OCTA provides quantitative vessel density measurement

February 25, 2019

Optical coherence tomography angiography (OCTA) may be changing the game when it comes to detecting glaucoma sooner, according to researchers.

“OCTA is non-invasive and of rapid execution; it could potentially elucidate vascular changes in glaucoma, and consequently detect glaucoma sooner,” according to Dr Ana I. M. Miguel, from the University of Portugal, and colleagues, who reported their finding in the British Journal of Ophthalmology.

Their analysis showed that vascular density is much lower in people with glaucoma than in people without glaucoma. OCTA is quicker, easier and more objective than visual fields in diagnosing glaucoma, according to the researchers.

While no one knows exactly what causes glaucoma, one theory is that axonal ischaemia causes decreased ocular blood flow and optic nerve perfusion.
So imaging methods evaluating microvascular changes in glaucoma have attracted interest.

In particular, OCTA provides quantitative vessel density (VD) measurement. Approaches differ by manufacturer:
• The RTVue of Avanti (Optovue) uses a split-spectrum amplitude-decorrelation algorithm to enhance the signal-to-noise ratio of the flow detection.
• The CIRRUS 5000 and the PLEX Elite (Carl Zeiss Meditec) detect differences in scattered light from moving particles, such as red blood cells in the vessels, to visualise and quantify the microcirculation within tissue beds in the eye, an approach called microangiography (OMAG). In theory, OMAG enhances contrast and provides a more precise segmentation and isolation of individual retinal layers.

To see how well OCTA can diagnose glaucoma, Dr Miguel and colleagues conducted a meta-analysis of all the studies they could find on the subject.

They defined VD as the percentage of area occupied by flowing blood vessels. They defined the peripapillary region as an annulus extending outward form the optic disc boundary, typically 750 µm wide.

They looked for prospective trials comparing VD in patients with glaucoma to VD in healthy controls. They found 2660 records of which 177 had full text available. Of these, 24 met their criteria.

As their primary outcome, they used the standardised mean difference (SMD) in the peripapillary VD between the glaucoma patients and the controls.

The secondary diagnostic outcomes included the whole-image optic nerve VD, the whole image macular VD and the parafoveal VD.

Only 18 of the studies met the researchers’ criteria for inclusion in their meta-analysis. These studies examined a total of 888 glaucomatous eyes and 475 healthy controls.

All the studies evaluated the radial peripapillary capillary segment. But the researchers found significant heterogeneity among the them (I2=59%).

In the studies included in the meta-analysis, the mean peripapillary VD in patients with glaucoma was 57.53%, with a 95% confidence interval (CI) of 52.60% to 62.46%. In the healthy controls the mean peripapillary VD was 65.47% with a 95% CI of 59.82% to 71.11%, yielding a SMD of –1.41 (95% CI 1.162 to -1.20, P<0.05).

The mean whole-image VD centred in the optic nerve head – a scan of 4.5 mm x 4.5 mm – had an SMD of -9.63 (95% CI -10.22 to -9.03, P< 0.05) and heterogeneity of I2 of 86%.
Two studies included the inside-disc VD. They had an SMD of -9.51 (95% CI -12.66 to -6.36, P< 0.05) with no heterogeneity. Two studies reported scans of 3 mm x 3 mm. They had an SMD of -4.81 (95% CI -5.97 to -3.66, P< 0.0001 and an I2 of 53%. In a subgroup analysis of OCTA devices, the 2 studies that used a Zeiss device had an SMD of -1.13. The other 17, using an Optovue device, had an SMD of -1.68.

And in a subgroup analysis of the risk of bias, those studies with the lowest risk had an I2 of 84%.

Overall, the area under the curve (AUC), when reported, was 0.983, suggesting that OCTA has a high diagnostic capacity. Studies that compared OCTA with OCT reached different conclusions: 4 found a superior AUC for OCTA, 5 found an inferior one and 2 an equivalent one.

Four studies found a better correlation between VD and visual fields than between OCT and visual fields. Those studies reporting the coefficient of variation of OCTA had values between 2.3% and 4.1%, which would give it better reproducibility than visual fields.

“This systematic review shows that peripapillary VD is decreased in patients with glaucoma compared with healthy controls,” the researchers concluded. “Future studies with a prospective longitudinal design are welcomed.”