Conjunctival Venular Hemodynamic Biomarkers in Glaucoma

Abstract

Purpose

Vascular dysfunction is considered a factor contributing to glaucomatous damage. Imaging the bulbar conjunctiva offers a noninvasive approach for directly assessing systemic microcirculation. The purpose of the current study was to test the hypothesis that there is an association between alterations in conjunctival venular hemodynamic biomarkers and presence of primary open-angle glaucoma (POAG).

Methods

This cross-sectional study included 26 normal control (NC) and 29 POAG subjects. Conjunctival microcirculation was assessed non-invasively by combining video microscopy and an automated software to measure the following conjunctival venular hemodynamic biomarkers: vessel diameter (D), blood velocity (V), blood flow (Q), wall shear rate (WSR), and wall shear stress (WSS).

Results

Age differed between NC and POAG groups (P < 0.001), while other ocular and systemic factors were not significantly different (P ≥ 0.05). Conjunctival venular V, WSR, and WSS were significantly reduced in POAG compared to NC, both with and without adjusting for covariates and accounting for multiple comparisons (adjusted P ≤ 0.03). In the adjusted models, the differences in conjunctival venular D and Q between NC and POAG groups were not significant (P ≥ 0.16).

Conclusions

The finding of reduced conjunctival venular wall shear stress in POAG supports the vascular theory of glaucoma, implicating endothelial dysfunction. Conjunctival wall shear stress may serve as a novel biomarker for glaucomatous vascular damage.

Translational Relevance

Assessment of microvasculature wall shear stress may serve as a novel biomarker of endothelium function with the potential to improve the understanding, diagnosis and monitoring of glaucoma and other optic neuropathies.

Introduction

Glaucoma encompasses a group of optic neuropathies that are characterized by progressive retinal ganglion cell degeneration and irreversible damage to the optic nerve. As the leading cause of irreversible blindness, it currently affects approximately 95 million people worldwide,¹ with projections estimating a rise to 111.8 million people by 2040.² This increase is due to aging populations and improved diagnostic capabilities, especially in low-resource settings.³ The gradual onset of glaucoma presents many clinical challenges. Early disease is frequently asymptomatic, and most patients often remain unaware of their condition⁴ until significant peripheral vision degeneration begins to impact their daily activities.⁵ By the time patients seek treatment, irreversible structural damage limits therapeutic efficacy. Conventional therapies focused on lowering intraocular pressure (IOP) fail to halt disease progression in approximately 30% to 45% of cases,^6,7 demonstrating the critical need for biomarkers other than IOP to improve early glaucoma detection.

Although increased IOP can directly apply mechanical compression of axonal and vasculature structures, the existence of normal-tension glaucoma (NTG) and patients with ocular hypertension (increased IOP without visible damage) suggests the existence of a nonmechanical influence in the pathogenesis of glaucoma.⁸ The vascular theory of glaucoma posits that vascular dysfunction and insufficient blood flow damages the optic nerve head independent of IOP elevation.⁸ Recent research has identified systemic and ocular microcirculatory factors including ocular systolic perfusion pressure flow as critical factors of glaucoma progression.^6,9

One method for assessing systemic microcirculation is nailfold capillaroscopy, which enables in vivo imaging of vessel morphology and capillary hemodynamics.¹⁰ Although nailfold capillaroscopy has been traditionally used to diagnose rheumatological disorders such as connective tissue diseases and systemic sclerosis, it has also emerged as a tool for assessing vascular dysfunction in non-rheumatological diseases including diabetes and glaucoma.^10–12 Another noninvasive approach for directly assessing systemic microcirculation is imaging of the bulbar conjunctiva. Several methods have been developed for the evaluation of conjunctival microcirculation including optical coherence tomography angiography,¹³ confocal microscopy,^14,15 video microscopy,^16–18 and retinal function imager.¹⁹ Relationships between conjunctival microcirculation hemodynamics and systemic diseases such as Alzheimer's, diabetes, hypertension, sepsis, and coronary microvascular disease have been reported.²⁰

Previous studies have reported alterations in microvascular morphology and hemodynamics due to glaucoma. Specifically, increased nailfold vessel tortuosity was detected in exfoliation syndrome with or without glaucoma (XFS/XFG) as compared to POAG.²¹ In NTG, alterations in nailfold capillary density and abnormalities in vascular pattern were reported.^22,23 In POAG, nailfold capillary abnormalities were associated with visual field loss and retinal thickness.^24,25 Nailfold vascular anomalies have also been associated with elevated plasma matrix metalloproteinase-9 level, a known contributor to endothelial dysfunction.^26,27 Conjunctival vessel density was shown to be associated with the use of a prostaglandin analog treatment of glaucoma.²⁸ With regard to hemodynamics, nailfold capillary blood flow was shown to be reduced in POAG, XFG, and NTG^9,29 and correlated with deep optic nerve head blood flow in POAG.³⁰

Wall shear stress (WSS) is the tangential force per unit area exerted on the vessel wall and can be estimated by determining the wall shear rate (WSR), defined as the gradient of blood velocity at the vessel wall. WSS is considered a key regulator of endothelial cells and determinant of cellular morphology, metabolism, and inflammatory phenotype for maintaining endothelial cell function.^31,32 In systemic diseases, WSS has been implicated in the development of atherosclerosis and plaque rupture in patients with high risk for coronary artery disease.³³ In the microvasculature, WSS has been shown to affect angiogenesis, vascular morphogenesis, vascular remodeling, and vascular tone.³⁴ Studies have elucidated the molecular mechanisms involved in sensing and response of endothelial cells to WSS. Therefore assessment of microvasculature WSS may serve as a novel biomarker of endothelium function, providing a more direct measure of hemodynamic properties that influence glaucoma risk compared to blood flow.

We have developed an imaging method for quantitative assessment of conjunctival hemodynamics biomarkers (vessel diameter, blood velocity and flow, wall shear rate and stress).^35–38 The method was previously applied to study microcirculation alterations in diabetic retinopathy, sickle cell disease (retinopathy and nephropathy), and stroke.^39–44 The purpose of the current study was to test the hypothesis that there is an association between alterations in conjunctival venular hemodynamic biomarkers and the presence of POAG.

Methods

Subjects

The study was conducted at the University of Illinois at Chicago and University of Southern California and was approved by their corresponding Institutional Review Boards. Informed consents were obtained in accordance to the tenets of the Declaration of Helsinki after the study was explained to the subjects. Systolic blood pressure (SBP) and diastolic BP (DBP) were recorded using a wrist cuff to determine the mean arterial pressure (MAP) by calculating 1/3 SBP + 2/3 DBP. Three MAP and heart rate (HR) measurements were averaged per subject. A blood sample was obtained via finger-prick and centrifuged in a micro-hematocrit centrifuge (Unico, Dayton, NJ, USA) to measure hematocrit (HCT). Intraocular pressure (IOP) was measured by Goldmann applanation tonometry. Data in one eye of each subject were selected for analyses.

Image Acquisition

Our previously established noninvasive optical imaging system (EyeFlow)^37–39 was used to capture conjunctival microcirculation images. Briefly, the conjunctiva was illuminated with a white light source of a slit lamp biomicroscope equipped with an optical filter with a transmission wavelength of 540 ± 5 nm. Image sequences one second in duration were acquired from the superficial conjunctival microcirculation using a digital charged coupled device camera at 50 frames per second. Imaging was repeated at multiple non-overlapping regions temporal to the limbus, encompassing a conjunctival area of approximately 10 mm by 13 mm.

Image Processing and Analysis

The automated technique for hemodynamic assessment of the conjunctival microvascular network consisted of multiple image processing steps, which have been described in detail elsewhere.^37–39 Briefly, the method consisted of image registration, vessel segmentation, centerline extraction and bifurcation detection, diameter and axial blood velocity measurement. Vessel diameter (D) was measured based on calculating the full width at half maximum of intensity profiles of lines perpendicular to the vessel centerline. A spatial-temporal image (STI) technique³⁷ was used to track the movement of red blood cells along the vessel centerline in the registered image sequences to calculate axial blood velocity. Using previously established formulas,^39,45,46 average cross-sectional velocity (V) was calculated from axial blood velocity and D measurements. Based on D and V, blood flow (Q = VπD²/4) and wall shear rate (WSR = 8V/D) were computed. Wall shear stress (WSS = ηWSR) was calculated as the product of WSR and the dynamic blood viscosity (η), which was determined based on measurements of HCT and D, as previously described.⁴⁶ Measurements were obtained in several venules which were selected for analysis due to their larger number and less sensitivity to pulsation variation, as compared to arterioles. Venules were identified by visual inspection of the direction of blood flow and characterized by the convergence of blood flow into other vessels.

Statistical Analysis

Normality of data distribution of continuous variables was assessed through Shapiro-Wilk tests and graphical visualization of quantile-quantile plots. Demographic and systemic variables (sex, race, age, eye examined, MAP, HR, HCT, IOP) were compared between NC and POAG groups using unpaired t-tests or Mann-Whitney U tests and χ² tests. Generalized linear mixed models (LMM) were performed to determine the associations between each hemodynamic biomarker (D, V, Q, WSR, WSS) and groups (NC, POAG), including subject as random intercepts to account for multiple measurements in vessels per subject. Adjusted models included fixed effects of covariates (sex, race, age, MAP, HR, HCT, IOP). Estimated slopes (β) and P values for unadjusted and adjusted models were reported. LMMs were also used to assess associations between V and D in NC and POAG groups in unadjusted and adjusted models. To account for multiple comparisons of the five hemodynamic biomarkers, the Benjamini-Hochberg procedure was applied to determine adjusted p-values. The potential effects of vasoactive medications and systemic vascular diseases was assessed by performing a secondary LMM analysis on a subgroup of POAG subjects (N = 19), which excluded data from those subjects using alpha-agonist medications, with a history of coronary disease, or diagnosis of hypertension (N = 10). Analyses were performed using SPSS software (version 29.0.2.0), and a two-sided P < 0.05 was considered statistically significant.

Results

Subject Characteristics

Subjects were stratified into healthy normal control (NC) (N = 26) and POAG (N = 29) groups based on clinical examination by a glaucoma or retina (NC only) specialist. According to a previously published method of glaucoma staging, there were 20, four, four, and one POAG subjects with early, moderate, advanced, and severe stages of glaucoma, respectively.⁴⁷ Among the POAG group, no subjects had diabetes, 10 had controlled hypertension, two had a history of heart disease, and one had a history of heart bypass. Regarding topical glaucoma medications, no patients were on a rho-kinase inhibitor, four were on an alpha-agonist, and 20 were on a prostaglandin analog. Although alpha-agonists are known to have vasodilatory properties, the effect of prostaglandin analogs on vessel diameters is minimal.^48–50 Table 1 presents the demographic and physiological characteristics of the subjects. There were no significant differences in sex, race, eye examined, MAP, HR, and HCT between NC and POAG groups (P ≥ 0.07). IOP was marginally different between groups (P = 0.05). Age was significantly lower in the NC group compared to POAG group (P < 0.001).

Table 1.

Comparison of Subjects’ Characteristics

	Total (N = 55)		NC (N = 26)		POAG (N = 29)
	N	%	N	%	N	%	P Value
Sex							0.15*
Male	18	33	6	23	12	41
Female	37	67	20	77	17	59
Race							0.07*
Asian	14	25	8	31	6	21
African American	11	20	6	23	5	17
White	13	24	2	8	11	38
Hispanic	17	31	10	38	7	24
Eye examined							0.70*
Right	26	47	13	50	13	45
Left	29	53	13	50	16	55
MD (dB)	—		—		−5.8 ± 6.2		—
Age (years)	62 ± 14		54 ± 10		69 ± 14		<0.001 †
MAP (mm Hg)	102 ± 13		102 ± 14		102 ± 12		0.86‡
HR (beats/min)	66 ± 10		69 ± 10		64 ± 10		0.08‡
HCT (%)	44 ± 5		45 ± 5		43 ± 5		0.18†
IOP (mm Hg)	13 ± 3		14 ± 3		13 ± 3		0.05‡

MD, mean deviation on Humphrey visual field test.

Bold signifies statistical significance. Age, MAP, HR, HCT, and IOP modeled as mean ± standard deviation.

^* P values derived by χ² test.

^‡ P values derived by unpaired t-test.

^† P values derived by Mann-Whitney U test.

Description of Conjunctival Hemodynamic Biomarkers

A total of 1192 and 1473 conjunctival venules were evaluated in the NC and POAG groups, respectively. The number of venules per subject was on average 46 and 51 in NC and POAG groups, respectively. Mean and standard deviation of conjunctival hemodynamic biomarkers averaged in all vessels of NC and POAG groups is presented in Table 2.

Table 2.

Descriptive Conjunctival Hemodynamic Bio-markers in NC and POAG Groups

Hemodynamic Biomarker	NC (N = 26)	POAG (N = 29)
D (µm)	19 ± 3	21 ± 3
V (mm/s)	0.35 ± 0.08	0.30 ± 0.08
Q (pl/s)	128 ± 59	124 ± 45
WSR (s−1)	158 ± 40	127 ± 41
WSS (dyne/cm2)	0.42 ± 0.13	0.31 ± 0.11

D, vessel diameter; V, cross-sectional velocity; Q, blood flow.

Values are reported as mean ± standard deviation.

Associations of Conjunctival Hemodynamic Biomarkers With POAG

Associations of hemodynamic biomarkers with the presence of POAG determined from models at vessel level are presented in Table 3. Compared to the NC group, conjunctival V, WSR, and WSS were decreased in the POAG group, with and without adjusting for covariates, and accounting for multiple comparisons (adjusted P ≤ 0.03). Conjunctival D and Q were not significantly associated with POAG in both unadjusted and adjusted models (adjusted P ≥ 0.09). In a subgroup of POAG subjects (N = 19) without use of alpha-agonists, a history of coronary disease and/or diagnosis of hypertension, reductions in conjunctival V, WSR, and WSS remained significant in the adjusted model and accounting for multiple comparisons (adjusted P = 0.02).

Table 3.

Associations of Conjunctival Hemodynamic Biomarkers With POAG Versus NC as Reference

	Unadjusted Model		Adjusted Model*
Hemodynamic Biomarker	β (LCL, UCL)	Adjusted P Value†	β (LCL, UCL)	Adjusted P Value†
D (µm)	1.48 (−0.124, 3.09)	0.09	1.63 (−0.47, 3.73)	0.16
V (mm/s)	−0.05 (−0.09, −0.10)	0.03	−0.07 (−0.13, −0.01)	0.03
Q (pl/s)	−4.88 (−33.4, 23.6)	0.73	−12.9 (−51.6, 25.8)	0.51
WSR (s−1)	−31.3 (−53.0, −9.49)	0.02	−40.6 (−69.3, −11.8)	0.02
WSS (dyne/cm2)	−0.10 (−0.17, −0.04)	0.005	−0.12 (−0.21, −0.04)	0.02

β, estimated slope; D, vessel diameter; LCL, lower 95% confidence limit; Q, blood flow; UCL, upper 95% confidence limit; V, cross sectional velocity.

^*Adjusted models included sex, race, age, MAP, HR, HCT, IOP as fixed effects.

^† P values were derived from linear mixed models and adjusted for multiple comparisons using Benjamini-Hochberg procedure.

Discussion

The current study demonstrated that reductions in conjunctival venular hemodynamics biomarkers, namely blood velocity, wall shear rate and stress are associated with the present of POAG. To our knowledge, this is the first study to quantify key regulators of conjunctival hemodynamics in POAG. The observed alterations of conjunctival hemodynamic biomarkers in POAG underscore the potential role of vascular factors in the pathophysiology of glaucoma. This aligns with the vascular theory of glaucoma which posits systemic vascular dysregulation as a contributor to optic nerve degeneration and extends previous research by implicating endothelial dysfunction.

Decreased wall shear stress is particularly compelling because of its direct association with endothelial dysfunction compared to its individual components (V, D, Q, and η). Li et al.⁵¹ demonstrated that WSS affects mechano-transduction pathways via a variety of ways including integrins, receptor tyrosine kinases, and ion channels. In addition, Sun et al.⁵² found that low WSS, typical of disturbed or turbulent blood flow, was associated with disordered endothelial cell arrangement, consistent with studies that showed WSS modulates endothelial cell migration and remodeling via Rho family small GTPases and FAK-associated pathways.⁵¹ Furthermore, WSS has been implicated in the regulation of endothelial apoptosis and proliferation via PI3K/Akt and MAPK pathways.⁵¹ Our finding of reduced WSS in the conjunctival microvasculature in POAG, suggests potential endothelial dysfunction and apoptosis. These factors can have an exacerbating effect on ischemic retinal and glaucomatous injury, because they potentially occur in parallel in the retinal vasculature.

The current study was focused primarily on conjunctival venules, which were selected based on their greater number and lower sensitivity to pulsation compared to arterioles. Although the arterial endothelium is a primary regulator of vascular tone, the post-capillary venule endothelium is the principal site for inflammatory processes such as leukocyte adhesion and trafficking.⁵³ Therefore venular microcirculation represents a highly relevant anatomical site to investigate pathological changes related to the vascular component of glaucoma. Future studies using cardiac synchronization or increased imaging duration could account for inherent arteriole pulsatility and allow for the assessment of arteriolar hemodynamics. Such analysis would provide complementary information on the regulation of vascular tone and local resistance, thus offering a more complete picture of microvascular dysregulation in glaucoma.

Assessment of conjunctival hemodynamics can be performed non-invasively in a clinical ophthalmic imaging setting. Although traditional hemodynamic parameters such as blood flow, velocity, and capillary density provide useful information, WSS provides a more direct measure of the damaging effect of the mechanical forces that are exerted on the vascular endothelium. Future studies are needed to substantiate the value of WSS for better understanding the pathogenesis of glaucoma and early glaucoma detection, especially in NTG cases in which traditional IOP-centric models fail.

In the current study, no significant reduction in conjunctival blood flow was found in POAG compared to NC, which may be contributed to the small sample size, whereas a previous study reported reduced nailfold capillary blood flow.⁹ Notably, the relationship between nailfold and conjunctival circulations is not known and has not been evaluated. Nevertheless, there are notable differences in vessel sampling and severity of glaucoma between the two studies. The current study sampled on average 50 venules per eye with an average diameter of 19 um, whereas in the previous study a significantly lower number of vessels (10 vessels) with a smaller diameter (10 um) was sampled. The cohort of POAG subjects in the current study had an average mean deviation of −5.8 dB, thus representing a lower visual field loss and glaucoma severity than the cohort in the previous study (−9.5 dB). Also, there is variability in blood flow among POAG subjects as both increased and decreased ocular blood flow was documented.⁵⁴ Future studies in the same cohort of subjects are needed to elucidate hemodynamic differences in different microcirculatory networks.

Our study had several limitations. First, the sample size was relatively small, which limited detection of differences in some hemodynamic biomarkers and subgroup analyses of changes in WSS according to stages of glaucoma. Second, the study design was cross-sectional, thus precluding causal relationships. Future longitudinal studies in a larger cohort are needed to address these limitations. Third, the imaging system was not synchronized with the cardiac cycle, although velocity changes caused by pulsatility are expected to be low in venules. Fourth, imaging of the microcirculation was limited to measurements in superficial vessels that were in the imaging system's focal plane. Fifth, there was a difference in age between the NC and POAG groups, although we accounted for age in the adjusted statistical models. Sixth, changes in ocular surface that may have affected conjunctival microcirculation imaging and may have contributed to the variability of data. Last, although secondary analyses were performed by excluding data from subjects using vasoactive medications and having systemic vascular diseases, in future studies, the confounding effects of local and systemic factors on conjunctival microcirculation should be considered.

In conclusion, we found that patients with POAG have significant changes in conjunctival hemodynamics biomarkers compared with normal controls, specifically in conjunctival vessel wall shear stress, a potential marker of endothelium dysfunction. Further studies are necessary to establish the clinical significance of wall shear stress as a risk factor for the development of glaucoma and a biomarker for predicting disease progression.