Evaluating the anterior chamber angle with OCT

Objective assessment of narrow angles correlates well with gold-standard gonioscopy

Example of anterior segment OCT which would provide rapid and objective angle assessment for anterior chamber glaucoma diagnosis
Source: Julian P. S. Garcia Jr., M.D.

Evaluating the anterior chamber angle is vexing for many clinicians due largely to the subjectivity of the task. Is that angle narrow? Is it occludable? There are potential consequences to making the wrong call, from an unnecessary procedure to a full-blown attach of angle-closure glaucoma. Nowhere is this issue more important than in Asia, where angle-closure glaucoma represents a much larger proportion of all glaucoma than in the United States.
“Gonioscopy is the current reference standard to detect angle closure,” said Arun Narayanaswamy, M.D., Singapore Eye Research Institute, Singapore. “However, subjective influences govern its interpretation.”
Lisandro Sakata, M.D., Ph.D., Curitiba, Brazil, agreed. “Gonioscopy is a simple and inexpensive examination. However, gonioscopy requires training and experience, and the procedure requires contact with the cornea.”

Gonioscopy vs. anterior segment OCT

To address this significant public health issue, a team of researchers has conducted a large, cross-sectional study comparing gonioscopy to anterior segment optical coherence tomography (AS-OCT) using the Visante AS-OCT (Carl Zeiss Meditec, Dublin, Calif.) to identify eyes at risk for angle closure.
“AS-OCT is capable of providing high-resolution images of the angle, which allows for quantitative analysis,” said Dr. Narayanaswamy. “Semi-automated analysis of these images could be useful in detecting eyes at risk for angle closure.”
The research team performed gonioscopy and AS-OCT evaluations on 2,047 phakic subjects aged 50 years and older. All subjects were recruited from a community general health clinic in Singapore, and all were free of any ocular symptoms at the time of evaluation.
“The gonioscopic definition of angle closure was the inability to visualize at least 180 degrees of posterior pigmented trabecular meshwork on non-indentation gonioscopy with the eye in the primary straight-ahead position,” said Dr. Narayanaswamy. “Gonioscopy was performed on all subjects by an ophthalmologist masked to the AS-OCT findings.”
AS-OCT imaging provided several parameters. Among them, the angle-opening distance (AOD) representing the distance between the posterior corneal surface and the iris 250, 500, and 750 microns from the scleral spur were determined. These were calculated with customized software using the scleral spur as the operator-defined landmark.
“We excluded 522 subjects due to poor localization of the scleral spur in the AS-OCT images and 60 subjects due to image artifacts,” said Dr. Narayanaswamy. This left an analyzable data set that included gonioscopy and AS-OCT findings on 1,465 subjects.
“Of those, 315 (21.5%) had angle closure on gonioscopy,” he said. “Using AS-OCT, the AOD750 parameter was the most useful angle parameter to identify subjects with gonioscopic angle closure.”
The AOD750 had an AUC (area under the curve) of 0.9, he said. The AUC is a measure that reflects both sensitivity and specificity and has a maximum value of 1.0 for a perfect test.

Improved performance from better modeling

The research team collaborated with Dr. Sakata to develop additional AS-OCT parameters in hopes of improving the diagnostic ability of the device. These parameters, such as iris thickness, iris curvature, and anterior chamber width were combined with AOD parameters to develop a novel classification system to assess the risk of angle closure.
The optimal cut-off values for the classification system were determined using a randomly selected group of 1,000 of the 1,465 eyes. “If anterior chamber measurements were over the predetermined cut-off points with specificity locked at 99% or 95%, the status of the classification system was ‘outside normal limits’ or ‘borderline,’ respectively,” said Dr. Sakata.
Once the cut-offs for borderline and abnormal tests were determined, “the classification system was subsequently tested on an independent sample of the remaining 465 eyes in the cohort,” said Dr. Sakata.
Consistent with the full cohort of 1,465 eyes, the prevalence of gonioscopic angle closure was 21.5% and 21.6% in the 1,000-eye and 465-eye samples, respectively.
In the initial sample of 1,000 eyes, the sensitivity and specificity of the novel classification system were 91.6% and 84.6%, respectively. In the independent sample of 465 eyes, the sensitivity and specificity were 83.2% and 88.8%, respectively.

Significance and next steps

While gonioscopy is cheap and effective, it is time-consuming and subjective. Based on the data described above, AS-OCT represents a potentially important part of the solution to Asia’s angle-closure glaucoma problem. Fast and objective angle assessment could significantly improve screening for this potentially blinding condition in the world’s most populous nation.
“This is the first community-based study to evaluate quantitative AS-OCT parameters for the detection of angle closure,” said Dr. Narayanaswamy. “Quantitative analysis of AS-OCT images is a promising approach for detecting angle closure.” However, he said, “the inability to define the scleral spur correctly in a significant proportion of images represents a major challenge.”
Dr. Sakata also pointed out that all of the data in these studies came from a single clinic in Singapore. “This classification system should now be tested in other validation samples from different geographic locations,” he said, in order to ensure the generalizability of their findings to the broader Asian population.

Editors’ note: Drs. Narayanaswamy and Sakata did not indicate any financial interests related to their comments.

Contact information

Narayanaswamy: answamy22@gmail.com
Sakata: zmsakata@yahoo.com.br