Potential of the Pentacam in Screening for Narrow Angles in Patients with Chronic Angle-Closure Glaucoma

Background: Chronic angle-closure glaucoma (CACG) is a visually destructive disease. Effective management of CACG requires identifying eyes with narrow angle. Objective: To compare pentacam with gonioscopy in detecting narrow angles in eyes with CACG. Method: We enrolled 101 eyes with glaucoma. Gonioscopy was performed on all eyes. Using Shaffer’s grading, subjects were classified into angle closure and open angle. Anterior chamber volume (ACV) and anterior chamber depth (ACD) were measured with the pentacam. Receiver operating curve was constructed for each parameter and the area under the curve (AUC) was calculated. Results: Ten eyes (9.9%) were classified as angle closure on gonioscopy. To detect narrow angles, ACV (AUC 0.956; 95% confidence interval (CI) 0.894– 0.987) performed similar to ACD (AUC=0.930, p=0.33). Using a cutoff of 102 mm, ACV had 100% sensitivity and 88.5% specificity for detecting narrow angles in CACG patients. With an ACV cutoff of 102 mm, the PPV for detecting angle closure was 48.9% (95% CI, 34.8–68.2), while the NPV was 100% (94.1– 100%), using 9.9% prevalence of angle closure from this study. Conclusion: ACV and ACD measured by the pentacam have the potential to determine narrow angles in eyes with CACG.


Introduction
Glaucoma is a group of diseases that affect the optic nerve of the eye. It is characterized by a specific optic neuropathy associated with visual field defect (1). Glaucoma causes irreversible blindness worldwide and accounts for 15% of blindness in Africa (2). Compared with the rates of blindness in primary open-angle glaucoma (POAG), studies show that the risk of bilateral blindness is three times higher in primary angle-closure glaucoma (PACG) (3). Based on clinical observations, PACG is classified into two: the acute, symptomatic phase which is dramatic but occurs in only a minority of those with PACG, and the chronic, asymptomatic form which predominates (4). Chronic angle-closure glaucoma (CACG) occurs as a result of gradual and silent closure of the anterior chamber angle, which eventually results in increased intraocular pressure and, finally, damage to the optic nerve (5). Currently, gonioscopy remains the reference standard for assessing anterior chamber angle (ACA). However, this technique requires a proficient examiner to provide PENTACAM IN CHRONIC ANGLE CLOSURE GLAUCOMA a confident diagnosis and involves contact with the surface of the eye (6). This necessitates anesthetic agents and can cause artifacts. And even among experienced examiners, there is variability in angle grading due to the subjective nature of the assessment (7). In recent years, a Scheimpflug camera, Pentacam, has been developed to evaluate anterior chamber characteristics. The Oculus Pentacam operates on the principle of Scheimpflug imaging. It eradicates the challenges that are encountered with goniolens (8). The Oculus Pentacam provides relevant and reproducible bio data from the anterior corneal surface to the posterior lens surface in a single scan without contacting the cornea (9). However, it is unable to visualize the most peripheral part of the iris and automatically calculates the assumed apex of the ACA (10). The concordance between gonioscopy and pentacam has been found to be very good (10). This study aimed to compare the diagnostic ability of the pentacam and gonioscopy to detect narrow angles in patients with chronic angle-closure glaucoma.

Methods
This was a hospital-based cross-sectional study. Consecutive glaucoma patients aged 40 years and above receiving treatment at the Komfo Anokye Teaching Hospital during the period of December 2017 to April 2018 were recruited for the study. All subjects received clinical examination including visual acuity, slit-lamp biomicroscopy, pentacam examination, Goldmann applanation tonometry, and gonioscopy, in that order at the glaucoma clinic. Data received from the glaucoma clinic were recorded on a data collection sheet for each subject. Glaucoma patients who have had any form of ocular surgery, congenital and secondary glaucoma, and corneal disorders, and patients who had any form of ocular infection at the time of study were excluded from the study.

Pentacam examination
The pentacam measurement was taken by a trained optometrist masked to the gonioscopy results of participants. The Pentacam HR (Typ 70900 ©Oculus 2013, Oculus Optikgerate GmbH, Germany), device takes pictures of the anterior chamber by a rotating Scheimpflug camera. The camera illumination system consists of a blue light-emitting diode which is ultraviolet free with a 475-nm wavelength. The rotating camera process takes pictures in three dimensions ( Figure 1) and can also measure the center of the cornea precisely. The machine can take 50 meridional sections through the center of the cornea. This allows the system to align with the central thinnest point of the cornea at each section before it constructs the corneal image, and thus any eye movement is eliminated during the exam. The whole measurement process does not take more than 3 seconds (11). The participants were asked to focus on the blue fixation target without any compensation for their refractive error. When the target was focused, the rotating Scheimpflug camera captured 50 images automatically around the optical axis of the eye. The following parameters were recorded from the pentacam overview: anterior chamber angle (ACA, degrees); anterior chamber depth from the endothelium (ACD, mm); anterior chamber volume (ACV, mm 3 ) and central corneal thickness (CCT, µm). The ACV is calculated using an integral calculus, which considers the anterior chamber as a solid bounded by the posterior surface of the cornea (12-mm diameter around the corneal apex), the iris, and the lens (12). The Pentacam ACA measurements were not used in this study, as their reliability in eyes with narrow angles has been questioned due to the inability of the pentacam to visualize the most peripheral part of the iris and base of the ACA (6).

Gonioscopy
Gonioscopy was performed by an ophthalmologist masked to the pentacam results of the subject. It was performed using a Goldman 3-mirror lens using standard examination technique in a moderately dark room. A topical anesthetic and hydroxyethyl cellulose were applied before examination. ACA was graded using Shaffer's grading system (13). Subjects with grade ≤1 in 2 or more quadrants were classified as angleclosure whereas subjects with grade 2-4 in 2 or more quadrants without indentation were classified as having open-angle. If necessary, the presence and extent of peripheral anterior synechia (PAS) were confirmed by compression gonioscopy. If both eyes were eligible for the study, one eye was chosen at random for analysis. Study data were entered into Microsoft Excel 2013. Statistical analysis was done using IBM® SPSS® Statistics for Windows, version 23.0 and MedCalc version 19.2.1 (MedCalc Software). Clinical characteristics of the study population were expressed in mean values and standard deviation. As data were not normally distributed, Mann-Whitney Utest for unpaired data was used to determine differences between groups. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and the receiver operating characteristic (ROC) curves for pentacam parameters were calculated using gonioscopy as the reference. PPV and NPV were calculated using a prevalence of 9.9% from this study. Optimum cutoff levels were determined using the Youden index (J) J=(sensitivity+specificity-1) (14). The parameter value with the maximum Youden index was used as the cutoff value. Ethical approval was sought from the Committee on Human Research, Publication & Ethics (CHRPE) of the Kwame Nkrumah University of Science and Technology, School of Medical Sciences with reference number CHRPE/AP/172/18. A written informed consent was taken from all participants before enrollment.

Results
A total of 101 eyes (50 right and 51 left) were selected. This included eyes of 47 males (46.5%) and 54 females (53.5%). Mean age of the study population was 57.79±13.50 years (range 40-87 years). Mean intra ocular pressure (IOP) was 22.18±8.25 mmHg and mean visual acuity in LogMAR was 1.14±1.21. The mean corneal thickness at the thinnest locale was 508±41 µm. Ten eyes (9.9%) were classified as having chronic angle closure on gonioscopy (Shaffer ≤1 in 2 or more quadrants with or without PAS ROC curves for predicting narrow angles were constructed for ACV and ACD from the pentacam, using gonioscopy as the reference (Figure 2). Comparing the sensitivity and sensitivity at different values of ACV, we found that using a cutoff of 102 mm 3 , ACV had 100% sensitivity and 88.5% specificity in detecting narrow angles with area under the curve (AUC)=0.956 (95% confidence interval (CI), 0.894-0.987). Although ACV had a higher AUC than ACD (AUC=0.930; 95% CI, 0.860-0.972), the difference was not significant (p=0.33). Table 2 presents the cutoff values and the corresponding sensitivity, specificity, PPV and NPV for ACV and ACD. With an ACV cutoff of 102 mm 3 , the PPV for detecting angle closure was 48.9% (95% CI, 34.8-68.2), while the NPV was 100% (95% CI, 94.1-100%), using 9.9% prevalence of angle closure from this study.

Discussion
Chronic angle-closure glaucoma usually characterized by a silent closure of the ACA leads to severe optic nerve damage and visual field loss. Laser peripheral iridotomy is prophylactically recommended as treatment and prevention for narrow and occludable angles. However, in doing this it is imperative to identify people with narrow or occludable angles (15). Gonioscopy is clinically the gold standard for identifying occludable angles. It is valuable and timetested clinical test to diagnose the type of glaucoma. However, its subjective nature of grading and lack of defined cut-off points between open and closed angles make comparison and follow-up difficult (16). Dynamic gonioscopy using four-mirror gonioprism can deepen the narrow angle due to applied force, leading to false assessment of the angle especially when there is PAS. Due to the contact nature of the procedure, the possibility of spread of ocular infections and corneal abrasions has been considered during gonioscopy. Abbreviations: CACG, chronic angle-closure glaucoma; POAG, primary open-angle glaucoma; IOP, intra ocular pressure; ACV, anterior chamber volume; ACD, anterior chamber depth; ACA, anterior chamber angle; CTmin, corneal thickness at the thinnest locale. a Independent samples t-test, b Chi-square test, c Mann-Whitney U-test, Data represent the mean±SD (range), except for gender. Table 2: Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), for Pentacam parameters compared with gonioscopy for identifying narrow angles and area under the receiver operating characteristic curve (AUC) for parameters measured with the pentacam in detecting narrow angles using gonioscopy as the reference standard.
In the quest to overcome these weaknesses, several noncontact methods of angle assessment have been developed. Ultrasound biomicroscopy (UBM), anterior segment optical coherence tomography (ASOCT) and Scheimpflug imaging devices (Pentacam) are objective methods of assessing the anterior segment parameters (17,18). The efficacy of UBM for evaluating the ACA is well established; however, the procedure is cumbersome and operator dependent and thus not suitable for quick assessment on a large scale (10). ASOCT is a promising  Figure 1: Three-dimensional image representation of the anterior segment obtained by the Pentacam (Oculus, Wetzlar, Germany). Note that the ACV is reduced with a corresponding quantitative value (Red box). Figure 2: Receiver Operating Characteristic curve for ACV and ACD by Pentacam non-contact imaging tool of angle assessment, but its efficacy in screening is not yet established (19). Pentacam is a non-contact tool that objectively evaluates the anterior chamber with a non-significant interobserver variability (20). It is less expensive than UBM and ASOCT. It provides quantitative, reproducible and repeatable assessment of anterior segment structures (20,21).
In the current study, we compared the discriminating ability of the pentacam parameters (ACV and ACD) in detecting angle closure using gonioscopy as the reference. The ACV had a high discriminating ability (AUC=0.956) in screening for occludable angles with a sensitivity of 100% and specificity of 88.5% at a cutoff of 102 mm 3  Our study has several limitations; the study group was not population based and had a relatively small sample size of 101 eyes. Validating these findings in a more diverse population-based study would be useful. Although the reported sensitivity values in our study is 100%, none of the specificity values recorded was in the high 90s, which should be the ideal characteristic of a diagnostic device for screening purposes (24). We did not verify the reproducibility of pentacam measurements, as this has been established (20,21). Although repeatability of ACV measurements from the pentacam has been demonstrated with an intraclass correlation coefficient of 0.991, there are concerns that as ACV is reported with no decimal places, it may be affected by rounding error (25).

Conclusion
ACV and ACD measured by the pentacam have the potential to determine eyes with chronic angle-closure glaucoma. It has the advantage of objectivity, repeatability and quantification. These criteria can be helpful when deciding to proceed for laser peripheral iridotomy in chronic angle-closure cases.