthalmol. Vis. Sci., December 1, 2005; 46(12): 4722 - 4727. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. D. Gilmore, C. Hudson, D. Preiss, and J. Fisher Retinal arteriolar diameter, blood velocity, and blood flow response to an isocapnic hyperoxic provocation Am J Physiol Heart Circ Physiol, June 1, 2005; 288(6): H2912 - H2917. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Jean-Louis, J. V. Lovasik, and H. Kergoat Systemic Hyperoxia and Retinal Vasomotor Responses Invest. Ophthalmol. Vis. Sci., May 1, 2005; 46(5): 1714 - 1720. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. G. Boehm, A. U. Koeller, and L. E. Pillunat The Effect of Age on Optic Nerve Head Blood Flow Invest. Ophthalmol. Vis. Sci., April 1, 2005; 46(4): 1291 - 1295. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. D. Gilmore, C. Hudson, S. T. Venkataraman, D. Preiss, and J. Fisher Comparison of Different Hyperoxic Paradigms to Induce Vasoconstriction: Implications for the Investigation of Retinal Vascular Reactivity Invest. Ophthalmol. Vis. Sci., September 1, 2004; 45(9): 3207 - 3212. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y Han, M A Bearse Jr, M E Schneck, S Barez, C Jacobsen, and A J Adams Towards optimal filtering of "standard" multifocal electroretinogram (mfERG) recordings: findings in normal and diabetic subjects Br J Ophthalmol, April 1, 2004; 88(4): 543 - 550. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C P Jonescu-Cuypers, A Harris, K U Bartz-Schmidt, L Kagemann, A S Boros, U E Heimann, B H Lenz, R-D Hilgers, and G K Krieglstein Reproducibility of circadian retinal and optic nerve head blood flow measurements by Heidelberg retina flowmetry Br J Ophthalmol, March 1, 2004; 88(3): 348 - 353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S L Hosking, A Harris, H S Chung, C P Jonescu-Cuypers, L Kagemann, E J Roff Hilton, and H Garzozi Ocular haemodynamic responses to induced hypercapnia and hyperoxia in glaucoma Br J Ophthalmol, March 1, 2004; 88(3): 406 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. A. Bearse Jr, Y. Han, M. E. Schneck, and A. J. Adams Retinal Function in Normal and Diabetic Eyes Mapped with the Slow Flash Multifocal Electroretinogram Invest. Ophthalmol. Vis. Sci., January 1, 2004; 45(1): 296 - 304. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Loukovaara, M. Harju, R. Kaaja, and I. Immonen Retinal Capillary Blood Flow in Diabetic and Nondiabetic Women during Pregnancy and Postpartum Period Invest. Ophthalmol. Vis. Sci., April 1, 2003; 44(4): 1486 - 1491. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A Harris, Y Ishii, H S Chung, C P Jonescu-Cuypers, L J McCranor, L Kagemann, and H J Garzozi Blood flow per unit retinal nerve fibre tissue volume is lower in the human inferior retina Br J Ophthalmol, February 1, 2003; 87(2): 184 - 188. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A Luksch, G Garhofer, A Imhof, K Polak, E Polska, G T Dorner, S Anzenhofer, M Wolzt, and L Schmetterer Effect of inhalation of different mixtures of O2 and CO2 on retinal blood flow Br J Ophthalmol, October 1, 2002; 86(10): 1143 - 1147. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J M Lopes de Faria, H Russ, and V P Costa Retinal nerve fibre layer loss in patients with type 1 diabetes mellitus without retinopathy Br J Ophthalmol, July 1, 2002; 86(7): 725 - 728. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. P Jonescu-Cuypers, H. S Chung, L. Kagemann, Y. Ishii, D. Zarfati, and A. Harris New neuroretinal rim blood flow evaluation method combining Heidelberg retina flowmetry and tomography Br J Ophthalmol, March 1, 2001; 85(3): 304 - 309. [Abstract] [Full Text]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
