Survey on retrobulbar blood flow in glaucomatous optic neuropathy (normotensive and hypertensive)

Introduction

Primary open-angle glaucoma (POAG) is a bilateral, chronic, multifactorial and progressive optic neuropathy characterized by morphological changes at the optic nerve (ON) head level and the retinal nerve fiber layer in the absence of other eye diseases or congenital abnormalities, the intraocular pressure being at present the main and most known risk factor [1-3].

POAG has been arbitrarily subdivided into high-tension glaucoma (IOP > 21mmHg without treatment) and normal tension glaucoma (IOP < 22mmHg without treatment).

Despite a normal IOP (i.e. Normal tension glaucoma - NTG), the proportion of patients with glaucomatous optic neuropathy (GON) seems to be increasing and varies considerably from one part of the world to another. Although extremely important, the relationship between IOP and GON is surprisingly weak at the bottom of the IOP spectrum (i.e. GTN), which indicates that other risk factors are involved [2-11]. Of these, it seems that vascular factors play an important role.

An argument to this is the 6th meeting of the “World Association of Glaucoma” (WGA) in 2009, which had the theme “Ocular blood flow in glaucoma”, and which was attended by over 200 ophthalmologists and researchers worldwide [8,12,13]. Also, at ESCRS 2014 in London a new device was launched by Optovue Company: Angio-OCT’s, and, at the EGS Congress in Nice, 2014, the Doppler-OCT device was announced, whose aim was just to measure the retinal blood flow.

Due to its size and location, retrobulbar circulation has been difficult to investigate. The occurrence of color Doppler ultrasound was achieved, thus opening new horizons in the investigation and diagnosis of the ocular vascular disease.

Objectives

The present paper analyzed the possible differences between retrobulbar blood flows in normotensive glaucomatous optic neuropathy patients versus hypertensive glaucomatous optic neuropathy patients, with intraocular pressure controlled treatment.

Material and Method

The study lot consisted of 102 patients (202 eyes) with a confirmed diagnosis of GON (normotensive and hypertensive), with a specific topical antiglaucoma treatment; mean IOP of 15,5mmHg (10,5-20,5mmHg); mean age of 66.41 years (sd = 9.57); women: 84 (82.4%) and men: 18 (17.6%).

The lot of patients with NTG = 16 (15.7%), 31 eyes:

-sex: women: 13 (81.25%) and men: 3 (18.75%)

- mean age: 66.87 years (sd = 9.78);

The lot of patients with HTG = 86 patients (84.3%), 171 eyes:

- Sex: women: 72 (83.72%) and men 14 (16.28%)

- mean age: 66.33 years (sd = 9.59)

All the patients in the study group underwent a color Doppler echography of retrobulbar vessels by using Acuson X300 device. Systolic (PSV) and diastolic (EDV) blood velocities were measured in both eyes in the ophthalmic artery (OA), central retinal artery (CRA) and posterior ciliary arteries (PCA), by using a linear transducer (VF10-5) with a frequency of 7.5 Mhz.

The Resistivity index Pourcelot (IR) was IR = PSV-EDV/ PSV, and it was calculated automatically by the device.

PSV (cm/ s) (peak systolic velocity) represents the highest speed of the blood flow during the systolic phase of the cardiac cycle.

EDV (cm/ s) (end diastolic velocity) represents the low speed of the blood flow obtained at the end diastolic phase of the cardiac cycle.

IR characterizes the peripheral vascular resistance of the blood vessels (RI reflects the resistance to the blood flow distal to the site of measurement) [14]. Its values ranges from 0 to 1 (higher values indicate a distal increased of vascular resistance) [15,16].

Data analysis

The data was collected and tested by using the SPSS software version 20. Further, the description of the data was assessed with the measures of central tendency for quantitative scales, the mean (the average), standard deviation, maximum, and minimum. After obtaining an indication of the typical scores in the data set, a normal distribution test was performed by using both a graphical representation and the Kolmogorov-Smirnoff test of Normality. The data sets had non-normal distributions, known in literature as non-parametric distributions.

Moreover, the analyses of the differences between the two conditions for a non-parametric distribution were conducted with the Mann-Whitney U test. The Mann-Whitney test is the non-parametric equivalent of the independent t-test. The Mann-Whitney test was selected due to its distribution-free assumption [17].

ROC curve

The ROC curve is an essential tool for diagnostic test evaluation. More exactly, in a ROC curve, the true positive rate (Sensitivity) is plotted according to the false positive rate (1-Specificity) for different cut-off points of a parameter [18].

Results

The values of hemodynamic parameters were obtained from measurements with color Doppler imaging (CDI) in the 2 groups of patients and were statistically analyzed.

The values obtained are summarized in two tables (Table 1,​22).

The only statistically significant differences between the patients with hypertensive glaucoma and the patients with normal-tension glaucoma were registered at the EDV of PCA level, respectively the decrease of EDV in subjects with NTG: Mann-Whitney U index being 2076 (p < 0.05) (Table 3).

ROC curve

As mentioned earlier, the Mann Whitney test revealed that there was a difference between the patients with hypertensive glaucoma and the ones with normal tension glaucoma registered at the end diastolic velocity (EDV) of the PCA.

Therefore, the ROC curve was employed to test whether the EDV in the PCA was a predictable parameter for patients with normal tension glaucoma. Additionally, to determine the power of the test, the area under the ROC curve was calculated. Table 2 illustrates the cut-off point when specificity and sensitivity have high values for the EDV parameter [19].

Fig. 1 illustrates all the ROC Curves, but only the end diastolic velocity had a significant area under the curve (p < 0.05) (Fig. 2).

Fig. 1

ROC curve for all the parameters

Fig. 2

EDV in PCA

The area under the curve is 0.70 which makes it a fair but acceptable test of prediction [20], meaning that EDV of PCA is a good predictor (Table 4).

The power to identify the predictive value of the patients with NTG by using EDV of PCA reaches 72% sensitivity at 33% specificity (cut-off > 5,95) (Table 5).

Discussions

The role of the vascular factor in glaucoma is not a new subject, having been discussed and researched for over 100 years, but in recent years, the techniques of visualization and measurement of ocular blood flow have developed in a spectacular way.

In 1858, Jaeger argued regarding the hypothesis that GON can have other, intrinsic causes, independent of the IOP, and in 1885 Smith suggested two factors involving the mechanical (IOP) and the vascular parts.

Over the time, retrobulbar vascularization and the blood flow at this level have been studied, investigated and various studies have been published by using different techniques.

While developing, CDI has proven useful in evaluating the ocular vascular flow by measuring the systolic and end diastolic blood velocity, too.

The initial studies compared the ocular blood flow in patients with glaucoma (normotensive or hypertensive) with the healthy subjects (Galassi et al., 1992, Harris et al., 1994, Butt et al., 1995, Nicolela, 1996, Chiou et al., 1999, Plange et al., 2009) [21,22].

In 1997, Kaiser et al. conducted a larger study that analyzed three categories of patients: with stable glaucoma, progressive glaucoma and normal tension glaucoma, finding significant changes in the hemodynamics of patients with progressive glaucoma and in those with normotensive glaucoma.

In a survey published in 1999, Rankin SJA did not find significant differences between the hemodynamic parameters of retrobulbar vessels in the NTG patients versus the HTG drug compensated patients [21]. Kuerten et al. and Plange underlined the same aspect in a study from 2015, in which an article published in 1997 by Butt et al. [23] was mentioned.

PCAs, which are branches of OA, directly or via peripapillary choroidal branches or short posterior ciliary arteries (or circle of Zinn-Haler when present) represent the only vascular source for the laminar and prelaminar regions and the main source (unless only) for the retrolaminar region of the nerve head [25].

The CDI characteristics for PCAs are the following: flow with low resistance, with a peak systolic, smoothly and continuously during diastole [25].

Due to the reduced caliber and direction, usually the variable insonation angles for the analysis of PCAs are required. Moreover, the large variability in their measurement is bigger than in the other vessels [26].

In healthy patients, these variations are between the following:

- 2,4% and 19% for PSV

- 2,7% and 38,8% for EDV

- 1,6% and 24% for IR

And in glaucoma patients:

- 15,8% and 26% for PSV

- 22,4% and 64% for EDV

- 4,2% and 69% for [25]

On the other hand, bigger sizes and locations more affordable of OA and CRA make their measurements easier and reproducible with CDI [26] and the variability of the measurements is much reduced both in healthy patients and in those glaucoma patients compared with the PCA [25].

EDV is considered more sensitive to hemodynamic changes than PSV and fluctuates, probably due to the low velocities of the blood flow to the end of the diastolic phase in retrobulbar circulation, while RI variability is lower compared with PSV and EDV in all the retrobulbar vessels [19].

Low diastolic ocular perfusion pressure (DOPP) is a major risk factor in glaucoma:

DOPP=DBP-PIO

DBP=Diastolic blood pressure

The low level of this optic nerve head leads to a compromise of the ocular blood flow and glaucomatous damage [24].

Perfusion pressure (PP) is influenced by the IOP and the pressure from OA [27]. Perfusion pressure may be decreased in conditions of increased IOP (increases the venous pressure at the outlet of the eye) or decreased arterial tension. However, the retinal blood flow and the prelaminar portion of the optic nerve head are less affected by a decrease in PP as long as it is at a level > 40mmHg, due to the autoregulation mechanism (metabolic and myogenic) of the blood flow at this level [24].

Chronic ocular ischemia can induce defects of vascular autoregulation and its inability to reduce high IOP and maintain adequate perfusion pressure. Circulatory insufficiency in PCA (main source of the vascularization for ON) is the main factor for the appearance of ischemic lesions of the ON [27]. The terminal nature of these vessels and their segmentary distribution explain typically ischemia which appears in a sector of the optic nerve head, called the “watershed zones” [27], localized in the temporal side of the optic disc, which also explains the susceptibility of upper and lower temporal areas of ON to ischemia in glaucoma and visual fields characteristic defects by default [24,28].

The fact that the perfusion of the optic nerve head is directly related to the retrobulbar circulation, accessible to direct evaluation with CDI, makes it a potential tool for the early evaluation of the changes in the vascular flow in glaucoma.

Conclusions

In past years, attention has been given increasingly to the vascular factor in glaucoma etiopathogeny. A proof of this are the papers presented at recent national and international glaucoma congresses and the improvement or identification of new investigative and measurement techniques for retrobulbar circulation and vascular flow at this level.

In the present study, changes of the mean values in all the retrobulbar hemodynamic parameters of the blood flow (except for PSV and EDV in CRA) have been found in NTG patients compared with HTG patients. Modifications with statistical significance have been recorded at the EDV in PCA respectively decreases in the NTG subjects: the Mann-Whitney U index was 2076 (p < 0.05).

The power to identify the predictive value of patients with NTG by using EDV of PCA reached 72% sensitivity at 33% specificity.

The vascularization of the outer zone of the retina was directly dependent on PCA, branches of OA, which induced the theory that the factors influencing the flow parameters of OA and CRA could influence the PCAs (the direct investigation being technically more difficult) in the same way. For these reasons, we believe that larger studies may find significant changes of the CDI parameters for the PCA in glaucoma.

CDI sensibility in glaucomatous optic neuropathy is influenced by both the inter-individual variability of the ocular hemodynamics in healthy eyes and in the glaucoma, and the reproducibility of the method of statistical analysis.

In addition, individual variability of ocular hemodynamics can be influenced by individual habits, systemic vascular disease, autoregulation capacity of retrobulbar vessels, etc. [19].

Considering these issues, the results of the present study need to be confirmed by further studies on larger groups of patients with NOG and, eventually, the evaluation of CDI parameters should be standardized.